M500T Carver help?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2007
Paid Member
i didn't find any bad joints. BUT when I attempted to set the bias, the right channel test points went great. Not so good for the left side. NO VOLTAGE SHOWING when trying to set the bias. Checked the trim pot resistance and it is working fine but adjusting it changed nothing. ALSO there is voltage to the to the collectors of the large transistors BUT there is no incoming signal to the bases and nothing at the emitters. They are on but nobody is home.

This has to be the symptom to work with initially. No bias current is achievable.

A big clue would be for you to measure the voltage across Q109 from C to E or if easier to access, the voltage across C123 which is also across this transistor.

With the preset at max resistance the voltage across Q109 will be at its lowest, however you could still read anything up to 2 to 2.5 volts depending on the thermistors (TH101) characteristics.

That setting is for minimum bias... but what voltage do you actually get across Q109?

With the preset turned the other way the voltage across Q109 should rise.

This is where the voltage checks Anatech mentioned becomes crucial. In order for the output transistors to conduct (for the bias to set correctly) there has to be sufficient voltage across Q109 to overcome ALL the (effectively) series connected base emitter junctions of the output transistors plus their drivers plus the pre-drivers.

Assuming a figure of say 0.65 volts for each B-E junction then that means you need to be able to see 0.65 multipled by six (so around 3.9 volts) across Q109. We multiply by six because there are six transistors in the chain. We discount the parallel output devices and count those as a single device.

So that has to be your first step, seeing what happens to the voltage across Q109 as you turn the preset.

You mentioned a volt drop of around 0.04 and 0.06 across R131 and R133. Splitting the difference and saying 0.05 gives a current of around 2.2 milliamps in those resistors which sounds about right.
 
Back on post #10 he states he has 2.9VDC across Q109 with -1.7VDC on the emitter and +1.2VDC on the collector. This brought us to low voltage drops across R123 and R131. Q101 and Q105 are not conducting enough causing the low voltage on the collector of Q105 me thinks. If the bias circuit, Q109 and associated components, was the cause wouldn't the collector and emitter voltages be equal but opposite since the bias circuit floats between Q105C and Q107C?

Craig
 
Administrator
Joined 2007
Paid Member
Its an interesting fault for sure :)

2.9v isn't enough to bias the complete output stage and once the stage drops out of conduction I suspect you will find the voltages on C and E of the bias transistor won't be all that symmetrical, at least not in the same way as when the stage is conducting correctly.

That said, the observed asymmetry of the voltages across the bias transistor and the slightly low bias voltage across that transistor could point to a problem with both upper driver and pre-driver and/or the components around them. It could be a clue...

Looking for this low bias condition is just a possible handle and way into the fault... although low bias on its own in an otherwise OK amp wouldn't account for the symptoms of unequal output between channels and the very poor sound quality.

Looking at the circuit and its possible a problem with the protection circuitry (Q129 and Q127) could conceivably give a basically working amp but with no bias. If one of the transistors was very leaky it would give distorted audio as it loads the driver stage. It might be worth just pulling those transistors out as a quick test. The amp would function normally without them.
 
Thank all of you again for your help. As an amateur this a lot for me to process. I hope that I have been smart enough to keep up with and correctly follow your instructions and ideas. With that in mind I can easily pull Q127 and Q129 if that is your suggestion. It might be a day or two before I get back with the results. Having said that, I am about to go into a 3 day stretch of long hours at work. If and when I pull these transistors, what results should I see? Or hear? Would bias voltage then be normal range again? Would the distortion still be present? Or clear audio? Hope my questions are not stupid. Thanks again gentlemen.
 
Administrator
Joined 2007
Paid Member
They are good questions :) and if you are not sure on anything then just ask.

Pulling those two transistors will eliminate them from suspicion and remove any effect they have from the equation. If one were leaky then it could give the symptoms you are experiencing. Removing them makes 100% sure they are not the problem.

Now at this point I'm going to throw another two parts into the mix and they are C125 and C127.

Do you see how they connect across the base and emitter of the two transistors we are talking about? Well if one of the caps was leaky (even very very slightly) then that leakage would turn on the transistor.

Now if the cap were 'fairly' leaky then the cap alone would disrupt the circuit, with or without those two transistors fitted and so......

......I'm going to suggest that instead of pulling the two transistors and the two caps, it is easier to just isolate one end of each diode D101 and D103. It achieves the same thing with less work for you, and that is total isolation of the protection circuitry from the driver stages.

If there happened to be a problem with those parts then doing this would allow the amp to operate normally.

If on the other hand there is still a problem then we are likely looking at a problem with the driver and/or pre-driver section. There would be no nagging doubt over the protection circuitry because its isolated.

In this scenario we look very carefully at the DC conditions present on the pre drivers and drivers. That means checking the base to emitter voltages carefully. Each should have no more than around 0.7 volts present. The 0.7 volts is measured across the junction. This is where Anatech's method (I'd never thought of doing it that way before tbh. I usually measure directly between the two pins on the device. Anatech's way is safer, less error for slipping with a probe which would be disastrous) comes into its own.

What comes next depends on the results of these checks.

Its important to realise that the low bias in itself isn't the real issue here, that wouldn't cause gross distortion and the big difference in output levels. The low bias is a symptom of another issue, one that hopefully these checks will reveal.

Fwiw, if the isolation of the protection circuitry doesn't 'fix' the issue immediately then I think you are looking at replacing the driver and pre-drivers and doing a quick check of the low value resistors connected to them.

These are the two diodes to lift:
 

Attachments

  • {8E199C76-F3C3-4D42-99AB-3A37AD9FF3FC}.png.jpg
    {8E199C76-F3C3-4D42-99AB-3A37AD9FF3FC}.png.jpg
    92.5 KB · Views: 138
I will get those diodes lifted and get back to you as soon as I can. As I said it may be a day or two. FWIW I would have done this tomorrow morning as soon as I get off of work but I just bought a complete Marantz stereo system from the 70s and I have to go pick it up when I get off work at 7 A.M. in the morning. Don't know if it works or not. Don't care. Got it cheap enough that it doesn't matter. One of those deals ya know. Even as parts it is worth far more than I gave. I will put it in the pile with the rest of the stuff I've got. Hope ya'll have a great day and thanks again.

Terry
 
Administrator
Joined 2004
Paid Member
Hi Mooly,
You've got this well in hand. As for the meter common being plugged into the hot speaker out terminal, I found that over the years it allows me to thing about voltage drops much more easily and it frees up one hand. You're right about the probe slipping too. This is how I taught my technicians to troubleshoot and output stage. Base-emitter drops become obvious and there is then only one probe to worry about slipping.

Hi Terry,
Just as a note, an open E-B connection or open resistor in series with a base or emitter will not be obvious. The action of the voltage map stage will make the voltage drops look like they are in the ballpark. So don't naturally assume these parts are okay just because they have approximately the correct voltage across them. If you suspect a component, power off, allow the voltages to drop, then measure the component with an ohmmeter or diode check function as appropriate.

To troubleshoot, send a 1KHz (approximately) signal in and adjust the controls and level so you are getting about 1 Vrms output. Connect a dummy load for this, then look at the waveforms with your 'scope.

-Chris
 
Thanks Anatech. I was just getting started up again this morning. Was a hectic few days at work. I was able to lift d101 and d103. No change in audio output. Audio still sounds muddled.

Now. Not exactly sure which are pre drivers and so on SO I checked the e to b voltages as follows.

Q101 = .58vdc
Q103 = .58vdc
Q105 = .58vdc
Q107 = .58vdc

Q119 = .17vdc
Q121 = .17vdc

Q123 = 0
Q125 = 0

Q117 = 1.84

Q113 = .55vdc

Did I miss anything?
 
Q101 and Q103 are level shifters. They shift the signal voltage to where the next stage, Q105 and Q107, can use the output of input IC101. Q105 and Q107 are the VAS, Voltage Amplification Stage. This stage is the last voltage amplification stage, every stage after this will be current amplification. Q111 and Q113 are the predivers, Q115 and Q117 are the drivers, and Q119, Q121, Q123 and Q125 are the outputs. And you already know that Q109 is your bias transistor, Vbe multiplier.

Craig
 
Oops sorry I left out Q 109. I actually took the reading and forgot to put it down.

Q109 = .60vdc.

I did attempt to rotate the bias voltage pot while measuring voltage on Q109. It only dropped from.60 to .59 when I completely closed the pot to zero.

I also did a comparison with the one on the right channel and it too was .60vdc.

I am now trying to figure out how I missed Q111 or if I just misnumbered it. Anyway here is Q111.

Q111 = .58vdc
 
In post two I was to understand that the output of the opamp is inside the feedback loop everywhere you check the signal it will look distorted. HOWEVER I was following the schematic which shows an enhanced dark line which I thought was the audio path and it shows the audio going into the opamp and out of a different pin. the audio BEFORE the opamp was clean. the audio AFTER was very distorted. But this was all just my thought process. Just thought I would share what it was I was thinking when I scoped this thing. Not sure that there is anything else I could offer as far as the scope goes as the right channel was a clean audio signal all the way through.
 
Administrator
Joined 2007
Paid Member
Its definitely (lol, famous last words) faulty then, that's assuming that there is no weird oscillation upsetting the meter readings.

So get it changed and be sure to reset the bias preset back to the position that gives lowest current before powering on again. That means preset on maximum resistance.
 
Administrator
Joined 2007
Paid Member
Hi Mooly,
You've got this well in hand. As for the meter common being plugged into the hot speaker out terminal, I found that over the years it allows me to thing about voltage drops much more easily and it frees up one hand. You're right about the probe slipping too. This is how I taught my technicians to troubleshoot and output stage. Base-emitter drops become obvious and there is then only one probe to worry about slipping.

-Chris

Thanks Chris :) This Carver is one I've never seen before... its quite an interesting design and I suspect capable of above average sonics.

My B-E radar detector is definitely twitching at the moment with this one ;) we have nearly two volts across one device...…
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.