That's generaly, regardless the cap, not something you can do without knowing what actual voltage the cap is exposed to.
The voltage rating of the cap is crucial: if you underate it the cap will blow... and perhaps more. The problem is you need to understand or measure what is going on, as it is true that some manufacturer simply buy in bulk and the rating is just that of the cap with little ressemblance to the actual situation. On the other hand, without knowledge, you can't find that out.
Caps in the signal path before power amplification are seldom a problem though.
The main problems are the PS caps, and there you absolutely need to make sure their rating is above any voltage they are likely to see under normal circumstances, including power on, shut down etc.
Here, I wouldn't advise lower rating of any of the high voltage PS caps. One could do it, but that would imply making dead sure you never enter the unit with more volts (voltage multipliers are unlikely here especialy for the big PS caps... and there aren't indeed any). That would probably mean stick to max regulated 22V feed and that will reduce significantly the performance of your amp unless your speakers are very low impedance and high efficiency (not seen often).
Regarding the other PS caps, for those who are exposed to less regulated volts inside the unit (OAP, digital etc.) it is of course not a problem but you need to know what they are exposed to. It is unlikely the manufacturer fitted there much higher ratings for nothing...
In a general way, I wouldn't advise lower ratings of PS caps unless you perfectly know what you are doing, whatever the unit. And here, considering the main PS caps, the rating margin is already quite thin if you use the potential of the unit...
The voltage rating of the cap is crucial: if you underate it the cap will blow... and perhaps more. The problem is you need to understand or measure what is going on, as it is true that some manufacturer simply buy in bulk and the rating is just that of the cap with little ressemblance to the actual situation. On the other hand, without knowledge, you can't find that out.
Caps in the signal path before power amplification are seldom a problem though.
The main problems are the PS caps, and there you absolutely need to make sure their rating is above any voltage they are likely to see under normal circumstances, including power on, shut down etc.
Here, I wouldn't advise lower rating of any of the high voltage PS caps. One could do it, but that would imply making dead sure you never enter the unit with more volts (voltage multipliers are unlikely here especialy for the big PS caps... and there aren't indeed any). That would probably mean stick to max regulated 22V feed and that will reduce significantly the performance of your amp unless your speakers are very low impedance and high efficiency (not seen often).
Regarding the other PS caps, for those who are exposed to less regulated volts inside the unit (OAP, digital etc.) it is of course not a problem but you need to know what they are exposed to. It is unlikely the manufacturer fitted there much higher ratings for nothing...
In a general way, I wouldn't advise lower ratings of PS caps unless you perfectly know what you are doing, whatever the unit. And here, considering the main PS caps, the rating margin is already quite thin if you use the potential of the unit...
Nine 10uF SMD caps TPA3255EVM BOM: https://www.ti.com/tool/TPA3255EVM
Output caps from opamp U5B, U6B: C17, C28, C55, C63
4 each- 10uF CAP, CERM, 10 μF, 16 V, +/- 10%, X7R, 1206 1206 GRM31CR71C106KAC7L MuRata
Input caps to opamp U5A, U6A: C20, C62, C66, C71.... Also C38 on 10uF is used on TLV1117 regulator
5 each- 10uF CAP, AL, 10 μF, 16 V, +/- 20%, 1.35 ohm, SMD SMT Radial B EEE-FK1C100R Panasonic
TPA3251EVM uses the same type of capacitors for the opamps as the TPA3255.
C17, C28, C55, C63
4each 10uF CAP, CERM, 10 μF, 16 V, +/- 10%, X7R, 1206 1206 GRM31CR71C106KAC 7L MuRata
C20, C38, C62, C66,C71
5each 10uF CAP, AL, 10 μF, 16 V, +/- 20%, 1.35 ohm, SMD SMT Radial B EEE-FK1C100R Panasonic
C17, C28, C55, C63
4each 10uF CAP, CERM, 10 μF, 16 V, +/- 10%, X7R, 1206 1206 GRM31CR71C106KAC 7L MuRata
C20, C38, C62, C66,C71
5each 10uF CAP, AL, 10 μF, 16 V, +/- 20%, 1.35 ohm, SMD SMT Radial B EEE-FK1C100R Panasonic
Excellent info. Thank you. I actually have some 10uf ceramics coming that I'll use for the opa's instead of the lytics I was going to use. Though I'm sure I'll end up trying both.
Hi daniboun, can you point to the JCCON /Sanyo part number?
From what i read in this and other threads, this is the list of suggested power caps that fit:
United Chemi-Con ELBK350ELL452AM25S 4500 uF 22 mOhms 3,36 €
Nichicon UBY1V302MHL 3000 uF 26 mOhms 2,28 €
Panasonic EEU-FS1V272 2700 uF 14 mOhms 1,49 €
JCCON /Sanyo ? 4700 uF
Sure here goes :
4 50 pieces/lot 16*25mm P55 faible ESR/impedance haute frequence 35v 4700UF condensateur electrolytique en aluminium 4700UF35V 20% | AliExpress
For the other, I disagree :
The ELNA is rated @ 35V /3000Uf and they really seem to be great. I have nothing to complain with )
What is the point of replacing them with others with lower values and which may be even worse?....
Sure here goes :
4 50 pieces/lot 16*25mm P55 faible ESR/impedance haute frequence 35v 4700UF condensateur electrolytique en aluminium 4700UF35V 20% | AliExpress
For the other, I disagree :
The ELNA is rated @ 35V /3000Uf and they really seem to be great. I have nothing to complain with )
What is the point of replacing them with others with lower values and which may be even worse?....
Yep, did it.
I posted all the details and reported on the benefits in this thread. Was amongst the last mods.
Post #151 is your friend
Not the best buck for the money IMHO, the existing coils are note the bottleneck for this amp.
Claude
Thanks as always. Based on your findings I will probably not bother. I dont even have access to the coils you used. I was looking at going with the Wurth recommended in post #1 but like I said, those arent through hole so Im not sure how I would connect them. Maybe Ill replace the coil capacitors with genuine KEMET's as there very cheap and presumable simple to replace? Or were they what were giving you the trouble to remove?
Would something like This work? Or This?
Last edited:
I dont' have the black board with ELNA 3000Uf but the green one with 2200uf ARC Audio
Definitely upgrade to bigger ones. Even the nice Nichicon 2200uf's I used didnt sound near as good as the 4500uf United's I have now. It is more about low ESR vs size but 2200uf seems like a bare minimum size for this chip, not the ideal one. Everything about the sound is better with the United's. The only other caps I would consider for this are the 3900uf Nichicon UBY's. Why 3900uf vs 3000? Lower ESR and higher ripple current.
Mello4, I have had to remove your image links because they were totally corrupting the page loading of this thread.
You can see just a small part of your links in the image attached, they were hundreds of thousands of characters in length.
Please attach images correctly using the forums own uploading utility
How to attach images to your posts.
Attachments
Quote: I was looking at going with the Wurth recommended in post #1 but like I said, those arent through hole so Im not sure how I would connect them. Maybe Ill replace the coil capacitors with genuine KEMET's as there very cheap and presumable simple to replace?
The Kemets (or whatever they are) were difficult to remove and little to no gain, I wouldn't bother. On one amp they came OKish, on the other were nightmare. I wouldn't bother again and I didn't in the 3rd amp. Would rather do the PS caps LOL
Other coils can be fitted provided you place them differently (verticaly perhaps) and solder pins to make them legs. You have to be inventive and Rob43 was on that. I didn't bother given the coil tuning turned out to be low ROI. As posted, we did it on 2 amps because we had already purchased the parts but on the 3rd I didn't. Given the cost of tuning 1 amp with coils, for the same price you are better off purchasing a second amp and go bi-amping!
Of course one could as we did go biamping andreplace 8 coils, but that's overkill IMHO.
Claude
The Kemets (or whatever they are) were difficult to remove and little to no gain, I wouldn't bother. On one amp they came OKish, on the other were nightmare. I wouldn't bother again and I didn't in the 3rd amp. Would rather do the PS caps LOL
Other coils can be fitted provided you place them differently (verticaly perhaps) and solder pins to make them legs. You have to be inventive and Rob43 was on that. I didn't bother given the coil tuning turned out to be low ROI. As posted, we did it on 2 amps because we had already purchased the parts but on the 3rd I didn't. Given the cost of tuning 1 amp with coils, for the same price you are better off purchasing a second amp and go bi-amping!
Of course one could as we did go biamping andreplace 8 coils, but that's overkill IMHO.
Claude
What about these? I’m sure two of the four leads probably are only there for anchoring (can be trimmed off)?
https://www.mouser.com/ProductDetail/Wurth-Elektronik/7444011715100?qs=GBLSl2AkiruGT3IIpBA3cw==
Cheers,
Pete
You mean the UBY1V392MHL6? ESR seems similar. Why not the Panasonic EEU-FS1V272? the lowest ESR of the bunch, but at 2700 uF.
United Chemi-Con ELBK350ELL452AM25S 4500 uF 22 mOhms 3,36 €
United Chemi-Con ELBK350ELL352AL25S 3500 uF 24 mOhms 2,80 €
Nichicon UBY1V302MHL 3000 uF 26 mOhms 2,28 €
Nichicon UBY1V392MHL6 3900 uF 25 mOhms 2,90 €
Panasonic EEU-FS1V272 2700 uF 14 mOhms 1,49 €