[TonyM]

I noticed a very definite fuzziness/distortion coming from one of my Aksonics last night, particularly the tweeter.  I switched the speaker cables over and its one of the channels on my 55w nirvana, not the aksonics that appears to be the problem.  I intend to get the DMM's out tonight and run through the diagnostic checks as per Hugh's instructions, I was just wondering whether anyone had come across this particular problem before or would have a considered likely cause, thus saving me some time with the hammer wielding. 

cheers

Tony

[Rom]

Hi Tony,

Cold solder was one I found before. Any solder joint the look fuzzy and dull, try to removed and place new solder . Other than that I will leave to the others on the board to suggest.
I hope this help.

Rom

[AKSA]

Tony,

Rom is likely on the money, usually this is a dry solder joint, but measure offset and bias, note anything out of the ordinary, and come back to us here.

Cheers,

Hugh

[TonyM]

Rom, Hugh thanks for the rapid response.  Glad i asked now as i did a bit of soldering on a shunt mode volume pot last night so i suspect my ham-fisted soldering is probably the culprit again

cheers

Tony

[TonyM]

Well i've gone over my soldering and the problem with the noise from one speaker persists, so i've done some checks on the offset and initial bias and these checked out ok.  However, when it came to measuring the final bias from the emitters TP's 1 and 2 beneath the modules, one of the modules gave no measurement whatsoever, so i'm thinking this may be the problem.  Any ideas chaps?

[AKSA]

It looks like your bias generator is kaput, Tony.

Measure the voltage from the base of the top driver (C4793) to the base of the bottom driver (A1837).

It should be around 2.44 volts for correct bias.

If it is much lower than this, then no quiescent current will flow, and the sound will be destroyed by crossover distortion.

Once we know the measurement, we can proceed.

Hope this helps,

Hugh

[TonyM]

no measurement on one of the modules Hugh,so looks like its kaput. I guess these need replacing then?

cheers

Tony

[AKSA]

Tony,

When amps fail, normally they go in the output devices, which are generally the most fragile parts of the entire chain.

My suggestion would be to:

1.  Remove both output devices from the pcb.
2.  Using a croc clip, connect the emitter of the left - or the right - driver to the output spade connector.
3.  Disconnect the speaker.

Now, fire it up, and measure the output offset wrt ground (power or signal, star earth is a convenient point!), and the base/emitter voltage of the two drivers (A1837, C4793), should each be 0.6V.

Once you've done this, we will talk again!!

Cheers,

Hugh

[TonyM]

Hugh, sorry to be thick, do you mean wrt ground of the ouput device of pcb ground?

[TonyM]

sorry Hugh just read your last post properly  

[TonyM]

Hugh, i get nothing like 0.6mv, perhaps i'm not measuring the right things; i took out the output devices, connected a croc clip to the emitter and to the output spade, then put the red probe of the dmm to the output spade and the black probe to star earth.  I have a hunch i'm being a bit thick still.  I did measure between the collector and emitter on each output device and both produced readings if this is any use.  Also i noticed when removing one of the output devices that it was a bit loose, could this be a problem?

Tony

[AKSA]

Tony,

That should be 0.6 volt, or 600mV!  The DRIVER base/emitter voltage (across the device, left terminal to right terminal, forget the centre terminal which is collector) has to be 0.6 volt for the device to be functioning as it should.

Yes, an output that's loose will definitely cause issues.

Now the outputs are out, check them with a meter.  Again, best bet is base to emitter, left to right terminal.  Use your DMM on ohms, the reading should measure DIFFERENTLY one way to the other, that is, red probe to base black to emitter should measure differently to red probe to emitter black to base.  If there is NO difference, the transistor is kaput.  There are other tests, but this one is definitive.  If there is no difference, measure collector to emitter and emitter to collector.  This is usually a short circuit IF the device is destroyed.

HOpe this helps,

Hugh

[stvnharr]

Also i noticed when removing one of the output devices that it was a bit loose, could this be a problem?

Tony

Tony,
I once had a lot of fuzziness, noise, etc. from one speaker. I found a loose output transistor. I tightened the screw/bolt, and all was well.  I'd try this if I were you before proceeding any further!!
Probes, clips, etc. can often innocently touch the wrong thing under power, and then you have real troubles.

Steve

[AKSA]

Tony,

Steve is 100% correct!  Check your outputs are strongly attached to the heatsink;  if not they get very hot and can fail.

Thanks Steve......

Hugh

[TonyM]

Hugh

Measured the output devices as you suggested and both appeared to be ok, so i soldered them back in, fired up and measured output wrt ground and get a reading from both modules:  -10 and -20 respectively.  I guess this means both output devices are working fine and as steve suggested, the loose driver may have been the culprit.  It only occurred to me after i had removed both devices that this may have been the cause .  Now with these measurements am i to assume that i need to do some tinkering with some resistors to get the offset up to 30mv? 

cheers

Tony

[AKSA]

Ah, Tony,

Good news, happy end, God is in his Heaven.......  ah, I love a Mary Poppins ending!

Yes, leave all adjustments as they are, switch on with speakers disconnected, wait half an hour to reach thermal equilibrium, and set up the bias for 30mV across ONE (left or right) emitter resistor, the large sand cast jobbies UNDER the pcb.

If you want to get cunning, you can set it up by measuring across BOTH resistors, that is, emitter of the top output to emitter of the bottom output, and adjusting to 60mV.

Either/or is just fine!

Happy listening!!

Cheers,

Hugh

[TonyM]

Hugh

Everything is back to working how it should, despite my interventions. 
Many thanks

Tony