First amp repair - using a dropped power transistor

[jacau]

I'm new to the forum and attempting my first amp repair on a NAD T765 with a blown front left channel.

I've carelessly dropped my 2 x Sanken 2SD2390's TO3P power transistors (in their sealed mylar bag) from a height of 70cm (table) onto bricks. Using a DMM both test fine (for what that is worth).

Now considering the result of a power transistor shorting and blowing the channel again and, all the effort of disassembling and compounding/mica 6  transistors again, I know I should just order more if I have any doubt but from a cost and time perspective, would be nice to avoid that.

The Sanken power transistor catalogue mentions a reliability check of 10 drop tests from 75cm but doesn't say to what surface.

Could I please get some thoughts on whether I might have damaged these transistors or are they fairly robust?

The annoying thing is the complimentary transistors (1560) came in a padded box so sod's law I had to drop the ones in the bag! 

Thanks

[bob91343]

I don't think the risk is too high, so go ahead and use the parts.  They are pretty rugged.

[aqibi2000]

They will be fine just fix you circuit

Use a series lamp to current limit initially

[shakalnokturn]

Very unlikely you'll have damaged them by dropping them, you can chip chunks of plastic from the packages and still have working transistors.

Study the schematic if available and check if other transistors linked to output stage aren't shorted and would cause a new failure.

[jacau]

Thanks very much guys; I feel much better.

I was lucky enough to find the service manual online and also that each discrete channel is identical so was good to to to be able to use the two other channels on the PCB to compare readings.

I'm also green enough to only just realize (after wasting time and replacing some components) that the voltage drop in transistor mode on a DMM also means that the missing power transistors would affect resistance readings in other areas of the circuit 

So two Sanken power transistors, 2 x 0.22 ohm 5 watt resistors and a couple of SMD components and I think I'm right.

I have a 100 watt light bulb which I'll use in a current limiting circuit.

If I could ask one more question, each channel (3 on this PCB) has 4 transistors that also are mounted on the heatsink side but they all use an exceedingly large amount of compound. Would this indicate that the top of each transistor possibly makes poor contact with the heatsink plate or is this a case of mass production excess for cost reasons that I've read about?

The attached image shows their outlines and both PCB and heatsink is covered in the goop.

Thanks again!



[ Specified attachment is not available ]

[shakalnokturn]

Transistors having a dedicated thermal exchange surface don't require much compound, just a thin film spread over the surface and sufficient tightening.
For packages with no thermal tab like TO92 it's worth over-doing it.

[jacau]

Well I've reattached the PCB to the heatsink with fresh mica/compound on the power transistors and plenty of goop on the smaller ones as Nad originally had.

I took aquibi's advice on the dim bulb tester which saved me as bulb was bright for a second and then amp turned off. Tested the power transistors on FL channel and they still read as the others.

The amp worked previously with the legs of both FL power transistors cut (and they had failed shorted) so I'm guessing something behind them...maybe bias circuit?

Time to do some reading.

[shakalnokturn]

I haven't looked at the schematics extensively, beware all the same that the series bulb itself isn't causing the shutdown.
If you switch back on repeatedly does it glow as bright every time?

Other components that would be worth checking are:
R5528, R5530, R5518, R5522, Q5512, R5519.

R5514, Q5508, Q5509, D5504, D5505 and equivalent in negative side.
There is very little chance anything got damaged further back by the power transistors blowing.

[jacau]

Wow not expecting your response, thankyou again.

I tried 4 times in a row and no change in the intensity of the bulb for the split second it's on.

I previously went through the schematic and tested everything using other channel as expected values and what I replaced from your first row was
R5528, R5530 (both open), Q5512 (Short C+B, B+E OK) and R5520 (139K Ohm instead of 1K) (I also replaced R5519 but it read 2K correctly) plus the two power trans (2390 all shorted, 1560 just C+E)

Of the 2nd row and equiv neg side, only changed C5512 (which didn't change subsequent reading) and remember C5508 gave strange readings but they both corrected after I soldered replacement power transistors in.

I did focus on C5508 for a long time before appreciating how the transistors would affect the circuit so will see if I can confirm values to make sure I didn't miss something.

[m3vuv]

sounds to me like the op needs droping on his head from 25ft!

[jacau]

I went through what I could again with the PCB mounted and the the only real difference that I could see from the two other channels is Q5510. I found front SMD components to again test Q5509/5510 that are located on the heatsink side and tQ5509 was consistent with other channels.

Q5510 B+E and B+C are identical to the other channels (0.635v approx) but C+E is a 1.267v drop stable while the other two are 1.305v.

That reading aside, is it possible the bulb current limit be triggering some protection circuit if the caps take too long to charge up?

I have access to an oscilloscope, could that help with diagnosis?

[shakalnokturn]

Yes the series bulb could be causing it to switch back to standby.
The other option would be putting the bulb in place of either R5518 or R5522, no load on amplifier output.
If the amp stays on and bulb doesn't glow or only faintly you should be good to go.
While the bulb is in you can play with the bias current adjustment to check it has a linear behaviour. Don't forget to turn it right down before removing the bulb. Once rewired normally adjust bias current as per service manual.
You should be ok without oscilloscope for that repair, the most important is being thorough on all the weak solders.

[jacau]

Yes the series bulb could be causing it to switch back to standby.

I took the chance turning it on without the bulb thinking I could monitor temps with a laser temp meter (or a clamp meter?) on the two 0.22 ohm resistors and switch off if I saw them starting to ramp up and all looking good so far!

Fans spin up and no noticeable temp rises after 60 seconds on those resistors.

I set the pot as advised in the service manual before turning on and haven't adjusted yet but I did meter the unloaded front channels : 2mv on the left and 8mv on the right.

Thanks very much again shakalnokturn and (most of) the others that offered their advice. Cheers guys!

[m3vuv]

Just to say i repaired my pioneer tuner/amp,it had one channel push pull npn output transistor sc and was blowing fuses,i didnt have a direct replacement,so looking at the specs used a 2n3055 mounted on a heatsing,all in all it works fine so wouldnt worry about a transistor that got dropped!,it works or it dont,just my thoughts,seems to me like your splitting hairs and looking for problems that dont exist,just fit the thing and try it for god sake!!.

[jacau]

..just fit the thing and try it for god sake!!.

As I'd never diagnosed an amp before let alone used thermal compound and mica I thought it was worth an ask especially because I read output transistors take out other components when they go and incorrectly mounted transistors are just as bad.

I don't know if the NAD design is common to other amps but 6 power transistors are behind the PCB attached to the heatsink and you need to bend the legs of a thermistor around 150degrees to remove so waiting for a clear cut response from knowledgeable people seemed prudent.

But yeah, I can see where you are coming from and there is an old fitter at work who rolls his eyes like I imagine you did 

BTW, if I saw someone plummet 25 feet, I wouldn't suggest they try walking it off