• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

NAD 326BEE amp problem

asolochek

New Member
Hello,

I have a 326BEE with a problem that's proving pretty difficult to pin down. The first thing I noticed a little pop from the unit and the left channel went out. The only damaged component I could find was R250, which looked like it had blown, but was maybe just scortched and was still testing as 10ohms. I replace it anyway and stood it off the PCB a bit for better cooling. There were also a number of electrolytics that looked cooked and I replaced all those (many tested bad). The film capacitor in series with R250 (C172) I also checked and it's ESR of 24ohms seemed high, so I replaced it with a wima MKP10 I had. Everything I replaced on the left channel I also replaced on the right to maintain symmetry.

After doing all this the left channel was still low. Feeding in a 50mV RMS 1kHz sine to both left and right main inputs, I was getting 1.3V on the right channel and 200mV on the left into 8ohm loads. In attempting to narrow down which part of the amp contained the fault, I swapped the vertical mounted boards between channels, and this is where things get confusing. I'll refer to the seperate PCBs as Board L and Board R, indicating their original positions in Channel L and Channel R.

With Board L in Channel L, as I mentioned previously the outputs are Channel L: 200mV, Channel R: 1.3V. The amp is drawing almost 120W in this configuration, and I'm unable to get the channel L bias voltage below 80mV. Channel R bias is fine. With the input signal removed the amp draws about 100W.

With Board L in Channel R, I get 1.3V on channel R, and 200mV on channel L. The amp draws about 42W with the signal, and 35W without. Bias voltages on both channels can be brought into spec. Note that the 35W idle power is where the amp is supposed to be, according to the manual.

So the low output problem follows Board L, but there is also something funny about Channel L which Board L prevents the bias from being set and causes a bunch of current to get dumped through Q151 and Q145.

I've gone over Board L multiple times at this point. I've pulled every transistor and tested them, checked every diode and resistor. I think I'm going to start swapping parts between Board L and Board R to see if I can narrow it to a component, but even if I find that, it's only half the problem, because what is going on with the rest of channel L that causes all that excess current dump. I have temporarily fitted sockets for Boards L and R so I can easily remove and swap them. I'm thinking I might want to also make a little extender cable so I can measure currents out on each pin.

Does anyone have any ideas here? Amplifier schematic attached
 

Attachments

Register to hide this ad
No great ideas I'm afraid, perhaps look for hot components that are consuming 100w. I did a recap on a 326bee and in the process managed to bridge two pads, they are so close, suddenly one of the regs was going nuke. Caught it in time and no issues, most of the caps are not great and if it's a early one they suffered from the capacitor plague, there's also an issue with the 5v supply for the controls which comes from 25v or something dropped by two diodes and a regulator 7805, they get roasting hot and the later units have them mounted higher to dissipate heat.
Incredible amp let down by caps.
 
Yeah, I saw those two diodes for the 5V getting crazy hot so I stuck a heatsink on them. Part of the problem debugging this is that so much of it runs hot by design it's hard to spot when something is misbehaving.
 
I don't think those vertical boards are part of the preamp on this unit. They are pretty squarely in the amp circuit.

I ended up swapping over every transistor from one board to the other to see if I could identify the problem, but they were all good. Then I started rechecking the passives and found that R260, a 1ohm resistor that's part of the feedback network, was measuring in the 10s of kohms. I replaced that and now both channels are reading the same voltage, I can adjust the bias current properly, and the THD is down in spec (-99dB on the right channel, -96dB on the left, but I'm not going to chase that difference down).
 
Great work, a bit out of the way that resistor but fairly central in what it does. Might be worth changing r460 too.
I'm sure that'll help someone else at some point.
 
Hi. where is r260, is it on the underside of the board?, have just purchased one of these with a known fault of intermittant cut out , hasn't done that yet but it does seem to be getting quite hot on the main and smaller heatsinks so probably overheating causing it to shutdown .Also when I checked the idle current it's fine on the right channel but it starts around 3mv on the left but then as it warms up the mv drop and when adjusting it suddenly jumps up to around 20mv so somethings obviously amiss .
 
Last edited:
Hi. where is r260, is it on the underside of the board?, have just purchased one of these with a known fault of intermittant cut out , hasn't done that yet but it does seem to be getting quite hot on the main and smaller heatsinks so probably overheating causing it to shutdown .Also when I checked the idle current it's fine on the right channel but it starts around 3mv on the left but then as it warms up the mv drop and when adjusting it suddenly jumps up to around 20mv so somethings obviously amiss .
R260/460 should be near the final output stage.
Screenshot_20251229-111207~2.png
 
I sold it on in the end , far too complicated and difficult to repair with all those tiny surface mounted components,
 
One of the features was short signal paths, I figure they dialled the pcb track spacing and component height right down. A lot of gear is basically disposable. I'm looking for a cheap c352, its older and might be better built.
 
Back
Top Bottom