You should be able to buy 22,000uF, 100V electrolytic capacitors of good reputable quality for about £25 each, unless you want either KEMET or Nichicon brand.
Printable View
You should be able to buy 22,000uF, 100V electrolytic capacitors of good reputable quality for about £25 each, unless you want either KEMET or Nichicon brand.
I want the best, Barry. The TDKs are the best option.
TDK EPCOS B41456 B9229M600 22,000uF 100V electrolytic capacitors can be obtained in the UK from either Mouser Electronics (£39 each), or from Enrgtech (£28 each) They have dimensions of 64.3mm dia. x 106.7mm high.Quote:
Originally Posted by User211
The ones you have bought are TDK EPCOS B41560 A229M600 22,000uF 100V. These do cost £79 each from Mouser and have larger dimensions of 76.9mm dia. x 105.7mm high.
Have you chosen the latter because they are of similar dimensions to those that you want to replace?
Yes.
Mouser was more expensive than DigiKey.
They have UK websites, but both ship from the states and the price includes the import tax. Mouser also stated they have significant shipping delays at the moment.
Then there's a dodgy Chinese website selling them for $75. Dead risky i.e. probably not genuine.
Grok and Gemini both came up with this cap as the best choice, followed by Vishay as the 2nd best.
So which ones have you bought?
Post #92 the ones I said I would.
And post #99 even supplies the link.
£145 each! :eek: Do you think they will better the 41456 versions? I hope so.
The M60s run very hot Barry. Part of that might be due to the aging ELNAs but I have no direct proof. These have a higher heat rating than the series you suggest. No bad thing.
The main concern is the sonics after what amounts to quite a big change being undertaken here. But they are screw terminal so they can be resold if they suck.
Fair enough, 105°C vs. 85°C might make all the difference.
Also, while they are a lot cheaper Barry they simply don't fit.
LOL. Just to say so you know - when I buy stuff I spend literally ages doing the research first. It was a well considered buy.Quote:
Originally Posted by Barry
For Barry - why the ELNAs have to go as you said and I also think given the below. I measured all of them at 90-100mOhms.
Here's what Grok said:
**Yes — 100 mΩ is high enough to cause noticeable heating**, and it is a clear sign the original Elnas are well past their best.
### Original (new) ESR estimate
For a large screw-terminal 22000 µF / 100 V 85 °C capacitor of that era and physical size (≈75 mm diameter), a typical new ESR at 100 Hz would have been roughly **10–20 mΩ** (most likely in the 12–18 mΩ range when fresh).
Modern equivalents of similar size and rating sit around 6–15 mΩ. Your measured ≈100 mΩ is therefore **roughly 5–10× higher** than new — classic ageing (dried-out electrolyte + increased internal resistance).
### Will 100 mΩ cause heating?
Yes, potentially quite a bit, depending on the ripple current in your amplifier.
Power dissipated as heat inside the capacitor is:
\[
P = I_{\text{rms}}^2 \times \text{ESR}
\]
- With a healthy new ESR of ~15 mΩ, even several amps of ripple produces only modest heat.
- At 100 mΩ the same ripple current produces **6–7× more heat**.
In a high-power amplifier (especially Class AB with heavy bass content or high continuous power), the original caps can easily run warm or even hot to the touch. That extra heat further accelerates ageing, which is why they degrade faster once ESR starts rising.
### Practical takeaway
100 mΩ is high enough that replacement is fully justified on both performance and reliability grounds. The new TDK/EPCOS B41560A9229M (≈6 mΩ) will run dramatically cooler and deliver much tighter regulation under load.
If you want a quick check before swapping: power the amp, play music with strong bass for 10–15 minutes, then carefully feel the old cans (power off first). If they are noticeably warmer than the chassis, the elevated ESR is definitely contributing to heating.
That paste didn't work well for the formula, which is basically power = RMS current squared times ESR.
It isn't a great answer, though, if you think about it. It doesn't actually give any idea of heat output in the real world based on the cap or caps in use.
New caps arrived. TBH the ELNAs test around 20mF but the new TDKs a mere 18mF.
The freaky thing is the ESR on the ELNAs is lower than the TDKs.
We are right at the limits of my LCR meter, though.
Gonna pop then in but I am not sure it will fix the issue.
The TDKs can be returned.https://uploads.tapatalk-cdn.com/202...c01d8dfdf1.jpg
In with minimal coercion. No time to test today.
Polarity is definitely correct.
https://uploads.tapatalk-cdn.com/202...842f2d969e.jpg
Your new electrolytic capacitors have a tolerance of +/- 20%, as indicated by the 'M' in the type number and the '(M)' marking on the body of the electrolytic.
That means their capacitance can be anywhere between 17,600 - 26,400uF. What is the accuracy of your LCR meter?
It's rated to 20mF but no accuracy is given at that limit.
It is obviously a joke reading the ESR. New TDK spec is 6mOhm. I see values 20-30% higher than the ELNAs which give about 90-100mOhm at 100Hz.
It's an East Tester ET430.
I was always aware that the meter was probably off by a fair amount at these cap values. But TBH it is producing joke results given what I just said.
The point is not whether the TDKs are in spec or not WRT capitance. They are. The obvious point is the ELNAs are better/closer to 22mF at 50 years of age.
That said the date code on the TDKs implies they have been sitting around doing nowt for a few years. Once they see some action I hear through the grapevine they may measure better after a while.
The theory is the ELNAs are failing with voltage applied. This will prove whether that's true or not. I won't be surprised if it's not the case but I have checked every goddam thing I can. I do notice watching experts like Mend It Mark and others on YT use a bit of guesswork too.
I have seen quite a few electronics guys on YT saying that modern capacitors aren't always better than those made decades ago, which is somewhat surprising you would have thought.
ESRs are measured at typically 120Hz. I would be surprised the ESR is as low as 6mOhm for your new electrolytics.
I take it back - looking at TDK's catalogue and component specification for your TDK EPCOS B41560 A 9229M600, the ESR at 100Hz is indeed 4mOhm typical, and 6mOhm maximum.
Given that data, and the assumption that your (modified) Duetta speakers have an impedance that is a near constant 4 Ohm, I should be able to calculate how hot the capacitors will get.
That sounds like fun Barry go for it.
I quite fancy a thermal imaging camera.