Well, you haven't explained why the negative power supply rail is lower in the fault condition. Looks to me like a PSU issue. Maybe the main filter caps need replacing? Maybe try to solder some temporary caps across the existing ones to see if it makes any difference?
He uses dim bulb tester in series with mains, also probably a resistor as load. So with dc offset and load present ,corresponding polarity supply is loaded more , and it voltage sags more. Op should look with scope what is coming to vokume control , and where oscillation present at first.
There is voltage sag on both power rails. The 44.7V on +UNREG in the lower picture in post #25 was a typo (actual value was 41.1V). This is when testing with a 200W lightbulb, with a 60W lightbulb the voltage sag is much larger.
The filter caps in the PSU have already been replaced.
I used a scope to have another look at the ripple voltages, see below.
+UNREG:
-UNREG:
-PWR:
+PWR:
+PWR (with volume at ~9 o'clock)
The filter caps in the PSU have already been replaced.
I used a scope to have another look at the ripple voltages, see below.
+UNREG:
-UNREG:
-PWR:
+PWR:
+PWR (with volume at ~9 o'clock)
I also checked the 1KHz input signal at various points in the pre and power amp circuits.
R107 (incoming signal in preamp):
R118 (preamp)
TR29 (power amp input)
R107 (incoming signal in preamp):
R118 (preamp)
TR29 (power amp input)
However, the 100MHz noise is visible even when feeding the signal into the Power Amp inputs and disconnecting the preamp and poweramp circuits using the dip switch on the main circuit board.
Given that the amp is drawing high current when bypassing the preamp, I guess this means the fault is either in the power amp circuit or in the power supply.
The ripple voltages in the power supply are ~300-400 mV pp for UNREG and 50-80 mV pp for the regulated rails (increasing with higher volume settings). Would that be within the expected range for solid state amplifiers?
Given that the amp is drawing high current when bypassing the preamp, I guess this means the fault is either in the power amp circuit or in the power supply.
The ripple voltages in the power supply are ~300-400 mV pp for UNREG and 50-80 mV pp for the regulated rails (increasing with higher volume settings). Would that be within the expected range for solid state amplifiers?
An early complaint was that current rose with increasing volume control rotation. Is this still true?
I think ripple voltage is in normal range , more interesting is where this RF signal coming from.
Volume control is connected at power amplifier input ,or somehow else ?
Volume control is connected at power amplifier input ,or somehow else ?
I believe volume control is part of the preamp circuit. When using the power amp inputs, volume control is bypassed.
Is there strong RF at your location? If you touch scope probe with fingertip, do you see RF on scope?
I'm trying to reconcile behavior you reported in earlier posts--- was ok until Riga volume control was raised.
You can tell I'm having trouble understanding the test setup/environmet. 😕 Thanks for details.
I'm trying to reconcile behavior you reported in earlier posts--- was ok until Riga volume control was raised.
"Increasing volume" is increased amplitude from an external signal generator? Is the external signal path well shielded?
You can tell I'm having trouble understanding the test setup/environmet. 😕 Thanks for details.
No worries, I should have provided better context...
#1: Integrated amplifier test
Results:
At zero volume setting all looks fine (10mV dc offset, 2.5-7.0mV bias voltage, low current draw)
At 8 o'clock volume setting: increasing bias current (both channels)
At 9 'clock volume setting: increasing dc offset (right channel first, then left channel)
#2: Power amplifier test (bypassing the preamp and volume control)
Results:
At zero volume setting all looks fine (10mV dc offset, 2.5-7.0mV bias voltage, low current draw)
With increased test signal amplitude bias current and dc offset quickly increase, large current draw
Hmm yes I think so. I just realized there is a radio transmission tower not far from here although not 100% sure it's still in use. However, there is no clear HF visible on the scope when I touch the probe. Also, the test signal at the beginning of the preamp stage looks fine, no significant HF noise.
As the Mira can be used as an integrated amp, preamp or power amp I've tried 2 different setups:
#1: Integrated amplifier test
- Fixed test signal (1KHz sine wave, 1V RMS from a Leader LAG27 audio generator) into AUX 1. Shielded RCA cables.
- Dim bulb tester with 60W bulb
- Pre-Power Link Switch (SW1) switched ON
- Starting with volume knob at zero, slowly raising to ~10 o' clock.
Results:
At zero volume setting all looks fine (10mV dc offset, 2.5-7.0mV bias voltage, low current draw)
At 8 o'clock volume setting: increasing bias current (both channels)
At 9 'clock volume setting: increasing dc offset (right channel first, then left channel)
#2: Power amplifier test (bypassing the preamp and volume control)
- Variable test signal (1KHz sine wave, 0-1V RMS from a Leader LAG27 audio generator) into POWER AMP IN. Shielded RCA cables.
- Dim bulb tester with 60W bulb
- Pre-Power Link Switch (SW1) switched OFF
- Volume control bypassed.
- Start with zero test signal amplitude from audio generator, slowly increasing.
Results:
At zero volume setting all looks fine (10mV dc offset, 2.5-7.0mV bias voltage, low current draw)
With increased test signal amplitude bias current and dc offset quickly increase, large current draw
Looks like amplifier oscillation then, maybe transistors replaced by slower types, causing stability problem ?
Bias current should be set with no signal applied; in this adjustment, you're setting up a relatively small class A current that exists in only the crossover region, operative before the amp transitions into class B. It would be normal for the voltage across the emitter resistors to grow with increasing signal amplitude.
That said, shift in DC offset with increased signal is concerning. It could be evidence of 2nd harmonic distortion. It could be associated with onset of an oscillation when drive is increased. But it might be an artifact of the measurement if the voltmeter doesn't adequately ignore the presence of large AC while trying to measure small DC offset voltage. Maybe this issue is the explanation for the change in measurement behavior you mentioned when changing your DVM from 300mV range to 3V? To test this possibility, use a low pass filter in front of your DVM (10K and 100nF film cap). You should see very minimal change with increasing 1kHz sine wave.
Lastly, with no signal applied, check for change in amp offset with the amp's internal volume control. There should be negligible shift. I got into the weeds thinking that this issue was appearing in the absence of applied signal. My bad.
That said, shift in DC offset with increased signal is concerning. It could be evidence of 2nd harmonic distortion. It could be associated with onset of an oscillation when drive is increased. But it might be an artifact of the measurement if the voltmeter doesn't adequately ignore the presence of large AC while trying to measure small DC offset voltage. Maybe this issue is the explanation for the change in measurement behavior you mentioned when changing your DVM from 300mV range to 3V? To test this possibility, use a low pass filter in front of your DVM (10K and 100nF film cap). You should see very minimal change with increasing 1kHz sine wave.
Lastly, with no signal applied, check for change in amp offset with the amp's internal volume control. There should be negligible shift. I got into the weeds thinking that this issue was appearing in the absence of applied signal. My bad.
Measure the ohmic ground connections troughout the amp with respect to the ground link. A bad ground might explain the change in measurement behaviour. Check again very toroughtly for bad solderings. Just some thoughts 🙂
Hugo
Hugo
Last edited:
Yes, I know that bias current should be set with no signal. But is it normal that bias voltages increase from ~2.5mV to 50-100mV with signal applied and volume turned up to a medium level?
One thing to mention about bias measurement. If you measure bias voltage across just one of emitter resistor, and then apply a signal and load , voltage will increase. But if you measure double bias voltage, between emitters , it will not increase with signal. If amplifier amplifies symmetrically, ac currents do cancel and you get just dc voltage . If about post #8 posted schematic ,it would be R7-R8 voltage.
So ,if you measured double bias , and you get heavy voltage increase when increasing signal amplitude , that may mean three possibilities:
1.Amplfier not amplify both halfwaves, maybe clipping early positive or negative.
2. Oscillation with any signal, then you may get anything and hard to predict what voltages would be.
3. Bias circuit not functioning, in example bad transistor, and bias voltage increases with signal.
Try to check amplifier output with scope, without any load, to exclude current draw by load. Also try to short bias circuit or shunt with three - four diodes in series and check if anything changes.
1.Amplfier not amplify both halfwaves, maybe clipping early positive or negative.
2. Oscillation with any signal, then you may get anything and hard to predict what voltages would be.
3. Bias circuit not functioning, in example bad transistor, and bias voltage increases with signal.
Try to check amplifier output with scope, without any load, to exclude current draw by load. Also try to short bias circuit or shunt with three - four diodes in series and check if anything changes.