yes it's around 45C.
I find it hard to adjust voltages and bias very precisely even if I think the amp is warmed up I go check after a couple of hours and it's slightly off (few mV). Possibly because the other channel of the amp is opened up and the amp is influenced more by the ambient temperature?
I find it hard to adjust voltages and bias very precisely even if I think the amp is warmed up I go check after a couple of hours and it's slightly off (few mV). Possibly because the other channel of the amp is opened up and the amp is influenced more by the ambient temperature?
It’s a living creature with moderate thermal feedback.
So don’t worry when it’s hovering a bit around the wanted bias current.
Hans
So don’t worry when it’s hovering a bit around the wanted bias current.
Hans
Having the same issue on my ml 33h.
Power cuts out after 10min of use and doesn’t turn back on until cooled which leaves me to believe it’s the thermal protection.
I can’t find any schematics online. Any place I should start probing first?
The other side was repaired before with the exact same issue under warranty. I didn’t get a invoice for it so I don’t even know what they replaced.
Power cuts out after 10min of use and doesn’t turn back on until cooled which leaves me to believe it’s the thermal protection.
I can’t find any schematics online. Any place I should start probing first?
The other side was repaired before with the exact same issue under warranty. I didn’t get a invoice for it so I don’t even know what they replaced.
Wow, just read through all the pages. Doesn’t look like anyone has solved the thermal protection issue.
A few years when I took my amp into the servicing place. I was told they didn’t have parts and didn’t know when they could get it. Instead of repairing I guess it’s easier to replace the entire board. Surprised there isn’t a service manual for my 33h
A few years when I took my amp into the servicing place. I was told they didn’t have parts and didn’t know when they could get it. Instead of repairing I guess it’s easier to replace the entire board. Surprised there isn’t a service manual for my 33h
Got it, good I can live with that. Was just considering a file I saw on another thread where it was advised to set pos and neg Vreg within 1mV difference. I can do that but with time they will drift somewhat
I don’t know about the 33H. Anyway my issue was quite different from yours since the protection never activated with the amp switched on.
This max allowable difference between pos and neg Vreg has to do with the slow start and fast switch off circuit.
When these voltages are too far apart, one of the transistors in said circuit start conducting and will interfere with the audio signal on P1006/P1007.
But since it's voltage divider between Pos and Neg has 0.1% resistors, this alone can already give a 85mV offset, so aiming for 1mV is a bit pointless,
Try to keep it within 50mV, that will do.
Hans
.
P.S. I'm not aware of any thermal protection issues in the 23.5.
This circuit is fully functional.
Dear Hans, I have news (good or bad news?).
The last components arrived yesterday and I reassembled the entire amp (left channel). Tested the main board and everything was ok. Then I mounted the top board and measured LS output, here I got very worried. Basically at turn on the voltage fluctuates a little as the other channel for a couple of seconds, but then it stabilized at around 100mV. Then after a couple of minutes it quickly drops to about 10mV, and in a few more minutes stabilizes at 0,4mV. At switch off the voltage also fluctuates a little like the other channel but for a few more seconds, then it goes to about 50 mV and it takes a few minutes to get back to 0,1-0,2mV. Not sure if it ever gets to 0V even with amp off. I thought all of this is not normal.
I anyway hooked up the speakers since what I saw on the scope didn't seem dangerous to me. Well, the amp seems to sound good, although I still have to listen to it carefully as I'm not sure about bass. So I don't know where this DC offset comes from. On this channel I replaced a lot of the transistors of the inverting input section that had been replaced with different ones. I tried to match as well as I could (within 0,5mV VBE) with this circuit https://dragonflyalley.com/synth/images/TransistorMatching/ianFritz-transmat0011_144.pdf. However I didn't check hfe since my meter doesn't measure it. Also the meter is very cheap and could be inprecise.
The last components arrived yesterday and I reassembled the entire amp (left channel). Tested the main board and everything was ok. Then I mounted the top board and measured LS output, here I got very worried. Basically at turn on the voltage fluctuates a little as the other channel for a couple of seconds, but then it stabilized at around 100mV. Then after a couple of minutes it quickly drops to about 10mV, and in a few more minutes stabilizes at 0,4mV. At switch off the voltage also fluctuates a little like the other channel but for a few more seconds, then it goes to about 50 mV and it takes a few minutes to get back to 0,1-0,2mV. Not sure if it ever gets to 0V even with amp off. I thought all of this is not normal.
I anyway hooked up the speakers since what I saw on the scope didn't seem dangerous to me. Well, the amp seems to sound good, although I still have to listen to it carefully as I'm not sure about bass. So I don't know where this DC offset comes from. On this channel I replaced a lot of the transistors of the inverting input section that had been replaced with different ones. I tried to match as well as I could (within 0,5mV VBE) with this circuit https://dragonflyalley.com/synth/images/TransistorMatching/ianFritz-transmat0011_144.pdf. However I didn't check hfe since my meter doesn't measure it. Also the meter is very cheap and could be inprecise.
ok thanks, good night.
After a few hours of use today, the voltages I described seem to have changed slightly (but maybe is just because the amp now is warm already). on switch on I now measured -70mV (it was negative before too although I wrote positive by mistake), it stays there for about 3 minutes and then quickly drops to below -5mV in 1-2 minutes and then stabilizes to -0,8mV and it stays there indefinitely.
on switch off the voltage now goes to 0V in a few seconds. Not sure if with the amp cold this would still be its behavior, will check tomorrow.
After a few hours of use today, the voltages I described seem to have changed slightly (but maybe is just because the amp now is warm already). on switch on I now measured -70mV (it was negative before too although I wrote positive by mistake), it stays there for about 3 minutes and then quickly drops to below -5mV in 1-2 minutes and then stabilizes to -0,8mV and it stays there indefinitely.
on switch off the voltage now goes to 0V in a few seconds. Not sure if with the amp cold this would still be its behavior, will check tomorrow.
ok, so now even with the amp completely cold when I switch it off the output goes to 0V in a couple of seconds (weird, maybe new components needed a little break in?).
on switch on however after initial fluctuations the voltage goes to -95mV and after 2-3 minutes starts to drop towards -0,8mV
on switch on however after initial fluctuations the voltage goes to -95mV and after 2-3 minutes starts to drop towards -0,8mV
Before going into details, this is the first quick check:
Take two ca 50k resistors preferably pricisely matched and connect each of them resp to Vreg+ and Vreg-.
Then join them together in the middle.
Now switch on the amp as before and measure the voltage where the resistors meet for the same amount of time that you looked at the changing LS output.
Hans
Take two ca 50k resistors preferably pricisely matched and connect each of them resp to Vreg+ and Vreg-.
Then join them together in the middle.
Now switch on the amp as before and measure the voltage where the resistors meet for the same amount of time that you looked at the changing LS output.
Hans
(170V)^2/100K = 0.28Watt.
So 1/8 watt is a bit meagre, but the input impedance of your multimeter is something between 1Meg and 10Meg, so that's why I have chosen 2 x 50K, giving an output impedance of 25K.
With a 1Meg Multimeter this already gives a 2.5% error in the measured voltage.
When your multimeter has a higher impedance, these 2 resistors may have a higher value and a lower wattage, but I wouldn't go for more than 2x100K.
Hans
So 1/8 watt is a bit meagre, but the input impedance of your multimeter is something between 1Meg and 10Meg, so that's why I have chosen 2 x 50K, giving an output impedance of 25K.
With a 1Meg Multimeter this already gives a 2.5% error in the measured voltage.
When your multimeter has a higher impedance, these 2 resistors may have a higher value and a lower wattage, but I wouldn't go for more than 2x100K.
Hans
Hi Hans, I did the test.
After a few initial fluctuations the voltage between the two 50k resistors sits at about 500mV and immediately starts to slowly decrease. In a couple of minutes it reaches -10mV, at this point LS output is 0V. After that the voltage between the two 50k resistors continues to decrease slowly until about -35mV and then slowly rises again, but I didn't observe further (I observed until it reaches -25mV).
p.s. I didn't have 50k resistors and used parallel 100k resistors. My meter measured as 48.6kOhms for both resistor couples, so for the precision and resolution of my meter is the closest I could match them.
After a few initial fluctuations the voltage between the two 50k resistors sits at about 500mV and immediately starts to slowly decrease. In a couple of minutes it reaches -10mV, at this point LS output is 0V. After that the voltage between the two 50k resistors continues to decrease slowly until about -35mV and then slowly rises again, but I didn't observe further (I observed until it reaches -25mV).
p.s. I didn't have 50k resistors and used parallel 100k resistors. My meter measured as 48.6kOhms for both resistor couples, so for the precision and resolution of my meter is the closest I could match them.
Last edited:
You don’t have two multimeters, have you, or even beter a 2 channel scope.
It would be nice to have this signal plus the LS signal at the same time.
The initial 500mV seems strange, almost as if the supply is oscillating shortly.
Can you do this same measurement on the other channel as a reference before we continue ?
Hans
It would be nice to have this signal plus the LS signal at the same time.
The initial 500mV seems strange, almost as if the supply is oscillating shortly.
Can you do this same measurement on the other channel as a reference before we continue ?
Hans
Yes I can try to do that, although my second meter is not reliable but for voltage measurements it should be ok. I have a two channel scope but only one probe. I will also do the test on the other channel.
The thing looks very strange to me. Yesterday after switch off the LS output quickly fell to 0V, today is behaving like the other day, fluctuating more and for longer and taking a few minutes to go to 0V.
One more strange thing I noticed: both transformers always used to buzz. I assumed it is DC in the polluted AC line. Disassembling and testing the amp I noticed the buzzing is dependent on the current draw (i.e. with bias set to 0 there's not buzzing), but I think this is normal. What is strange is that since I reassembled the amp the buzzing coming from one of the transformers (I believe the one for this weird acting channel), sometimes gets intermittent brief interruptions.
The thing looks very strange to me. Yesterday after switch off the LS output quickly fell to 0V, today is behaving like the other day, fluctuating more and for longer and taking a few minutes to go to 0V.
One more strange thing I noticed: both transformers always used to buzz. I assumed it is DC in the polluted AC line. Disassembling and testing the amp I noticed the buzzing is dependent on the current draw (i.e. with bias set to 0 there's not buzzing), but I think this is normal. What is strange is that since I reassembled the amp the buzzing coming from one of the transformers (I believe the one for this weird acting channel), sometimes gets intermittent brief interruptions.
I repeated the test on the other channel. The behavior is similar but the voltage sits initially at around 200mV rather than 500mV. However this difference could also depend on the amp not being completely cold. The LS output goes relatively quickly to 0,8mV but there is no relation between its decrease rate and the voltage between the 50k resistors, as this remains a 200mV for much longer after the output goes to 0V. Also weird is today the output after switch off remains to 40mV for quite a long time. By the way, I'm taking this measurements with no load on the output terminals, is this ok? I don't have dummy load resistors, but if necessary and safe I will hook up the speakers.
I will repeat the measurements on the left channel, but with the two meters to see if there seem to be any relation between the two voltages.
Anything you would like to check with the scope?
I will repeat the measurements on the left channel, but with the two meters to see if there seem to be any relation between the two voltages.
Anything you would like to check with the scope?
Repeated the test on the left channel (the one acting strange) with both meters. The voltage between the 50k resistors again initially was 500mV and LS output -90mV. The voltage between the resistors immediately starts to slowly decrease, by the time it reaches 250mV (approximately 2 minutes), the LS output starts to quickly decrease. Output gets to about 5 mV in less than two minutes, at this point the voltage between the two resistors is 100mV.