Nice that it works! My understanding is that they need rectified AC, with just a little smoothing so as not to make it a pure DC voltage, as they extinguish during the portion of the voltage when it is to low to maintain the glow, and they are re-ignited when the voltage rise again as long as a trigger electrode has the correct voltage to initiate a glow, much like a thyratron.
Maybe you can find the reason for it working now compared to earlier failures if you check the voltage across the Anode-Cathode-Ignition electrode with an oscilloscope? I've also had some problems with CMOS ic's when I have had poor, or no, connection to VCC/GND on the power pins - the circuit worked quite well until I touched some pins. /Martin On Thursday, 11 February 2021 at 21:50:49 UTC+1 [email protected] wrote: > I have solved the problem but I don't understand why. > Having tried all the suggestions except this, I went to a full wave > rectifier. It is not quite the same as the data sheet but does the same > thing. > It also has the capacitor connected per the data sheet, the same as my > half wave circuit that did not work. > This is rock solid with no indication of any other flickering. I am using > the first (worst) tube. Touching the glass has no effect. > Apparently the tube does not like half wave. Perhaps it was my 1N4007 > diode? They can switch too fast and maybe it needed a snubber. > I also tried DC from a variable regulated power supply (HP 6448B up to 600 > [email protected] amp, and no, I did not go that far). That locked the display on > one digit. The data sheet says it needs rectified line, not DC. > I am running my tests from an HP supply for the 5v and an isolation > transformer driven by a Variac for the AC so I have full control of all > voltages. > This is my final schematic. It is driven by a PIC with parallel connectors > for each display. I used a PIC to generate the BCD because CMOS would not > drive the three 74141 for the Nixies. The PIC counts up and down at varying > speeds which would have been more difficult with CMOS. > [image: 8453 Circuit.jpg] > > > > On Thu, Feb 11, 2021 at 10:33 AM Dekatron42 <[email protected]> wrote: > >> Touching the glass means that you get a capacitive coupling which will >> upset them somewhat, there was a special socket made for this by Philips >> which has a partial metal screen surrounding the tube. The socket is >> nicknamed "der Kuss" , "The kiss", due to its form. I couldn't find a photo >> on the Internet now but I know it exists as I have some in my storage and >> in an instrument that uses them. That instrument is called PW4261 Timer, >> some photos of the externals can be found on the Internet. On this socket >> all of the resistors are mounted flush to the pins to minimize the >> distance, but the capacitors and power supply is mounted some 40cm from the >> tubes themselves, likewise there are long wires to the drivers. >> >> You can also have a look at the manual for the PW4231 which I scanned >> that can be downloaded from here: >> https://frank.pocnet.net/other/sos/Philips_PW4232.pdf if that can help >> you with the voltages for the drivers. >> >> /Martin >> >> >> >> On Thursday, 11 February 2021 at 15:50:55 UTC+1 [email protected] wrote: >> >>> I meant dekatron in my last reply to gregebert. >>> I have 3 of these tubes, all apparently new (NOS). >>> Per your suggestion I tried the other two. One is almost perfect but >>> flashes the 2 a bit when 8 is selected. The third is perfect unless I hold >>> the tube in my fingers by the glass where I get lots of random flashing. >>> Remember that the really bad one works perfectly with the capacitor >>> moved as I mentioned. I don't know if touching the glass affects it. >>> More later, I will be away this morning. >>> Peter. >>> >>> On Thu, Feb 11, 2021 at 2:10 AM Dekatron42 <[email protected]> wrote: >>> >>>> Have you tried the circuit in the J.B Dance book below? Here they use a >>>> center tapped transformer and also a much lower value capacitor, only 33nF >>>> versus 250nF in your circuit - I've only evere seen 33nF used in real >>>> instruments using these indicators. >>>> >>>> Quite a few of the Z550M/ZM1050 are broken internally, I have a box of >>>> them, and that seems to be due to the welding of the internal parts coming >>>> loose when shaken or hit hard (the same problem exists with the >>>> B9012/NL9012 tubes - I have a few broken ones of those too with internal >>>> pieces that have come loose) - sometimes you can hear these loose pieces >>>> if >>>> you shake the tube very carefully close to your ear. What usually happens >>>> is that a rather large round center piece comes loose and in the worst >>>> case >>>> shortens some of the electrodes but usually only makes it hard or >>>> impossible to get all digits to light up poperly and some of them trigger >>>> easier than others due to the distances between electrodes are differing, >>>> shaking the tube a little moves that effect around so other digits will >>>> start to work and vice versa. >>>> >>>> /Martin >>>> >>>> [image: JBDance-Z550M.JPG] >>>> >>>> On Thursday, 11 February 2021 at 06:02:24 UTC+1 gregebert wrote: >>>> >>>>> If you have an isolation transformer, can you put a scope on the >>>>> cathode line to verify the ripple is not excessive ? Rk and Ck create an >>>>> RC >>>>> filter around 77Hz, and the line is 50-60Hz, so there could be some >>>>> ripple. >>>>> >>>>> On Wednesday, February 10, 2021 at 8:26:51 PM UTC-8 [email protected] >>>>> wrote: >>>>> >>>>>> I tried varying it with absolutely no effect. It runs perfectly at 5v >>>>>> with the capacitor moved to the other side of the resistor. However I >>>>>> was >>>>>> clocking slowly. It may not run at spec speed. I seem to remember >>>>>> reading >>>>>> of using 5 to 8 volts for the logic. >>>>>> The data sheet shows the following which I'm inclined to believe. I >>>>>> will try Ck directly on the pins tomorrow. Someone must have tried this >>>>>> circuit. >>>>>> [image: 8453 Cct.JPG] >>>>>> >>>>>> On Wed, Feb 10, 2021 at 11:12 PM gregebert <[email protected]> >>>>>> wrote: >>>>>> >>>>>>> I suspect the differential voltage between lit vs non-lit numerals >>>>>>> is too low. The CMOS device is basically driving grids to determine >>>>>>> which >>>>>>> cathode will be illuminated. I've seen similar behavior with an A-101 >>>>>>> dekatron. What voltage are you using for VDD ? Is it 8V as indicated on >>>>>>> the >>>>>>> schematic, or a more-conventional 5V ? >>>>>>> >>>>>>> On Wednesday, February 10, 2021 at 7:22:25 PM UTC-8 >>>>>>> [email protected] wrote: >>>>>>> >>>>>>>> I think you are wrong. The literature explains this tube is >>>>>>>> designed to be driven by 5v logic and it does work. The steering >>>>>>>> electrodes >>>>>>>> are close to the Anode voltage which is grounded and only 5v pulses >>>>>>>> (square >>>>>>>> waves) are needed. Note all the cathodes are connected together >>>>>>>> internally. >>>>>>>> I got to thinking about the C1 position. The data sheet shows it >>>>>>>> per my schematic and I can't believe they made that mistake several >>>>>>>> times >>>>>>>> including the hand drawn notes of the designer. I think maybe I needed >>>>>>>> to >>>>>>>> have C1 directly on the socket pins and will try that tomorrow. >>>>>>>> Thanks for your interest though but you are thinking Nixie Tubes, >>>>>>>> this is a special tube with steering electrodes.. >>>>>>>> >>>>>>>> >>>>>>>> On Wed, Feb 10, 2021 at 9:42 PM chuckrr <[email protected]> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> That schematic is so wrong in so many ways. You need high >>>>>>>>> voltage transistors operating the tube cathodes. >>>>>>>>> >>>>>>>>> You need high resistance drving the transistor bases. You need a >>>>>>>>> buffer such as 4049 or 4050 driving the >>>>>>>>> >>>>>>>>> resistor, which in turn drives the transistor base. Only then >>>>>>>>> would I dare to use the 4028....to operate the buffer, which in turn >>>>>>>>> operates the >>>>>>>>> >>>>>>>>> transistor base via appropriate high resistance. That is the only >>>>>>>>> sure fire way I know of to attain noise-free performance from CMOS >>>>>>>>> logic >>>>>>>>> >>>>>>>>> driving cold cathode tubes. That schematic there is a noisy deal. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> ---- Original Message ---- >>>>>>>>> From: "peter bunge" <[email protected]> >>>>>>>>> Sent: 2/10/2021 8:15:21 PM >>>>>>>>> To: "neonixie-l" <[email protected]> >>>>>>>>> Subject: [neonixie-l] 8453/Z550M erratic >>>>>>>>> >>>>>>>>> Using the schematic from the data sheet >>>>>>>>> [image: 8453 Circuit.jpg] >>>>>>>>> This works with the count moving around but other numbers flash >>>>>>>>> erratically, especially close to the number that is supposed to be >>>>>>>>> lit. >>>>>>>>> *When I moved the bottom of C1 to the other side of R1 it works >>>>>>>>> perfectly. It is rock steady and does not care about line voltage or >>>>>>>>> the >>>>>>>>> 5v supply (shown above as 8v but used at 5v)* >>>>>>>>> If this is an error it is continued through all the documentation >>>>>>>>> and is consistent. Changing the value of C1 up and down by 10 had >>>>>>>>> little >>>>>>>>> effect but a smaller C1 helps a bit. My Rst are all directly on the >>>>>>>>> socket >>>>>>>>> pins and the wires are all about 5 inches long. >>>>>>>>> I have varied the line voltage with little effect. >>>>>>>>> Any suggestions??? >>>>>>>>> >>>>>>>>> -- >>>>>>>>> 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 [email protected]. >>>>>>>>> To view this discussion on the web, visit >>>>>>>>> https://groups.google.com/d/msgid/neonixie-l/b996ef0b-edbb-4c77-8ec1-57255dd58a82n%40googlegroups.com >>>>>>>>> >>>>>>>>> <https://groups.google.com/d/msgid/neonixie-l/b996ef0b-edbb-4c77-8ec1-57255dd58a82n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>>>>> . >>>>>>>>> >>>>>>>>> -- >>>>>>>>> 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 [email protected]. >>>>>>>>> >>>>>>>> To view this discussion on the web, visit >>>>>>>>> https://groups.google.com/d/msgid/neonixie-l/202102101842701.SM05295%40ADT-IM1 >>>>>>>>> >>>>>>>>> <https://groups.google.com/d/msgid/neonixie-l/202102101842701.SM05295%40ADT-IM1?utm_medium=email&utm_source=footer> >>>>>>>>> . >>>>>>>>> >>>>>>>> -- >>>>>>> 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 [email protected]. >>>>>>> >>>>>> To view this discussion on the web, visit >>>>>>> https://groups.google.com/d/msgid/neonixie-l/9543e23e-ebb0-46e7-8624-ddcf4a843667n%40googlegroups.com >>>>>>> >>>>>>> <https://groups.google.com/d/msgid/neonixie-l/9543e23e-ebb0-46e7-8624-ddcf4a843667n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>>> . >>>>>>> >>>>>> -- >>>> 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 [email protected]. >>>> >>> To view this discussion on the web, visit >>>> https://groups.google.com/d/msgid/neonixie-l/7085da7d-27dd-46f3-9bba-9e2d369dcc4fn%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/neonixie-l/7085da7d-27dd-46f3-9bba-9e2d369dcc4fn%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> -- >> 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 [email protected]. >> > To view this discussion on the web, visit >> https://groups.google.com/d/msgid/neonixie-l/ad3deb14-52ce-4b11-b048-dd6dea83bdban%40googlegroups.com >> >> <https://groups.google.com/d/msgid/neonixie-l/ad3deb14-52ce-4b11-b048-dd6dea83bdban%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > -- 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 [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/dcd99e19-3b9c-406c-8057-d69f1547ce03n%40googlegroups.com.
