13 Jul
2012
13 Jul
'12
8:49 p.m.
I'm getting around to doing some work on my veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB
goodness inside, the first thing that jumps out at me is a bulging
electrolytic. It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not
50V, but 40V. I'm going to substitute a pair of 470uF, 50V units
paralleled as a substitute, but this had me wondering if anyone knows
why the strange values, particularly since +/-20 percent tolerances
are common on electrolytic caps.
--Chuck
14 Jul
14 Jul
1:34 a.m.
On Fri, 13 Jul 2012, Chuck Guzis wrote:
I'm getting around to doing some work on my
veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB
goodness inside, the first thing that jumps out at me is a bulging
electrolytic. It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not
50V, but 40V. I'm going to substitute a pair of 470uF, 50V units
paralleled as a substitute, but this had me wondering if anyone knows
why the strange values, particularly since +/-20 percent tolerances
are common on electrolytic caps.
The 940uF 40V capacitor just dates to before we had the modern preferred
values system. If it is just a filter capacitor, a 1000uF 50V should work
just fine, and would likely be more reliable (and have less leakage) than
two smaller capacitors in parallel.
I replace odd value capacitors all the time in older SMPSUs and as long as
the values are close, substitutions are generally not a problem. The hard
part I've found is finding modern parts with the same lead spacing.
Sometimes you have to go up a case size and/or voltage rating to find an
electrolytic that will fit the existing location without putting excessive
bends in their leads. This is especially true with old 3-lead "computer
grade" capacitors (typically two ground leads).
12:09 p.m.
On Sat, 14 Jul 2012, William Donzelli wrote:
Are you sure about that? I thought with /electrolytics/ it was the 70s?
I've seen these "odd value" electrolytics in gear which was made up
through the early 80s (using old stock parts).
The 940uF 40V
capacitor just dates to before we had the modern preferred
values system.
No. The preferred values system dates back to the 1930s in the US. 
1:38 p.m.
Are you sure about that?
I thought with /electrolytics/ it was the 70s? I've
seen these "odd value" electrolytics in gear which was made up through the
early 80s (using old stock parts).
No one said the manufactures needed to stick to the standard, and with
a giant like HP, values could be tweaked while working with the cap
makers.
--
Will
1:21 p.m.
On 14 Jul 2012 at 0:34, Tothwolf wrote:
I replace odd value capacitors all the time in older
SMPSUs and as
long as the values are close, substitutions are generally not a
problem. The hard part I've found is finding modern parts with the
same lead spacing. Sometimes you have to go up a case size and/or
voltage rating to find an electrolytic that will fit the existing
location without putting excessive bends in their leads. This is
especially true with old 3-lead "computer grade" capacitors (typically
two ground leads).
My error--it's a 950 uF cap, not 940 and it still is 40 VDC.
Fortunately, it's an axial-lead cap, so replacement with a physically
smaller size won't be an issue.
It seems to be a filter cap on the +22 supply line, so it's not
critical as regards timing. But the odd value had me wondering--I've
been placing 1000 uF filter caps virtually since there were
transistors. I'd never seen a 950 uF before.
--Chuck
4:55 p.m.
On Sat, 14 Jul 2012, Chuck Guzis wrote:
On 14 Jul 2012 at 0:34, Tothwolf wrote:
Who's the maker of the original cap? I'm curious now given what Will
mentioned if it might very well have been a custom made part for HP. I
think I've actually seen 950uF before but I can't remember where.
Capacitors with axial leads have been getting harder and harder to find,
and when you can actually find them, they tend to be so much more
expensive than an equivalent radial too. If all you need is an 85C rated
part, I think Nichicon still has them in their VX series. According to my
Nichicon book, the P/N for a VX series 1000uF 50V in a 16x31.5mm case
would be TVX1H102MCD. I don't know of anyone who currently makes an axial
leaded 105C rated part though, which is what I generally prefer for
radials in SMPSU rebuilds.
I replace odd value capacitors all the time in
older SMPSUs and as
long as the values are close, substitutions are generally not a
problem. The hard part I've found is finding modern parts with the
same lead spacing. Sometimes you have to go up a case size and/or
voltage rating to find an electrolytic that will fit the existing
location without putting excessive bends in their leads. This is
especially true with old 3-lead "computer grade" capacitors (typically
two ground leads).
My error--it's a 950 uF cap, not 940 and it still is 40 VDC.
Fortunately, it's an axial-lead cap, so replacement with a physically
smaller size won't be an issue.
It seems to be a filter cap on the +22 supply line, so it's not
critical as regards timing. But the odd value had me wondering--I've
been placing 1000 uF filter caps virtually since there were
transistors. I'd never seen a 950 uF before.
5:41 p.m.
On 14 Jul 2012 at 15:55, Tothwolf wrote:
Who's the maker of the original cap? I'm
curious now given what Will
mentioned if it might very well have been a custom made part for HP. I
think I've actually seen 950uF before but I can't remember where.
It's a Mallory; part number 0180 2357-1, date 9-73.
It's one of the few capacitors on the PSU board that actually has a
namufacturer's name on it--the others do not, although they appear to
be Sprague "Vitamin Q" models by their construction.
--Chuck
15 Jul
15 Jul
4:55 p.m.
On 14 Jul 2012 at 15:55, Tothwolf wrote:
0180-2357 ia the right format for an HP part number. This means it might
be a custom part, or just that it's been custom-labelled.
-tony
Who's the maker of the original cap? I'm
curious now given what Will
mentioned if it might very well have been a custom made part for HP. I
think I've actually seen 950uF before but I can't remember where.
It's a Mallory; part number 0180 2357-1, date 9-73. 
14 Jul
14 Jul
2:35 p.m.
Tothwolf wrote:
On Fri, 13 Jul 2012, Chuck Guzis wrote:
If you are replacing capacitors in switching power supplies you must use
replacement caps that are high frequency and (usually, but no always)
low ESR. Same goes with recapping monitors - the caps on the B+ lines
all care a lot about high frequency response (heat factor) and the caps
in the horizontal circuit are also usually fussy.
We use (after some research which variety is needed for the specific
application) the Panasonic caps from Digi-Key or Mouser rather than
buying them from the corner supplier as generic caps just don't cut it
in a great many applications.
John :-#)#
--
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
Call (604)872-5757 or Fax 872-2010 (Pinballs, Jukes, VideoGames)
www.flippers.com
"Old pinballers never die, they just flip out"
I'm getting around to doing some work on my
veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB
goodness inside, the first thing that jumps out at me is a bulging
electrolytic. It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not
50V, but 40V. I'm going to substitute a pair of 470uF, 50V units
paralleled as a substitute, but this had me wondering if anyone knows
why the strange values, particularly since +/-20 percent tolerances
are common on electrolytic caps.
The 940uF 40V capacitor just dates to before we had the modern
preferred values system. If it is just a filter capacitor, a 1000uF
50V should work just fine, and would likely be more reliable (and have
less leakage) than two smaller capacitors in parallel.
I replace odd value capacitors all the time in older SMPSUs and as
long as the values are close, substitutions are generally not a
problem. The hard part I've found is finding modern parts with the
same lead spacing. Sometimes you have to go up a case size and/or
voltage rating to find an electrolytic that will fit the existing
location without putting excessive bends in their leads. This is
especially true with old 3-lead "computer grade" capacitors (typically
two ground leads).
4:10 p.m.
The 940uF 40V capacitor just dates to before we had
the modern preferred
values system. If it is just a filter capacitor, a 1000uF 50V should work
I don't think that;s the whole story, for several reasons :
The old preferrd values, at least in the UK, were a cobling series (1, 2,
4, 8, 16, 32, 64 uF) and decades multiples therefor. (for example, 640uF
was a common Philips one). The HP values have no particualr pattern
There are hP boards with (offiical, as documented in the service manual)
odd-vlaue capactiors which are close in value, like 5600uF and 6000uF
I have a schematic here for the HP9826/9836, from the 1980s. The colour
monitor PSU (for the 9836C)( has a 900uF smoothing caapcitor on the input
to the 12 regulator (C22), not 1000uF. The main smoothing capacitor in
the computer is 21000uF, not 22000uF. This machine was designed and sold
long after the adoption of 'modern' preferred values.
-tony
4:02 p.m.
I'm getting around to doing some work on my veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB
goodness inside, the first thing that jumps out at me is a bulging
electrolytic. It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not
50V, but 40V. I'm going to substitute a pair of 470uF, 50V units
paralleled as a substitute, but this had me wondering if anyone knows
why the strange values, particularly since +/-20 percent tolerances
are common on electrolytic caps.
I'e wondersd about this before (I may even have posted about it). I
can;'t see any electrical rason why they would use such capacitors. I've
seen a PSU board where the smoothing capacitors for the inputs to various
regulators were a mix of (I think( 5600uF and 6000uF, again +/-20% (at
least). Why theey didn't specify and ft al lthe same type is beyond me.
I hae no idea what the 1000uF capacitor in your counter is for, but I'll
guess it's i nthe PSU. I would be _very_ suprised if a 1000uF didn't work
correctly there too.
-tony
5 p.m.
On Sat, 14 Jul 2012, Tony Duell wrote:
Maybe in large volume it resulted in a cost savings with a very tiny cost
savings per unit for them to use a capacitor with slightly less aluminum
in it?
I'm
getting around to doing some work on my veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB goodness
inside, the first thing that jumps out at me is a bulging electrolytic.
It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not 50V, but 40V. I'm
going to substitute a pair of 470uF, 50V units paralleled as a
substitute, but this had me wondering if anyone knows why the strange
values, particularly since +/-20 percent tolerances are common on
electrolytic caps.
I'e wondersd about this before (I may even have posted about it). I
can;'t see any electrical rason why they would use such capacitors. I've
seen a PSU board where the smoothing capacitors for the inputs to
various regulators were a mix of (I think( 5600uF and 6000uF, again
+/-20% (at least). Why theey didn't specify and ft al lthe same type is
beyond me. 
15 Jul
15 Jul
5:26 a.m.
Tony Duell wrote:
Maybe simple size considerations are the cause for this. Maybe the
Condenser manufacturer would'nt guarantee that 1000?F wil fit in a can of
the desired size for the pcb, so they moved the specs down to 950.
One of my customers here is an electrolytic condenser manufacturer, Frolyt.
Regards,
Holm
--
Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe,
Freiberger Stra?e 42, 09600 Obersch?na, USt-Id: DE253710583
www.tsht.de, info at tsht.de, Fax +49 3731 74200, Mobil: 0172 8790 741
I'm getting around to doing some work on my veneered and generated
HP5307A frequency counter. Looking at all the gold-plated PCB
goodness inside, the first thing that jumps out at me is a bulging
electrolytic. It's a 940 uF, 40V unit. Not 1000 uF, but 940. Not
50V, but 40V. I'm going to substitute a pair of 470uF, 50V units
paralleled as a substitute, but this had me wondering if anyone knows
why the strange values, particularly since +/-20 percent tolerances
are common on electrolytic caps.
I'e wondersd about this before (I may even have posted about it). I
can;'t see any electrical rason why they would use such capacitors. I've
seen a PSU board where the smoothing capacitors for the inputs to various
regulators were a mix of (I think( 5600uF and 6000uF, again +/-20% (at
least). Why theey didn't specify and ft al lthe same type is beyond me.
I hae no idea what the 1000uF capacitor in your counter is for, but I'll
guess it's i nthe PSU. I would be _very_ suprised if a 1000uF didn't work
correctly there too.
-tony
4424
days inactive
4425
days old
13 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
holm＠freibergnet.de -
pinball＠telus.net -
tothwolf＠concentric.net -
wdonzelli＠gmail.com