power supply cap replacement: NAD C370

[nick78]

I have a NAD C370, which has a complement of 4x 10,000uF 80V power supply caps. They are known to fail in this application. NAD literature alludes to the failure being a thermal situation, though i have read a comment or two from random sources that indicate high mains voltage in some locals can nudge the voltage these caps see high enough to cause failures. Either way, they are a known failure spot and they are coming out.

The question becomes, what to replace them with?

They are 4-terminal caps, of maximum diameter 35mm. I can go a bit taller; so there is some leeway to get slightly larger capacitance or go with a higher voltage rating (if indeed they are close to being under-specced in this design)..

Of the easily available options, there is:

Panasonic T-UP 10,000uF 80v (41mOhm ESR) rated to 85*C
Panasonic T-UP 12,000uF 80v (35mOhm ESR) rated to 85*C
Panasonic T-UP 10,000uF 100v (33mOhm ESR) rated to 85*C
Panasonic T-HA 10,000uF 80v (41mOhm ESR) rated to 105*C

The T-HA unit is attractive because of the higher temp ratings and it also has higher ripple current multipliers so I assume it will last longer; it's also an "industrial grade" series. One possible concern: the digikey catalog page says the inactive terminals MUST be isolated from the circuit design; but in my amp the inactive terminals are all soldered to ground traces?

Going up the food chain, there are the Jensen caps, and they have a 10,000uF 80v unit; but it's 5x the price ($60 each); is it even remotely worth it?

I'm aware of FTcap/F&T but was unable to find an online source selling a 10,000uF 80v unit.

Finally, any recommendations/commentary on bypass caps on these?

Any comments/input/suggestions welcome.

Thanks!

[nick78]

rails are +/- 65v.

[nick78]

after more research it would seem the jensens are out because they are true 4-pole caps; whereas my PCB design is simply for 4 pins - 3 of which are common; it seems that putting the jensens on my pcb would short the PS due to their 4 pole design.

[jaycee]

If you have 65V rails then 80V should be adequate, unless its something like 110v input and you normally get 120V going up to 130V at times.

I'd go with the 10,000uF 100V Panasonics myself.

[reddish75]

i think i maybe wrong but im pretty sure that the failures were attributed to the units that had the white power supply caps in them? if yours are purple and the amps working fine i would just leave them.

another under documented fault with these amps is the capacitors on the protection pcb (the one stood up on the board) dry out from the heat off the zener diode, causing a host of intermittent problems. the solution, take the zener diode off the front of the pcb and put it on the back taking care to put it on the right way, and replace the caps with high temp ones.

other problems are very poor heat flow in the unit, as you've touched on. take a look at all of the caps in the amp and look for heat damage, i've seen a few where the caps are so black you cannot make out the values.

chris

[nick78]

Actually the protection PCB failure is the reason I tore into the amp in the first place, and through that research is when I discovered the "white cap" issue.

I've already re-capped the protection PCB with 105*C caps and moved the zener to the back side of the board, (as well as re-routed a wire bundle that was too close for my comfort to the zener and it's heat output); and the amp is now functioning again; but I might as well take care of the white cap issue at the same time.

I'll look for any other temp-damaged caps. I've always kept the unit on a flat surface with nothing on top of it; so it's had the best airflow possible.

jaycee, with regard to the recommendation for the panasonic 100v caps, is that because they have the lowest ESR of the bunch?

[nick78]

just thought I'd see if there was any other recommendations/suggestions out there?

[reddish75]

panasonic t-ha is the way to go, 100v and 105*C easily the best for this unit from your list as long as they are not too big.