advertisement


HackerNAP troubleshooting help please

L

Lokidtc99

What does this button do...?
Hi all,

I wonder if someone might be able to point me in the right direction with steps to troubleshoot a HackerNAP please?

I built it some 10 years ago, and it was in daily use for probably the first 7 of those. Never had any problems with it. Then I moved, and took it out of my system. It has been on a shelf since.

Today I wanted to try it back in my system, so plugged it in, and got no sound. On closer inspection I find that one channel's speaker protection board is 'tripped' (more accurately, starts open circuit as normal, but never closes circuit.) The other channel works fine. But only when the problematic channel is not connected to a source.

Oddly, when I plug in the interconnect of the affected non-working channel, it trips the speaker protection board of the unaffected working channel, as well as remaining tripped on the affected channel.

I've measured voltages coming from the PSU boards, and all are normal, so I'm assuming something's died on the amp PCB itself. But I don't know how to work out what. And I don't understand why connecting the affected PCB to a source would cause the unaffected working channel's protection board to trip.

I only have rudimentary electronics understanding and built the HackerNAP by following the instructions closely and asking questions, so please assume no knowledge on my part!!

Many thanks in advance.
 
Thanks for replying.

This is the amp (click for large image):

Soft start module operating normally:

Lt channel Velleman speaker protection circuit operating normally:

Rt channel Velleman speaker protection circuit 'tripped':
Note, the amp is powered on, but there are no other cables plugged in (ie no speaker cables and no interconnects)

This is the Rt channel amp PCB:

Rt channel from an angle to see what's below power cables:
 
Sounds like some sort of grounding error. However it is worth checking:

a) With power off measure all 0V points e.g. input 0V, output 0V, power star ground etc to check that they are all connected together (Ohms setting or continuity setting on the DVM).
b) With power off measure from 0V to + supply and 0V to - supply using the continuity/diode test setting. You should get a beep then silence each time you check.
c) With power on check the output voltage of each channel (at the amp board, not at the amp output terminals). See if this changes when the interconnect is plugged in.
 
Thank you. Results are:

a) All connected
b) Continuous tone when black probe on 0V and red probe on -V supply on both FE and OP PSU boards on both channels, but not when probes reversed.
Continuous tone when black probe on +V and red probe on 0V supply on both FE and OP PSU boards on both channels, but not when probes reversed.
c) Power on/no interconnect or speaker connected: 0.03V Lt (good) channel, 15.8V Rt (bad) channel.
No change when interconnect connected to good channel.
No change to voltages when interconnect connected to bad channel, but the speaker protection circuit on the good channel triggers (despite still only seeing 0.03V at amp board output)
 
Lokidtc99 said:
when interconnect connected to bad channel, but the speaker protection circuit on the good channel triggers (despite still only seeing 0.03V at amp board output)
Click to expand...
Is this with both interconnects connected back to a pre-amp? If so it is likely that the DC offset (with 16V on the output, there will be a marked DC offset on the input too) is being fed from the bad channel back into the good channel via the interconnect as I see you have no DC blocking caps on the hackernap boards. That will trigger the protection on the good side.
Leakage in your feedback cap is probably the first culprit, or has it possibly become disconnected as they sit a long way above the boards. This isn't good as it creates a loop area that degrades sound (I know this to my own cost...)
 
Yes, as you say, both interconnects connected to preamp (hope I haven’t damaged that)

I have the feedback caps (I think they’re tants IIRC) on leads to facilitate cap rolling, but if the lead length is critical then I’ll just put the replacement caps straight on the board.

I’ll need to go back through the forum to see what character different feedback caps imbue. As I’m getting older I’m preferring a smoother sound (relatively speaking) compared to the razor sharp sound I preferred when I was younger.

Do you have any recommendations by any chance?

thanks again
 
Wet tans are supposed to give the best overall compromise. I wouldn't make any personal suggestions as my hearing is pretty sha99ed anyway....rolls off from about 5kHz and gone by 9kHz
 
Good Afternoon,

Wet tantalums do seem to be a firm favourite as feedback, as stated elsewhere I'm using 300uF devices.

Regards

Richard
 
Lokidtc99 said:
Yes, as you say, both interconnects connected to preamp (hope I haven’t damaged that)

I have the feedback caps (I think they’re tants IIRC) on leads to facilitate cap rolling, but if the lead length is critical then I’ll just put the replacement caps straight on the board.

I’ll need to go back through the forum to see what character different feedback caps imbue. As I’m getting older I’m preferring a smoother sound (relatively speaking) compared to the razor sharp sound I preferred when I was younger.

Do you have any recommendations by any chance?

thanks again
Click to expand...

Next steps:
a) Measure the dc voltage on the input of each channel.
b) Remove (or disconnect one end) of the feedback cap in the bad channel and measure the output dc voltage at the amp board. Don't have speakers connected!!

In situation b) the amp will not amplify but it will work at dc and will identify if the feedback cap is leaky.
Theoretically test b) could cause the amp to oscillate. I have never had this happen, but I've never had a NackerHAP. Keep you finger on the zobel resistor, if it gets hot switch off ASAP.
 
@S-Man you could alternatively just short the feedback cap out so you have full DC feedback, rather than this just 'floating' by removal of the cap (as you would do if you wanted an op-amp circuit to work down to DC). I did this inadvertently once and it resulted in about 1.5V dc offset on the output, but this will depend on the balance of the input transistor pair. Shouldn't cause oscillation either.
 
Yes, that was my thought after writing it.
Edit: a further thought. Not familiar with the Hackernap, but basically it is a souped-up NAPA circuit. If you take 16V and divide by 28 (the gain if same as NCC200 etc) you get circa 0.6V: a Vbe voltage drops worth. does this have cascode or similar?
 
S-Man said:
Next steps:
a) Measure the dc voltage on the input of each channel.
b) Remove (or disconnect one end) of the feedback cap in the bad channel and measure the output dc voltage at the amp board. Don't have speakers connected!!

In situation b) the amp will not amplify but it will work at dc and will identify if the feedback cap is leaky.
Theoretically test b) could cause the amp to oscillate. I have never had this happen, but I've never had a NackerHAP. Keep you finger on the zobel resistor, if it gets hot switch off ASAP.
Click to expand...

a) Good channel = -0.029V, Bad channel = 0V
b) With C2 shorted = 16.4V initially but dropping to 16.0V over about 10 seconds and then staying at 16.0V
 
29mV across 27K is 1uA which is about right for the base current of the input transistor. I.e. 1uA X ~500 (TR1 hFE) = 0.5mA which is about right for TR1 collector current.

0V indicates that TR1 is "off". This stacks up with the dc offset... TR2 is fully on and conducting all of the tail current.

I don't know the NackerHAP but for an NCC200 I would suggest checking the equivalent voltages to those in post #41 here:
https://pinkfishmedia.net/forum/threads/avondale-ncc200-mod.235972/page-3
The schematic is lower down the page in Alan's post.
1st check the current sources by measuring the voltages across the 680R and 68R.
Then check the LTP current balance (which seems to be wrong in your faulty channel).

If you measure the Vbe of every transistor you can often figure out where the problem is. They should all be about 0.6V (the ZTXs run at about 0.5V).
 
Had a look at the schematic for the Hackernap and it is essentially the same as the NCC200 apart from minor extras. If Tr1 is essentially 'off' then there will be a low voltage across the 1k that turns the VAS transistor on, so the VAS transistor will also be 'off'. If that is the case then there should be a large negative offset and not a positive one. Assuming that none of the major components are damaged, I wonder if there could be a problem with the bias pot. If for some reason the wiper isn't contacting the track, poor solder joint or similar, then there may be insufficient bias on the output stage, which could also essentially be 'off', but just provide enough current to drive the feedback network. Can you measure the voltages across each of the 0R33 output emitter resistors?
 
I may have missed something but a 16v offset in 'normal' and a 16v offset with the feedback cap shorted tends to suggest an internally shorted feedback cap.

Just measuring voltages around the front end can make this clear, in working condition both ends of the 1k feedback resistor would be at about -0.6v. Right now I'm guessing the transistor end is positive, I guess about +0.6. So emitter is at 0v and input transistor cannot turn on.
 
S-Man said:
1st check the current sources by measuring the voltages across the 680R and 68R.
Click to expand...

R18 (68R) = 0.84V
R19 (680R) = 0.36V

a.palfreyman said:
Can you measure the voltages across each of the 0R33 output emitter resistors?
Click to expand...

Can you link to the schematic you've looked at please. I think I may have a different version (v1.6) as I can't see any 0R33 resistors on it. If you mean R26 and R27 then those are 0R22 on my schematic. But if it's those you mean, I measure 33.4V across R26 (TR9 side), and 0V across R27 (TR10 side).

Paul R said:
I may have missed something but a 16v offset in 'normal' and a 16v offset with the feedback cap shorted tends to suggest an internally shorted feedback cap.

Just measuring voltages around the front end can make this clear, in working condition both ends of the 1k feedback resistor would be at about -0.6v. Right now I'm guessing the transistor end is positive, I guess about +0.6. So emitter is at 0v and input transistor cannot turn on.
Click to expand...

Assuming you mean R6, I measure 4.04V on the transistor end, and 3.59V on the other end (though these seem to move around by +/-0.05V when I measure repeatedly)
 
I don't have the schematic for your power amp board, is there an upload anywhere? Then we're all on the same page.

33.4v across an 0R22 resistor means 250W, that resistor is white hot. Or more likely it's not a resistor any more. But we're just fishing.

If you can leave the power on without anything getting hot then measure all the voltages around all the devices, ideally print the schematic and write them on and post an image Then the whole picture can be considered. You can get a really long way with Ohms law. voltage measurements and knowing just a little about transistors, mostly that you expect 0.6v ish between base and emitter for one that's working.
 
You must log in or register to reply here.


advertisement


Back
Top