Ok now I get it. That's pretty neat.
I have already be caught lurking around the flat and pulling out table lamps as possible adhoc bulb testers.
60 or 100 should be good.
Do not connect speakers to begin with when using a bulb. See how it all behaves first. Bulb will flash brightly and should then go dim/out after a second or two.
Even turning the bias current up will be enough to light the bulb somewhat as the amp draws more power.
Once you know all is good with a bulb then the final bias adjustment is done on full mains again.
Playing music with a bulb might cause problems because the rails will suddenly collapse as you turn the volume up and that might cause unpredictable behaviour... or the relay might drop out 😉 but very low volume is fine.
The bulb behaves as a serious non linear limiter... that is what it is meant to do.
Do not connect speakers to begin with when using a bulb. See how it all behaves first. Bulb will flash brightly and should then go dim/out after a second or two.
Even turning the bias current up will be enough to light the bulb somewhat as the amp draws more power.
Once you know all is good with a bulb then the final bias adjustment is done on full mains again.
Playing music with a bulb might cause problems because the rails will suddenly collapse as you turn the volume up and that might cause unpredictable behaviour... or the relay might drop out 😉 but very low volume is fine.
The bulb behaves as a serious non linear limiter... that is what it is meant to do.
60watts is what I found. The last one. Plus a suitable lamp (don't tell my wife).
I need to finish some writing now (deadline), so I'll check in later today. As I see it, with the quick-DBT in place I could put in the second diode D302, unsolder the short of the right-channel preset, and dial in both bias presets (starting with minimum resistance on both).
Right? And if all is ok, then hook it to speakers? I had it running yesterday for 20 mins on headphones as it is now.
Then I could do the D303/353 diodes – and the new supply cans – in the next go when the parts have arrived.
I need to finish some writing now (deadline), so I'll check in later today. As I see it, with the quick-DBT in place I could put in the second diode D302, unsolder the short of the right-channel preset, and dial in both bias presets (starting with minimum resistance on both).
Right? And if all is ok, then hook it to speakers? I had it running yesterday for 20 mins on headphones as it is now.
Then I could do the D303/353 diodes – and the new supply cans – in the next go when the parts have arrived.
60 watts should be OK to get it running. As the amp is working now, just add the bulb tester and check its happy with that in place.
If it is OK you can swap the other diode having turned the bias down first.
------------------------------
If you short out the other diodes (D303/353) then there is no reason not to try the amp for real on full mains and with speakers. With no bias it can never go into thermal runaway. Providing everything else is good the amp should perform absolutely normally.
-----------------------------
Just for info. Try measuring the resistance of the cold bulb.
Now watts (its a 60 watt bulb remember) equals voltage squared divided by R. So what do you get 😀 A whole lot more than 60 watts.
When the filament is hot the resistance rises dramatically and so the current falls away. That is why the bulb works so well as a protective device.
I'll look in later, good luck 🙂
If it is OK you can swap the other diode having turned the bias down first.
------------------------------
If you short out the other diodes (D303/353) then there is no reason not to try the amp for real on full mains and with speakers. With no bias it can never go into thermal runaway. Providing everything else is good the amp should perform absolutely normally.
-----------------------------
Just for info. Try measuring the resistance of the cold bulb.
Now watts (its a 60 watt bulb remember) equals voltage squared divided by R. So what do you get 😀 A whole lot more than 60 watts.
When the filament is hot the resistance rises dramatically and so the current falls away. That is why the bulb works so well as a protective device.
I'll look in later, good luck 🙂
ok! I love these infos, learning along!
my son and I were guessing how it worked.
so you'd suggest to short out these diodes (D303/503) from the get-go? As they might just fail under my hands? Then I'll better do that.
Thanks!
Yes, although perhaps on balance just do as you have on the other channel and short the bias pot instead.
Essentially the same but it leaves R314 and R315 still doing their thing. We know shorting the preset didn't cause any issues. I'm just trying to cover every possibility of things that could go wrong and those resistors just might possible influence stability (unwanted oscillation).
So thinking a bit more, short the preset as you have done already.
Essentially the same but it leaves R314 and R315 still doing their thing. We know shorting the preset didn't cause any issues. I'm just trying to cover every possibility of things that could go wrong and those resistors just might possible influence stability (unwanted oscillation).
So thinking a bit more, short the preset as you have done already.
So:
1. short other preset
2. put in substitute for D302
3. power up on DBT
4. if ok, try with speakers
and then later
5. dial in right preset when the rubber diodes are fitted
so
6. substitute old presets by a 1k resistor?
I hope i did not make any confusion.
1. short other preset
2. put in substitute for D302
3. power up on DBT
4. if ok, try with speakers
and then later
5. dial in right preset when the rubber diodes are fitted
so
6. substitute old presets by a 1k resistor?
I hope i did not make any confusion.
1/ Yes.
2/ Yes
3/ Yes
4/ Now check there is no DC offset on that channel. Assuming the speaker relay closes normally you can measure the offset at the sockets.
5/ If you try with speakers and a DBT then don't turn it up loud. The light will start to illuminate and the rails will collaspe, maybe unpredicatably.
6/ If everything above is OK then you can run it on full mains at any volume.
7/ Leave the old presets fitted. If you are doing the official Sony mod then they will be needed to set the bias when its all done. If you have doubts over their condition then they should be replaced.
2/ Yes
3/ Yes
4/ Now check there is no DC offset on that channel. Assuming the speaker relay closes normally you can measure the offset at the sockets.
5/ If you try with speakers and a DBT then don't turn it up loud. The light will start to illuminate and the rails will collaspe, maybe unpredicatably.
6/ If everything above is OK then you can run it on full mains at any volume.
7/ Leave the old presets fitted. If you are doing the official Sony mod then they will be needed to set the bias when its all done. If you have doubts over their condition then they should be replaced.
ok great. good list!
I will then - after completion of the procedure – try with speakers which are fully protected as they have caps in series with both tweeter and woofer, so no DBT.
If I'll leave in and use the old preset pots, why is it better to short them out now instead of dialling in the right bias from the get-go? (They look good, quality parts).
Shorting the preset on the yet to be worked on channel just gives a bit extra safeguard in case anything went wrong with the replacement diodes you are going to fit.
Also, unless you could be 100% certain that the biasing diodes D303 and D353 are good and are not going to fail intermittently then we may as well short the preset and remove that ever present danger.
Also, unless you could be 100% certain that the biasing diodes D303 and D353 are good and are not going to fail intermittently then we may as well short the preset and remove that ever present danger.
ok thanks got it!
I'll report back. Need to get on writing now and will hopefully manage to get back at the amp afterwards.
(and there's the cat to be walked as well).
I'll report back. Need to get on writing now and will hopefully manage to get back at the amp afterwards.
(and there's the cat to be walked as well).
I am at point 4) of your list now, checking offset (DBT in place). Offset is drifting wildly on both channels, left (good one) went up to 200 mV over 5 minutes, right (repaired one) was at -20 mV, now I am measuring again and it started at 30 mV, going down now.
Now I am back at the bottom of my my learning curve!
Now I am back at the bottom of my my learning curve!
After 10 more minutes it seems to settle at -10 mV (right) and -110 mV (left). so the repaired channel seems better than the old channel. I had also replaced the input differential there, which was a SONY double BJT in a 6 pin casing.
Mooly gave very detail instructions and steps to help you to repair the amp.
Now you learn how to repair an amp and it is great fun to do that.
Best regards.
Patrick
Now you learn how to repair an amp and it is great fun to do that.
Best regards.
Patrick
Last edited:
That's a bit unexpected tbh, however...
So you have replaced the double diode in the same way as the other channel. -20 to +30 isn't that horrendous in the scheme of things. Is the offset changing smoothly with time or is it jumping rapidly?
Jumping around would point to a problem, smoothly changing is more like thermal drift somewhere. The input pair are in the same package and so should track well. The only thing would be the diode voltage changing with temperature.
Not easy to pin down is this. You could swap the input pair around and see if the problem follows the transistor.
If that proved nothing then I would swap those pairs of diodes around to see if the problem was there.
A bit odd is that.
Oh, also try this on full mains before we go looking to deep. See if the offset is steady... it could be. Try it.
Thanks Patrick 🙂
That was a fantastic instruction. Thanks for your help as well, Patrick!
And it IS fun. I succeeded at other times, but this one was tough 🙂
It's changing smoothly. R-CH drifts slowly around -10 mv, from -5 to -15. L-CH settled pretty much about 110 mV.
I'll try with full mains. I tested with DBT and speakers at low volume, all good. Bassy sounding amp, that's what I remember. Treble could be more refined, but that's a first assessment. And there is no bias 🙂
Gonna do this first, yes.
No bias will cause possibly faintly audible distortion at low levels. A pure (it must be pure to tell) sine wave of 1khz or higher is good test.
Many years ago... how old is this amp 😀 we used to say that -/+100mv or lower was acceptable as an offset.
If you think about it, 100mv across 8 ohms is only 12.5 milliamps and the power dissipated in 8 ohms is just 1.25 milliwatts