Re: [Emc-users] ATS-667 advice

[Gene Heskett]

On Saturday 15 October 2016 02:55:39 Chris Albertson wrote:

> Actually it's easy.  You use a hot air re-work station.   Basically a
> hair drier with a 1/8th inch diameter nozzle.
> Hold the little part in place with a wood tooth pick and aim the hot
> air generally at the part.  The solder on the part and the PCB melt,
> release the toothpick and surface tension then aligns the part on the
> PCB pad before the solder resolidifies.   You can do it with a solder
> iron but it's harder.  Hot air stations have gotten cheap, as low as
> $40.

Already have one, sitting about 3 feet away ATM. :)

> > >They are cheaper than the through hole caps that I normally
> > > buy. Hand soldering them will be a royal PIMA!
> >
> > At .2mm x.4mm, you have that right, the pain will be in My Ass. My
> > hands are just starting to get nervous,


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Gene Heskett]

On Saturday 15 October 2016 02:38:59 Chris Albertson wrote:

> A source for 0.1 bypass caps?   I buy most everything fro eBay now.
> But you don't want a film cap, you want a "Multilayer Monolithic
> Ceramic Capacitor".  I just happen to have one here on my desk and a
> copper US penny too.  I place the cap on the penny and see it is about
> 2/3 the size of Mr. Lincoln's head   or maybe 4x6 millimeters.   They
> typically sell for 1 cent each.
> Here is a random vendor, the first one Google found
> http://www.ebay.com/itm/0-1uF...
> pacitor-100-pcs-LW-SZUS-/181932472979?hash=item2a5c053a93:g:NxMAAOSwiwV
>WT0bI>

Thats about right, I put a bag of them in my cart for $0.99.

> This shows how an RPM sensor from a Chinese mini lathe is mounted.  I
> build these using perf board for robot joints and wheels.  In this
> case their sensor points up from the PCB but you can bend the leads
> and have it look parallel to the PCB.  The picture shows an optical
> sensor but hall effects mount the same way  There is always room for a
> tiny passive part http://littlemachineshop.com...
> roductID=1945>

I didn't see any all effect sensors, and that optical is too big unless I 
put one of Lester Caine's wheels clear out on the rear of the spindle.  
And there's nothing out there but air to bolt it to. The swingout cover 
on the left is empty now. I'll put a sheet metal roof over the z drive 
belt thats using an inch of that space.  Ditto the X drive belt but I've 
not quite visualized that one, yet. Possibly route a cover profile in 
1/4" alu, with a groove to hold some 24 gauge alu.

At least I don't have to sweat carving an optical encoding wheel with the 
ATS667's watching that 60 tooth bull gear.  Ack the docs, quadrature 
accuracy is off a couple percent, but I have a hal file kit that does 
a /4 to that error if needed.

Hint, use it only in the feedback path to the PID driving the spindle. I 
need to do that in both the G0704 and TLM's hal files after standing 
back to see the whole picture.

> One of my goals for my CNC mini-mill is to make PCBs by routing the
> space between the coper traces then I can convert to using surface
> mount parts even for one-off prototypes but for now I use these little
> radial lead 0.1 capacitors.  The surface mount equivalents are less
> than 1/4 the size.

I do that routing by hand, almost in a varoni style. 100x faster than 
trying to do something like that in eagle. I'll need, on the left face 
of the indexing 667 bracket, wire wrap pads for 5 surface connections, 
expanded out to where I have pads big enough I can paste the 
interconnect cable to it. Then I can use wrapping wire the last 1.5" to 
the devices.  And it all comes off as a unit with one 5mm bolt thru the 
front of the spindle houseing when I'm done. Where the index 667 is, I 
still need to carve a path for its 3 wires to that pad on the pcb. Its 
mounted in reverse in the trench, but the trench wasn't cut clear thru 
to the left end so the wires will have to go back by it to get out. I 
made the trench wide enough I can stuff the wrapping wire into the gap 
and spot them in place with a bit of superglue.

Next subject;
Time to check the gaskets in my dishwasher, it put a couple ounces of 
water on the floor during this run.  Real life is always bothering me 
with piddly stuff. :(

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Chris Albertson]

Actually it's easy.  You use a hot air re-work station.   Basically a hair
drier with a 1/8th inch diameter nozzle.
Hold the little part in place with a wood tooth pick and aim the hot air
generally at the part.  The solder on the part and the PCB melt, release
the toothpick and surface tension then aligns the part on the PCB pad
before the solder resolidifies.   You can do it with a solder iron but it's
harder.  Hot air stations have gotten cheap, as low as $40.



> >They are cheaper than the through hole caps that I normally buy.
> > Hand soldering them will be a royal PIMA!
> >
> At .2mm x.4mm, you have that right, the pain will be in My Ass. My hands
> are just starting to get nervous,


-- 

Chris Albertson
Redondo Beach, California
--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Chris Albertson]

A source for 0.1 bypass caps?   I buy most everything fro eBay now.
But you don't want a film cap, you want a "Multilayer Monolithic Ceramic
Capacitor".  I just happen to have one here on my desk and a copper US
penny too.  I place the cap on the penny and see it is about 2/3 the size
of Mr. Lincoln's head   or maybe 4x6 millimeters.   They typically sell for
1 cent each.
Here is a random vendor, the first one Google found
http://www.ebay.com/itm/0-1uF...


This shows how an RPM sensor from a Chinese mini lathe is mounted.  I build
these using perf board for robot joints and wheels.  In this case their
sensor points up from the PCB but you can bend the leads and have it look
parallel to the PCB.  The picture shows an optical sensor but hall effects
mount the same way  There is always room for a tiny passive part
http://littlemachineshop.com...


One of my goals for my CNC mini-mill is to make PCBs by routing the space
between the coper traces then I can convert to using surface mount parts
even for one-off prototypes but for now I use these little radial lead 0.1
capacitors.  The surface mount equivalents are less than 1/4 the size.



On Fri, Oct 14, 2016 at 12:01 PM, Gene Heskett  wrote:

> On Friday 14 October 2016 14:45:31 Chris Albertson wrote:
>
> > Typically 0.1 bypass caps REALLY to have to be physically close to the
> > load they are bypassing. Within a few millimeters.  Cut the leads
> > short and bridge the power pins with the cap.
> >
> > Placing a larger cap farther away can't cause any harm but it is not
> > going to do the same job.
> >
> > Reason:  The real world is not the same as theory.   All caps have an
> > internal series resistance (ESR). Think of them as a cap and resister
> > in series and model them that way too when you design the circuit.
> > The  ESR of a .1uF ceramic cap is nearly zero, but for the tantalum,
> > you need to look it up in the data sheet.
> >
> > Why physically close to the load.  Wire has non-zero impedance.
> > Mainly in this case it is inductive.   You need to minimize the value
> > of the inductor between the 0.1 cap and the device.  Apparently the
> > manufacturer wants a low impedance power supply.
> >
> > Draw the circuit schematic with a the lead wire inductance and cap's
> > ESR shown then you can see that needs to be minimized.  Don't worry
> > about space you candy 0.1 caps that are the size of a large grain of
> > sand
>
> Anyplace of note to source them?
> >
> > The usual way to mount these is to make a tiny PCB that has the device
> > on to and a mounting hole and the passive parts like the capacitor(s)
> > and a connector for the cable.
> >
> > All that said, it will likely work OK if you can get that small ca
> > reasonably close.   Better to get the 0.1 closer then a larger one
> > farther way as the goal is to minimize the amount of wire
> >
> > And using a larger cap is no substitute for the .01 either because I
> > bet the .1 was spec'd mainly for its low ESR, not because it needed to
> > be exactly 0.1uF
> >
> As a CET, I am well aware of all that, including the inductance of 2" of
> wire.
>
> As for the ESR, tantalums are generally sub ohmic too.  For half a
> century or more. Its reverse bias that nails them to the fence, even a
> few millivolts will do it eventually.
>
> Any caps on the leads will have to be not longer than the 667 is wide,
> and not more than half its thickness.
>
> Thanks Chris.
>
> > On Fri, Oct 14, 2016 at 9:51 AM, Gene Heskett 
> wrote:
> > > Greetings everybody;
> > >
> > > A bit too cold this morning to go out to the shop and saw out the
> > > thicker spacers I am going to need to permanently mount the Z motor
> > > and screw anchor.
> > >
> > > So, not wanting to waste the day waiting for it to warm, I went back
> > > to work on the encoder thing this morning.
> > >
> > > I now have an additional tab of alu extending to the left to hold
> > > one of these ATS667's as an index pulse generator, and have that one
> > > s-glued in a bit beyond where the center of the glued on screw
> > > passes by, hoping it won't trigger on a regular tooth.
> > >
> > > But, I've never used any of these before.  No clue as to their bad
> > > habits.
> > >
> > > The spec sheet I have shows a "required" .1 bypass to ground on the
> > > VCC lead, but has anyone had any problems using just one for all 3
> > > devices? This thing is cramped for space for a bulky mylar .1.
> > >
> > > So I intend to put an even larger tantalum on the pcb, which will be
> > > on the left face of this added tab as I can let it project inward
> > > nearly 20 mm's in that space, lots of space before encountering
> > > anything that moves there.  That means I'll have to paste some
> > > wrapping wire 

Re: [Emc-users] ATS-667 advice

[Gene Heskett]

On Friday 14 October 2016 23:40:08 Erik Christiansen wrote:

> On 14.10.16 18:41, MC Cason wrote:
> >Here is one in a 01005 package (Metric 0402)   0.016"L x 0.008"W
> > - (0.40mm x 0.20mm) :
> > http://www.mouser.com/ProductDetail/AVX/01016D104KATUA/?qs=sGAEpiMZZ
> >Ms0AnBnWHyRQAwkDkDzLNakJDHmuCcZvLc%3d
> >
> >They are cheaper than the through hole caps that I normally buy.
> > Hand soldering them will be a royal PIMA!
>
> Gene, there appears to be room for an 0805 sized cap:
>
> https://en.wikipedia.org/wiki/Surface-mount_technology#Package_sizes
>
> and they're a darn sight easier to wrangle¹, I find. An 0805 readily
> solders directly between pins 1&4 of the ATS-667, if you can move pins
> 2&3 one grid out of line. There's then about 0.6 mm (0.024") to bridge
> on each component end, according to preferred technique. (And not
> enough path length for any quibbling about inductive Z)
>
> If not inclined to buy a plastic box of assorted capacitor values,
> then most retailers have them in little strips of 10 or so. (They're
> worth about 0.00 individually, at the discount places, so we're lucky
> if they're available in less than 100 quantity.)
>
> Erik
> (Who bought a magnifier lamp at the same time as the SMD tweezer-type
> soldering station, and still doesn't want to muck with smaller than
> 0805.)
>
Chuckle, I am on my 3rd or 4th one of those.

> ¹ Not only big enough to see, but there's enough meat on 'em to grab
>   with a pair of tweezers, or hold down with a toothpick while
> soldering. (They can "tombstone", i.e. rear up when the first end is
> soldered, otherwise.)

Surface tension, one end only.

I may see if I can source some solder paste, turn the air flow down to 
zip and see how that works.  Done right, surface tension of the solder 
should pull it dead center across the two traces. But you have to 
liquify the paste simultaneously on both ends.  And it is still not 
fun...

Cheers, Erik.

Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Erik Christiansen]

On 14.10.16 18:41, MC Cason wrote:
>Here is one in a 01005 package (Metric 0402)   0.016"L x 0.008"W - 
> (0.40mm x 0.20mm) :
> http://www.mouser.com/ProductDetail/AVX/01016D104KATUA/?qs=sGAEpiMZZMs0AnBnWHyRQAwkDkDzLNakJDHmuCcZvLc%3d
> 
>They are cheaper than the through hole caps that I normally buy. Hand 
> soldering them will be a royal PIMA!

Gene, there appears to be room for an 0805 sized cap:

https://en.wikipedia.org/wiki/Surface-mount_technology#Package_sizes

and they're a darn sight easier to wrangle¹, I find. An 0805 readily
solders directly between pins 1&4 of the ATS-667, if you can move pins
2&3 one grid out of line. There's then about 0.6 mm (0.024") to bridge
on each component end, according to preferred technique. (And not enough
path length for any quibbling about inductive Z)

If not inclined to buy a plastic box of assorted capacitor values, then
most retailers have them in little strips of 10 or so. (They're worth
about 0.00 individually, at the discount places, so we're lucky if
they're available in less than 100 quantity.)

Erik
(Who bought a magnifier lamp at the same time as the SMD tweezer-type
soldering station, and still doesn't want to muck with smaller than 0805.)

¹ Not only big enough to see, but there's enough meat on 'em to grab
  with a pair of tweezers, or hold down with a toothpick while soldering.
  (They can "tombstone", i.e. rear up when the first end is soldered,
  otherwise.)

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Gene Heskett]

On Friday 14 October 2016 19:41:23 MC Cason wrote:

> Gene,
>
>Inline
>
> On 10/14/2016 02:01 PM, Gene Heskett wrote:
> > On Friday 14 October 2016 14:45:31 Chris Albertson wrote:
> >> Typically 0.1 bypass caps REALLY to have to be physically close to
> >> the load they are bypassing. Within a few millimeters.  Cut the
> >> leads short and bridge the power pins with the cap.
> >>
> >> Placing a larger cap farther away can't cause any harm but it is
> >> not going to do the same job.
> >>
> >> Reason:  The real world is not the same as theory.   All caps have
> >> an internal series resistance (ESR). Think of them as a cap and
> >> resister in series and model them that way too when you design the
> >> circuit. The  ESR of a .1uF ceramic cap is nearly zero, but for the
> >> tantalum, you need to look it up in the data sheet.
> >>
> >> Why physically close to the load.  Wire has non-zero impedance.
> >> Mainly in this case it is inductive.   You need to minimize the
> >> value of the inductor between the 0.1 cap and the device. 
> >> Apparently the manufacturer wants a low impedance power supply.
> >>
> >> Draw the circuit schematic with a the lead wire inductance and
> >> cap's ESR shown then you can see that needs to be minimized.  Don't
> >> worry about space you candy 0.1 caps that are the size of a large
> >> grain of sand
> >
> > Anyplace of note to source them?
>
>Here is one in a 01005 package (Metric 0402)   0.016"L x 0.008"W -
> (0.40mm x 0.20mm) :
> http://www.mouser.com/ProductDetail/AVX/01016D104KATUA/?qs=sGAEpiMZZMs
>0AnBnWHyRQAwkDkDzLNakJDHmuCcZvLc%3d
>
>They are cheaper than the through hole caps that I normally buy.
> Hand soldering them will be a royal PIMA!
>
At .2mm x.4mm, you have that right, the pain will be in My Ass. My hands 
are just starting to get nervous, which won't help a bit.  And to see 
them well, I might have to get some cataracts removed.  That is about 
the size of a comma in the newspaper!  My hot air wand could blow it 
away, never to be found again by anything but the vacuum cleaner.  That 
would fit across the gap between 2 traces, cut about as narrow as my 
mill could make them.  This is going to be fun, for some definition of 
fun. I may even see if a bigger one, say about 1/8" long might be had.  
Those I have played with a wee bit.

Thanks MC Cason.

> >> The usual way to mount these is to make a tiny PCB that has the
> >> device on to and a mounting hole and the passive parts like the
> >> capacitor(s) and a connector for the cable.

I considered that, but the magnets in the 667's are strong enough to 
wiggle and crack the joints, so the 667s are super-glued to the bottom 
of their protective trench's.

I'll have to reach a much bigger board a couple inches away with wire 
wrapping wire before it graduates to a 5 conductor cable to snake its 
way out of the back rear of the spindle housing.  Stress relief & all 
that twaddle. :-)
> >>
> >> All that said, it will likely work OK if you can get that small ca
> >> reasonably close.   Better to get the 0.1 closer then a larger one
> >> farther way as the goal is to minimize the amount of wire
> >>
> >> And using a larger cap is no substitute for the .01 either because
> >> I bet the .1 was spec'd mainly for its low ESR, not because it
> >> needed to be exactly 0.1uF
> >
> > As a CET, I am well aware of all that, including the inductance of
> > 2" of wire.
> >
> > As for the ESR, tantalums are generally sub ohmic too.  For half a
> > century or more. Its reverse bias that nails them to the fence, even
> > a few millivolts will do it eventually.
> >
> > Any caps on the leads will have to be not longer than the 667 is
> > wide, and not more than half its thickness.
> >
> > Thanks Chris.
> >
> >> On Fri, Oct 14, 2016 at 9:51 AM, Gene Heskett
> >> 
> >
> > wrote:
> >>> Greetings everybody;
> >>>
> >>> A bit too cold this morning to go out to the shop and saw out the
> >>> thicker spacers I am going to need to permanently mount the Z
> >>> motor and screw anchor.
> >>>
> >>> So, not wanting to waste the day waiting for it to warm, I went
> >>> back to work on the encoder thing this morning.
> >>>
> >>> I now have an additional tab of alu extending to the left to hold
> >>> one of these ATS667's as an index pulse generator, and have that
> >>> one s-glued in a bit beyond where the center of the glued on screw
> >>> passes by, hoping it won't trigger on a regular tooth.
> >>>
> >>> But, I've never used any of these before.  No clue as to their bad
> >>> habits.
> >>>
> >>> The spec sheet I have shows a "required" .1 bypass to ground on
> >>> the VCC lead, but has anyone had any problems using just one for
> >>> all 3 devices? This thing is cramped for space for a bulky mylar
> >>> .1.
> >>>
> >>> So I intend to put an even larger tantalum on the pcb, which will
> >>> be on the left face of this added tab as I can let it project
> >>> inward nearly 20 mm's in that space, lots of space before

Re: [Emc-users] ATS-667 advice

[MC Cason]

Gene,

   Inline

On 10/14/2016 02:01 PM, Gene Heskett wrote:
> On Friday 14 October 2016 14:45:31 Chris Albertson wrote:
>
>> Typically 0.1 bypass caps REALLY to have to be physically close to the
>> load they are bypassing. Within a few millimeters.  Cut the leads
>> short and bridge the power pins with the cap.
>>
>> Placing a larger cap farther away can't cause any harm but it is not
>> going to do the same job.
>>
>> Reason:  The real world is not the same as theory.   All caps have an
>> internal series resistance (ESR). Think of them as a cap and resister
>> in series and model them that way too when you design the circuit.
>> The  ESR of a .1uF ceramic cap is nearly zero, but for the tantalum,
>> you need to look it up in the data sheet.
>>
>> Why physically close to the load.  Wire has non-zero impedance.
>> Mainly in this case it is inductive.   You need to minimize the value
>> of the inductor between the 0.1 cap and the device.  Apparently the
>> manufacturer wants a low impedance power supply.
>>
>> Draw the circuit schematic with a the lead wire inductance and cap's
>> ESR shown then you can see that needs to be minimized.  Don't worry
>> about space you candy 0.1 caps that are the size of a large grain of
>> sand
> Anyplace of note to source them?

   Here is one in a 01005 package (Metric 0402)   0.016"L x 0.008"W - 
(0.40mm x 0.20mm) :
http://www.mouser.com/ProductDetail/AVX/01016D104KATUA/?qs=sGAEpiMZZMs0AnBnWHyRQAwkDkDzLNakJDHmuCcZvLc%3d

   They are cheaper than the through hole caps that I normally buy. Hand 
soldering them will be a royal PIMA!


>> The usual way to mount these is to make a tiny PCB that has the device
>> on to and a mounting hole and the passive parts like the capacitor(s)
>> and a connector for the cable.
>>
>> All that said, it will likely work OK if you can get that small ca
>> reasonably close.   Better to get the 0.1 closer then a larger one
>> farther way as the goal is to minimize the amount of wire
>>
>> And using a larger cap is no substitute for the .01 either because I
>> bet the .1 was spec'd mainly for its low ESR, not because it needed to
>> be exactly 0.1uF
>>
> As a CET, I am well aware of all that, including the inductance of 2" of
> wire.
>
> As for the ESR, tantalums are generally sub ohmic too.  For half a
> century or more. Its reverse bias that nails them to the fence, even a
> few millivolts will do it eventually.
>
> Any caps on the leads will have to be not longer than the 667 is wide,
> and not more than half its thickness.
>
> Thanks Chris.
>
>> On Fri, Oct 14, 2016 at 9:51 AM, Gene Heskett 
> wrote:
>>> Greetings everybody;
>>>
>>> A bit too cold this morning to go out to the shop and saw out the
>>> thicker spacers I am going to need to permanently mount the Z motor
>>> and screw anchor.
>>>
>>> So, not wanting to waste the day waiting for it to warm, I went back
>>> to work on the encoder thing this morning.
>>>
>>> I now have an additional tab of alu extending to the left to hold
>>> one of these ATS667's as an index pulse generator, and have that one
>>> s-glued in a bit beyond where the center of the glued on screw
>>> passes by, hoping it won't trigger on a regular tooth.
>>>
>>> But, I've never used any of these before.  No clue as to their bad
>>> habits.
>>>
>>> The spec sheet I have shows a "required" .1 bypass to ground on the
>>> VCC lead, but has anyone had any problems using just one for all 3
>>> devices? This thing is cramped for space for a bulky mylar .1.
>>>
>>> So I intend to put an even larger tantalum on the pcb, which will be
>>> on the left face of this added tab as I can let it project inward
>>> nearly 20 mm's in that space, lots of space before encountering
>>> anything that moves there.  That means I'll have to paste some
>>> wrapping wire to extend the leads about 1" to get to the pcb.  So
>>> this bypass cap may be as much as 2" of wire away from the device
>>> body.
>>>
>>> Can anyone testify that the extra lead length is a problem?
>>>
>>> Thanks all.
>>>
>>> Cheers, Gene Heskett
>>> --
>>> "There are four boxes to be used in defense of liberty:
>>>   soap, ballot, jury, and ammo. Please use in that order."
>>> -Ed Howdershelt (Author)
>>> Genes Web page 
>>>
>>> 
>>> -- Check out the vibrant tech community on one of the world's
>>> most engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>>> ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> Cheers, Gene Heskett


-- 
MC Cason
Eagle3D - Created by Matthias Weißer
github.com/mcason/Eagle3D



--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot

Re: [Emc-users] ATS-667 advice

[Gene Heskett]

On Friday 14 October 2016 14:45:31 Chris Albertson wrote:

> Typically 0.1 bypass caps REALLY to have to be physically close to the
> load they are bypassing. Within a few millimeters.  Cut the leads
> short and bridge the power pins with the cap.
>
> Placing a larger cap farther away can't cause any harm but it is not
> going to do the same job.
>
> Reason:  The real world is not the same as theory.   All caps have an
> internal series resistance (ESR). Think of them as a cap and resister
> in series and model them that way too when you design the circuit.
> The  ESR of a .1uF ceramic cap is nearly zero, but for the tantalum,
> you need to look it up in the data sheet.
>
> Why physically close to the load.  Wire has non-zero impedance.
> Mainly in this case it is inductive.   You need to minimize the value
> of the inductor between the 0.1 cap and the device.  Apparently the
> manufacturer wants a low impedance power supply.
>
> Draw the circuit schematic with a the lead wire inductance and cap's
> ESR shown then you can see that needs to be minimized.  Don't worry
> about space you candy 0.1 caps that are the size of a large grain of
> sand

Anyplace of note to source them?
>
> The usual way to mount these is to make a tiny PCB that has the device
> on to and a mounting hole and the passive parts like the capacitor(s)
> and a connector for the cable.
>
> All that said, it will likely work OK if you can get that small ca
> reasonably close.   Better to get the 0.1 closer then a larger one
> farther way as the goal is to minimize the amount of wire
>
> And using a larger cap is no substitute for the .01 either because I
> bet the .1 was spec'd mainly for its low ESR, not because it needed to
> be exactly 0.1uF
>
As a CET, I am well aware of all that, including the inductance of 2" of 
wire.

As for the ESR, tantalums are generally sub ohmic too.  For half a 
century or more. Its reverse bias that nails them to the fence, even a 
few millivolts will do it eventually.

Any caps on the leads will have to be not longer than the 667 is wide, 
and not more than half its thickness.

Thanks Chris.

> On Fri, Oct 14, 2016 at 9:51 AM, Gene Heskett  
wrote:
> > Greetings everybody;
> >
> > A bit too cold this morning to go out to the shop and saw out the
> > thicker spacers I am going to need to permanently mount the Z motor
> > and screw anchor.
> >
> > So, not wanting to waste the day waiting for it to warm, I went back
> > to work on the encoder thing this morning.
> >
> > I now have an additional tab of alu extending to the left to hold
> > one of these ATS667's as an index pulse generator, and have that one
> > s-glued in a bit beyond where the center of the glued on screw
> > passes by, hoping it won't trigger on a regular tooth.
> >
> > But, I've never used any of these before.  No clue as to their bad
> > habits.
> >
> > The spec sheet I have shows a "required" .1 bypass to ground on the
> > VCC lead, but has anyone had any problems using just one for all 3
> > devices? This thing is cramped for space for a bulky mylar .1.
> >
> > So I intend to put an even larger tantalum on the pcb, which will be
> > on the left face of this added tab as I can let it project inward
> > nearly 20 mm's in that space, lots of space before encountering
> > anything that moves there.  That means I'll have to paste some
> > wrapping wire to extend the leads about 1" to get to the pcb.  So
> > this bypass cap may be as much as 2" of wire away from the device
> > body.
> >
> > Can anyone testify that the extra lead length is a problem?
> >
> > Thanks all.
> >
> > Cheers, Gene Heskett
> > --
> > "There are four boxes to be used in defense of liberty:
> >  soap, ballot, jury, and ammo. Please use in that order."
> > -Ed Howdershelt (Author)
> > Genes Web page 
> >
> > 
> >-- Check out the vibrant tech community on one of the world's
> > most engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> > ___
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ATS-667 advice

[Chris Albertson]

Typically 0.1 bypass caps REALLY to have to be physically close to the
load they are bypassing. Within a few millimeters.  Cut the leads
short and bridge the power pins with the cap.

Placing a larger cap farther away can't cause any harm but it is not
going to do the same job.

Reason:  The real world is not the same as theory.   All caps have an
internal series resistance (ESR). Think of them as a cap and resister
in series and model them that way too when you design the circuit.
The  ESR of a .1uF ceramic cap is nearly zero, but for the tantalum,
you need to look it up in the data sheet.

Why physically close to the load.  Wire has non-zero impedance.
Mainly in this case it is inductive.   You need to minimize the value
of the inductor between the 0.1 cap and the device.  Apparently the
manufacturer wants a low impedance power supply.

Draw the circuit schematic with a the lead wire inductance and cap's
ESR shown then you can see that needs to be minimized.  Don't worry
about space you candy 0.1 caps that are the size of a large grain of
sand

The usual way to mount these is to make a tiny PCB that has the device
on to and a mounting hole and the passive parts like the capacitor(s)
and a connector for the cable.

All that said, it will likely work OK if you can get that small ca
reasonably close.   Better to get the 0.1 closer then a larger one
farther way as the goal is to minimize the amount of wire

And using a larger cap is no substitute for the .01 either because I
bet the .1 was spec'd mainly for its low ESR, not because it needed to
be exactly 0.1uF

On Fri, Oct 14, 2016 at 9:51 AM, Gene Heskett  wrote:
> Greetings everybody;
>
> A bit too cold this morning to go out to the shop and saw out the thicker
> spacers I am going to need to permanently mount the Z motor and screw
> anchor.
>
> So, not wanting to waste the day waiting for it to warm, I went back to
> work on the encoder thing this morning.
>
> I now have an additional tab of alu extending to the left to hold one of
> these ATS667's as an index pulse generator, and have that one s-glued in
> a bit beyond where the center of the glued on screw passes by, hoping it
> won't trigger on a regular tooth.
>
> But, I've never used any of these before.  No clue as to their bad
> habits.
>
> The spec sheet I have shows a "required" .1 bypass to ground on the VCC
> lead, but has anyone had any problems using just one for all 3 devices?
> This thing is cramped for space for a bulky mylar .1.
>
> So I intend to put an even larger tantalum on the pcb, which will be on
> the left face of this added tab as I can let it project inward nearly 20
> mm's in that space, lots of space before encountering anything that
> moves there.  That means I'll have to paste some wrapping wire to extend
> the leads about 1" to get to the pcb.  So this bypass cap may be as much
> as 2" of wire away from the device body.
>
> Can anyone testify that the extra lead length is a problem?
>
> Thanks all.
>
> Cheers, Gene Heskett
> --
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> Genes Web page 
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



-- 

Chris Albertson
Redondo Beach, California

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] ATS-667 advice

[Gene Heskett]

Greetings everybody;

A bit too cold this morning to go out to the shop and saw out the thicker 
spacers I am going to need to permanently mount the Z motor and screw 
anchor.

So, not wanting to waste the day waiting for it to warm, I went back to 
work on the encoder thing this morning.

I now have an additional tab of alu extending to the left to hold one of 
these ATS667's as an index pulse generator, and have that one s-glued in 
a bit beyond where the center of the glued on screw passes by, hoping it 
won't trigger on a regular tooth.

But, I've never used any of these before.  No clue as to their bad 
habits.

The spec sheet I have shows a "required" .1 bypass to ground on the VCC 
lead, but has anyone had any problems using just one for all 3 devices? 
This thing is cramped for space for a bulky mylar .1.

So I intend to put an even larger tantalum on the pcb, which will be on 
the left face of this added tab as I can let it project inward nearly 20 
mm's in that space, lots of space before encountering anything that 
moves there.  That means I'll have to paste some wrapping wire to extend 
the leads about 1" to get to the pcb.  So this bypass cap may be as much 
as 2" of wire away from the device body.

Can anyone testify that the extra lead length is a problem?

Thanks all.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users