Re: [neonixie-l] Re: Counterfeit RTC modules

[Luka C]

I have written previously about a similar incident with ATmega328p chips 
ordered from Aliexpress. They did not arrived in a reel, but just packed in 
a small plastic bag. Two of them worked fine, but Atmel Studio was unable 
to even read the ID of the third one, thus rendering it unprogrammable and 
useless..and since it's an SMD, it was not as easy as just popping it out 
of the socket. I guess they are chips which failed QC and somehow, these 
companies got them, performed a somewhat "basic check" and sold them as new 
ones.

I have also ordered a lot of 10 DS3231 chips from some other Aliexpress 
seller (they had considerable 5* ratings so I decided to give it a try). 
They were like 14$/10pcs. The chips worked, the time is still very precise 
after two years of operation of the older clock I made. However, I did 
notice one "bug" in the chip. If you keep querying the chip too often (like 
in an unconditional loop) for seconds, it will eventually hang the I2C 
line. The solution was to set a timeout for I2C read operation, and if 
timeout does occur, the master (microcontroller) has to keep clocking the 
SCL line until the DS3231 releases the bus. However, I've read some people 
reporting the same "bug" on multiple sites, so I'm not sure if it's a "bug" 
specific to possibly counterfeit chips or just a general bug that affects 
genuine ones as well.

On Thursday, January 3, 2019 at 1:41:52 AM UTC+1, gregebert wrote:
>
> For now I'm going to hold off on any experiments that require 
> code-changes. I'll keep this RTC chip online for additional experiments; 
> known-good RTC will go into the second board-set.
> It appears the alarm registers are testable with software; I dont have the 
> INT/SQW pin connected in my system.
>
> I'm really itching to get the NIMO lit-up and running. So close to 
> actually firing it up, but I wont attempt that unless everything is 100% 
> stable. Mindset is that I have only 1 chance to get it right and it must be 
> right the first time. Anything that can go wrong and remain uncorrected for 
> more than a few milliseconds is assumed to result in a dead NIMO tube, and 
> I only have 1. 
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/9529a437-2084-41b9-af82-158aa95ff505%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Increasing output current limit on MAX1771 power supply design

[Luka C]

@gregebert Thank you for the anode current limit suggestion. However, I 
have already produced the board so I might try it out in some future 
prototype. I have, however, scaled the current in each cathode down so that 
the sum of them all (when all segments light up) will not exceed the 
maximum current per tube. I do realize this will probably mean that the 
tube will have somewhat dimmer appearance, but hopefully not too much. I'll 
post results once I get to soldering the parts and trying it out!

@Nick Your site is a very valuable resource, this is my second clock using 
the HV power supply based on your design, the last one is now some two 
years in operation and works very well.

On Wednesday, January 16, 2019 at 4:29:23 PM UTC+1, Luka C wrote:
>
> Greetings everyone,
>
> It's been a long time since I posted here last time and I've completed my 
> modular clock design since then (photo attached). The main idea is to have 
> a compact smart single "control board" which you can then add different 
> number of daisy chained "display boards" for various tubes. Now, for the HV 
> power supply, I've used a MAX1771 design presented here: 
> https://desmith.net/NMdS/Electronics/NixiePSU.html . Tested it with 6x 
> IN-18 tubes and it drives them at 5mA per tube fine. Now, I'd like to 
> modify the power supply (without changing the board layout, considering all 
> components are SMD) so that it can power 6x B7971 tubes (at ~17mA per tube 
> with all segments lit is what I have in plan). The link with the power 
> supply description contains a paragraph on how to expand the design so that 
> it can handle such currents. The thing is that since the website was 
> published, the FET proposed there is no longer in production and the 
> inductor is not compatible with the SMD layout (the once that author 
> proposes for higher current is a bobbin type).
>
> After doing a little digging on replacement components, I've come up with:
>
> 1. Replacement FET candidate: *FQP22N30 
> (http://www.mouser.com/ds/2/149/FQP22N30-1009534.pdf 
> <http://www.mouser.com/ds/2/149/FQP22N30-1009534.pdf>) , *the difference 
> being that this FET has a maximum drain current of 21A, while the 
> discontinued one the author suggests has 32A.
> 2. Replacement inductor candidate: *SRP1265A-470M* 
> (*https://www.digikey.com/product-detail/en/bourns-inc/SRP1265A-470M/SRP1265A-470MCT-ND/4876720
>  
> <https://www.digikey.com/product-detail/en/bourns-inc/SRP1265A-470M/SRP1265A-470MCT-ND/4876720>*)
>  
> , *Irms = 6.5A, Isat = 9.5A  , *original one suggested by the author has 
> Idc(max) = 8.5A. The new replacement candidate has a max DC resistance of 
> 90mΩ, while the original one has 20mΩ (this would decrease efficiency 
> somewhat, but there doesn't seem to be any footprint-compatibile inductor 
> with lower DC resistance?).
> 3. Rsense *0.025Ω 2010 rated at 1W*.
>
> Do you think this would work?
>
> Thanks in advance and best regards,
> Luka
>
> [image: IMG_20190116_154939.jpg]
>
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/46821861-adb6-47d3-989a-2d75c4714436%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Increasing output current limit on MAX1771 power supply design

[Luka C]

Thank you very much for providing valuable information. 

As for B7971, what I meant to say is that 17mA would be the maximum current 
per tube (sum of current from each cathode) in case all segments are lit up 
at the same time (although this will rarely happen, obviously) and thus, 
that the power supply should handle 6x 17mA current when modified. I am 
aware that each segment has a specific maximum current defined as per 
datasheet, I have implemented current control in each cathode, as per 
@gregebert's suggestion last time we discussed it. I'm talking about a 
different "display board" for B7971 here, I don't plan to use the same one 
I use for IN-18 of course.

Dana srijeda, 16. siječnja 2019. u 18:12:28 UTC+1, korisnik SWISSNIXIE - 
Jonathan F. napisao je:
>
> Why do you plan to drive all the segments of a B7971? Do you realized that 
> this would need a resistor on each cathode? If you use one anode resistor, 
> the current changes depending on how many cathodes are grounded. This 
> results either in tube damage or not working/darker display.  I suggest to 
> multiplex this tube.
>
> About the mosfet, 21A or even 32A is overkill, but the problem is, that 
> usually mosfets that have a low RDSon and GateChange, have a lot of current 
> rated. If you use a Rsense of 25mOhm, the maximum current would be areound 
> 3.5A@5V, so a mosfet with a few amps more should work fine, but low current 
> fets usually have higher RDSon. This is why most DC/DC Converters use 
> overkill Mosfet.
>
> Rdson and GateChange influence the efficiency of the circuit, so lower is 
> better, your chosen mosfet has 160mOhm, you could go lower.  I've used the 
> FDPF44N25 a few times:
> https://www.mouser.ch/ProductDetail/512-FDPF44N25T
>
> The Inductor looks fine. 
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/861f31bc-ec52-4225-b24a-2bbc24e08771%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: PCB Panelization

[Luka C]

@Allen Thank you for all the help you have provided. I will let everyone 
know how the project turns out once I'm done with acquiring the BOM parts 
and assembling the boards, most likely in January :)

Dana četvrtak, 14. prosinca 2017. u 00:45:34 UTC+1, korisnik Allen Dutra 
napisao je:
>
> @Luka
>
> That is good spacing. Some designers put in to few mouse bites but you 
> have good spacing.
>
> Allen
>
> On Wednesday, December 13, 2017 at 3:38:37 PM UTC-8, Luka C wrote:
>>
>> @Allen I was thinking about 1mm spacing between the centers of mouse bite 
>> holes, each one of 0.5mm diameter? As for the distance between breakaway 
>> tabs themselves, 25mm between the breakout tabs, this gives me space for 3 
>> tabs connecting the two PCBs
>>
>> Dana srijeda, 13. prosinca 2017. u 04:37:05 UTC+1, korisnik Allen Dutra 
>> napisao je:
>>>
>>> Hi Luka,
>>>
>>> That panel configuration looks good. What is your mouse bite spacing?
>>>
>>> I general don't use a stencil when hand soldiering a board. Don't find 
>>> then useful. That said there is a fine art to design to prevent defects 
>>> during mass assembly. Elongating paste shapes, changing paste shapes, 
>>> getting a stencil with the correct thickness, etc.
>>>
>>> Seeing your not making hundreds or thousands of these, this isn't a huge 
>>> concern.
>>>
>>> Allen
>>>
>>> On Tuesday, December 12, 2017 at 11:49:15 AM UTC-8, Luka C wrote:
>>>>
>>>> Thank you very much for your valuable feedback. 
>>>> Just to make sure, this would be alright:
>>>>
>>>> --
>>>> |   SMALLER PCB   |
>>>> --
>>>>)¤¤¤(  )¤¤¤(  )¤¤¤(
>>>> --
>>>> |  LARGER PCB  |
>>>> --
>>>>
>>>> the shorter PCB above the longer one connected by mouse bite tabs?
>>>>
>>>> Oh, and since I will be ordering a stencil along with the PCBs, while 
>>>> googling something related to it, I found this: 
>>>> https://learn.adafruit.com/smt-manufacturing/laser-cut-stencils
>>>>
>>>> The only thing I find interesting is where they suggest to make 
>>>> modifications to the gerbers in order to shrink the size of the SMD pads: 
>>>> "You'll 
>>>> now see that your pads are thinner. This prevents bridging since the laser 
>>>> is not perfectly precise and tends to 'go over' the boundaries by a few 
>>>> mils."
>>>> Is it really needed? I have some QFN components with small pitch 
>>>> between pads, I will be ordering the electropolished stencil for added 
>>>> precision. 
>>>>
>>>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/f55a1f31-52bf-48cd-81cb-edf8a7f157ea%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: PCB Panelization

[Luka C]

@Allen I was thinking about 1mm spacing between the centers of mouse bite 
holes, each one of 0.5mm diameter? As for the distance between breakaway 
tabs themselves, 25mm between the breakout tabs, this gives me space for 3 
tabs connecting the two PCBs

Dana srijeda, 13. prosinca 2017. u 04:37:05 UTC+1, korisnik Allen Dutra 
napisao je:
>
> Hi Luka,
>
> That panel configuration looks good. What is your mouse bite spacing?
>
> I general don't use a stencil when hand soldiering a board. Don't find 
> then useful. That said there is a fine art to design to prevent defects 
> during mass assembly. Elongating paste shapes, changing paste shapes, 
> getting a stencil with the correct thickness, etc.
>
> Seeing your not making hundreds or thousands of these, this isn't a huge 
> concern.
>
> Allen
>
> On Tuesday, December 12, 2017 at 11:49:15 AM UTC-8, Luka C wrote:
>>
>> Thank you very much for your valuable feedback. 
>> Just to make sure, this would be alright:
>>
>> --
>> |   SMALLER PCB   |
>> --
>>)¤¤¤(  )¤¤¤(  )¤¤¤(
>> --
>> |  LARGER PCB  |
>> --
>>
>> the shorter PCB above the longer one connected by mouse bite tabs?
>>
>> Oh, and since I will be ordering a stencil along with the PCBs, while 
>> googling something related to it, I found this: 
>> https://learn.adafruit.com/smt-manufacturing/laser-cut-stencils
>>
>> The only thing I find interesting is where they suggest to make 
>> modifications to the gerbers in order to shrink the size of the SMD pads: 
>> "You'll 
>> now see that your pads are thinner. This prevents bridging since the laser 
>> is not perfectly precise and tends to 'go over' the boundaries by a few 
>> mils."
>> Is it really needed? I have some QFN components with small pitch between 
>> pads, I will be ordering the electropolished stencil for added precision. 
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/4c40dbb1-d765-478b-9135-07c6019b079b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: PCB Panelization

[Luka C]

Thank you very much for your valuable feedback. 
Just to make sure, this would be alright:

--
|   SMALLER PCB   |
--
   )¤¤¤(  )¤¤¤(  )¤¤¤(
--
|  LARGER PCB  |
--

the shorter PCB above the longer one connected by mouse bite tabs?

Oh, and since I will be ordering a stencil along with the PCBs, while 
googling something related to it, I found this: 
https://learn.adafruit.com/smt-manufacturing/laser-cut-stencils

The only thing I find interesting is where they suggest to make 
modifications to the gerbers in order to shrink the size of the SMD pads: 
"You'll 
now see that your pads are thinner. This prevents bridging since the laser 
is not perfectly precise and tends to 'go over' the boundaries by a few 
mils."
Is it really needed? I have some QFN components with small pitch between 
pads, I will be ordering the electropolished stencil for added precision. 

Dana subota, 9. prosinca 2017. u 22:36:39 UTC+1, korisnik Allen Dutra 
napisao je:
>
> Hi Luka,
>
> You are correct. Everyone that I've ever talked to, looks at that mess of 
> board outlines you've circled and says "Mouse Bite". Ultimately any board 
> outline is provided to a CNC technician and *interpreted *to create tool 
> routing. Tool routing that is offset from the board outline by the drill 
> radius and concave acute angles rounded. Often a good PCB Fabricator (fab 
> shop) will point out areas that can't be CNC routed and work with the 
> customer on a solution.
>
> Take way: board outlines and mouse bites are always interpreted by a 
> technician to make sure the CNC routing works correctly. I've yet to see 
> mouse bites done incorrectly though I've have seen the drill size, spacing, 
> and count adjusted based on board thickness. If your building a thinner 
> than standard boards a fab shop may adjust the mouse bite specification for 
> reliability. Translation, make sure the mouse bite doesn't break in the 
> factory but can be snapped latter by the customer.
>
> Done correctly you shouldn't need cutting tools to snap mouse bites.
>
> Allen
>
> On Sat, Dec 9, 2017 at 5:26 AM, Luka C <luka.culi...@gmail.com 
> > wrote:
>
>> @Allen I'm planning on assembling them "by hand", manually pick and place 
>> SMD components and bake them inside a reflow oven. I guess I'll try 
>> following the Dangerous Prototyping post today.
>> The only thing I do not understand is the part I've marked in yellow on 
>> the picture here: https://i.imgur.com/HmpFPwv.jpg
>> Those break tabs contain outlines of both the first and the second board 
>> as well as the arc. I suppose the technician at the fab house then alters 
>> this in a way that the machine only cuts the arcs and not the board outline 
>> directly around it as it would cut the two boards apart?
>>
>> @gregebert I'm was thinking about cutting them with something like that 
>> indeed, I'd just like to make it easier by having empty space between them 
>> and having mousebites to simplify cutting them apart.
>>
>> Dana subota, 9. prosinca 2017. u 04:17:25 UTC+1, korisnik Allen Dutra 
>> napisao je:
>>>
>>> Hi Luka,
>>>
>>> Allen Dutra, professional PCB designer currently working for Apple and 
>>> with a lot of first hand experience with PCB panels.
>>>
>>> In your case you want to use mouse bites to combine two different sized 
>>> rectangles. V-Scores won't work here. The mouse bite instructions from 
>>> dangerous prototypes will work well for standard 0.064" designs. Question, 
>>> are you soldering these boards by hand or having them built on an assembly 
>>> line? (I could guess your answer but no assumptions from me) Assembly lines 
>>> will want tooling rails 
>>> <https://www.pcbuniverse.com/articles/Tooling%20Rails,%20Holes%20and%20Fiducials.png>
>>>  
>>> for best results but these are easily forgotten.
>>>
>>> I haven't use the PCB Panelizer tool that you linked to. Generally I 
>>> design my panel in the ECAD tool (Eagle CAD, Allegro, Altium etc.) I'm 
>>> already using to ensure Gerber accuracy.
>>>  
>>> Let me know if you have other questions,
>>>
>>> Allen
>>>
>>> On Friday, December 8, 2017 at 5:07:06 PM UTC-8, Luka C wrote:
>>>
>> -- 
>> You received this message because you are subscribed to a topic in the 
>> Google Groups "neonixie-l" group.
>> To unsubscribe from this topic, visit 
>&

[neonixie-l] Re: PCB Panelization

[Luka C]

@Allen I'm planning on assembling them "by hand", manually pick and place 
SMD components and bake them inside a reflow oven. I guess I'll try 
following the Dangerous Prototyping post today.
The only thing I do not understand is the part I've marked in yellow on the 
picture here: https://i.imgur.com/HmpFPwv.jpg
Those break tabs contain outlines of both the first and the second board as 
well as the arc. I suppose the technician at the fab house then alters this 
in a way that the machine only cuts the arcs and not the board outline 
directly around it as it would cut the two boards apart?

@gregebert I'm was thinking about cutting them with something like that 
indeed, I'd just like to make it easier by having empty space between them 
and having mousebites to simplify cutting them apart.

Dana subota, 9. prosinca 2017. u 04:17:25 UTC+1, korisnik Allen Dutra 
napisao je:
>
> Hi Luka,
>
> Allen Dutra, professional PCB designer currently working for Apple and 
> with a lot of first hand experience with PCB panels.
>
> In your case you want to use mouse bites to combine two different sized 
> rectangles. V-Scores won't work here. The mouse bite instructions from 
> dangerous prototypes will work well for standard 0.064" designs. Question, 
> are you soldering these boards by hand or having them built on an assembly 
> line? (I could guess your answer but no assumptions from me) Assembly lines 
> will want tooling rails 
> <https://www.pcbuniverse.com/articles/Tooling%20Rails,%20Holes%20and%20Fiducials.png>
>  
> for best results but these are easily forgotten.
>
> I haven't use the PCB Panelizer tool that you linked to. Generally I 
> design my panel in the ECAD tool (Eagle CAD, Allegro, Altium etc.) I'm 
> already using to ensure Gerber accuracy.
>  
> Let me know if you have other questions,
>
> Allen
>
> On Friday, December 8, 2017 at 5:07:06 PM UTC-8, Luka C wrote:
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/67036d33-d5a0-465e-8f20-57f96a9acbe2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] PCB Panelization

[Luka C]

I'm done with routing two PCBs which I'll be sending to fab house this 
weekend hopefully. While using the price calculator, I found out that if I 
am able to put both of them in the same order, it would turn out to be like 
50% cheaper and I could use the saved money for BOM parts. Now the thing is 
these are two different boards, one of size 42x84mm and the second one 
42x92mm (I entered the total value of 100x100mm in the calculator for price 
quote). They are both of rectangular shape, nothing irregular there. Since 
I have never made a panel before, I'd like to ask if anyone has any 
experience with it that he would like to share?

This is what I've found so far. There are two frequently mentioned options, 
V-scoring and break tabs + mouse bites. I kind off like the latter, so I 
browsed on how to do it properly in Eagle.

I've found this post:

"I make a slot of 50 mill wide (which is well bigger then what is allowed: 
32mil (0.8mm)). To add mousebites I use the line tool with layer dimension, 
width 0 and draw an arc to make the end of a slot round. the width of the 
mouse bite I use is 75 mil. Next I place the smallest hole possible 25 mill 
apart on both sides."

http://dangerousprototypes.com/forum/download/file.php?id=11775=view

I've also taken a look at PCB Panelizer tool:

http://blog.thisisnotrocketscience.nl/projects/pcb-panelizer/

https://hackaday.io/project/19202-small-pcbs-for-panelizing-tutorial/log/59210-panelization-using-gerberpanelizer-on-windows-linux-possible

Has anyone used this tool?
I downloaded it and tried it out, the included gerber viewer seems to 
display everything correctly in the final panel layout. I'm just a bit 
worried whether there may be any bugs that could cause displacement of 
inner layers (power+ground planes) in relation to the outer layers as I 
place components manually a lot using a very fine grid (which means X and Y 
positions of components typically have many decimal places), but I guess 
someone would have already reported such problems.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/855d9db7-c624-423e-8033-2795c227afdf%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: Global Shipping Program disaster, need help

[Luka C]

The only Nixie tubes that I have ordered from the U.S. are B-7971, I think 
all went through GSP, arrived safely and had no problems clearing their 
export control, fingers crossed in case of future orders :)

Dana petak, 8. prosinca 2017. u 09:26:41 UTC+1, korisnik Manuel Azevedo 
napisao je:
>
> As I mentioned in another thread about GSP, they disposed 6 DA-2021 
> numitrons because “they contained mercury”. 
> As someone pointed out, it was maybe the shinny getter on the glass, which 
> might have triggered a “oh my god, mercury” reaction, which made GSP 
> dispose of the tubes, which in this case are real lamps, as numitrons just 
> have glowing filaments, like ordinary light bulbs. 
> I was reimbursed, but I could never find numitrons again at that prce. It 
> was an excellent deal.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/c9d250d7-94cf-448a-a210-5e6c4cace089%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] New cazyness...

[Luka C]

I didn't notice until now, but this was his listing as well:
https://www.ebay.com/itm/Sockets-driver-boards-acrylic-case-for-huge-B-7971-nixie-tubes-/253280344586?_trksid=p2047675.l2557=true=%252BwLb3fpLJSR3Pq2fJfKBqd%252BbX4I%253D_cvip=true=nc

He basically asks a minimum bid of 300$ for 2 untested Ultronic driver 
boards and a piece of plastic with six sockets inside? He must be joking.

Dana ponedjeljak, 4. prosinca 2017. u 17:30:34 UTC+1, korisnik Jeff Walton 
napisao je:
>
> Yep.  Same vendor.  Selling same batch of B7971's with a couple obviously 
> bad tubes right in the photo.  Listing is pickup only and he will try to 
> get you to purchase outside of eBay.  
>
> He's been reported to eBay a couple times but keeps coming back... 
>
>  Original message 
> From: Nicholas Stock  
> Date: 12/4/17 10:23 AM (GMT-06:00) 
> To: neoni...@googlegroups.com  
> Subject: Re: [neonixie-l] New cazyness... 
>
> He's at it again!
>
>
> https://www.ebay.com/itm/1-pc-HUGE-B-7971-B7971-NIXIE-TUBE-FOR-COOL-CLOCKS/253293292291?_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D49452%26meid%3D8b64270446fa48a49f5ad7895b23ee2e%26pid%3D100675%26rk%3D1%26rkt%3D15%26sd%3D253293292291&_trksid=p2481888.c100675.m4236&_trkparms=pageci%253A71202c1b-d90f-11e7-b3a4-74dbd180abe0%257Cparentrq%253A2256c6101600aa479d89ece8fff1feaf%257Ciid%253A1
>
> On Sun, Dec 3, 2017 at 10:46 AM, Jeff Walton  > wrote:
>
>> I reported this one too.
>>
>>  
>>
>>  
>>
>> *From:* neoni...@googlegroups.com  [mailto:
>> neoni...@googlegroups.com ] *On Behalf Of *Nicholas Stock
>> *Sent:* Sunday, December 3, 2017 12:35 PM
>> *To:* neoni...@googlegroups.com 
>> *Subject:* Re: [neonixie-l] New cazyness...
>>
>>  
>>
>> Ouch...
>>
>>  
>>
>> Jeff, did you relay that message to eBay? What an ass
>>
>>  
>>
>> On Sun, Dec 3, 2017 at 10:29 AM, Jeff Walton > > wrote:
>>
>> Fun!  Used tubes with 0.3” digits for $999.J
>>
>>  
>>
>> This is the same vendor that was selling obviously defective B7971 tubes 
>> about a month ago.  There’s something off with this dealer.  He seems to 
>> offer many items for pickup only in the Washington DC area.  
>>
>>  
>>
>> On this listing for used BD-200 tubes @ $999, he offers the tubes for 
>> shipping or:  *“Or cash on PICKUP in Woodstock Virginia”*.  EBay 
>> probably just hasn’t yanked the listing yet.
>>
>>  
>>
>> When he was selling the defective B7971 tubes, I sent him a PM regarding 
>> selling items that were known defective and suggested that folks in the 
>> nixie community look out for each other.  He got no bids and this was the 
>> message that I got from him:
>>
>>  
>>
>> We all know the expression, “There’s a sucker born every minute”  but it 
>> is unfortunate that there are listings on eBay that rely on that principle.
>>
>>  
>>
>>  
>>
>>  
>>
>> *From:* neoni...@googlegroups.com  [mailto:
>> neoni...@googlegroups.com ] *On Behalf Of *SWISSNIXIE - 
>> Jonathan F.
>> *Sent:* Sunday, December 3, 2017 11:43 AM
>> *To:* neonixie-l
>> *Subject:* [neonixie-l] New cazyness...
>>
>>  
>>
>>
>> We all know some rare tubes are sold for a lot of money but this is 
>> probably a little to enthusiastic from the seller :)
>>
>> https://www.ebay.com/itm/253280307539
>>
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "neonixie-l" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to neonixie-l+...@googlegroups.com .
>> To post to this group, send email to neoni...@googlegroups.com 
>> .
>> To view this discussion on the web, visit 
>> https://groups.google.com/d/msgid/neonixie-l/cdae9b5f-bd59-4604-8c09-0bece06cc95e%40googlegroups.com
>>  
>> 
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "neonixie-l" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to neonixie-l+...@googlegroups.com .
>> To post to this group, send email to neoni...@googlegroups.com 
>> .
>>
>> To view this discussion on the web, visit 
>> https://groups.google.com/d/msgid/neonixie-l/5a2442db.4dad240a.fbbf7.ffd5%40mx.google.com
>>  
>> 
>> .
>>
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>>  
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "neonixie-l" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to neonixie-l+...@googlegroups.com .
>> To post to this group, send email to neoni...@googlegroups.com 
>> .
>> To view this discussion on the web, visit 
>> 

[neonixie-l] Re: Global Shipping Program disaster, need help

[Luka C]

Update: I have contacted the seller sending pictures and the unboxing video 
and explaining the situation. The seller was extremely helpful and 
contacted eBay first to explain the situation to them. After that, I called 
the eBay's Customer Service and explained everything again and upon looking 
into the photos, the unboxing video and the previous conversation with the 
seller, the customer support agents were able to conclude that it was 
neither his (seller's), nor my fault and they have issued me a full refund 
stating it was the fault of their courier's bad handling. I will now try to 
find someone who can repair the damage since I'm not really skillful when 
it comes to working with wood.

@J Forbes: The seller had actually put a large "pillow" filled with foam 
peanuts inside the case of the radio but, most likely, when eBay/Pitney 
Bowes repackaged the item in their sorting center in Erlanger, they have 
turned the radio around, so that the front face with the speaker was facing 
upwards (probably thinking that it would protect the glass dial better that 
way). This combined with how packages are usually tossed around when 
loading/unloading on planes (and the fact that the package is quite large 
and heavy, could have easily dropped from someone's hands in the process) 
at some point caused the weight of the speaker to break the entire front 
part off.

Just a note to buyers, when you receive a package, get your cell 
phone/camera and record the unboxing (and make sure you record the shipping 
label with the tracking number!), this seems to help a lot with speeding up 
the resolution of your case as it is a definitive proof that you did not 
damage the item yourselves. I have recorded every delivery so far and have 
never had a problem before this, but taking a few more seconds to record 
can save a lot more dollars!

@blkadder: I think Global Shipping Program is only available to sellers 
from the US and UK.

Dana ponedjeljak, 4. prosinca 2017. u 23:43:02 UTC+1, korisnik J Forbes 
napisao je:
>
> The description of the damage, and the picture, both make it look to me 
> like this is an item that needed extra care to get it ready for shipment. 
> It probably was not obvious to the seller that the radio either had a basic 
> structural design flaw, or else was missing fasteners. The heavy speaker 
> appears to have been supported only by two flimsy strips of wood. If the 
> speaker board had been securely screwed to the inside front of the case, it 
> probably would have survived.
>
> It's a shame it was damaged...but we all can at least learn to be more 
> careful with stuff that we ship. Heavy things need to be securely fastened! 
> It's a basic principle of packing things for shipment, that many folks seem 
> to be oblivious to. I've received several damaged items because of the same 
> problem. If something heavy can move inside a package, it will break stuff.
>
> Hopefully the damage can be repaired without too much loss of appearance. 
> thanks for sharing your experience!
>
> On Monday, December 4, 2017 at 4:26:13 AM UTC-7, Luka C wrote:
>>
>> this morning, I was excited that I'll finally enjoy it only to find out 
>> that the radio is heavily damaged, the front wooden face has an entire 
>> speaker portion broken off. "Luckily", I always record the entire process 
>> of opening 
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/a13e82cf-3cc3-4538-955c-7630168c22f6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Global Shipping Program disaster, need help

[Luka C]

I'm sorry if this is considered off topic, I just know Global Shipping 
Program was discussed many times here and I believe this could serve as 
sort of a warning on what could happen to international buyers. Of course, 
it depends on how this will be resolved, I have sent the seller the picture 
and the video of unpacking and hopefully will start a refund case soon.
I have previously purchased B-7971 tubes using GSP, it was rather slow in 
terms of "priority" that it claims to be, but the package has arrived 
without damage...I am not sure whether I will order anything of such value 
again.

As for forbidden items, I'm not even sure they do such "diagnostics" on 
materials, they probably just use "approximation" of whether something 
might be problematic based on previous items that were problematic, just my 
guess.

In the end, I just feel some sort of "guilt" because I have asked the 
seller to enable GSP (thinking it would be the safest option for him 
because eBay/Pitney Bowes takes responsibility for damage in transport) and 
now this happens. I wouldn't even mind that much if it was a laptop or 
something that can just be easily returned and replaced, but items such as 
these and Nixies will never be produced again and seeing them destroyed in 
this way just breaks my heart.

Dana ponedjeljak, 4. prosinca 2017. u 13:00:33 UTC+1, korisnik Shaun 
Merrigan napisao je:
>
> Not strictly related to nixies but cautionary nonetheless is what follows. 
> I have had the following items deemed restricted and thus could not be 
> shipped from the USA to me (in Canada):
>
> Heathkit W-7A amplifier 
> Tektronix 130 LCR Meter
> HP 410B VTVM
>
> This has occurred within the last two months. My guess as to why this 
> occurred would be (in order):
>
> Selenium Rectifier
> VR Tube with isotope OA2
> VR Tube with isotope OB2
>
> Of course no explanation is given and no appeal is possible. 
>
> Shaun M
>
> Sent from my iPad6
>
> No Regret
>
>
> On Dec 4, 2017, at 04:26, Luka C <luka.culi...@gmail.com > 
> wrote:
>
> Hi everyone,
>
> As I live in EU, I've used Global Shipping Program quite a few times to 
> order things from the US. Recently, I saw a beautiful old RCA Victor 
> tombstone radio that was being sold on eBay. I asked the seller if he could 
> enable GSP on the item so I could bid on it, after he contacted eBay 
> representative he decided it's ok and enabled it. In the end, I won the 
> auction and the seller shipped the package. Long story short, it has just 
> arrived this morning, I was excited that I'll finally enjoy it only to find 
> out that the radio is heavily damaged, the front wooden face has an entire 
> speaker portion broken off. "Luckily", I always record the entire process 
> of opening the package so I have proof that the item was indeed damaged in 
> transport and no by my fault in any way.
>
> Has anyone had any such problems before? Do I contact the seller, open a 
> case or contact eBay (I can't seem to find their telephone any more on the 
> site?).
>
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+...@googlegroups.com .
> To post to this group, send email to neoni...@googlegroups.com 
> .
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/f32ea6ff-6d74-4e6e-9d1d-840a03c2bcaf%40googlegroups.com
>  
> <https://groups.google.com/d/msgid/neonixie-l/f32ea6ff-6d74-4e6e-9d1d-840a03c2bcaf%40googlegroups.com?utm_medium=email_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>
> 
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/04be1f44-4b0f-4be6-88a8-5fe469940652%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] PCBWay

[Luka C]

I will be placing a PCB order soon so I have checked Elecrow that was 
mentioned in this topic. They have a pdf with details on SMD stencil 
production ( 
https://www.elecrow.com/download/Readme%20before%20ordering%20stencil.pdf ) 
. In the document they state that they will use the "paste" layer to 
manufacture the stencil except in case there is an "obvious RF antenna". 
Considering my PCB contains an UFL antenna connector, what would this mean? 
Perhaps someone knows?

Dana petak, 1. prosinca 2017. u 22:11:03 UTC+1, korisnik Tomasz Kowalczyk 
napisao je:
>
> At the place I work at we're ordering our PCBs from PCBway, about 25 
> different designs in last 12 months. I generally recommend them, however it 
> seems that they have more than one plant and/or are outsourcing the job to 
> smaller companies if they get overloaded - shades of soldermask are 
> inconsistent and once we received really low quality PCBs (it looked like 
> the drilling machine was working during an earthquake, drills were 
> misplaced by as much as 0,3mm).
> Their silkscreen accuracy is consistently poor, there is always a slight 
> offset. And the order number...  My boss allows me to fill the PCBs up to 
> 10cm x 10cm with my designs, so I used some extra space to make a board for 
> a B-5445 clock. I took some effort to hide the designators under the 
> sockets, but PCBway didn't see it my way and they placed the order number 
> in plain sight. Usually they hide it under some IC.
> I don't mind the shipping cost, we're using DHL anyway :)
>
> I have to try Elecrow. I'd like to compare the quality of those two 
> companies. What were the smallest holes/traces/clearances that you've tried?
>
>> I am still running with Elecrow. 
>>
>> Elecrow's 10 boards are fractionally cheaper at $4.90, not $5.00. Elecrow 
>> do any profile milling and sub boards within the cost, PCBway want to 
>> charge per sub board. PCBway load on the shipping, which for me in UK is 
>>  $12.00. A total of $17.00. 
>>
>> My last Elecrow order for the 10 boards was $4.90 + $4.80 shipping, a 
>>  total of $9.70. 
>>
>> Elecrow send me a picture of what they are sending me when it ships. I've 
>> had one issue with a board, out of several dozen, and they remanufactured 
>> immediately without any argument. They had misunderstood a plated-through 
>> tag slot pad and joined it to a ground plane. 
>>
>> Elecrow do not put any of their own annoying identification codes on the 
>> PCBs and if they manufacture any extra PCBs these will be included free 
>> also. My last order yielded 13 rather than the 10 I paid for. 
>>
>> John S 
>>
>>
>>
>>
>> On 1 Dec 2017, at 17:29, Terry S wrote: 
>>
>> > PCBway has a sale on boards -- 10 boards for $5 + shipping. Not sure 
>> what size that goes to, and you do need to use their default parameters. 
>> But if you can live with them this is a great deal. The default parameters 
>> are pretty much OK and what you might expect. 2 layer, 6 & 6, 1.6 thick, 
>> .3mm drill, green mask, white silk, 1 oz copper, HASL. 
>> > 
>> > I plugged in 100 x 100 mm and for 5 boards the price came to $5. Not $5 
>> each, but $5 total. 
>> > 
>> > 
>> > Terry 
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/49f25cc7-6fc8-45d2-a8d7-764d195869bc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Nixie Voltmeter - Vaping machine

[Luka C]

I have been vaping for more than 3 years now. I'm not sure where you've 
found the 2-15ohm information on coils, but for most people that vape, 
including myself, sub-ohm vaping is the way to go these days. This pretty 
much includes coils of around 0.4ohm-0.8ohm. Also it is important to 
mention the difference between the non-regulated (so called "mech" / 
"mechanical") mods and regulated mods. The mechanical ones are pretty much 
like a basic circuit with a 18650 battery and a coil, when you press the 
switch, the circuit closes and the current heats up the coil...these are 
however dangerous as in the event of a short circuit the battery will heat 
up rapidly and might even blow up (this unfortunately happens to 
inexperienced users and brings bad publicity to these life-saving 
devices!). On the other hand, regulated ones both have the protection 
circuit in place (reverse polarity protection, short circuit, overcurrent, 
etc) and allow the user to apply specific wattage or voltage to the coil 
(referred to as VV/VW - Variable Voltage/Wattage), thus influencing the 
vapor production, taste, temperature. There is also a thing called TC 
(temperature control) which allows the user to apply specific temperature 
to the coil, thus allowing tight regulation of flavor and preventing the 
possible emission of formaldehyde by not overheating the e-liquid or 
causing dry burns on the cotton wick in the coil.

All in all, I think you could design a regulated mod with nixies to display 
variable voltage / wattage setting. Of course, keep in mind that these 
devices need to be able to fit in our palm and will be exposed to constant 
movement which is sometimes a challenge for nixies' lifespan :)

Dana četvrtak, 19. listopada 2017. u 17:39:45 UTC+2, korisnik Paul Parry 
napisao je:
>
> Hi All,
>
> I've been asked if I can make a Vape machine with a little 2 digit IN-17 
> volt display built in. After doing some research ( I know nothing about 
> Vaping machines ) there are some coils anywhere between 2 - 15 ohm and a 
> couple of Lithium Ion cells. They heat up the liquid and then create the 
> vapour.
> From an electronics point of view it will need a fancy power supply as it 
> will have to be powered from the 4 - 7v DC produced by the Lithium cells 
> that it is also trying to manage.
>
> Has anyone experimented with a Nixie Voltmeter, or Vaping machines that 
> can help out a little?
>
> Cheers,
> Paul
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/b578efd4-2c39-43a6-ac28-3e27a8eec7a5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: BR355 GPS receivers for $30 on EBay

[Luka C]

But what if the NSA had radio transmitter and bug secretly planted in your 
GPS module?

Just kidding when you mentioned the tinfoil hat. But...I've recently 
finished reading the book about how the US and/or Israel undermined Iranian 
nuclear enrichment program. One of the interesting things is, they 
(NSA/CIA) do actually intercept electronic devices and plant so called 
"implants" in them, which pretty much collect data (usually the "bugged" 
device will be connected to the local network) and send it out over 
specific radio frequency. The NSA then sets the antenna in the surrounding 
buildings and receives it. This way admins looking through the local 
firewall/intrusion prevention system logs can't detect any signs of data 
extraction.

Dana nedjelja, 9. srpnja 2017. u 16:03:36 UTC+2, korisnik MichaelS napisao 
je:
>
> Yeah - to be honest, I feel better using a GPS receiver over a foreign 
> made WiFi module inside my "safe" firewall.  Excuse me while I don my 
> tinfoil hat!  :D
>
> These GPS modules may be old enough to justify Fire Sale pricing, since 
> they're limited to 4800 baud and you can get newer modules for less.
>
> The auction ends in 2-3 hours, if anyone's on the fence.  Who knows if 
> they'll be relisted, but they haven't exactly been selling like hotcakes. 
>  Only three sold so far.
>
> One is scheduled to arrive tomorrow.  'will report later.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/814aef3a-501e-4f3b-90e9-5ec91cbbca03%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: IN-8 Digit Failure

[Luka C]

Not sure what you mean by "spacer bar at the base of the digit", but you 
should move the tube around and observe if there are any sounds coming from 
it while you rotate it. I have one IN-14 tube in which the decimal point 
"cathode" has ripped off and is moving inside the tube freely...also, the 
small metal parts are sometimes used to attach the "shadow paper" on the 
back of the tube. In such case, it might be possible for the metal piece 
for short the digit and the anode (not necessarily the digit cathode to the 
anode mash if it's not the first digit stacked, but perhaps the wires below 
the mica/plastic disk)..this would result in anode and the shorted cathode 
to be on the same potential, thus not igniting the digit.

Dana četvrtak, 1. lipnja 2017. u 22:42:52 UTC+2, korisnik Roddy Scott 
napisao je:
>
> Folks,
>
> I have a little IN-8 clock that was unplugged a couple of days ago and 
> when powered up again today the #1 digit on one of the tubes does not 
> illuminate. I swapped the tubes around and it goes with the tube so not a 
> driver issue.
> On observing the numeral change I can see a glow from the numeral spacer 
> bar at the base of the digit but that is all. It runs on a PV IN-8 board 
> and has shown no sign of cathode poisoning at all.
> Possible fracture of the digit itself?
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/42a77243-cbb5-4f6a-93db-97fa884d3034%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Pulsed DC vs direct DC in cathode poisioned

[Luka C]

@greg, I might add that I have actually healed the IN-18 tubes (which I 
have purchased with cathode poisoning for a discount)  in less than a day 
per tube, but increased the current like 50% above maximum specified in the 
datasheet...at that point, the tubes run warm (not hot that you'd burn your 
finger on them), but I guess it won't kill them if run for less than a day.

Dana četvrtak, 1. lipnja 2017. u 18:38:32 UTC+2, korisnik gregebert napisao 
je:
>
> I will add that:
>
> 1. Not all nixies will respond to depoisoning attempts. I have a few 
> Burroughs 5031's (very early nixie) that are hopelessly gone.
> 2. It takes weeks to  depoison at normal current.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/8d897bb3-4c07-4f5c-a9bc-ab5c28a6a1fd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Pulsed DC vs direct DC in cathode poisioned

[Luka C]

@terry, I agree, but it seems like his tubes got poisoned over three months or 
so as far as I understand. Could be a slow gas leak from a broken seal, but I'm 
not sure he could have healed them (and have them working for some time clean 
of cathode poisoning without raising the voltage) if that was the case. Gas 
leak typically irreversibly progresses until the tube completely fails.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/7cbc7560-6ad8-4be0-b08c-680490472df3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Pulsed DC vs direct DC in cathode poisioned

[Luka C]

@Ira, I think he already mentioned that he does not own a scope. If we knew the 
model of the clock, maybe there would be someone who knows such data from the 
firmware of the clock. Btw, if the tubes are NOS and have been running since 
January 2017, then there is something seriously wrong with the mentioned clock? 
Does it have anti-cathode poisoning routines activated and how often?

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/62b1343e-8fab-4a3b-a94a-ec8360ba3bdb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Interesting INS-1 behavior...

[Luka C]

the clock that I have built uses four INS-1 tubes, first few times when the 
clock was running for half a day (testing purposes) the INS-1 did not 
flicker. But after the clock was left to run for a few days, pretty much 
all of them started to flicker at some point. But..what is important is 
that now, half a year later, the clock is still running and all the 
flickering on INS-1 tubes has stopped, almost though if it was some sort of 
a burn-in period for them. I drive them with direct DC and within the 
datasheet specs, all the tubes are NOS.

Dana ponedjeljak, 22. svibnja 2017. u 07:18:11 UTC+2, korisnik Ben W 
napisao je:
>
> 
> So the INS-1 indicator neons are very fickle, anyone who has used them 
> knows they tend to flicker. Usually you have to switch them out until you 
> find a stable pair. I installed new INS-1's inside my clock recently, and 
> one of them showed very odd behavior, the neon glow seems to be precessing 
> in a circular pattern inside the unit like a decatron.
>
> obviously I'm very interested to know what is promoting this behavior! 
> Anyone have any ideas? the clock is multiplexed, maybe that has something 
> to do with it, but the colons are as I understand driven directly from the 
> HV power.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/9e04a546-3dcb-40dc-980b-0b0330661eda%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Constant current source design

[Luka C]

@Nick - I know that cathode current control per segment might seem a bit of 
an overkill, but as @greg said, considering the price of the tubes, I think 
we should do our best to meet all of the datasheet specs if possible. And 
if you carefully look at the datasheet, consider the case you want to 
display a character like "*" (asterisk), then add together the currents of 
each segment from the datasheet and it will exceed the maximum anode 
current defined in the same datasheet, this is why each cathode current 
should be adjusted so that if you add them all together, the sum will be 
lower than the maximum anode current (of course, how much this really 
impacts the life of the tube will remain a mystery I guess, but we should 
still be careful)

This could be achieved by adjusting the resistor value in each of the 
segment lines but this carries a side effect. These tubes will presumably 
run in our clocks for years and during the that time, their characteristics 
will probably change. The voltage drop on the tube will probably change and 
if the supply voltage is fixed this will also result in changes to the 
segment current (Is = (Ua - Um) / Rs). If Um rises over time, this would 
result in lower voltage drop across the resistor and in turn lower the 
segment current (perhaps to the point where brightness might not be what we 
would like it to be) and this is why a constant current sink might come 
handy to make sure current in each segment remains fixed over the years of 
operation.

Dana četvrtak, 11. svibnja 2017. u 17:36:02 UTC+2, korisnik gregebert 
napisao je:
>
> The problem is that if the segment currents stated in the Burroughs 
> datasheet are summed-up for various characters, many characters exceed the 
> max-spec for anode current. However, the average character-current is 
> on-par with the max anode-current spec. Unfortunately, we'll never know 
> what Burroughs intended.
>
> I took a different approach and gathered a lot of data on my 7971's, 
> including a guesstimate of the current that gives good visual illumination 
> for each segment.
> When I scaled-back the spec-value for segment currents to get most 
> characters within the 22mA max anode-current spec, I found that the 
> 'visually good' current was on-par with the scaled current. What I ended up 
> with is segment currents that are visually good, and within the 
> max-anode-current spec limit (slightly over for 2 characters).
>
> Hopefully I will have time this weekend to finish my board and try it out.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/a4d95431-e69e-4e94-98be-d782587d70fd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Constant current source design

[Luka C]

@greg, I have taken a look at 
this: http://tayloredge.com/storefront/1386_B7971SmartSocket/1386.pdf . 
When using 175V as supply voltage and assuming 140V voltage drop on the 
tube, the currents should be: 

S15, S4, S1 = 1.59mA
S14, S10 = 1.06mA
S13, S11, S9, S7 = 1.46mA
S12, S8, S6, S5, S3, S2 = 1.30mA

When you add them all together, you get 20.53mA which is just right under 
the specs of the datasheet so I adjusted the currents according to it. I 
know this will result in less brightness than the maximum possible when all 
segments are not lit up because plenty of margin will be left then, but I 
see no other way, other than implementing the anode current limiter as well 
as you have described in earlier posts.

Dana srijeda, 10. svibnja 2017. u 20:26:24 UTC+2, korisnik gregebert 
napisao je:
>
> Luka - What segment currents are you using for the 7971 ?
>
> If you turn all segments on at the rated datasheet current, there are 
> several characters, such as 8 and Q, that will cause the total current to 
> significantly exceed the max rating of 22mA. I analyzed current-draw for 
> most characters, and even the average (~26mA) exceeds spec.
>
> I scaled-back the current in my design so I never exceed the max 22mA.
>
> 7971's are really expensive & rare, so I'm not taking any chances.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/e39361f0-3003-43c9-8612-78e17df62b11%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Constant current source design

[Luka C]

I'm glad you've done the test with BSS139 and it turned out to work fine. 
I've used the same transistor for the cathode current control on my B7971 
clock, completed routing the PCB and currently sparing some money to send 
it to the fab house :)

Dana utorak, 18. travnja 2017. u 11:08:58 UTC+2, korisnik Tomasz Kowalczyk 
napisao je:
>
> I've ran some tests with BSS139 + a potentiometer as a variable resistor.
> Conclusions:
>  - this design offers cheap and stable current limiting
>  - it varies with voltage dropped over the circuit, so it should be 
> tunable - the difference is about 14% change between 5V dropped and 105V 
> dropped across it (4,37mA vs 4,91mA).
>  - it is thermally stable - at 100V dropped and 4,93mA cold current I am 
> breaking the allowed power dissipation (allowed 0,36W, here - 0,5W), it 
> heats up to a bit over 50°C (many probe cables connected close to the 
> transistor act like a radiator), the current has changed by 40µA 
>  - while testing it I found out that striking voltage of tubes is a max 
> value - I've tested one Z567M and one LC-631, they both strike with 
> voltages lower than their normal maintaing voltage! While powered correctly 
> they sit at 140V across the tube, and I was able to lower power supply 
> voltage to 135V, disconnect tube, connect it again and it worked. I 
> suspected that due to large resistors and a cap in the feedback I have high 
> pulses over 170V in my supply, but it is not true - at 135V the spikes 
> measured with an oscilloscope were less than 5V. Of course, with voltage 
> lower than 140V + minimum drop over BSS39 and resistor, the current was 
> lowered.
>  - due to that it is possible to create a clock wasting very low power on 
> anode limiting - classic operation uses 170V, 180V or 200V supply + 
> resistor, which wastes quite a lot of power on those resistor. here we can 
> tune the power supply just above striking voltages of all tubes and be sure 
> that we don't have to recalculate the resistors. This will have a very 
> positive effect in clocks which run on batteries or USB supplies - with 
> both lower voltage required and lower power consumption.
>  - after tuning to desired current it is very stable - drift is max 20µA 
> I wonder if this low striking voltage is common among different tubes or 
> does the striking voltage change with temperature.
>
> In my opinion the only drawback of this design is that it is changing the 
> current with voltage drop across the limiter. This change in normal 
> operation shouldn't be noticable, because normally power supply doesn't 
> change much - maybe 10V under bad circumstances. With such voltage change 
> the current will be pretty much the same, but a resistor calculated once 
> isn't universal for every design. 
> Everything else (simplicity, price) says that this is it, I'll be using 
> this instead of resistors.
>
> I'm attaching photos of setup - left multimeter measures voltage across 
> the tube, middle measured power supply voltage, and the right one measures 
> current (in mA). Voltage across limiter is of course the difference between 
> middle and left.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/935caee2-bcee-4d8b-ae6d-901e10314c13%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Constant current source design

[Luka C]

@greg

I'd like to ask something related to the B7971 since I'm designing a PCB 
for the clock at the moment. I have implemented the controlled current 
sinks for each cathode and will fine tune the current for each segment 
(already discussed in the HV control chips thread with the schematic). I'm 
wondering if you implement both the cathode sink drivers and the anode 
driver you proposed here:

1) If sum of all lit cathodes currents exceeds the maximum value in the 
datasheet (the current set on the PMOS controlled source) what will happen 
to the currents of the individual segments, will they proportionally go 
down for each activated cathode? If so, will this cause the tube to reduce 
the glow?

2) If the sum of all lit cathodes is less than the maximum defined in the 
datasheet, it's supposed to stay that way because the PMOS source only 
limits the upper maximum current but does not enforce it constantly to flow 
trough the tube when it's activated?

3) How important do you think the maximum current per tube control is? 
Before deciding to go with the sink controller for each cathode, I browsed 
for other solutions to drive the B7971 and most of them use resistors, but 
these resistors have to limit the current to rather low values, are these 
enough to light the tube properly?

For 
example, http://tayloredge.com/storefront/1386_B7971SmartSocket/1386A.pdf

Segment 11 in the datasheet is defined for a max current of 5.5mA, but the 
above schematic would produce (170V - 140V) / 24 kOhm = 1.25mA. This seems 
rather low?

Dana nedjelja, 26. ožujka 2017. u 09:01:56 UTC+2, korisnik gregebert 
napisao je:
>
> Some other things I forgot to mention...
>
> 1. Generally, you would use a current-regulator on the anode side for 
> non-segmented tubes (0-9), where all cathodes use the same current.
>
> 2. Segmented displays (b7971) have different currents for various 
> segments, so you will need cathode-side current-regulation. I chose to 
> include an anode-side current-limiter as well because the Burroughs 
> datasheet specifies a max total current. This max current is about half the 
> sum of all the spec'd individual segment currents.
>
> 3. If you use MOS devices, be very careful not to exceed the max-voltage 
> specs. Doing so will damage the device (impaired reliability, or outright 
> destruction). I've done some research on MOS device-failures and they often 
> fail-shorted, which is disastrous. Fortunately, those kinds of failures are 
> mostly caused by overstress (voltage or temp).
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/bea2bf0a-852e-4d8a-a6f1-57e7ffe79835%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: HV control chips

[Luka C]

Small update:

After some further research, I've decided that those current limiting 
diodes are a bit too expensive for the clock (6 x 15  = 90 diodes). I've 
implemented a current source using a depletion type MOSFET along with a 
resistor like in the attachment, I'll test it out and see how it works. 
Since these types of MOSFET have some margin in the Vgs between batches 
(apparently the Vgs is written on each reel), I'll adjust the resistor size 
after purchasing a lot of them first.

Dana ponedjeljak, 27. veljače 2017. u 18:32:10 UTC+1, korisnik gregebert 
napisao je:
>
> I see no reason why it wouldn't work; the voltage-drop across the diode 
> will vary as needed to limit the current. This is the typical behavior of a 
> current source; even a resistor behaves somewhat in this manner.  The 
> datasheet states the device can handle 600mW, so if you are running at 6mA 
> or below, you wont have a thermal problem as long as the POV spec (100V) is 
> followed. 
>
> Regarding constant voltage across the nixie tube, it's a consequence of 
> how it operates. You want to limit the current to the datasheet spec; this 
> can be done with a resistor or a current-source. Whichever approach you 
> take is a matter of choice; I prefer current-limiting because it gives 
> tighter control and can be more energy-efficient.
>
> Glow tubes have a nonlinear I-V curve that becomes steep after ionization, 
> which means that a small change in voltage will cause a large change in 
> current. Even LED's behave this way. That's why you can't just drive them 
> from a constant voltage, like an incandescent bulb; there needs to be 
> something to limit the current.
>
> If you still feel nervous, test it out with a spare or expendable nixie 
> tube (IN-1's are cheap...). Just make sure the diode is installed 
> correctly; if it's backwards...bye bye nixie.Unfortunately the 
> datasheet doesn't show the test-circuit, which is unusual.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/9083295b-be82-45ed-a9a4-62e353110c3c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: HV control chips

[Luka C]

Thank you for the suggestion.

But shouldn't this device, after the initial "setup" of the voltage across 
itself and the current flowing trough remains mostly on the same voltage 
drop? And since the maximum voltage drop that's specified in the datasheet 
is cca 4.5V, shouldn't (anode voltage - nixie tube voltage drop = 30V) this 
leave enough margin for the "diode" to absorb changes while allowing the 
tube to maintain constant voltage drop across itself?

Dana ponedjeljak, 27. veljače 2017. u 01:39:35 UTC+1, korisnik nixiebunny 
napisao je:
>
> Luka, 
>
> This part works by increasing the voltage drop across itself until the 
> current is as low as its design current. Therefore, it needs the ability 
> to do this, which means that the device that it controls needs to be 
> able to respond appropriately. 
>
> A Nixie tube is a constant-voltage device, so it will have the same 
> voltage across it, independent of the current flowing. This means that 
> the Nixie tube is a very poor load for such a two-terminal 
> constant-current device. 
>
> I would recommend using a voltage-dropping resistor instead. It is known 
> to work properly. 
>
> On 2/26/2017 1:14 PM, Luka C wrote: 
> > I have spent a day or two looking for a final solution and found 
> > something I honestly didn't know existed. Although they seem rare, there 
> > are so called current limiting diodes (CLD) which you can pretty much 
> > put in series with a load and they'll limit the current trough it and 
> > keep is stable as voltage changes across them. 
> > 
> > https://sub.allaboutcircuits.com/images/05269.png 
> > 
> > https://www.centralsemi.com/PDFs/selection/leaded/CLD_Standard.pdf 
>
>
> -- 
> David Forbes, Tucson AZ 
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/ddc3f462-ac99-49fa-be55-ecfae6247af2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Some guy offering Rodan CD47 tubes

[Luka C]

I took a look as well, there doesn't seem to  be the mercury capsule 
inside, at least from the angles on the pictures available 
(http://www.leapsecond.com/pages/CD47/cd47-2h.jpg)

Also, you can search for an auction on eBay where the seller sold one 
defective CD47, this is what he said:

"I have had mine for about 8 years now; carefully stored in a display 
cabinet in a temperature and humidity controlled lab (No kidding!). Every 
year, I take it out and connect it to demonstrate this beauty in 
operation.  This year, she did not light, she did nothing at all!!  In a 
panic,  I grabbed my sparker, surely I'm just connecting it up wrong?  I 
sparked her, but she was gone;  her neon gas has completely leaked out!"

So the state and quality of the tubes being sold is kinda questionable 
considering the price.

Dana petak, 24. veljače 2017. u 15:24:25 UTC+1, korisnik MichaelS napisao 
je:
>
> Are CD47's long life tubes?  I haven't noticed any mercury capsules in 
> them.  I've seen some heavily silvered tubes that were smaller versions. 
>  It would make sense if they were made in the early to mid 60's.
>
> It seems that prices of bigger tubes aren't exactly skyrocketing now that 
> Dalibor's tubes are available.
>
> IN18's went up recently, but seem to be coming back down a bit...
>
>
>
> On Wednesday, February 22, 2017 at 6:42:02 AM UTC-6, Ian Sparkes wrote:
>>
>> There's some guy called Andre offering a number of CD47 tubes here:
>>
>>
>> https://www.tubeclockdb.com/component/kunena/7-buy-sell-trade/7336-rodan-cd47-gr-414-nixie.html
>>
>> Still waiting for further information...
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/00de2849-9e49-40a3-97f5-9e774aa8bdd1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: Some guy offering Rodan CD47 tubes

[Luka C]

I don't know how you found that item on eBay, but that appears to be the 
only nixie tube "seller" from my country that I have ever seen. Actually, 
he lives like 20 minutes by car away...shame he doesn't have any "more 
interesting" tubes in stock, these +/- (at least that's what they look 
like) are not really useful for a clock, heh.

Dana četvrtak, 23. veljače 2017. u 16:25:36 UTC+1, korisnik Pramanicin 
napisao je:
>
> Speaking off odd/rare nixies
>
>
> http://www.ebay.com/itm/2-x-Siemens-ZM-1043-Special-Symbol-Nixie-Rohre-Tube-Tested-NOS-Clock-Rare-/122368750757?&_trksid=p2056016.l4276
>
> I haven't seen this type before
>
> On Thu, Feb 23, 2017 at 7:20 AM, Nicholas Stock <nick...@gmail.com 
> > wrote:
>
>> I concur Luka. However, there will always be someone who wants to collect 
>> rare things at any cost, so best of luck to him/her, I'm sure they'll find 
>> a buyer somewhere...personally, I think nixies should be used for what they 
>> were intended, not stored away in a drawer etc...but that's just my 
>> opinion, nothing more.
>>
>> With those kinds of prices, it's just a matter of time before Dalibor 
>> flexes his tube making skills and makes an extra large Z568 
>> variant..hopefully!!!
>>
>> Best to all,
>>
>> Nick
>>
>> On Thu, Feb 23, 2017 at 6:42 AM, Luka C <luka.culi...@gmail.com 
>> > wrote:
>>
>>> I pretty much enjoy "collecting" different types of Nixie tubes and 
>>> sometimes spend like 100$ or 150$ per tube, but paying ~2500 euros for a 
>>> Nixie tube is just ridiculous in my honest opinion. I mean, if such amount 
>>> is something you earn easily and you really feel like you "just have to 
>>> have one or more", then I can kinda understand it, but if you have to to 
>>> really spare for quite some times to buy these, you should think about how 
>>> many other things or experiences there are that you could enjoy and enrich 
>>> your life with...at least here in Europe, you can buy a brand new car or 
>>> take a really exotic trip somewhere you'd always wanted to go to. I look at 
>>> this the same way as with people spending tons of money on some rare 
>>> Elvis's clothes or similar "collectibles" and wonder do they still feel it 
>>> was worth it after all the years they own those. Besides, I doubt you can 
>>> be sure these tubes are NOS or that they have been properly stored for all 
>>> those years, so you could be buying something that might malfunction in a 
>>> few years due to gas leaking caused by broken air-seals on the pins and 
>>> similar defects (of course, you could say this for other tubes as well, but 
>>> for them you're paying like 15 times less money so you can pretty much 
>>> replace them in the future or stock up on spares).
>>>
>>> Just my 2 cents :)
>>>
>>> Dana četvrtak, 23. veljače 2017. u 13:37:46 UTC+1, korisnik ten kowal 
>>> napisao je:
>>>>
>>>> I've got an answer from him. When asked for price for 4 tubes, he 
>>>> answered with 10 000 EUR. 
>>>> Well, he really does know how rare and valuable they are.
>>>>
>>>> W dniu środa, 22 lutego 2017 13:42:02 UTC+1 użytkownik Ian Sparkes 
>>>> napisał:
>>>>>
>>>>> There's some guy called Andre offering a number of CD47 tubes here:
>>>>>
>>>>>
>>>>> https://www.tubeclockdb.com/component/kunena/7-buy-sell-trade/7336-rodan-cd47-gr-414-nixie.html
>>>>>
>>>>> Still waiting for further information...
>>>>>
>>>> -- 
>>> You received this message because you are subscribed to the Google 
>>> Groups "neonixie-l" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to neonixie-l+...@googlegroups.com .
>>> To post to this group, send email to neoni...@googlegroups.com 
>>> .
>>> To view this discussion on the web, visit 
>>> https://groups.google.com/d/msgid/neonixie-l/714e796c-1a4b-411a-b79e-12837efb02eb%40googlegroups.com
>>>  
>>> <https://groups.google.com/d/msgid/neonixie-l/714e796c-1a4b-411a-b79e-12837efb02eb%40googlegroups.com?utm_medium=email_source=footer>
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/54554d53-05df-4fc2-9a29-54b39a2e0409%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Some guy offering Rodan CD47 tubes

[Luka C]

I pretty much enjoy "collecting" different types of Nixie tubes and 
sometimes spend like 100$ or 150$ per tube, but paying ~2500 euros for a 
Nixie tube is just ridiculous in my honest opinion. I mean, if such amount 
is something you earn easily and you really feel like you "just have to 
have one or more", then I can kinda understand it, but if you have to to 
really spare for quite some times to buy these, you should think about how 
many other things or experiences there are that you could enjoy and enrich 
your life with...at least here in Europe, you can buy a brand new car or 
take a really exotic trip somewhere you'd always wanted to go to. I look at 
this the same way as with people spending tons of money on some rare 
Elvis's clothes or similar "collectibles" and wonder do they still feel it 
was worth it after all the years they own those. Besides, I doubt you can 
be sure these tubes are NOS or that they have been properly stored for all 
those years, so you could be buying something that might malfunction in a 
few years due to gas leaking caused by broken air-seals on the pins and 
similar defects (of course, you could say this for other tubes as well, but 
for them you're paying like 15 times less money so you can pretty much 
replace them in the future or stock up on spares).

Just my 2 cents :)

Dana četvrtak, 23. veljače 2017. u 13:37:46 UTC+1, korisnik ten kowal 
napisao je:
>
> I've got an answer from him. When asked for price for 4 tubes, he answered 
> with 10 000 EUR. 
> Well, he really does know how rare and valuable they are.
>
> W dniu środa, 22 lutego 2017 13:42:02 UTC+1 użytkownik Ian Sparkes napisał:
>>
>> There's some guy called Andre offering a number of CD47 tubes here:
>>
>>
>> https://www.tubeclockdb.com/component/kunena/7-buy-sell-trade/7336-rodan-cd47-gr-414-nixie.html
>>
>> Still waiting for further information...
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/714e796c-1a4b-411a-b79e-12837efb02eb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: HV control chips

[Luka C]

Thanks for the response.

I wasn't thinking of multiplexing anyway exactly for the reason of excess 
currents, I only implement direct drive in my designs :)

But what I wanted to ask is, I only saw these constant current regulators 
operating as for example, LM317 
(http://diyaudioprojects.com/Technical/Current-Regulator/Precision-Current-Limiter-Schematic-Circuit-LM317-LM338-LM350.png)..so
 
would this have to be implemented in series of each cathode?

Dana srijeda, 22. veljače 2017. u 21:08:04 UTC+1, korisnik gregebert 
napisao je:
>
> Given the high cost + rarity of b7971 tubes, I would only recommend a 
> current-regulated driver for each segment. Using individual cathode 
> resistors will limit the current, but it's not well-regulated and will vary 
> tube-to-tube and with age. Yes, the HV5530 will *work*, but I would be 
> concerned about reduced tube life.
>
> Also, I would not multiplex these tubes because the brightness will be 
> reduced. Boosting the current to compensate for this will likely exceed the 
> datasheet limits and shorten tube life. In fact, driving all segments of a 
> b7971 tube at their rated current *at the same time* will exceed the spec 
> limit for maximum anode current. So, along with limiting the current on 
> each segment, you should also have a current-limit for each anode.
>
> I spent a lot of time gathering & analyzing data on my 7971 tubes to get a 
> tradeoff between segment-current, max-current, and brightness.
>
> I know it sounds like overkill, but a few extra $ for components to 
> protect a $100 tube is well worth it. Besides, I had a lot of fun designing 
> and simulating the driver circuitry.
>
> The HV series drivers aren't necessary for current-regulators; you can use 
> TTL or LVTTL shift-registers (74HC595) to control the driver transistors 
> (PMBTA42DS).
>
> I haven't had the time to fab the PCB for my 7971 tubes yet; once I get 
> the board running I'll post schematics. Just too doggone busy with my day 
> job.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/253bee53-8aff-46d5-a9d6-44bca55f5131%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: HV control chips

[Luka C]

If I may ask, what's the reason HV chips would not be suitable for driving 
segmented tubes like the B7971? Is there some device specific "glitch" when 
multiple cathodes are enabled and current flowing trough each one is 
different? Of course, suppose we put the appropriate resistor between the 
each output pin of the HV chip and the pin of the segment of the B7971 tube.

Dana utorak, 21. veljače 2017. u 17:57:12 UTC+1, korisnik gregebert napisao 
je:
>
> Are you driving segmented tubes (like the 7971) ? If so, I would advise 
> against the HV-series drivers because you will need several different 
> segment-currents.
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/5370ca2b-6552-45fb-9ae0-4a9886edd6fd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Interested in B7971

[Luka C]

wow, so many posts here. I was busy working on my bachlor's degree and 
forgot to check this out, thanks for the answers everyone!

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/48edb740-7348-43c3-b6fb-4e631924de40%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Interested in B7971

[Luka C]

I'm actually planning on designing my own PCB for it with exact current 
limiting for each segment as per datasheet. Did you do the same or did you 
install just one anode resistor and how important is it to stick to the 
datasheet in that sense?
Dana nedjelja, 22. siječnja 2017. u 06:13:02 UTC+1, korisnik gregebert 
napisao je:
>
> Are you designing your own board, or using an existing design ? I recently 
> finished my 7971 PCB design, but I'm in the 'countdown hold' before I send 
> out the board for fabrication while I make design tweaks and other 
> improvements.
>
> Sorry, I dont have any extra tubes. Grthey are getting expensive 
> and rare.
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/f51a3f14-1cef-4e08-906e-718c936b76fc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Interested in B7971

[Luka C]

I'm interested in purchasing B7971 tubes for a project of mine, if anyone 
is willing to sell some, feel free to send me a PM and we'll see if we can 
sort something out, thanks!

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/c0f5bca3-d715-47ae-8032-be3a30835e18%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Have a great Christmas...

[Luka C]

Best wishes to all of you! May the New Year be filled with even more 
glowing Nixies and clocks :)

Dana subota, 24. prosinca 2016. u 19:59:24 UTC+1, korisnik Nick napisao je:
>
> ...and a Happy and Healthy New Year !
>
> Just as importantly, start planning those projects !
>
> Cheers
>
> Nick
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/bcd46f08-5c6f-45f3-ade3-3e9ce4e8a86c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Faulty Atmel Microcontrollers from Aliexpress

[Luka C]

I agree with what you and the others have said. I should mention, the 
reason why I want to warn the others about this is because the listings on 
sites such as Aliexpress/Alibaba actually do say that the component is 
"original" and "new". If it was marked as refurbished or something like 
that, I would have never ordered it in the first place. I should also 
mention that I had many other orders from other sellers for components such 
as capacitors, resistors, inductors, power ICs, etc...and all of them work 
100% fine and came packaged in a sealed reel. I guess it's the 
microcontrollers that are mostly prone to such scams. 

Dana nedjelja, 18. prosinca 2016. u 00:22:01 UTC+1, korisnik nixiebunny 
napisao je:
>
> I buy parts from Digikey for this reason. They always work. Think how 
> much money and time you would have saved by spending more on legitimate 
> components. 
>
> On 12/17/2016 1:26 PM, Luka C wrote: 
> > I'm writing this to warn others or ask if anyone had similar 
> > experiences. I purchased a lot of Atmel ATMEGA328P microcontrollers from 
> > a seller on Aliexpress. The lot was listed as "new" in the description 
> > and had a picture of the microcontroller in the reel so I thought it's a 
> > legit new sealed lot. After the package arrived, I noticed the 
> > microcontrollers were not in a real and were just randomly taped on a 
> > piece of some material with some semitransparent tape. I sent the boards 
> > to the local PCB soldering company and they have soldered 
> > microcontrollers on the boards. I flashed the program and the first 
> > board and it worked just fine so I thought I made a great deal because 
> > the price was really good for the lot. 
> > 
> > But this is where things became strange, after I was done programming 
> > the first board, I tested the other boards. The results were strange to 
> > say at least. Some of the microcontrollers came in a "state" where any 
> > fuse reprogramming was impossible (btw, SPIEN was not disabled in the 
> > fuses!). Two particular microncontroller samples were really strange. 
> > 
> > One seemed to execute really strange sequences of commands without any 
> > reason and my nixie clock would get frozen every now and then. Since I 
> > own the debugger (Atmel ICE), I decided to debug the firmware on the 
> > chip. It turns out that the chip would go really crazy when, for 
> > example, 0 and 5 were displayed on the two middle tubes on my clock. The 
> > debugger call stack showed that one function was executed when it should 
> > not have been and the values of variables in the programmed had values 
> > that in no way could be there in the normal program operation. 
> > 
> > Second one had trouble outputting data to the LED controller. Debugging 
> > this one's firmware showed that the microcontroller was not frozen and 
> > in fact was sending data to the LED controller but I guess the data was 
> > not properly formatted or something. 
> > 
> > There were some boards with perfectly fine chips so I decided to do a 
> > simple tested. Since my clock consists of 2 boards, one for the 
> > microcontrollers and power supply circuitry and the other one for the 
> > nixie tubes and the LEDs, I decided to do a test and swapped the board 
> > with the tubes and LEDs across both "working" and "faulty" 
> > microcontroller boards. The working ones never produced not a single 
> > fault or glitch, I tried to replicate the bugs on them with no success. 
> > On the other hand, the faulty ones were impossible to fix even by 
> > reflashing the microcontrollers multiple times with the exact same hex 
> > filed used to flash the working ones. 
> > 
> > At the end, I am confused. I am not sure what to conclude from this 
> > really. I believe the fault is not in the board itself (PCB layout or 
> > connections) but that it comes from the faulty microcontrollers I have 
> > purchased. After doing a little research on the internet, I found some 
> > people saying that these Chinese companies basically buy used equipment 
> > and remove the microcontrollers from them or that they simply purchased 
> > large quantities of chips that have failed quality control and sells 
> > them at lower prices. I will try to find a way to remove the faulty ones 
> > from the boards and replace them with new ones purchased from RS 
> > Components and then do the tests again. 
> > 
> > Anyone ever had similar experiences or has any idea why would this 
> happen? 
> > 
>
>
> -- 
> David Forbes, Tucson AZ 
>
>

-- 
You received this message be

[neonixie-l] Faulty Atmel Microcontrollers from Aliexpress

[Luka C]

I'm writing this to warn others or ask if anyone had similar experiences. I 
purchased a lot of Atmel ATMEGA328P microcontrollers from a seller on 
Aliexpress. The lot was listed as "new" in the description and had a 
picture of the microcontroller in the reel so I thought it's a legit new 
sealed lot. After the package arrived, I noticed the microcontrollers were 
not in a real and were just randomly taped on a piece of some material with 
some semitransparent tape. I sent the boards to the local PCB soldering 
company and they have soldered microcontrollers on the boards. I flashed 
the program and the first board and it worked just fine so I thought I made 
a great deal because the price was really good for the lot.

But this is where things became strange, after I was done programming the 
first board, I tested the other boards. The results were strange to say at 
least. Some of the microcontrollers came in a "state" where any fuse 
reprogramming was impossible (btw, SPIEN was not disabled in the fuses!). 
Two particular microncontroller samples were really strange. 

One seemed to execute really strange sequences of commands without any 
reason and my nixie clock would get frozen every now and then. Since I own 
the debugger (Atmel ICE), I decided to debug the firmware on the chip. It 
turns out that the chip would go really crazy when, for example, 0 and 5 
were displayed on the two middle tubes on my clock. The debugger call stack 
showed that one function was executed when it should not have been and the 
values of variables in the programmed had values that in no way could be 
there in the normal program operation.

Second one had trouble outputting data to the LED controller. Debugging 
this one's firmware showed that the microcontroller was not frozen and in 
fact was sending data to the LED controller but I guess the data was not 
properly formatted or something.

There were some boards with perfectly fine chips so I decided to do a 
simple tested. Since my clock consists of 2 boards, one for the 
microcontrollers and power supply circuitry and the other one for the nixie 
tubes and the LEDs, I decided to do a test and swapped the board with the 
tubes and LEDs across both "working" and "faulty" microcontroller boards. 
The working ones never produced not a single fault or glitch, I tried to 
replicate the bugs on them with no success. On the other hand, the faulty 
ones were impossible to fix even by reflashing the microcontrollers 
multiple times with the exact same hex filed used to flash the working ones.

At the end, I am confused. I am not sure what to conclude from this really. 
I believe the fault is not in the board itself (PCB layout or connections) 
but that it comes from the faulty microcontrollers I have purchased. After 
doing a little research on the internet, I found some people saying that 
these Chinese companies basically buy used equipment and remove the 
microcontrollers from them or that they simply purchased large quantities 
of chips that have failed quality control and sells them at lower prices. I 
will try to find a way to remove the faulty ones from the boards and 
replace them with new ones purchased from RS Components and then do the 
tests again.

Anyone ever had similar experiences or has any idea why would this happen?

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/c3f5f01f-0494-4912-a8f6-3d5b4be2d564%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: OT: Rant - eBay Sucks ! Shipping Calculator

[Luka C]

eBay's Global Shipping Program is just terrible. I purchase stuff from the 
U.S. all the time (I'm in Croatia) and it's typically marked as 
"International Priority Shipping By GSP" in the description.

1. I'm not sure what "priority" means because it first has to go trough the 
whole U.S.A. to arrive in Erlanger, KY. After that, it's shipped to the 
U.K., which is followed by Germany where it's taken by some commercial 
shipping provider (GLS, DPD) and transferred to Croatia. I should mention 
that, once it reaches Europe, all further transfer of the item is done by 
trucks, not by plane which is kinda strange for something called 
"priority"? Anyway, shipping times are terrible, a typical USPS First Class 
Intl Mail arrived at the same time, if not sooner.

2. The shipping costs are just ridiculous, as somebody mentioned earlier, 
shipping typically makes around 50% of the total price I pay for the items 
shipped by GSP.

3. They claim to act as a "customs broker" for you and charge you with 
customs fees in advance. I believe this is a pure scam since there has 
never been any evidence that Croatian Customs Service received any money 
for my incoming packages (they always need to provide you an invoice of the 
customs procedure when importing goods, it's give to you along with the 
package). What I believe is that they illegally accept packages in the U.K. 
(they don't have to report them to customs there because they're intended 
to continue their journey to the destination country - Croatia), they then 
relabel them as if they're coming from the U.K. (no customs fees in the 
European Union) and transfer freely to the destination and customs never 
even realizes the true origin of the package, so they make 100% profit on 
the "calculated import costs".

All in all, it seems they provide the lowest and cheapest possible service 
so they can take most of the money for themselves which has been bad news 
for the international customers since the day it was introduced.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/7e0b8125-1cd3-42e0-8209-17e2fff4207f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: two digit glass dome clock

[Luka C]

this seems as if someone used Google Translate

Dana utorak, 26. srpnja 2016. u 06:02:40 UTC+2, korisnik westdave napisao 
je:
>
> nice and clean two digit clock made from a clause urbach ,
> http://www.nixieclocks.de, two digit clock kit from long in the past with 
> the brand new R/Z568  from Dalibor Farny.and a kundo glass dome clock ,
>
> (VINTAGE KUNDO ELECTRONIC GLASS DOME PENDULUM CLOCK MADE IN GERMANY WORKS 
> GREAT.)*there is always one of these type clocks for parts,at a resonable 
> price. Do*  I like these tubes ? why ,yes i do, they plug into the old 
> style sockets ,and are priced right and have a great #4 if you were 
> wondering if you should have a set ,you could not do better ,i used a 8.66k 
> anode resistor at 170 Vdc
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/51acf5bd-0207-4059-a510-4d9b05803bd2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Cathode poisoning and tube's current flow

[Luka C]

Dana četvrtak, 14. travnja 2016. u 16:56:43 UTC+2, korisnik gregebert 
napisao je:
>
> I had no luck trying to depoison 1 or 2 Burroughs 5031 tubes, even after 
> several hours. I considered these tubes expendable and put so much current 
> thru them that they got hot. Darn! I didn't measure it, though.
>
> I'm amazed how quickly you were able to depoison your tubes; I was 
> expecting days or even weeks.
>
> Did you purchase your tubes in this condition, or did it occur while they 
> were in your project? I'm curious about the conditions and timeframe that 
> cause poisoning. My 14-digit clock has several static digits, so I run a 
> 1-hour depoisoning every night; maybe it's overkill ?
>



I bought the tube with cathode poisoning so I don't really know how long 
was it operating and under which conditions. To be honest, I was also 
surprised to see most of them recover in such shorter periods. I'm not sure 
if it's of any significance but I digits that took less time to recover 
were closer to the anode grid, perhaps there is something in that. 

>From what I read on the web, the reason why your tubes didn't recover may 
be that tubes have a certain point after which they no longer seem to be 
repairable because the current that would be needed for healing to 
effectively take place is way too high and would cause the destruction of 
the tube. I don't know if your tubes still operated after cooling down, 
mine seemed to be warm to the touch, but not hot?

Regarding my clock, I have the first version of it running for around half 
a year now and there are no signs of cathode poisoning, probably both due 
to the fact that the time was way too short for it to occur and also, I 
enjoy the slot machine effect that I made for cathode poisoning prevention 
so I let it spin for ~5 seconds every 1min. So I guess letting them run the 
routine for 1 hour at night can't be a bad idea, I mean, you're sleeping 
anyway and the tubes will sure benefit from it (I guess there is no damage 
from exceeding the "standard" amount of time dedicated to anti-poisoning 
routines) :) 

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/b071cbcc-c29d-4e14-96b4-4a70502f1349%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Cathode poisoning and tube's current flow

[Luka C]

I've recently performed healing on cathode poisoned IN-18. I used the 
method of increasing the current (never more than 8mA, datasheet specifies 
it as max) to the point when all the parts of the digit light up and then 
letting it "bake" until the digit is evenly lit on nominal current of 4mA. 
I succeeded on every digit and noticed something in the process. I was 
measuring the current flowing trough the tube and it seems to increase 
gradually while the tube is healing. I guess this is perhaps related to the 
fact that removing the coating from the cathode increases it's "conductive" 
surface and thus resulting in increasing current. What seemed interesting 
was also, that unrelated to how large the poisoned segment was on the tube, 
the rate at which the current was increasing corresponded to the amount of 
time it took to heal the tube. In my case, digit "3" had almost all of the 
upper part poisoned, but once setting the current to 6.0 mA, over the next 
15min it went to 6.9mA and after that the digit was healed when returning 
to 4.0mA. On the other side, digit "8" was lightly poisoned on two places 
(both areas probably 3-4mm each) and the current was really slowly 
increasing (I had to set it to 7mA to start with) to 7.6mA and it took over 
6 hours to regenerate those areas.  So those are my observations, I wonder 
if anyone else noticed something similar?

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/78c1242e-0e3d-4497-9837-88ae17da1429%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Nixie Power Supply MK.II Schematic?

[Luka C]

Reading about the power supplies for the clock, I found the MK 1.5 power 
supply (schematic: 
https://threeneurons.files.wordpress.com/2011/08/mc34063_mk15cm1.gif) and 
have already built a clock based on it, but the coil seems to be buzzing 
pretty noticeably, especially when all the tubes turn OFF -> ON or change 
numbers at the same time, you can hear the buzzing frequency change at 
those moments. I suppose that changing the inductor could fix it so I'll 
give it a try. But, on multiple threads I saw people mentioning MK.II 
version of the power supply but wasn't able to find the schematic for it. 
So, does anyone have the schematic for it and would be kind to share it or 
was it never even made publicly available?

Thanks in advance :)

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/8e97346d-04b4-41ba-ad0f-1293e22e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with HV5622 drivers

[Luka C]

@greg

Well, I inspected the power line to the HV5622 with the oscilloscope and it 
shows a straight 12V line, no noises. On the clock line, I do sometimes see 
to have longer rise so it could be a problem I guess. 
I have made the second version of the board with 4505 level shifter, both 
Vcc and Vdd on it decoupled and separated the clock lines so I guess it 
should work as expected this time?
Also, since I'm directly driving the tubes and HV5622 gives around 60V on 
the inactive cathodes, the anode switching circuits shouldn't be needed to 
blank the tube? 

Dana ponedjeljak, 2. studenoga 2015. u 21:51:02 UTC+1, korisnik gregebert 
napisao je:
>
> I took another look at the level-translator circuit, and I noticed there 
> is a resistive pulldown (10K). This could be the problem, but without any 
> scope traces I cant be certain. You could do a SPICE simulation and get a 
> decent idea what the timing is.
>
> The CLK on the HV5622 is falling-edge, and with a resistive pulldown it 
> will have a slow edge-rate. The datasheet does not specify an 
> input-capacitance, but it's probably around 10pF. With two HV5622 devices 
> and the 10K pulldown, the time-constant is on the order of 200nsec. I've 
> seen two kinds of timing problems with slow edge-rates.
>
> 1. Noise susceptibility. With a slow clock-edge, any noise that occurs 
> while the clock is near the threshold point can cause an extra clock edge 
> (ie, a glitch).
>
> 2. Timing-skew. Each component will have a slightly different threshold 
> voltage (the point where it distinguishes a '1' from a '0'). With a slow 
> clock-edge, the difference in threshold-voltage (dv) causes a 
> timing-difference (dt). A slow clock-edge has a low dv/dt, and you can 
> actually calculate the timing-uncertainty if you know the variation in 
> threshold voltage. The risk here is that the first HV5622 clocks slightly 
> 'early', and the second one clocks slightly 'late'. If that happens, the 
> serial data being shifted will skip a bit. This is a hold-time violation 
> and also called 'shoot-thru'.
>
> One option as you said is to use an IC level-translator. A possible 
> quick-and-dirty option is to change the circuit so it uses an NPN pulldown, 
> rather than a resistor. If you go that route, you should change the PNP 
> pullup to a resistor unless you carefully simulate the circuit and optimize 
> the design. If you dont optimize, you will get 'crowbar' current between 
> the +12V supply and GND while both transistors are on, and that will create 
> tons of noise at the worst possible time -- when your clock is changing. 
> (Trust me, I've designed I/O pads on ICs before).
>
> You should be able to get-by with a slow rising-edge on the clock line 
> (remember: the HVxxx device is using the falling-edge, not the rising edge) 
> *as 
> long as you dont have any noise*. Any significant noise will cause 
> another clock-glitch.
>
> The other signals (LE, DATA) dont need "clean" edges as long as you 
> provide enough setup and hold margin, so you can keep the level-shifter 
> as-is for these signals.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/cc6f73f1-4225-4b8b-93e8-6bec6b51d4ee%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Help with HV5622 drivers

[Luka C]

Hi everyone,

I have made a prototype of my nixie clock using HV5622 nixie drivers, 2 of 
them chaned (DATAOUT of first one connected to DATAIN of the second one). 
They're both clocked trough the same pulse and also both latched trough the 
same pulse. 
The problem I'm dealing with now is the appearance of random numbers while 
the data shifts to the register at higher speeds, please take a look here: 
*https://youtu.be/QHljLFrYjss*   . 
The video was slowed down *50%* for you to actually see these glithces.

Now, this is an extremely annoying problem when I do a code that looks like:

void loop
{
   FreezeLatch();
   ShiftDataToHV5622(x, y, z, q);
   ShiftDataToHV5622(x, y, z, q);
   UnfreezeLatch();
}

if I add a delay of >100ms

void loop
{
   FreezeLatch();
   ShiftDataToHV5622(x, y, z, q);
   ShiftDataToHV5622(x, y, z, q);
   UnfreezeLatch();
   delay(100);
}

x, y, z, q - bytes of data

Then I have a situation where the occurrence appears to be pretty rare. 
Also, if I do both shifts once and don't loop it, it either makes the 
glitch and stays that way or it doesn't and tubes continue to display 
normal digits.

Now, for me it pretty much seems to be a problem in which data is 
incorrectly placed inside the shift registers of HV5622 quite randomly and 
it causes the outputs of the HV5622 to connect multiple digits and light 
them up.

This is how everything is connected:




Now, I guess there could be 2 possible problems:


1.) Level shifters are too slow, I'm this type and it could be too slow 
(TTL): http://obrazki.elektroda.pl/4179751000_1362756714.png

2.) Level shifters output possibly not compliant with CMOS inputs of the 
HV5622?



What do you think, what are the possible solutions? An IC level shifter of 
some kind?


Anyway, sorry for the longer post here, would be very thankful for any help.


PS. thanks a lot for the previous advice on power supply, it works like a 
charm! :)


Luka

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/70facb89-8730-4cc7-9cef-a8d1aa0c99b5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with HV5622 drivers

[Luka C]

Dana nedjelja, 1. studenoga 2015. u 21:36:27 UTC+1, korisnik gregebert 
napisao je:
>
> Problems that occur or worsen at higher speeds, and disappear at lower 
> speeds, are usually data-setup-time problems.
> On the other-hand, if the problem persists at all speeds, it's a hold-time 
> problem.
>
> As you pointed-out, the level-shifters are slow, so that is another clue.
>
> Have you sketched-out the data and clock paths to the first HV device and 
> analyzed the timing ?
>

Thanks for your reply.

I see from the datasheet that data setup time is 25ns and data hold time is 
10ns.
I have tried the following:

1. Put clock HIGH
2. Put one bit on the DATA line
3. Wait for 5 microseconds
4. Put clock LOW
5. Wait for 5 microseconds

and repeat the steps again.

At the end of the these cycles, I put latch HIGH, wait for 5 microseconds 
and put latch LOW.

Then I removed the other delay which was in the main loop and these random 
flickers started happening like crazy. I guess it should indicate that the 
time it takes the level shifting circuit to move from HIGH/LOW or LOW/HIGH 
state is too long, which then shortens the time these impulses stay still 
on the inputs of the HV5622 and that causes undesired random sequences of 
bits ending in the register and being moved to the output of the chip? If 
so, do you think this might be a good level shifter between the Atmega328 
and HV5622: http://www.onsemi.com/pub_link/Collateral/MC14504B-D.PDF ?
  

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/a5c7c20c-d84b-4d59-b040-6b6aebdeee3e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with HV5622 drivers

[Luka C]

Dana nedjelja, 1. studenoga 2015. u 21:36:27 UTC+1, korisnik gregebert 
napisao je:
>
> Problems that occur or worsen at higher speeds, and disappear at lower 
> speeds, are usually data-setup-time problems.
> On the other-hand, if the problem persists at all speeds, it's a hold-time 
> problem.
>
> As you pointed-out, the level-shifters are slow, so that is another clue.
>
> Have you sketched-out the data and clock paths to the first HV device and 
> analyzed the timing ?
>
Thanks for your reply.

I see from the datasheet that data setup time is 25ns and data hold time is 
10ms.
I have tried the following:

1. Put clock HIGH
2. Put one bit on the DATA line
3. Wait for 5 microseconds
4. Put clock LOW
5. Wait for 5 microseconds

and repeat the steps again.

At the end of the these cycles, I put latch HIGH, wait for 5 microseconds 
and put latch LOW.

Then I removed the other delay which was in the main loop and these random 
flickers started happening like crazy. I guess it should indicate that the 
time it takes the level shifting circuit to move from HIGH/LOW or LOW/HIGH 
state is too long, which then shortens the time these impulses stay still 
on the inputs of the HV5622 and that causes undesired random sequences of 
bits ending in the register and being moved to the output of the chip? If 
so, do you think this might be a good level shifter between the Atmega328 
and HV5622: http://www.onsemi.com/pub_link/Collateral/MC14504B-D.PDF ?

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/b9129aca-572a-4398-b055-19763e99fa58%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Help with HV5622 drivers

[Luka C]

@Spencer

Indeed that probably was the case with your clock because the HV5222 has OE 
that can either turn all the outputs LOW or push the data from the shift 
register to them from what I saw in the datasheet, but HV5622 has no such 
function, the only thing it has is LE (latch) and BL (blank). Correct me if 
I'm wrong.

Dana nedjelja, 1. studenoga 2015. u 23:44:37 UTC+1, korisnik Spencer 
napisao je:
>
> I had the same issue with the HV522PJ (which I believe is the same chip 
> but just shifts counter clockwise) with flickering when updating the 
> display.
>
> I found I had to pull OE LOW when updating and HIGH when finished.
>
>
>
> Sent from my Verizon Wireless 4G LTE smartphone
>
>
>  Original message 
> From: Luka C <luka.culi...@gmail.com > 
> Date: 11/01/2015 1:12 PM (GMT-06:00) 
> To: neonixie-l <neoni...@googlegroups.com > 
> Subject: [neonixie-l] Help with HV5622 drivers 
>
> Hi everyone,
>
> I have made a prototype of my nixie clock using HV5622 nixie drivers, 2 of 
> them chaned (DATAOUT of first one connected to DATAIN of the second one). 
> They're both clocked trough the same pulse and also both latched trough the 
> same pulse. 
> The problem I'm dealing with now is the appearance of random numbers while 
> the data shifts to the register at higher speeds, please take a look here: 
> *https://youtu.be/QHljLFrYjss* <https://youtu.be/QHljLFrYjss>  . 
> The video was slowed down *50%* for you to actually see these glithces.
>
> Now, this is an extremely annoying problem when I do a code that looks 
> like:
>
> void loop
> {
>FreezeLatch();
>ShiftDataToHV5622(x, y, z, q);
>ShiftDataToHV5622(x, y, z, q);
>UnfreezeLatch();
> }
>
> if I add a delay of >100ms
>
> void loop
> {
>FreezeLatch();
>ShiftDataToHV5622(x, y, z, q);
>ShiftDataToHV5622(x, y, z, q);
>UnfreezeLatch();
>delay(100);
> }
>
> x, y, z, q - bytes of data
>
> Then I have a situation where the occurrence appears to be pretty rare. 
> Also, if I do both shifts once and don't loop it, it either makes the 
> glitch and stays that way or it doesn't and tubes continue to display 
> normal digits.
>
> Now, for me it pretty much seems to be a problem in which data is 
> incorrectly placed inside the shift registers of HV5622 quite randomly and 
> it causes the outputs of the HV5622 to connect multiple digits and light 
> them up.
>
> This is how everything is connected:
>
>
>
> <https://lh3.googleusercontent.com/-F39TI3n-UfY/VjZifjx2xQI/AGM/F1volFi4euA/s1600/sch_simplified.png>
>
> Now, I guess there could be 2 possible problems:
>
>
> 1.) Level shifters are too slow, I'm this type and it could be too slow 
> (TTL): http://obrazki.elektroda.pl/4179751000_1362756714.png
>
> 2.) Level shifters output possibly not compliant with CMOS inputs of the 
> HV5622?
>
>
>
> What do you think, what are the possible solutions? An IC level shifter of 
> some kind?
>
>
> Anyway, sorry for the longer post here, would be very thankful for any 
> help.
>
>
> PS. thanks a lot for the previous advice on power supply, it works like a 
> charm! :)
>
>
> Luka
>
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+...@googlegroups.com .
> To post to this group, send email to neoni...@googlegroups.com 
> .
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/70facb89-8730-4cc7-9cef-a8d1aa0c99b5%40googlegroups.com
>  
> <https://groups.google.com/d/msgid/neonixie-l/70facb89-8730-4cc7-9cef-a8d1aa0c99b5%40googlegroups.com?utm_medium=email_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/7a107c28-0c2c-4d81-99c9-050f13c92f61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with Nixie partially lighting up

[Luka C]

Okay, so I read all the advices and thank you all for explaining the 
problems with this power supply. Since the board is extremely small and I 
will be installing it in a very thin housing, I don't really have much 
space to do modifications on it. So, I'll implement PS in my PCB design and 
solder its components on it.
I'd like to know if you could check this PS and see if it should be ok for 
my clock (4x IN-14 and 4x INS-1 tubes)?

Thanks a lot in advance.



Dana utorak, 22. rujna 2015. u 21:23:39 UTC+2, korisnik gregebert napisao 
je:
>
> Why not just use a totem-pole pre-driver for the MOSFET ?
>
> BTW, this touches on a challenge when designing HV drivers, whether they 
> are for DC-DC converters or driving cathodes. Most high-voltage NMOS 
> devices (~400V)  need 5-6 volts for solid gate turn-on, yet logic-levels on 
> most micro-controllers and FPGAs are only 1.2 to 3.3volts.  That leaves a 
> few options:
>
> 1. Use high-voltage logic, such as 4000-series CMOS, to drive HV NMOS. My 
> first nixie clock does this.
>
> 2. Use NPN drivers; you can easily design constant-current drivers running 
> from 3.3V. I've used surface-mount devices with 2 NPNs inside, and they are 
> rated at 250V.
>
> 3. Use a level-shifter between your logic and the driver. I'm doing this 
> on my current design with a HV5530.
>
> 4. Find a device with a low-voltage gate. Good luck on this; most 
> logic-levels NMOS devices only handle 60V or less. If you also want low Rds 
> (necessary for DC-DC converters), you have even fewer options.
>
> 5. Pull-up resistor. Ugghhh... It will work, but you will burn more power. 
> Could be dangerous for a DC-DC converter if the logic/software turning off 
> the gate goes haywire, which would lead to large current/burnout.
>
>
> I advise against relying upon the voltage-drop of the nixie tube to 
> protect the driver, though this is commonly done with designs using the 
> 74141. I always select my driver to handle the maximum-possible anode 
> supply, and add margin above that. I've never had a fried driver, ghosting, 
> etc. Parasitic capacitance and leakage can stress drivers above their 
> maximum ratings, which will lead to premature failure.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/b6b785cc-160d-43c2-b8e1-4502d37a8c20%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with Nixie partially lighting up

[Luka C]

<https://lh3.googleusercontent.com/-v0bjuN0XQWg/VgMGNJhH7bI/AF0/vLSC_u2tz3M/s1600/mc34063_mk15cm1.gif>


The picture seems to be moved, so this is the PS I'm planning to implement 
into my design. Do you think this one is fine for 4x IN-14 and 4x INS-1?
Thanks.

Dana srijeda, 23. rujna 2015. u 20:39:23 UTC+2, korisnik Luka C napisao je:
>
>
> Okay, so I read all the advices and thank you all for explaining the 
> problems with this power supply. Since the board is extremely small and I 
> will be installing it in a very thin housing, I don't really have much 
> space to do modifications on it. So, I'll implement PS in my PCB design and 
> solder its components on it.
> I'd like to know if you could check this PS and see if it should be ok for 
> my clock (4x IN-14 and 4x INS-1 tubes)?
>
> Thanks a lot in advance.
>
>
> <https://lh3.googleusercontent.com/-b9BU5T8kpY4/VgLxiEHdmZI/AFk/x7hJ0aVd_dA/s1600/mc34063_mk15cm1.gif>
>
> Dana utorak, 22. rujna 2015. u 21:23:39 UTC+2, korisnik gregebert napisao 
> je:
>>
>> Why not just use a totem-pole pre-driver for the MOSFET ?
>>
>> BTW, this touches on a challenge when designing HV drivers, whether they 
>> are for DC-DC converters or driving cathodes. Most high-voltage NMOS 
>> devices (~400V)  need 5-6 volts for solid gate turn-on, yet logic-levels on 
>> most micro-controllers and FPGAs are only 1.2 to 3.3volts.  That leaves a 
>> few options:
>>
>> 1. Use high-voltage logic, such as 4000-series CMOS, to drive HV NMOS. My 
>> first nixie clock does this.
>>
>> 2. Use NPN drivers; you can easily design constant-current drivers 
>> running from 3.3V. I've used surface-mount devices with 2 NPNs inside, and 
>> they are rated at 250V.
>>
>> 3. Use a level-shifter between your logic and the driver. I'm doing this 
>> on my current design with a HV5530.
>>
>> 4. Find a device with a low-voltage gate. Good luck on this; most 
>> logic-levels NMOS devices only handle 60V or less. If you also want low Rds 
>> (necessary for DC-DC converters), you have even fewer options.
>>
>> 5. Pull-up resistor. Ugghhh... It will work, but you will burn more 
>> power. Could be dangerous for a DC-DC converter if the logic/software 
>> turning off the gate goes haywire, which would lead to large 
>> current/burnout.
>>
>>
>> I advise against relying upon the voltage-drop of the nixie tube to 
>> protect the driver, though this is commonly done with designs using the 
>> 74141. I always select my driver to handle the maximum-possible anode 
>> supply, and add margin above that. I've never had a fried driver, ghosting, 
>> etc. Parasitic capacitance and leakage can stress drivers above their 
>> maximum ratings, which will lead to premature failure.
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/5abe0c61-3c71-43e7-8196-41a1a0e5eee9%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Help with Nixie partially lighting up

[Luka C]

Hi everyone,

I need help with my newly built clock. The thing is, after I turn it on and 
the clock runs for like 5-10mins, some of the digits on tubes become 
partially lit like on the picture. Now, I noticed the power supply gets 
very hot after running for 5-10mins and I measured the voltage on +HV 
(+185V) pin of the power supply while the clock is running. It seems like 
the voltage keeps dropping as the supply gets hotter (last time I measure 
it, it went down to 148V and these artifacts became showing on the tubes. 
The digit usually lights up normally at first, then after the cathode 
poisoning routing it becomes partially lit like on the first picture, then 
after the second cathode poisoning routine, it gets partially lit like on 
the second picture and so on, it shows on all of the tubes at some times. 
I'm using 27k anode resistors on IN-14 tubes and I guess the power supply 
is of the design from the picture I attached.

Thanks in advance!



Power supply design (I suppose)










-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/7e4a8d74-0446-4bf1-829e-787929329f51%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] Re: Help with Nixie partially lighting up

[Luka C]

Thank you both.

@gregebert
I would be very thankful if you could explain it or send a link here so I 
can see how I could implement it into my design. My clock uses 4x IN-14 
tubes + 4x INS-1 tubes

@Ian
I can't say for sure since the supply is on a very small board so the whole 
thing gets very hot, but yes, I'd say the MOSFET is the hottest component. 
And, to be honest, I bought the PS from eBay from EU seller because I 
though Chinese ones could be problematic (such irony!) and there was only 
the picture I attached here. After I noticed the problems occurring in the 
PS, I inspected the PS and concluded it should be the one from the 
schematic I posted in my previous message (unfortunately).










Dana ponedjeljak, 21. rujna 2015. u 22:22:52 UTC+2, korisnik gregebert 
napisao je:
>
> I stick with linear power supplies, especially for the anodes 
> (high-voltage), whenever possible. They are much simpler to design, 
> least-likely to overheat, and very reliable. In fact, the main reliability 
> concern is the HV filter cap. You can mitigate that risk by over-design. If 
> anyone's interested, let me know and I'll post my best-known-methods that I 
> use.
>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/10f5d1ce-d8eb-44b7-8e10-12be632c5131%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.