LC audio TEM distorsion?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hoping for some help, I have a LC Audio The End Millenium that distorts audible with a 1khz test signal on one channel. Oddly the amp sounds fine and will play very loud without one thinking there is anything wrong.

Since the amp plays fine at very high levels and with lower frequencies I am thinking or should I say hoping that its over all function/transistors are still good but maybe there are some other non linearity introduced that should not be there. I have cleaned the board a few times but it did not make a difference.
When I first noticed the distorsion it was fine on low volumes but now it's there at any level.
There is a glass substrate resistor (R4) that according to the documentation lowers 3rd order distorsion compared to a traditional resistor. The distorting channel has a lot of of 3rd order and I cant help thinking maybe something has gone wrong with the glass substrate.
Any thoughts or ideas on how I can figure this out is appreciated.

dxOROSi.jpg
 
The first post isn't clear to me. What does "plays fine at very high levels" mean? That it is functional but with audible distortion or that at high volumes, there is no audible distortion? In the latter case, what is meant by "now it's there at any level"?

In any case, I agree with OldDIY that it could be related to the quiescent current setting (P1 or something near P1). Crossover distortion due to an incorrect quiescent current will be less noticeable at high levels.
 
I only hear distortion using test tones but with music it's not obvious and the amp will sound ok with music even at really high output levels. On bias it is on spec or running a bit higher and I even bumped it up as much as I could to see if it went away on low volume but it did not.
I measured both channels (using a cheap computer ADC) and this is what I get:
Re1CqSN.jpg

CHzTyVF.jpg


What about R4 and glass substrate, are they prone to break and if they did would this be a possible result?
 
Last edited:
Bumping this, any one have any ideas or suggestions?

Question, can transistors be partially broken?

Listening to the distorsion with a 50-200 sinus signal it's like a ticking sound from the tweeter following the lower freq. A bit like something switching on and off.
 
Bias is ok at above 0,5 Ah, currently running at 1 Ah. I called LC audio (did not expect them to pick up) and they suggested one of the transistors had overheated which thinking about it I might have done with the soldering iron re-bulding the amp into a new enclosure etc.
 
As you have two channels and one works as intended, you could try measuring the node voltages of both channels with inputs shorted and outputs open or loaded with power resistors. Maybe differences between the channels could give an idea of what is going on.

Multimeters and their leads can have a high enough input capacitance to cause instability in amplifiers, so it may be necessary to connect a resistor between the node under test and the lead to the multimeter. Anything from 1 kohm to 10 kohm will do with a typical digital multimeter.
 
Question, can transistors be partially broken?

Listening to the distorsion with a 50-200 sinus signal it's like a ticking sound from the tweeter following the lower freq. A bit like something switching on and off.

Yes, for example, avalanche breakdown, particularly of the base-emitter junction, can lead to increased 1/f noise and decreased hFE at low currents. I doubt if that has anything to do with your problem, though.
 
As you have two channels and one works as intended, you could try measuring the node voltages of both channels with inputs shorted and outputs open or loaded with power resistors. Maybe differences between the channels could give an idea of what is going on.

Multimeters and their leads can have a high enough input capacitance to cause instability in amplifiers, so it may be necessary to connect a resistor between the node under test and the lead to the multimeter. Anything from 1 kohm to 10 kohm will do with a typical digital multimeter.

This was a good suggestion but I had the exact same voltages from T1 to T8, I also looked at the signal on these and it showed the same distorsion. This got me thinking it's created on the input side and after mulling over it for a bit I realized I could feed the - and + inputs independently while measuring.

It turns out the distorsion does not show feeding the - input and shorting + to ground while its there feeding the + input.

Am I correct in thinking T17 on + input is broken or could it still be something else?

dxOROSi.jpg
 
T15, or incorrect biasing of T17 or T15. If one of them would be biased too close to saturation, signal at the positive input might just drive it into saturation at the signal peaks, while signal at the negative input just doesn't.

Do the collector, emitter and base voltages of T17, T15, T3 and T4 more or less match between channels?
 
T15, or incorrect biasing of T17 or T15. If one of them would be biased too close to saturation, signal at the positive input might just drive it into saturation at the signal peaks, while signal at the negative input just doesn't.

Do the collector, emitter and base voltages of T17, T15, T3 and T4 more or less match between channels?
Yes they do, does that then leave T15 as the broken part?
 
I'm not sure. For reasons of symmetry, a malfunctioning T15 should have about the same effect as a malfunctioning T17, so if it can be T15, it can also be T17. In either case, it would be unusual that a malfunctioning transistor causes increased high-order distortion but has no impact on the bias point. It could also be a rotten contact right at the positive amplifier input.
 
I'm not sure. For reasons of symmetry, a malfunctioning T15 should have about the same effect as a malfunctioning T17, so if it can be T15, it can also be T17. In either case, it would be unusual that a malfunctioning transistor causes increased high-order distortion but has no impact on the bias point. It could also be a rotten contact right at the positive amplifier input.

Ok I'll check the connection but to run it on unbalanced inputs I have now shorted + to earth in the chassis xlr and if the connection was dodgy I assume I would still have issues even though they would be different?

I'll order T15/17 replacements and we'll see how it goes.
 
Yes, for example, avalanche breakdown, particularly of the base-emitter junction, can lead to increased 1/f noise and decreased hFE at low currents. I doubt if that has anything to do with your problem, though.

Maybe I shouldn't have written that last sentence, because it looks now like something is wrong in the input section and the input transistors can be driven into base-emitter avalanche breakdown. Their base-emitter reverse breakdown voltage is specified as 5 V minimum. Wild guessing that the typical value will be somewhere in the 7 V to 9 V range, a peak input signal of about 8 V to 10 V with the amplifier switched off or 11 V to 13 V peak with the amplifier on will drive them into avalanche. Has the amplifier ever been driven very hard or has its input by mistake been connected to a microphone input with 48 V phantom supply?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.