I finally made use of your link and discovered the power supply is more complex than I imagined. I believe they sag far more severely than they should. From 1st post: Rega Mira 2000
Would you monitor and advise the drops in supply voltages you nudge the amp into the problem state. i.e. +UNREG, -UNREG, +PWR, -PWR. If they appear to fall in lockstep, would you measure ACV on secondary of transformer. (J4 and J5). Also be alert to suspicious drops rectifier diodes, D1 --- D4 and diode bridge BR1.
Would you monitor and advise the drops in supply voltages you nudge the amp into the problem state. i.e. +UNREG, -UNREG, +PWR, -PWR. If they appear to fall in lockstep, would you measure ACV on secondary of transformer. (J4 and J5). Also be alert to suspicious drops rectifier diodes, D1 --- D4 and diode bridge BR1.
In #8 the output voltage is +280mV and stuck, the loopcorrection cannot pull it to near 0mV. Yet the base of TR33 is at (almost) the right setting.
As if the loop or feedback path is broken (0.3Vdc openloop offset?), TR33 not 'noticing' the output offset.
In #20 soft shorting of the supply rails, something is drawing serious current. It is not surprising now that the feedback cannot correct the output offset.
An output offset of 0.3V in this unexpected condition is rather good! It resembles a startup/shutdown sequence - dc protection should kick in.
Check R42/R32, C56/C43, C53/C57 also.
Both +PWR/-PWR and +Unreg/-Unreg are affected, from two different rectifiers. It is a power supply for both channels, both channels should show the same problems.
Agreed with #22.
I'm afraid I have to confess a rookie mistake 😳. The previous measurements were done with a dim bulb tester (75W bulb) and apparently that was the cause for the extreme power rails voltage sag. With additional bulbs switched on (total 225W) the power rails sag is looking much better (45.0V to 41.4V). Sorry for the misinfo, that was rather embarrassing.
The DC offset problem still remains to be solved however, see below for the new measurements.
Green: volume at zero (<10mV dc offset in both channels).
Red: fault state (~100mV dc offset in right channel).
One thing that stood out is the OL readings at the TR13/14 and TR15/17 emitters.
I've investigated a bit further: when increasing volume, the multimeter in manual 300mV range shows a quickly rising voltage until it overloads. In the manual 3V range however, the multimeter shows a constant 0.007V. Wondering if there is an explanation for the different readings in 300mV and 3V ranges and if this could point to a potential fault.
The DC offset problem still remains to be solved however, see below for the new measurements.
Green: volume at zero (<10mV dc offset in both channels).
Red: fault state (~100mV dc offset in right channel).
One thing that stood out is the OL readings at the TR13/14 and TR15/17 emitters.
I've investigated a bit further: when increasing volume, the multimeter in manual 300mV range shows a quickly rising voltage until it overloads. In the manual 3V range however, the multimeter shows a constant 0.007V. Wondering if there is an explanation for the different readings in 300mV and 3V ranges and if this could point to a potential fault.
Better confess an hiccup then hide a sin.
What kind of DMM do you use? Everything looks ok sofar - in another topic it appeared as if the probe was not pried deeply enough into the track for proper contact.
A Gossen Metrahit 24S. I used a mini grabber to hook onto the transistor lead so I'm sure there was proper contact (multiple measurements with the same result).
Curious to note that the OL readings are around the protection circuits (both channels), whereas other readings looks fine (output stages, input differential, even the bias circuits).
What happens on the bases of TR33 and TR26? Is anything heating up?
All measurements are ok, except the OL's.
What happens on the bases of TR33 and TR26? Is anything heating up?
All measurements are ok, except the OL's.
Taking another stab at repairing the Rega Mira...
I took some more measurements and noticed that the bias voltages increase significantly when turning up the volume (at zero volume the bias voltages across R5-R6 and R7-R8 are set to 7mV). As a result the amp is drawing a lot of current when raising the volume, as indicated by the dim bulb tester starting to glow. So it seems that there is a problem in the bias circuit and the DC offset problem is just a symptom of this.
However, what I don't understand is that the problem manifests itself in both the left and right channels. The power supply seems to be the only generic circuit that affects both channels but I cannot find any fault in the power supply and the rail voltages and ripple seem to be fine. There is a 3-4V voltage sag on both power rails on higher volume but I assume this is due to the amp drawing high current at that point.
Any tips on how to proceed?
I took some more measurements and noticed that the bias voltages increase significantly when turning up the volume (at zero volume the bias voltages across R5-R6 and R7-R8 are set to 7mV). As a result the amp is drawing a lot of current when raising the volume, as indicated by the dim bulb tester starting to glow. So it seems that there is a problem in the bias circuit and the DC offset problem is just a symptom of this.
However, what I don't understand is that the problem manifests itself in both the left and right channels. The power supply seems to be the only generic circuit that affects both channels but I cannot find any fault in the power supply and the rail voltages and ripple seem to be fine. There is a 3-4V voltage sag on both power rails on higher volume but I assume this is due to the amp drawing high current at that point.
Any tips on how to proceed?
Last edited:
The volume pot is AC coupled but does affect the DC setting of the amp?
C59/C49 to replace first. C48/C36 are the next candidates.
All electrolytic caps have already been replaced and double checked for polarity.
I also replaced C49/59 but to no effect.
I also replaced C49/59 but to no effect.
Hi. Did you tried to measure R84 /R64? If they are faulty , input capacitors can't charge and just passing any DC that is present after volume control. I would desolder completely input capacitors C49/C59 and perform voltage checks without them. Also , how much dc is passing to volume control potentiometer from preamplifier?
Also your symptoms looks known to me , had something like low frequency oscillation in a early design , maybe ground track burned somewhere ?
Also your symptoms looks known to me , had something like low frequency oscillation in a early design , maybe ground track burned somewhere ?
Do you have a scope? Check for HF oscillation using x10 scope probe.
A wild guess is HF content in earlier stage; advancing volume eventually generates overload rectification presenting itself as offset error. If you don’t have a scope, as a test try installing a a 0.1uF cap in parallel with C54. If there’s improved immunity to volume control rotation, that would be evidence of earlier stage problem.
A wild guess is HF content in earlier stage; advancing volume eventually generates overload rectification presenting itself as offset error. If you don’t have a scope, as a test try installing a a 0.1uF cap in parallel with C54. If there’s improved immunity to volume control rotation, that would be evidence of earlier stage problem.
If there are shorts on C49 and C59 (L+R power amp inputs) to GND and offset voltage is within 20-30mV, then an unwanted oscillation is present as mentioned in previous post.
If that's correct, I don't assume a faulty part, but an error in the design of the circuit or the conductor path routing.
I hope you've identified the problem if not yet the remedy. But be alert to the possibility you're seeing local radio transmissions. 100MHz is pretty fast for an internally generated oscillation. Try to confirm the signal originates within the circuit, not with external fields.
Good luck.
Good luck.
Without a regular input signal, take a 100NF cap (>100V), connect one leg to ground (wire ~ 30cm^), and use the the other leg to 'short' the oscillation on several places. When it stops, there is the problem.
@All: this just pops up in my mind, but I've never done/used this method - are there second thoughts?
@All: this just pops up in my mind, but I've never done/used this method - are there second thoughts?
@Citizen124032
Probing with a bypass cap to ground could work, but my misgiving is that a cap to ground can also provoke oscillation in otherwise stable circuit, eg. cap loading on an opamp output can induce oscillation.
Assuming that spurious oscillation in a stage preceding the power amp is proven to be the underlying problem, I would poke with a x10 probe to try locating the offending circuit. With luck, an appropriate remedy will be evident.
Best,
Steve
Probing with a bypass cap to ground could work, but my misgiving is that a cap to ground can also provoke oscillation in otherwise stable circuit, eg. cap loading on an opamp output can induce oscillation.
Assuming that spurious oscillation in a stage preceding the power amp is proven to be the underlying problem, I would poke with a x10 probe to try locating the offending circuit. With luck, an appropriate remedy will be evident.
Best,
Steve
Using a capacitor to ground is possible if signal path is separated by resistor, like input filter of power amplifier. If you by incident touch ground a fedback input , heavy oscillation will occur , so better be careful.