All the resistors check out, though my meter wasn't good enough to get a reading on the 0R22 ohm resistors. However I doubt all four would go faulty, as the mV reading is the same per channel at approx. 17mV, raising to 18mV as the amp warms up.
I guess the next thing is to check exactly what voltages are getting into the power amp section. Measured the SMPS output and it was 97.7-98V DC. The manual says to set for 98-99V so perhaps it would help if I raised it slightly to 99V instead. Really want to get this sorted. The service manual mentions 22mV many times throughout so it definitely isn't a typo in this instance.
I guess the next thing is to check exactly what voltages are getting into the power amp section. Measured the SMPS output and it was 97.7-98V DC. The manual says to set for 98-99V so perhaps it would help if I raised it slightly to 99V instead. Really want to get this sorted. The service manual mentions 22mV many times throughout so it definitely isn't a typo in this instance.
The PSU voltage should have no effect even if were right down.
I can't really suggest much because the bias generator is internal in the IC (and they won't be faulty). Can't even hazard a guess at what the min-max range would be.
Can we be certain your meter is accurate on low millivolt readings ? Just trying to cover all bases. Perhaps rig up a simple divider across a 9 volt battery. 47k in series with 100 ohms will give around 19mv across the 100 ohm.
The other way is to remove one of the 0.22 ohms and replace with a meter set to DC current and confirm then reading. Much care needed though and good connections required... no holding probes with a chance of intermitent contact or slipping
Are those four zeners original in value D304 and 305 etc. Looks as though there should be around 22 volts across each going off the emitter voltages of the transistors they connect to but you need to be ever so careful measuring... one slip and all.
Did you mention some other forums right at the start... maybe ask on those where some have rebuilt them if the current adjusted OK to value given .
I'll keep thinking...
I can't really suggest much because the bias generator is internal in the IC (and they won't be faulty). Can't even hazard a guess at what the min-max range would be.
Can we be certain your meter is accurate on low millivolt readings ? Just trying to cover all bases. Perhaps rig up a simple divider across a 9 volt battery. 47k in series with 100 ohms will give around 19mv across the 100 ohm.
The other way is to remove one of the 0.22 ohms and replace with a meter set to DC current and confirm then reading. Much care needed though and good connections required... no holding probes with a chance of intermitent contact or slipping
Are those four zeners original in value D304 and 305 etc. Looks as though there should be around 22 volts across each going off the emitter voltages of the transistors they connect to but you need to be ever so careful measuring... one slip and all.
Did you mention some other forums right at the start... maybe ask on those where some have rebuilt them if the current adjusted OK to value given .
I'll keep thinking...
Pretty sure the meter is fine as I've used it to set bias accurately on other amps before with no problems.
Those four zeners are original, but due to the way they're mounted - one end butted right up to the body on the PCB, the only other visible leg bent over - it makes taking a reading impossible. I guess I could flip the whole power amp section over and take a reading from underneath though without doing any damage (touch wood!).
Wish muggins here had taken a bias reading before I replaced all the caps and those trimmers... foollll!
Those four zeners are original, but due to the way they're mounted - one end butted right up to the body on the PCB, the only other visible leg bent over - it makes taking a reading impossible. I guess I could flip the whole power amp section over and take a reading from underneath though without doing any damage (touch wood!).
Wish muggins here had taken a bias reading before I replaced all the caps and those trimmers... foollll!
I sadly can't remember back to when I did my amp up
Ah many thanks for that offer - much appreciated indeed! ONLY if it's no hassle for you though.
Cheers!
Here's a better quality scan of the power amp circuit, plus the conductor side of the PCB:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Hi John. It's what's internal within the IC (the actual circuitry as if it were discrete) that hold the clue as to exactly how the bias pot operates. It's an unusual arrangement using the transistors marked "cascode amp". The base of these are held at a fixed voltage by the zeners and they are controlled by altering the conditions on the emitter.
It would be a help to know what the adjustment range is like on another amp.
It would be a help to know what the adjustment range is like on another amp.
How do!
Hopefully 'thefragger' might be able to shed some light on the adjustment range with his Sony.
However being as all the passive components appear to be fine perhaps it's something as simple as the new adjustment pots not having quite as much range as the originals. They're 100K vishay types:
VISHAY SFERNICE|T7YA104MB40|TRIMMER, 100K | Farnell United Kingdom
Turned fully clockwise I can just hit 18mV (after a bit of warming up) on both channels. Perhaps with just a smidgen more adjustment range I'd hit the 22mV - they're only 20% tolerance. Perhaps the original ALPS trimmer pots were more tightly toleranced? I still have them so perhaps I should reinstall them...
Hopefully 'thefragger' might be able to shed some light on the adjustment range with his Sony.
However being as all the passive components appear to be fine perhaps it's something as simple as the new adjustment pots not having quite as much range as the originals. They're 100K vishay types:
VISHAY SFERNICE|T7YA104MB40|TRIMMER, 100K | Farnell United Kingdom
Turned fully clockwise I can just hit 18mV (after a bit of warming up) on both channels. Perhaps with just a smidgen more adjustment range I'd hit the 22mV - they're only 20% tolerance. Perhaps the original ALPS trimmer pots were more tightly toleranced? I still have them so perhaps I should reinstall them...
Last edited:
To do that I take it I'd need to put the negative meter probe onto any part of the PCBs groundplane to get a reading? (sorry - electronics for beginners type question I know!).
Any ground would do, the chassis, even the sockets on the back
Cheers Mooly. Well taken the readings and they're approx. (fluctuate slightly) 6.8V per channel. A little short of the expected 8.8v
I don't know much about amp design, but looking at that circuit on the previous page, do I need to check R306 / R307 / R356 / R357 to make sure they're the correct values & dropping the voltage properly? Or shall I measure to make sure I have the 49 / -49V entering the PCB?
EDIT: Just measured the voltages just before those four resistors mentioned above, and I get a steady 48.8V reading. So with me layman's brain, I guess that points to those four dropping resistors being suspect?
Last edited:
Just re-tested and 6.8v at pin.2 of ICs 301 and 351. At pin 14 I get 8.8v.
Also tested the voltages before/after R306/307/356/357 and all fine so that rules out those resistors as being suspect.
So does this point to the 100k trim pots (RT302 & 352) as not having quite enough range then?
Also tested the voltages before/after R306/307/356/357 and all fine so that rules out those resistors as being suspect.
So does this point to the 100k trim pots (RT302 & 352) as not having quite enough range then?
Last edited:
OK just taken those readings:
Pin 14: 8.7v
Pin 4: -7.1v
Pin 2: -7.1v
Both channels measure the same.
Must admit that the -7.1v sounds a bit low. That voltage is internally derived form the IC... like a zener. If R307 and R340 are OK that feed that rail then I would hazard a guess its within the range of normal. Again reference to another unit would be helpful.
Just clutching at straws but is the voltage at pin 7 actually zero as shown or is the IC trying to drive current through it which would show as a slightly negative voltage at that point. Even that doesn't make sense really as the amp offset as a whole is correct. but it would be a worthwhile check non the less.
Just clutching at straws but is the voltage at pin 7 actually zero as shown or is the IC trying to drive current through it which would show as a slightly negative voltage at that point. Even that doesn't make sense really as the amp offset as a whole is correct. but it would be a worthwhile check non the less.
Voltage at pin 7 on each IC is -25mV and -30mV.
Also took the following readings which might be useful:
Voltages at the B- end of R311 and R361 (27k) are -49.7 and -49.8.
Voltages at the negative terminals of C310/C360 (1000uF) are -10.8 and -11.
Voltages on the 'left side' (when looking at schematic) of R340/R390 are 7.24 and 7.27v.
Also took the following readings which might be useful:
Voltages at the B- end of R311 and R361 (27k) are -49.7 and -49.8.
Voltages at the negative terminals of C310/C360 (1000uF) are -10.8 and -11.
Voltages on the 'left side' (when looking at schematic) of R340/R390 are 7.24 and 7.27v.
Last edited:
Hmmm... so it is slightly negative.
Lunch time... but if you want to try this in the meantime.
With no speakers connected, see if altering the offset pot (the DC balance pot) actually alters the voltage on pin 7. Can you bring pin 7 nearer to zero disregarding that the output will now have a DC error. If you can is the DC bias (the 22mv) any better. Does the -8 volt rail to the IC increase a little.
Lunch time... but if you want to try this in the meantime.
With no speakers connected, see if altering the offset pot (the DC balance pot) actually alters the voltage on pin 7. Can you bring pin 7 nearer to zero disregarding that the output will now have a DC error. If you can is the DC bias (the 22mv) any better. Does the -8 volt rail to the IC increase a little.
Altering the DC offset pots did change the readings at pin 7 but only very slightly i.e. from -25.8mV to -25.3mV
Altering the DC offset pots and checking the 22mV bias test points, again there is a very slight variation - i.e. 17.3mV becomes 17.9mV if I turn the DC offset pot fully clockwise.
So it appears the DC offset pots are VERY slightly affecting the other readings.
So to sum up, the main problem as it stands at present is the fact I cannot get -8.8v at pins 4/13/2. Wondering if R340 and R390 are to blame? Getting approx. -11v before, and 7.2v after them.
Looking at the service manual again it definitely shows an 8-ohm load connected to the speaker terminals when setting DC offset, so perhaps I'll give that a try even though it shouldn't ordinarily make a difference?
Altering the DC offset pots and checking the 22mV bias test points, again there is a very slight variation - i.e. 17.3mV becomes 17.9mV if I turn the DC offset pot fully clockwise.
So it appears the DC offset pots are VERY slightly affecting the other readings.
So to sum up, the main problem as it stands at present is the fact I cannot get -8.8v at pins 4/13/2. Wondering if R340 and R390 are to blame? Getting approx. -11v before, and 7.2v after them.
Looking at the service manual again it definitely shows an 8-ohm load connected to the speaker terminals when setting DC offset, so perhaps I'll give that a try even though it shouldn't ordinarily make a difference?
Last edited: