Weird NAD 208 THX protection loop

I'm sorry for delay in responding.

If I interpret correctly, you find 0 voltage drop across R393 and no drop in the corresponding resistor in the other (right) channel. R358 in the left channel is very hot, but you don't find the corresponding resistor in the right channel is hot? You find the left output is 0V; what is the output on the right channel? With no drop across R393 or its brother in the other channel, I would not expect the heatsinks to be hot. Are you able to find sources of heat in addition to R393 and R502?

Would you provide a bit of history on the amp? Eg. acquired not working vs. failed while you were using, etc? Has it been worked on before?

I would like to estimate some of the voltages I would expect to see, assuming the amp output is 0V. (The voltages in post 13 and the 0V on Feedback 1 & 2 are very encouraging.) I assume there is no load connected to either channel.

The NAD Circuit Description indicates the VL supply is typically +87V in 8 Ohm mode, so I'll assume 87V. Given 0V at amp output, I expect D326 to conduct and D325 to forward bias, yielding about 85.8V at the cathode of D325.

Given 0V at amp output, and that the power FETs are biased near threshold, the Q327 gate voltage would be roughly about +4V. I don't have an accurate handle on Q308 collector voltage, but I'm guessing it's +10V or less. This would yield +11.2V on the emitter of Q320, resulting in about +26.2V at the anode of D319. This guesswork would yield 59.6V across R358. You reported 74.4V in post 12. Perhaps the estimate vs. actual can be reconciled if the actual supply voltage is higher than 87V and my estimate of 10V on Q308 collector is too high.

Details of the bias spreader are also of interest. I estimate the collector to emitter across Q309 is about 1.92V. The voltage from collector of Q310 to anode of D308 is very uncertain because of the large adjustment range of VR301; also, I could not find any data for the D308 1V26 diode.

To help resolve some of these uncertainties, would you measure several voltages, all with respect to ground.

Power supplies:
+V VH
+VH
+VL
-VL
-VH

Amp test points:
Cathode D325
Cathode D319
Anode D319
Collector Q309
Anode D308
Feedback 1
Feedback 2
R326
Junctions of:
R313,314
R314,315
R315,316

Some of these Amp points you've already posted.

Are both channels misbehaving? If you find that corresponding points in the other channel are dramatically different, would you tabulate both channel readings side-by-side for easy comparison?

Any idea about why R502 is too hot?

Thank you.
 
Hello BSST! Nice that you are back.🙂 ok i'm ready for new challenges.
Both channels behave the same, left and right. R393 has 0VDC on the right and left channels.
R358 is hot on the right and left channels.
R393 does not get hot on the right or left channel.
R502 may get hot due to imbalance on +VH and -VH but I'm not sure.
I have had the amplifier since it was new and there has never been anything wrong with it. The error occurred when I started the amp and it went into protection mode. That's when I did some troubleshooting and discovered that the BR501 was broken.
There is nothing connected to the inputs or outputs, only the mains voltage 230VAC
When I've used the amplifier, I've had it in 4 ohm mode, so I'll include those measurements as well. VL in 4ohm is 75,9VDC. 8ohm 87VDC

I will return with more measured values tomorrow. Thanks for tonight,here from Sweden.😊
 
If the protection circuit is still triggering, it may act as a signboard pointing to the defective area. The test points listed in post 21 may still prove necessary, but clues from the protection may be quicker/less tedious.

I've attached a data sheet for the TA7317 IC that handles much of the protection. I would love to find a better data sheet, but maybe it's sufficient.

Referring to the data sheet, it appears that fault protection is activated when current on pin 1 is sufficient to drive Q6 (one Vbe), or alternately when there's current to drive 2*Vbe differentially between pins 2 and 3. Either of these events pulls the anode of D7 to near ground and disables both speaker relays. So the test is to follow any activity on pins 1 through 3 to find its origin. With luck, the fault in the amp will be evident. It's conceivable the protection circuitry itself has failed and the amp is OK, but far less likely.

Let us know what you encounter.
 

Attachments

protection mode is not in after I change BR501 so it works. Relay 301 pulls.
I noticed after a while the voltage on VL varies. When the amplifier is cold, it is 87VDC and after about 30 seconds it increases to 88VDC

When I measure with a thermal camera, I see that R358 gets hot first then R357 gets hot. When R358 starts to approach 100 degrees, R357 is 80 degrees celsius then I turn off the amplifier so it doesn't break. It takes about 30 seconds for the resistors to heat up.
 
On the amplifier, there is a switch called SW501 that can be changed from 4ohm to 8ohm, so when I have previously made measurements it has been at 4ohm but now I have changed it to 8ohm.

+VL 87 to 88VDC L-channel not stable. R-channel 87 to 88VDC not stable.
+V VH 123VDC L-channel not stable. R-channel 123VDC not stable.
+VH 119 to 120VDC L-channel not stable. R-channel 119 to 120VDC not stable.
-VL -85 to -86VDC L-channel not stable. R-channel -85 to -86VDC not stable.
-VH-119 to -120VDC L-channel not stable. R-channel -119 to -120VDC not stable.
 
  • Like
Reactions: BSST
L-Channel R-Channel
Cathode D325 84,2 to 84,6 VDC 84,2 to 84,6 VDC
Cathode D319 25VDC and increases over time 25VDC and increases over time
Anode D319 10,06 VDC 10,11 VDC
Collector Q309 8,9VDC and increases overtime 8,9VDC and increases over time
Anode D308 3,29 VDC and increases overtime 3,3VDC and increases over time
 
The left vs right voltages are remarkably similar, yet you observe R358 (right channel?) is hotter and its temperature rises more quickly than the corresponding R357 in the left(?) channel. The DC voltages across R358 and R357 seem to be identical, per your data, i.e. about 59.2 to 59.6V:

L-Channel R-Channel
Cathode D325 84,2 to 84,6 VDC 84,2 to 84,6 VDC
Cathode D319 25VDC and increases over time 25VDC and increases over time

You might measure the voltages directly across the two resistors with your DVM as a sanity check, but I tend to trust they are correct. I'm led to suspect oscillation is responsible for the larger dissipation in R358. Do you have access to a 'scope to look for oscillation? Be certain to use x10 probes to avoid capacitive loading of the amp internal nodes--- it can sometimes provoke oscillation.

Does your thermal camera suggest any culprit for the hotter R358?
 
Both channels look good with respect to output offset.

Both R358 and R357 appear to have ~59VDC across them, so they should get equally hot and equally quickly, assuming similar environments for dissipation--- that is to say similar air flow and similar paths for heat conduction. 59VDC across 1.5k implies about 2.3W dissipation, but actual dissipation could be larger if there's a large oscillation present and it wouldn't be evident on a DC voltmeter. Oscillation could also explain different behavior in the two channels. This suspicion prompted my to ask if you have an oscilloscope. My suspicion might be wrong, but a scope is the easiest way to check. (IMHO, R358 and R357 are inadequately sized, but more on this later.)

The DC checks on the amp are perfunctory at best, but so far I don't see anything abnormal in the readings you've reported. I'm not discounting your familiarity with your amp, but asking you to confirm that the temperature phenomena are definitely abnormal and not just to-be-expected paranoia following repairs.

In the bias spreaders, you might check that collector to emitter voltage at Q309 is about 1.92V. Another check is confirm collector to emitter voltage at Q310 rises as you advance the bias pot. Be cautious and increase pot only enough to confirm the circuit responds, then return to previous minimum position.

Any thoughts about present state?
 
I think R358 gets too hot because the soldering point melts and cracks, so you have to solder it again with a soldering iron. The circuit board is starting to turn black also at those components
The same with R502.
I have an oscilloscope but don't know if I have x10 probes. Otherwise I can buy those.
Will return with more measurement data.
Thanks for your patience. 🙂
 
  • Like
Reactions: BSST
R-Channel L-Channel

Junctions of:

R313,314 0,818VDC and increases over time. 0,816VDC

R314,315 0,830VDC and increases over time. 0,516VDC

R315,316 -0,833 VDC and increases over time. -0,823VDC

Could it be a problem at measuring point R314, R315?
 
No oscillation is a happy outcome.

I remain convinced R358 is undersized. It dissipates ~2.3W merely from the DC bias; in addition, it must dissipate any audio voltages present during operation. If it were under test in a magazine review, it would see AC sufficient to test at rated 8 Ohm full power. Power into R358 would exceed 5W.

Fine to continue testing, but I will suggest a revision to R358 shortly.
 
I'm pasting a Digikey search for some metal-oxide resistors to use in place of R358, R357.

I suggest a 5W, 1.5k resistor. Even 5W is not very conservative re its rating, but it's easier installation. Place the resistor at least a body diameter above the surface of the board to allow better airflow and to lessen thermal abuse of the board substrate. I believe the board is single sided, but if is double-sided, solder the resistor lead on both sides to improve thermal transfer of heat.

The link also includes part numbers for 750 and 3K resistors so that you could consider series or parallel implementations of 1.5k as alternate ways to distribute the heat.

I'm not certain the search link will post correctly. Let me know...

https://www.digikey.com/en/products/filter/through-hole-resistors/53?s=N4IgjCBcoExaBjKAzAhgGwM4FMA0IB7KAbRBgAYBmcsSkfMAdjBlvvEcrAA4AWdpjABsAThEDGQoTH4NGjXt24TuVcgNWsI+GDCW8h7CtICsdHeRMjyyi0N4UjNybZC8TL2WV4jpXmFzC-pRc3IYAuvgADgAuUCAAyjEATgCWAHYA5iAAvvh00CBIkGhYeIQkZCZUApS86pEgsfFJaVm5cvyFxaU4+ESQpJQKItQgjc2QiSkZ2XngItzi3SgYfRWDIA3zFNwm8EWrZf2VjNUABADyABYAtpgCAHQm5wDWN-fslG8fDxNxUwAqulUjFLsgALLYVCYACuyWwHRAAFo4CtIClYeUBqR9vhxOEckSgA