I have during the last couple of weeks changed transformers in my M2X mono blocks to more silent ones. I had some mechanical noise I did not like. Now I realized that the new also suddenly can start making noise and I found out that while it is making noise the DC offset gets unstable. My theory in the moment is that reason for making noise is a current draw somewhere. When noise is there it goes a bit up and down and I can see DC offset instability follows the noise (more noise more instability) . Now it is silent and everything is stable as it should be and I have made some measurements and will see how they change when transformer noise starts again. I have measured:
Bias: 1.3A (0.62V over R47)
Rail voltage: +-23.9 VDC
Main voltage: 244 VAC
Main current: 0.4 A
I think reason for DC offset being unstable is that rail voltage drops caused by current draw somewhere. First I have to prove that reason for transformer noise etc. is extra current draw.
I have rail fuses = 4A (F-type) and they have not burned until now. It could be some fault in PSU that causes extra current draw. So maybe it is not the amp itself.
If it is the amp it might be a beginning short of Drain to heatsink of one of the IRFP240?
Can that kind of error be so cruel so it is just a partly short and it can come and go?
First I have to see if the first part of the theory is correct.
0.4 x 244 = 98W
1.3 x 48V = 62W
Question is if I burn more than 30W in normal conditions in PSU as heat. Maybe I could measure rail current as it should be almost equal to bias current I guess.......to check if some current is going elsewhere.....
Bias: 1.3A (0.62V over R47)
Rail voltage: +-23.9 VDC
Main voltage: 244 VAC
Main current: 0.4 A
I think reason for DC offset being unstable is that rail voltage drops caused by current draw somewhere. First I have to prove that reason for transformer noise etc. is extra current draw.
I have rail fuses = 4A (F-type) and they have not burned until now. It could be some fault in PSU that causes extra current draw. So maybe it is not the amp itself.
If it is the amp it might be a beginning short of Drain to heatsink of one of the IRFP240?
Can that kind of error be so cruel so it is just a partly short and it can come and go?
First I have to see if the first part of the theory is correct.
0.4 x 244 = 98W
1.3 x 48V = 62W
Question is if I burn more than 30W in normal conditions in PSU as heat. Maybe I could measure rail current as it should be almost equal to bias current I guess.......to check if some current is going elsewhere.....
Rail current draw is about 1.4A in moment which fits a "normal" case. Now I just have to wait until the noise condition comes again and I have some numbers to compare with.
Thank you!
I think transformer mount is ok.
I have some loss in the diodes and a bit in the transformer and NTC I use as current limiter. But loss seems a bit high?
The main voltage and main current is measured by a cheap unit I have put on power connector. It can also show Watts directly. It may be good to check what it says in Watt mode.
When I switch to watt mode it says 75W (voltage 247V and current 0.4A).....so I was fooled by the "angle" between voltage and current.....
So I think amp in the moment when no transformer noise is running as it should and I must wait and see what happens when it makes noise again. Now it is night and I don't know if special conditions on main voltage when people are active can cause transformer noise (I have isolation transformer and filter etc.....so no DC).
So I think amp in the moment when no transformer noise is running as it should and I must wait and see what happens when it makes noise again. Now it is night and I don't know if special conditions on main voltage when people are active can cause transformer noise (I have isolation transformer and filter etc.....so no DC).
Now at this time at day (before noon) everything is stable and DC offset variation is with 1 mV on both amps. What I noticed when the instability was there then on distortion measurement 50 Hz peak raised from like -115 dB to like -100 dB and went op and down in synch with how much transformer noise there was. Now both are dead silent.
During the evening when many cooks and use a lot of power.....if curve form of sinus changes.....can this cause transformer noise.....and then make the DC to amp change a little bit which would affect DC offset?
When DC offset if instable it varies up and down maybe some 50 mV.
During the evening when many cooks and use a lot of power.....if curve form of sinus changes.....can this cause transformer noise.....and then make the DC to amp change a little bit which would affect DC offset?
When DC offset if instable it varies up and down maybe some 50 mV.
man, you're really overthinking it, if you're making a fuss about 50mV, if and when you're sure that everything is done as it needs to be 🙂
anyway, try being more precise and explanatory - is "strange behavior" being same for both monoblocks, what kind of "noise" is that, how long in duration?
be aware of fact that impedances of mosfet gates to respective rails are never the same, simply because Ugs voltages of N and P are not the same ...... if mains voltage jump or sag for, say, 10% - it must result in some DC offset sliding ........ and remember - there is no overall NFB to serve as servo mechanismus, so that's Nature of the Beast
your speaker doesn't care for that - mild changes even up to 1V5 are irrelevant
anyway, try being more precise and explanatory - is "strange behavior" being same for both monoblocks, what kind of "noise" is that, how long in duration?
be aware of fact that impedances of mosfet gates to respective rails are never the same, simply because Ugs voltages of N and P are not the same ...... if mains voltage jump or sag for, say, 10% - it must result in some DC offset sliding ........ and remember - there is no overall NFB to serve as servo mechanismus, so that's Nature of the Beast
your speaker doesn't care for that - mild changes even up to 1V5 are irrelevant
I have seen the strange behavior on both mono blocks but not at same time so far. But there seems to be a correlation between transformers gets noisy to the instability of DC offset. Overall distortion did not change but noise changed a bit. Normally the noise figure is stable about 0.005 % or so. But during this period of instable DC offset there was some variation of noise figure. It is strange that it is not both mono blocks on same time. The max. DC offset I have seen when adjusted to zero in silent period is about 0.1 V for a short period.
Yesterday evening it made troubles for a couple of hours (as I recall only one of the mono blocks) but from 8PM to 1PM they where dead silent and DC offset within 1mV. So a theory is that it is caused by some "bad" things on main voltage. Now I am an analyse period so I will get more input to the problem the coming time and I can make some measurements.
I have played for some hours now....and everything perfect so far......
Yesterday evening it made troubles for a couple of hours (as I recall only one of the mono blocks) but from 8PM to 1PM they where dead silent and DC offset within 1mV. So a theory is that it is caused by some "bad" things on main voltage. Now I am an analyse period so I will get more input to the problem the coming time and I can make some measurements.
I have played for some hours now....and everything perfect so far......
75W is way much better, so let presume that this mode of measurement is closer to truth than previous one
so, considering that amps are working for some time, and taking 75W figure, nothing is wrong with your amps
look for culprit on mains side - is isolation xformer really isolating and what else you have connected to it
leave just amps on isolating xformer and observe; be aware of fact that isolating xformer is good for several other things, but not exactly intended for DC isolating, if same is present in your mains - it'll not transfer DC on secondary side obviously, but it can prevent blocking transients resulting from DC changes in mains ....
maybe is time to make cheapest possible DC blocker and put it in front of your isolation xformer, for test and see what you can get
if that solves or decrease your problems, then make it properly
take care of all safety precautions, of course , even if just testing
There is a noise filter in front of the isolation transformer and I assume this filter has a DC blocker. Now this filter is quite old so maybe caps in it are not good anymore......who knows....
When noise arise again I will have a look at the waveform on scope to check. I have a TEK P5100 x100 probe which I will use.....but it will be with banana plugs out of the outlets 🙂 ....or I may build something more safe for this purpose. Something with fixed contact points good for a probe.
On one of the mono blocks I adjusted DC offset with +-23.9 VDC rail. Now I checked rails and they showed -+23.0 and DC offset was 90 mV.
So it seems a change of 1V rails changes offset 90 mV or so.....does that make sense?
But this is not my instability problem......it is just an observation.....and I have written it down to remember.
Now with both mono blocks running + Whammy + streamer the total consumption is about 172 Watts. When transformer noise occurs I will check wattage also.
When noise arise again I will have a look at the waveform on scope to check. I have a TEK P5100 x100 probe which I will use.....but it will be with banana plugs out of the outlets 🙂 ....or I may build something more safe for this purpose. Something with fixed contact points good for a probe.
On one of the mono blocks I adjusted DC offset with +-23.9 VDC rail. Now I checked rails and they showed -+23.0 and DC offset was 90 mV.
So it seems a change of 1V rails changes offset 90 mV or so.....does that make sense?
But this is not my instability problem......it is just an observation.....and I have written it down to remember.
Now with both mono blocks running + Whammy + streamer the total consumption is about 172 Watts. When transformer noise occurs I will check wattage also.
You may have intermittent oscillations. Download the sinewave test signals (*.wav) and then load the amplifiers with 8ohms / 50W or-so resistors, just make sure that the speaker cables are "included", i.e. place the resistors at the end of the speaker cables. Then, use that oscilloscope to check for oscillations at the amps' binding posts.
The DC offset variations do seem excessive, I do not have M2X... maybe someone else who owns it can chip in.
The DC offset variations do seem excessive, I do not have M2X... maybe someone else who owns it can chip in.
My M2 made funny noises and had unstable bias intermittently when I first built it. First noticed it when I moved it to a different room.
Turned out one of the wires coming from the power supply was in the through hole on power supply PCB, but I had failed to solder it. 😱
Russellc
for me, M2 is known territory
other iterations are not, speaking of possible oscillations
though , DC offset is matter of unchanged part of circuit, OS biasing part
other iterations are not, speaking of possible oscillations
though , DC offset is matter of unchanged part of circuit, OS biasing part
I am sure it is not oscillation as I had REW on while the problem was there and harmonic distortion was as low as when problem was not there. But offset jumped a bit up and down and 50 Hz peak larger and more noise from 20 to 50 Hz. I can also play nice music when the problem is there. But I don't like if I can hear the transformer in listening position. I have also checked with oscilloscope.
Now problem is there this evening and I measured sine at mains (not output from isolation transformer.....I can do that later). Mains voltage a bit higher than this morning (237 VAC vs. 233 VAC). Wattage about same and bias ok.
This is how sine looked like (AC coupling and DC coupling). AC coupling is the sine where there is "uphill" at the top.
Is that a lot distortion and enough to cause transformer noise and unstable DC offset?
I will check sine when transformer is silent to see if it looks different.
This is how sine looked like (AC coupling and DC coupling). AC coupling is the sine where there is "uphill" at the top.
Is that a lot distortion and enough to cause transformer noise and unstable DC offset?
I will check sine when transformer is silent to see if it looks different.
Attachments
