I found the sony service manual first ...and then I found a different circuit in a forum ....that was newer and the amp is under repair according to its circuit. The difference just the protection transistors the amount of resistance next to the bias pot ... which is not Important.
compared the voltages of both channels by focusing on the bases of the transistors. They are all the same, only when connect a 4.7 Ohm 5 watt resistor to the output of both channels, the voltage of the defective channel changed ....in picturs can see v+of c... vt 404 is decrease and -v of e is increase...so bias of vt408 go up and bernt.
The first time..... the Amp was working. And the CD was finished .... it was silent for an hour.Connected two speaker wires for a very short time....The amp turned off .... Fuses and a one of pnp power transistors ... burned. 😢
Alas...
The power pnp sinks to much current, heats up and a meltdown happen - in a very short time.
Given the small difference between the two power pnp's, the best conducting goes first.
Connecting a load shifts the output dc level a minor bit, but sufficient to topple the stability of the amp.
The bases of 401 and 402 are at -15mV (0.3µA*51k).
The output will not be exactly on 0.0Vdc, but at the value that is forced by the feedback loop.
My guess is that it is just a little bit from true 0.0Vdc, several mV's only.
As soon as this output level is changed (by a serious load), the loop will try to correct that.
It fails.
Why? The loop 'overacts', because it cannot correct the error.
It tries to do so, and will not stop correcting until the balance is set again.
This causes weird settings at 403, 411, 406 and is lethal for 408 or 410.
Best advice for now, follow:
Replace all the power devices (407/408/409/410) with small signal TO92 with a 1kΩ at their emitters in advance.
The only other cause I can think of for now is a shorting between the load return terminal ('black') and the positive rail supply: as soon as you connect the load, the output ('red') terminal is lifted and the amp will die...
The power pnp sinks to much current, heats up and a meltdown happen - in a very short time.
Given the small difference between the two power pnp's, the best conducting goes first.
Connecting a load shifts the output dc level a minor bit, but sufficient to topple the stability of the amp.
The bases of 401 and 402 are at -15mV (0.3µA*51k).
The output will not be exactly on 0.0Vdc, but at the value that is forced by the feedback loop.
My guess is that it is just a little bit from true 0.0Vdc, several mV's only.
As soon as this output level is changed (by a serious load), the loop will try to correct that.
It fails.
Why? The loop 'overacts', because it cannot correct the error.
It tries to do so, and will not stop correcting until the balance is set again.
This causes weird settings at 403, 411, 406 and is lethal for 408 or 410.
Best advice for now, follow:
Or start with a 10kΩ to ground and see what happens.
Replace all the power devices (407/408/409/410) with small signal TO92 with a 1kΩ at their emitters in advance.
The only other cause I can think of for now is a shorting between the load return terminal ('black') and the positive rail supply: as soon as you connect the load, the output ('red') terminal is lifted and the amp will die...
Assuming your sketch in post 42 is showing -1.1V at base of Q411, please report base and emitter of Q406, and Output--- without load.
Thanks!
Thanks!
w
. yesterday i remod vt411.... with no loade dc v =good . sound = good and normal
but if use speaker 4 ohm 1.1 v go down + .78 ......bias of pnp power transistor go up and burnt.
. yesterday i remod vt411.... with no loade dc v =good . sound = good and normal
but if use speaker 4 ohm 1.1 v go down + .78 ......bias of pnp power transistor go up and burnt.
I don't understand your notation: "0/52" at Q407, and "0/08" at Q408. Would you explain, please? In the end, I would like to know voltage at output, and base voltages, emitter voltages, all with respect to ground.
The voltage across Q404 is +1.25 plus 1.1V = 2.35V. This raises the question is the spreader 2.35V controlled by conduction through Q404, or is Q404 not conducting but clamped by parallel path Q405 to Q407 to Q408 to Q406 to Q411? Would you observe where the wiper of VR401 is positioned? An alternate method is to measure the base-emitter voltage across Q404.
The voltage across Q404 is +1.25 plus 1.1V = 2.35V. This raises the question is the spreader 2.35V controlled by conduction through Q404, or is Q404 not conducting but clamped by parallel path Q405 to Q407 to Q408 to Q406 to Q411? Would you observe where the wiper of VR401 is positioned? An alternate method is to measure the base-emitter voltage across Q404.
Last edited:
I don't understand your notation: "0/52" at Q407, and "0/08" at Q408. Would you explain, please? In the end, I would like to know voltage at output, and base voltages, emitter voltages, all with respect to ground.
1,25 / -1,1 / -0,53 / ON / off are readable.
-908... +152... ???
You don't have to draw the meter.
There are five BE's in series. For proper operation (class AB) one should expect 5*0.6=3V as bias voltage: +1.25 to "-1.7" is 2.95 which is ok; 1.25 to -1.1 is 2.35 is to little.
In #1, the base of 406 is at -1.2V If the base of 411 is at -1.1V, it's cutoff completely.
-908... +152... ???
You don't have to draw the meter.
There are five BE's in series. For proper operation (class AB) one should expect 5*0.6=3V as bias voltage: +1.25 to "-1.7" is 2.95 which is ok; 1.25 to -1.1 is 2.35 is to little.
In #1, the base of 406 is at -1.2V If the base of 411 is at -1.1V, it's cutoff completely.
As a sanity check/confidence builder, in the working channel would you measure the amp output and the base of VT311? Expected voltage is about -1.7V.
Thank you.
Thank you.
please....... The main problem is the voltage change of
+ 1.25 in the transistor 403 ... when connecting the load resistance of 4 ohms and less. Yes. I wish Sony didn't make this device. 😢 I decide to add a current source to the bias circuit.
+ 1.25 in the transistor 403 ... when connecting the load resistance of 4 ohms and less. Yes. I wish Sony didn't make this device. 😢 I decide to add a current source to the bias circuit.
Yes, we've noticed, but it's still unclear what causes this.
Can you clearify #47 & #49?
No one can found the defect of this amplifier. I couldn't either. No results can be obtained from searching on Google. The output transistor of the left channel may burn after a short period of time. The right channel, which has already been repaired, is biased in the wrong state. Both outputs are off without a signal, but the transistor does not burn. It seems that the bias current is low.
Your probably have excessive crossover distortion because the bias current isn't set properly. No doubt, you already know this.
On the working channel (since you know it can safely deliver audio to a 4 Ohm load), try to adjust the bias current in accordance with section 3.1. Be sure to test without a load, since you are trying to adjust the current passing from the positive rail to the negative rail. With no load, carefully advance the VR301 pot until you observe a few mV developed across the emitter resistor.
Then remeasure amp output and voltage at base of of VT311.
Good luck!
On the working channel (since you know it can safely deliver audio to a 4 Ohm load), try to adjust the bias current in accordance with section 3.1. Be sure to test without a load, since you are trying to adjust the current passing from the positive rail to the negative rail. With no load, carefully advance the VR301 pot until you observe a few mV developed across the emitter resistor.
Then remeasure amp output and voltage at base of of VT311.
Good luck!
I removed the 311 transistor from pcb and connected the 330 Ohm resistor to 306 base. adjusted the bias. Everything became normal. Connecting a 4 ohm load did not make any changes.good ab class amp.😱😳😡 ..(311 transistor inL chanel.. L chanal has problem )
Last edited:
I don't understand this last comment. If I understand correctly, the Left channel is the working channel and has the 3xx series of part numbers. You have removed 311, rebiased current, and left channel behaves as desired. Did you intend " R channel has problem"?
This is an interesting experiment, but you avoided discovering whether VT311 base (=VT303 emitter) would have settled at -1.7V once bias current was set correctly. Perhaps this mod is OK, but it's not what Sony intended and not a revision I would make. But this is your project, not mine.
You could try making the same change to the problem channel, but my fear is that the same change in the right channel will not resolve the problem.
If you remain stuck with the problem, I'll suggest a debugging procedure.
P.S. Another thought. The spreader transistor should be attached thermally to the output power transistors so that it can sense their temperature and adjust spreader voltage appropriately. I can't tell from the manual how the transistor is mounted.
Last edited:
Finally, .................... It is clear that the device is vintage and it was not possible to change a large number of parts at once. I changed the transistors of the mirror circuit one by one. There was no change. Even used ransistors of that channel that The power transistors did not burn.....I put them one by one in the mirror circuit, but I decided to change first three transistors at once.
to replace ...............I tested several different pir transistor for mirror and v amp.
..the last one seemed to make a change in the bias 😍 . I adjusted the bias carefully. Finally, I reached a point of setting which was biased as follows and there were no other changes with each load connection. played a music device for three days. The bias is fixed.🙂frinds thank.
to replace ...............I tested several different pir transistor for mirror and v amp.
..the last one seemed to make a change in the bias 😍 . I adjusted the bias carefully. Finally, I reached a point of setting which was biased as follows and there were no other changes with each load connection. played a music device for three days. The bias is fixed.🙂frinds thank.