advertisement


HackerNAP troubleshooting help please

Oh, and I tried swapping the feedback cap (C2) for another to see if that made any difference, but it didn't...
 
I think you have a measurement error around Tr4. Assuming that then TR1, 2, 3, 4, 5, 6 look OK, to a first approximation.

We'd expect about + and - 1.4v either side of the bias generator, a difference of somewhere between 2.5 and 3. You have a plausible difference but offset by 30v, which suggest this part of the circuit may be working. In the driver and output section Tr7, 8, 9 could be OK but Tr10 looks knackered. And the 0R22 resistors don't seem likely to still have those values. It's possible or likely that Tr9 is also bad.

You could measure the 0R22 resistors in circuit quite happily I think (obviously power off....) and also look for shorts in the output transistors. Your meter may have 'diode mode' and you should be able to measure the base-emitter junctions.

That's my guess, it will be interesting to see what others think.
 
Thanks Paul. Yes, the 0R22 resistors don't seem right. R27 (TR10 side) is showing as 0.4 ohms, and R26 (TR9 side) is showing a negative value that changes and contradicts itself as I change measurement range on the meter. I've never seen anything like that before - could it be an effect of being in the circuit with the blown transistors?

As for the transistors, using the continuity tester mode on the multimeter I can say that:
TR7 has a closed circuit from collector to emitter in that direction only and nothing else
TR8 has a closed circuit from collector to base in that direction only and nothing else
TR9 has closed circuits between all terminals in both directions
TR10 doesn't have any closed circuits at all

So it looks like I need to take the board out, desolder some components and test and replace them as necessary.

Just out of interest, I'm curious as to the cause of all this. The amp worked flawlessly for several years, and was doing so when I last used it. It then lived on a shelf, and has suffered no physical or environmental insult that I'm aware of. When I plugged it back in to the system, I did so observing all the usual rules - I never connect or disconnect interconnects or speaker cables with the amp on, and the amp is always the first thing to be turned off, and the last thing to be turned on. When I turned it on, there was no thump, pop, crackle or magic blue smoke, and none of the components look burnt or damaged in any way. I'm not aware of solid state components expiring of old age while off. Any ideas what might have caused this?? I'm just concerned that since I've no idea, I have no confidence that replacing the components and putting the board back in won't just 'expire' a bunch more!
 
Good suggestion - thanks:

Non electronic suggestion - when I was working as a computer programmer, most books and printouts were black and white. I always used red, green/blue pens.

Red for "highlighting" (or errors, or really important stuff), and green/blue for actual (you know) writing.

Using a black pen would have meant stuff I'd written didn't stand out, and (being an egomaniac) I considered anything I'd written to be FAR more important than anything printed. :)

BugBear
 
33.4v across an 0R22 resistor means 250W, that resistor is white hot. Or more likely it's not a resistor any more.
This. It must be open circuit. Can you check with DVM? Sorry I see I'm a little late...
(Wasn't sure if they were 0R22 or 0R33.)
This thread here, scroll down
https://pinkfishmedia.net/forum/threads/hackernap-advice-thread.108130/page-12
Double check your transistor readings against the good board as that will tell you what they ought to be in circuit. Same with resistors.
DVM on lowest resistance can be +/-0.2 as it is difficult to measure very low values. Mine typically measure +0.2 ish with probes shorted.
 
This. It must be open circuit. Can you check with DVM? Sorry I see I'm a little late...
(Wasn't sure if they were 0R22 or 0R33.)
This thread here, scroll down
https://pinkfishmedia.net/forum/threads/hackernap-advice-thread.108130/page-12
Double check your transistor readings against the good board as that will tell you what they ought to be in circuit. Same with resistors.
DVM on lowest resistance can be +/-0.2 as it is difficult to measure very low values. Mine typically measure +0.2 ish with probes shorted.
I recently watched a YouTube video about using a regulated 100 mA current source as a driver, so that milliohms become tenth-of-millivolts, readable with a multimeter.

To do this you need a 4 wire connect (AKA Kelvin connection) so that the wires carrying the current are NOT the wires detecting the voltage.

The milliohmeter thus created can even tell the difference between a fully shorted component, and the same component open circuit, but shorted via PCB tracks. PCB tracks have "some" resistance.
 
Looks like both output devices are shot as base and emitter are same voltage but should be circa 0.7vdc difference. I'd strip out tr7 to tr10 plus associated resistors r22 to r25 and d3 just in case and definitely the 0r22s. Rebuild leaving the outputs tr9 and tr10 out of circuit (standard practice with NCC amps) and check you have satisfactory bias adjustment, think it's about 1.6vdc to 2.5vdc or thereabouts (edit: across the common point of r9-r10 and the common point of r19-r20 in your image above) then set this low and replace tr9 and tr10. Check all is good and adjust bias.
 
Last edited:
On my Naim if I put the DVM across the 0R22 I get a reading of 0R2. Anything low is plausible, we're looking for gross issues.

So the OP reads 0R4 on R27, probably OK. But negative and varying on R26, probably open circuit and seeing C13 and C14 through a shorted output transistor.

I think it's reasonable to run these amps without the output transistors fitted, not to develop power of course, but to establish DC conditions and that the feedback and bias generator are working. That might be an approach. Would be good to have a sanity check from someone more dialled in though.
 
Thanks Paul. Yes, the 0R22 resistors don't seem right. R27 (TR10 side) is showing as 0.4 ohms, and R26 (TR9 side) is showing a negative value that changes and contradicts itself as I change measurement range on the meter. I've never seen anything like that before - could it be an effect of being in the circuit with the blown transistors?

As for the transistors, using the continuity tester mode on the multimeter I can say that:
TR7 has a closed circuit from collector to emitter in that direction only and nothing else
TR8 has a closed circuit from collector to base in that direction only and nothing else
TR9 has closed circuits between all terminals in both directions
TR10 doesn't have any closed circuits at all

So it looks like I need to take the board out, desolder some components and test and replace them as necessary.

Just out of interest, I'm curious as to the cause of all this. The amp worked flawlessly for several years, and was doing so when I last used it. It then lived on a shelf, and has suffered no physical or environmental insult that I'm aware of. When I plugged it back in to the system, I did so observing all the usual rules - I never connect or disconnect interconnects or speaker cables with the amp on, and the amp is always the first thing to be turned off, and the last thing to be turned on. When I turned it on, there was no thump, pop, crackle or magic blue smoke, and none of the components look burnt or damaged in any way. I'm not aware of solid state components expiring of old age while off. Any ideas what might have caused this?? I'm just concerned that since I've no idea, I have no confidence that replacing the components and putting the board back in won't just 'expire' a bunch more!


The R26 thing doesn't make any sense - maybe you have some residual charge stored somewhere, otherwise DVMs don't give negative values of resistance.
If the resistor is damaged, it is possible to get varying readings especially if you wiggle it around.

As for the TR7-10 readings, I would use the resistance setting to test. On the contunuity setting you will be checking diodes. Not sure if your amps have catch diodes across the outputs, but this will further confuse things when doing this test.
Bear in mind that a transistor looks like a pair of diodes (B-C and B-E) to a diode tester. NPN obs has the diodes the other way round compared to PNP.

Root cause: difficult to say, however hand soldering is far from perfect (10X less reliable than automated soldering, even for experienced builders). Maybe a joint went iffy during storage?
The only other thing that comes to mind is electroytics needing reforming, but this usually takes years of storage and usually just blows fuses on power-up (Not that there are any fuses to blow in these amps :rolleyes:).
 
Ok. A huge thanks for all your help chaps. I’ll take the board out and replace transistors and associated components as suggested and see if that resurrects it. It’ll be a while before I can get the parts, but I’ll report back once I’ve swapped them out.

Thanks again for all your input and patience!

David
 


advertisement


Back
Top