I'm after a supplier for 4 off the following.
Electrolytic caps.
~25,000uF 100v
Termination can be; Pin, Snap in, Tag. Not screw.
Pitch of 10mm to 28mm
Normal two pin. Not multi pin. (double sided board with tracks under the cap)
Maximum diameter of 53mm
Hight up to 140mm
I've tried; rs, e14, digikey, mouser, dalbani, hifi collective.
Any other suggestions or links appreciated.
Electrolytic caps.
~25,000uF 100v
Termination can be; Pin, Snap in, Tag. Not screw.
Pitch of 10mm to 28mm
Normal two pin. Not multi pin. (double sided board with tracks under the cap)
Maximum diameter of 53mm
Hight up to 140mm
I've tried; rs, e14, digikey, mouser, dalbani, hifi collective.
Any other suggestions or links appreciated.
Last edited:
Yes.
The original is 25,000uf 80v.
But they run right on 80v.
If (when) the mains goes up obviously they'll be over stressed.
So I'd really prefer to replace them with a higher voltage.
But it looks like I'm trying to find unicorn tears.
I don't want to go to screw terminals as that means a whole lot of lateral thinking to make them fit.
But I may be forced down that road yet.
The original is 25,000uf 80v.
But they run right on 80v.
If (when) the mains goes up obviously they'll be over stressed.
So I'd really prefer to replace them with a higher voltage.
But it looks like I'm trying to find unicorn tears.
I don't want to go to screw terminals as that means a whole lot of lateral thinking to make them fit.
But I may be forced down that road yet.
Still, whatever you do, 22000 will be easier to find. OEMs can have any value they want made up for their production, but we rely on off the shelf products.
Caps have "working voltage" ratings, so an 80v cap is OK at 80v. They also all have a surge rating, which is higher. Often that is to cover momentary higher voltage as circuits warm up, but running slightly over would fall under that.
Caps have "working voltage" ratings, so an 80v cap is OK at 80v. They also all have a surge rating, which is higher. Often that is to cover momentary higher voltage as circuits warm up, but running slightly over would fall under that.
I don't agree.
I would be happy running a smoothing capacitor at 99.9% of it's rated maximum voltage.
I would not operate it at 101% of maximum rated voltage.
I consider the openers concern at using 80V caps at >80Vdc when the mains voltage is high to be worthy. The overvoltage could last many hours and in exceptional circumstances maybe even exceed days.
Yes I may have to resort to a lower capacity.
I'd rather not if I can help it though.
It's a nice amp designed by a great designer so I'd like to it justice.
But the reality of physical size, mounting restrictions, and what can be bought for realistic sums will all determine what actually gets installed.
The caps in question are ps filter caps in an amplifier.
The owner (a friend) asked me to investigate several faults.
When I opened it up I saw the caps bulging. (Not the main issue, but still need to be fixed in my opinion)
He and I are a bit ocd so I need to replace them for my own sanity 😉
Note that I have Not measured the secondary or filtered voltages.
These are from the schematic, and I have no reason to doubt them.
Heres an excerpt from an email I sent him about the caps.
......................................
amp is supposed to be 230vac.
transformer out to board is listed as 55vac.
then shows the main supply rail listed as 80vdc.
so with some quick maths and generalizations applied.
230vac input -> 55vac x 1.4 (rectification and filtering factor) = 77vdc
Thats to close to the 80v cap rating for my liking.
If we throw in the fact of it being on 240vac into the mix.
230vac + 4.5% = 240.35vac
55vac + 4.5% = 57.5vac x 1.4 = 80.4vdc
Thats right on the edge of what the cap is rated for.
Now, caps can, and will, survive with higher voltages.
But, they will run hotter and their life will be reduced.
If for fun we now look at what I would allow for as 'normal'.
230vac + 9% = 250vac
55vac + 9% = 60vac x 1.4 = 84vdc
Ok so thats survivable. Albeit with a shorter life span.
I have an article on an amp i built many years ago.
The designer did his research and came up with this..
His office in waroonga sydney ranged from 245-255vac.
and various councils he contacted said he should allow for 265vac.
230vac + 15% = 264.5vac (pretty common for w.a.)
55vac + 15% = 63.25vac x 1.4 = 88.55vdc
.....................................
As a side note.
Where he and I live the mains is normally around 245vac
This kit amp I refer to (AEM6000), I built it in 1988 using caps with a higher voltage than specified.
I have never needed to repair it.
And it still has the original caps in it.
Obvious solution is to use higher voltage caps.
which means lower capacitance units.
which means more of them to get the capacitance needed.
which means more space to fit them.
which means more cost at production.
The amp has outlived it's warranty period.
Thats all that matters.
Right ???? 😉
I'd rather not if I can help it though.
It's a nice amp designed by a great designer so I'd like to it justice.
But the reality of physical size, mounting restrictions, and what can be bought for realistic sums will all determine what actually gets installed.
The caps in question are ps filter caps in an amplifier.
The owner (a friend) asked me to investigate several faults.
When I opened it up I saw the caps bulging. (Not the main issue, but still need to be fixed in my opinion)
He and I are a bit ocd so I need to replace them for my own sanity 😉
Note that I have Not measured the secondary or filtered voltages.
These are from the schematic, and I have no reason to doubt them.
Heres an excerpt from an email I sent him about the caps.
......................................
amp is supposed to be 230vac.
transformer out to board is listed as 55vac.
then shows the main supply rail listed as 80vdc.
so with some quick maths and generalizations applied.
230vac input -> 55vac x 1.4 (rectification and filtering factor) = 77vdc
Thats to close to the 80v cap rating for my liking.
If we throw in the fact of it being on 240vac into the mix.
230vac + 4.5% = 240.35vac
55vac + 4.5% = 57.5vac x 1.4 = 80.4vdc
Thats right on the edge of what the cap is rated for.
Now, caps can, and will, survive with higher voltages.
But, they will run hotter and their life will be reduced.
If for fun we now look at what I would allow for as 'normal'.
230vac + 9% = 250vac
55vac + 9% = 60vac x 1.4 = 84vdc
Ok so thats survivable. Albeit with a shorter life span.
I have an article on an amp i built many years ago.
The designer did his research and came up with this..
His office in waroonga sydney ranged from 245-255vac.
and various councils he contacted said he should allow for 265vac.
230vac + 15% = 264.5vac (pretty common for w.a.)
55vac + 15% = 63.25vac x 1.4 = 88.55vdc
.....................................
As a side note.
Where he and I live the mains is normally around 245vac
This kit amp I refer to (AEM6000), I built it in 1988 using caps with a higher voltage than specified.
I have never needed to repair it.
And it still has the original caps in it.
Obvious solution is to use higher voltage caps.
which means lower capacitance units.
which means more of them to get the capacitance needed.
which means more space to fit them.
which means more cost at production.
The amp has outlived it's warranty period.
Thats all that matters.
Right ???? 😉
Last edited:
Realistically, 22000 vs 25000 with a 20% tolerance on these caps mean they're essentially fungible. And very very doubtful to materially change anything about the PSU performance. Possibly beneficial if you have a modern electrolytic capacitor with a bit lower ESL ESR. And who knows how much the value of your old cap has changed over its lifetime.
ooh yes ta for reminding me about that.
I did have super quick search and couldn't find anything about the original caps.
I'll have another better search when I get home tonight.
Old caps are Elna CE-W 85 Degree
Sadly none of my meters are able to measure the old ones so I can't say just how close or far out they now are.
I did have super quick search and couldn't find anything about the original caps.
I'll have another better search when I get home tonight.
Old caps are Elna CE-W 85 Degree
Sadly none of my meters are able to measure the old ones so I can't say just how close or far out they now are.
While 25m (25,000uF) isn't a standard value, 33m (33,000uF) should be. Not sure how old your chassis is, but chances are that if it's 20yrs or more, a modern 33m would easily fit where an old 25m did. And 33m shouldn't be all that much more than 22m.
I'm currently fluffing-up a 35yrs old Hafler DH-200, which uses 2" (51mm) computer-grade cans rated 10m @ 75V. After some searching, it seems the sweet spot for price / quality is Nichicon 10m @ 100V, in a much smaller 1-3/8" (35mm) snap-cap package. Sure, I'll have to adapt the to the mounting clamps with shims, but hey - why not? Sure, I could use 1-3/8" clamps, but that would require drilling holes.. and that's always to be avoided.
The pair of 10m cost $17US shipped, and the 22m @ 100V were only about 50% more in price..
I'm currently fluffing-up a 35yrs old Hafler DH-200, which uses 2" (51mm) computer-grade cans rated 10m @ 75V. After some searching, it seems the sweet spot for price / quality is Nichicon 10m @ 100V, in a much smaller 1-3/8" (35mm) snap-cap package. Sure, I'll have to adapt the to the mounting clamps with shims, but hey - why not? Sure, I could use 1-3/8" clamps, but that would require drilling holes.. and that's always to be avoided.
The pair of 10m cost $17US shipped, and the 22m @ 100V were only about 50% more in price..
Last edited:
Oh, odd. Now I see what you meant by 'tag leads'. Those caps are meant to be clamp-mounted and wired in, rather than mounted directly to the PCB. It's the sort of lead you see on 1960s style caps and some conventional photo flash types.
Depending on where the traces are, which side of the board, etc., it might be an issue dropping-in regular snap-caps, if you have to drill to make the leads pass though.
You'll figure it out, cheers.
Depending on where the traces are, which side of the board, etc., it might be an issue dropping-in regular snap-caps, if you have to drill to make the leads pass though.
You'll figure it out, cheers.
I have one similar, had conventional mounting for the big caps.
That one came with rubycon standard types, in fact nearly all the capacitors were replaced except the silver mica ones. After a year of listening to different parts, the stuff that was original wasn't at the top of the heap.
That may require some adaptation surgery to tackle that one, you'd think that there would at least be the option of using a standard configuration when they did weird stuff like that.
That one came with rubycon standard types, in fact nearly all the capacitors were replaced except the silver mica ones. After a year of listening to different parts, the stuff that was original wasn't at the top of the heap.
That may require some adaptation surgery to tackle that one, you'd think that there would at least be the option of using a standard configuration when they did weird stuff like that.
Pic is worth a thousand words as they say 🙂
Yes the originals are the old school solder tags that I remember from the 70's
Yet, these have the vent on the other side of the cap facing away from the terminals.
The board is double sided plate through.
The cap 'slots' on the board will take 11mm - 29mm pin spacing.
So with a tiny nick in one end of the board I can fit in the modern 10mm snap in.
Previously I've straightened the pins on snap in's to fit older gear.
Replacing caps only ever becomes an issue for these larger values.
Why ? Why I say !
So now I shall hunt for the obsolete elna docs.
Yes the originals are the old school solder tags that I remember from the 70's
Yet, these have the vent on the other side of the cap facing away from the terminals.
The board is double sided plate through.
The cap 'slots' on the board will take 11mm - 29mm pin spacing.
So with a tiny nick in one end of the board I can fit in the modern 10mm snap in.
Previously I've straightened the pins on snap in's to fit older gear.
Replacing caps only ever becomes an issue for these larger values.
Why ? Why I say !
So now I shall hunt for the obsolete elna docs.
Right so the best looking option I can find is this one.
https://www.digikey.com/product-det...cs-cde/382LX273M100B102VS/338-1991-ND/2256599
Cornell dubilier. 27,000uf 100v 85degree.
Sadly it has extra pins. But I can sort that out.
Does anyone have anything good or bad to say about CD as a brand ?
I know the name, but have never used them.
https://www.digikey.com/product-det...cs-cde/382LX273M100B102VS/338-1991-ND/2256599
Cornell dubilier. 27,000uf 100v 85degree.
Sadly it has extra pins. But I can sort that out.
Does anyone have anything good or bad to say about CD as a brand ?
I know the name, but have never used them.
Cornell was traditionally a hi-quality manufacturer. Most of my experience with their products is limited to 2"-3" computer-grade cans in the 10m-30m range, 160-200V. These parts were always excellent, took a beating and lasted many years in large OTL amps.
Funny you mentioned CD, I was just looking for some 100u @ 100V today, and these were the only ones on eBay that I'd trust not being fakes.. 4 pcs - 100uf 100v electrolytic capacitors 105c FREE SHIPPING
Funny you mentioned CD, I was just looking for some 100u @ 100V today, and these were the only ones on eBay that I'd trust not being fakes.. 4 pcs - 100uf 100v electrolytic capacitors 105c FREE SHIPPING
I would certainly think so.
Let's put it this way - presuming they are genuine Cornell Dubilier, quality is well assured. Far as I see, the CD name doesn't have any real cachet with the hi-fi audio crowd, so one may infer it's not an obvious target for counterfeiting in the same way that the well-known types are (Black Gate, Cerafine, MUSE, etc.).
Update for anyone thats interested 🙂
I used cornel dubiler 27,000uf 100v 85 degree from digikey.
Seem to work. Glad I went for 100v not the original 80v as the supply rails are right on 80v.
There a 5 pin unit, so I cut of the three unwanted pins.
Made an insulating shim from 1mm blank fiberglass to go between the cap and the pcb,
Used a smear of neutral clue silicon between cap and shim, and shim to pcb.
Did this to provide arc over insulation and provide stability.
At 80vdc im not expecting any sort of arc over risk.
But I'm a bit ocd about things like this 🙂
I used cornel dubiler 27,000uf 100v 85 degree from digikey.
Seem to work. Glad I went for 100v not the original 80v as the supply rails are right on 80v.
There a 5 pin unit, so I cut of the three unwanted pins.
Made an insulating shim from 1mm blank fiberglass to go between the cap and the pcb,
Used a smear of neutral clue silicon between cap and shim, and shim to pcb.
Did this to provide arc over insulation and provide stability.
At 80vdc im not expecting any sort of arc over risk.
But I'm a bit ocd about things like this 🙂
- Status
- Not open for further replies.