The original design took 20 mins to stabilise - why? Only air thermally couples the bias transistor to the output stage.
The thermal design is always poor with quasi-comp designs but this a bad case and this particular kit seems like
the bias is not working correctly, as you say, istari knight.
There are 2 kits there - are they both doing the same thing, measuring similarly?
Check the bias transistor for orientation by checking pinouts against the schematic. Dont be fooled by substitute TO92
transistors with wrong pinouts - it has happened before with these kits. Check Vbe (Voltage between base and emitter of
that and the driver transistors is close to 0.65V so that you know the bias
circuit is capable of operation.
It should be after 10 mins or so, if the bias holds => 5.5 mV across either 0.22R resistor, according to your measurements
but is the heatsink near the transistors and the driver tansistors getting even slightly warm?
BTW, what are you using for a heatsink? Anything goes I guess, but that looks a tad thin, unless the pic lies.
The thermal design is always poor with quasi-comp designs but this a bad case and this particular kit seems like
the bias is not working correctly, as you say, istari knight.
There are 2 kits there - are they both doing the same thing, measuring similarly?
Check the bias transistor for orientation by checking pinouts against the schematic. Dont be fooled by substitute TO92
transistors with wrong pinouts - it has happened before with these kits. Check Vbe (Voltage between base and emitter of
that and the driver transistors is close to 0.65V so that you know the bias
circuit is capable of operation.
It should be after 10 mins or so, if the bias holds => 5.5 mV across either 0.22R resistor, according to your measurements
but is the heatsink near the transistors and the driver tansistors getting even slightly warm?
BTW, what are you using for a heatsink? Anything goes I guess, but that looks a tad thin, unless the pic lies.
Thanks for that, I hadn't refreshed the thread so missed your reply when I posted the update. It looks like faulty NOS caps were the cause.
The heatsink is 6mm copper plate with a 3mm ally "channel" coupled with thermal compound.
I dont think so as both boards do it at exactly the same time. I used these exact boards in the past in another amplifier so I know they're working... Or at least should be.
Its now the same as before... Power up, set bias, bias will slowly drop, as the bias drops DC offset goes up, driver & output heatsinks stone cold even when bias is cranked up to 10mv across R29.
Last edited:
Yes, bias is sick. Measure Vbias (across both driver bases). 'should be 3 diode junction drops or ~2.0 V and rock steady. If not, you'll have to check the semis in the bias circuit, including Q4,6. Check voltage across Q6 emitter resistor (68R) to see what bias current is running through the VAS and bias genenerator circuit (should be about 6 mA)
That should keep you occupied for 5 mins or so.
edit: we'll have to gather a bit of data for a picture of what's going wrong - maybe this will simplify the search or not but we need to establish the basics are there first.
That should keep you occupied for 5 mins or so.
edit: we'll have to gather a bit of data for a picture of what's going wrong - maybe this will simplify the search or not but we need to establish the basics are there first.
Last edited:
Well I couldn't leave well alone so continued to mess around with it when I remembered last time I used these boards I had similar issues... I did a search & found them on page 64 of this very thread where I said "connecting SGND+GND to star earth on the cap bank & Speaker 0v to DC- sorted it" ... Well I just tried it out of desperation & its worked. Bias has sat at 4.4mv for 20 minutes, DC offset is under 50mv & it produces sound perfectly.
I wonder if this is due to an error on the PCB ? As back on page 64 user "umfowenu" had exactly the same boards with exactly the same problem.
So, if anyone out there has blue "140C" boards using Sanken 2SC3858 outputs with this fault that's what you need to do to get it working.
I wonder if this is due to an error on the PCB ? As back on page 64 user "umfowenu" had exactly the same boards with exactly the same problem.
So, if anyone out there has blue "140C" boards using Sanken 2SC3858 outputs with this fault that's what you need to do to get it working.
We've been here before, Barry - around page 70, IIRC.
You were looking for the ground lead connection for the input stage, which had 2 connectors, one for the input lead shield and the other to return direct to star ground. The speaker return is to power ground (on the PCB) which works as designed but stereo and multi-channel amps are better wired direct back to star ground too.
Several people on other forums and probably here too, have questioned Naim's grounding arrangement but bearing in mind the success of the design and how many decades it was like that with few problems, who can argue? Ref. Neil McBride's audio pages archive: http://www.acoustica.org.uk/ or via: http://urlm.co.uk/www.neilmcbride.co.uk (the acoustica.org archive link was dead) if you want to glean some bits of info. 'Good read anyway, but with your gear, I guess you've already been there - done that
You were looking for the ground lead connection for the input stage, which had 2 connectors, one for the input lead shield and the other to return direct to star ground. The speaker return is to power ground (on the PCB) which works as designed but stereo and multi-channel amps are better wired direct back to star ground too.
Several people on other forums and probably here too, have questioned Naim's grounding arrangement but bearing in mind the success of the design and how many decades it was like that with few problems, who can argue? Ref. Neil McBride's audio pages archive: http://www.acoustica.org.uk/ or via: http://urlm.co.uk/www.neilmcbride.co.uk (the acoustica.org archive link was dead) if you want to glean some bits of info. 'Good read anyway, but with your gear, I guess you've already been there - done that
Last edited:
I should add that the 140C clone boards are just that, including the requirement to wire them up as the Naim amp design required. That wiring diagram isn't included with the kits any more apparently and only the schematic is shown on the website of those few sellers still offering the kits. That's a problem
Edit: so was my link above -sorry! try http://www.neilmcbride.co.uk/
Edit: so was my link above -sorry! try http://www.neilmcbride.co.uk/
Last edited:
Good question, nasty challenge though.
I hope that's not the real reason the caps took a while to charge and the bias drifted. If it is the case, then you won't get a lot of power out of that.
You do actually need a transformer with centre tapped or dual, in-phase secondary windings to derive balanced supply rails. 'Odd though, most are dual wound
I hope that's not the real reason the caps took a while to charge and the bias drifted. If it is the case, then you won't get a lot of power out of that.
You do actually need a transformer with centre tapped or dual, in-phase secondary windings to derive balanced supply rails. 'Odd though, most are dual wound
An externally hosted image should be here but it was not working when we last tested it.
I thought at first this was a clone . Helpful perhaps .
Does the power supply on these clone's have to have a center tapped transformer ? 'Cos mine isn't