I hope some of you guys can help:
I have a Rotel RA 930 amp. I’m a bit of an electronics noob, but to further my understanding wanted to check the bias adjustment and offset…
I first checked the DC offset by probing the speaker terminals with a multimeter – readings were 0.02mV for both channels.
I then tried checking the bias by measuring the voltage across the emitter resistors for each of the output transistors. There are four output transistors (2 pairs) and they were reading 3.9 and 0.8mV on the left and 3.9 and 0.9mV on the right. I thought it odd that there were differences between the readings i.e. 3.9 and 0.8 (presumed they should all be the same) so I checked again and ended up bridging the emitter resistor with the left leg of one of the transistors (base?) on the right channel. There was a pop from the speaker that was still connected, and when I measured the voltage again, the bias had gone up to approximately 90mV on both the transistors of that channel.
The DC offset measured at the speaker terminals has now also gone to about 90mV and the heatsink where the affected transistors are got really hot really quickly so I shut the amp down, although it does seemed to be working for the brief time I tested it even after my slip with the probe.
I then found a service manual and see that I should have been measuring across TP1/3 and TP4/2 – So I measured across the TP points – left channel now measures 4.3mV and right 87mV.
Can anyone please diagnose what I’ve done and what if anything I can do to fix?
RA930 schematic is here: http://www.bwgroup-support.com/rotelservicemanuals.html
I think I bridged resistor R630 to Base of Q630.
Thanks.
I have a Rotel RA 930 amp. I’m a bit of an electronics noob, but to further my understanding wanted to check the bias adjustment and offset…
I first checked the DC offset by probing the speaker terminals with a multimeter – readings were 0.02mV for both channels.
I then tried checking the bias by measuring the voltage across the emitter resistors for each of the output transistors. There are four output transistors (2 pairs) and they were reading 3.9 and 0.8mV on the left and 3.9 and 0.9mV on the right. I thought it odd that there were differences between the readings i.e. 3.9 and 0.8 (presumed they should all be the same) so I checked again and ended up bridging the emitter resistor with the left leg of one of the transistors (base?) on the right channel. There was a pop from the speaker that was still connected, and when I measured the voltage again, the bias had gone up to approximately 90mV on both the transistors of that channel.
The DC offset measured at the speaker terminals has now also gone to about 90mV and the heatsink where the affected transistors are got really hot really quickly so I shut the amp down, although it does seemed to be working for the brief time I tested it even after my slip with the probe.
I then found a service manual and see that I should have been measuring across TP1/3 and TP4/2 – So I measured across the TP points – left channel now measures 4.3mV and right 87mV.
Can anyone please diagnose what I’ve done and what if anything I can do to fix?
RA930 schematic is here: http://www.bwgroup-support.com/rotelservicemanuals.html
I think I bridged resistor R630 to Base of Q630.
Thanks.
Soudns like an output transistor has gone.
Check for shorts across the transistor pins.
Its not unusual to have different voltages across the emitter resistors becasue of the output DC offset sending more current through half of the output pairs.
Check for shorts across the transistor pins.
Its not unusual to have different voltages across the emitter resistors becasue of the output DC offset sending more current through half of the output pairs.
Thanks for the reply.
I haven't taken the transistors off the board but just checked the continuity across the output transistor pins on the bad channel and seems the same as the good one, plus the amp still seems to play ok.
Do they need to be removed from the board to check?
I haven't taken the transistors off the board but just checked the continuity across the output transistor pins on the bad channel and seems the same as the good one, plus the amp still seems to play ok.
Do they need to be removed from the board to check?
OK I checked CE, BE and CB with a digital multimeter using continuity/diode test and got the following:
BE - No continuity
CE - No full continuity but would get a beep for one second from the meter when swapping black and red leads across CE, but stabilised as no continuity
CB - Same results as CE
BE - No continuity
CE - No full continuity but would get a beep for one second from the meter when swapping black and red leads across CE, but stabilised as no continuity
CB - Same results as CE
Strange. At the very least you should have seen the emitter/base resistance in series when you did the BE test.
Anyway ...
Pos probe on B of Q619(Q620) should conduct to E and C
Neg probe on B of Q621(Q622) should conduct to E and C
If not, check meter against any handy diode etc.
The unit should of course be switched off when doing a continuity test!
Good Luck,
Rob
Anyway ...
Pos probe on B of Q619(Q620) should conduct to E and C
Neg probe on B of Q621(Q622) should conduct to E and C
If not, check meter against any handy diode etc.
The unit should of course be switched off when doing a continuity test!
Good Luck,
Rob
never do measurements with speakers connected, if anything goes wrong you will be cooking them with DC.
The 87mV across 0.22ohm in your right channel means there is about 0.4A of quiescent current flowing instead of 18mA, and that the two output transistors dissipate about 30W (instead of 1.35W).
Q608 and Q610 form a 6mA current source; the 6mA current from this source flows through Q614 and Q612 to the negative rail. The voltage drop across Q614, which normally should be 2.2V according to the schematic, sets the quiescent current of the output pair. The voltage drop can be adjusted by VR602, but DON'T TOUCH IT for now.
If the amp still works, I would suspect any of Q608, Q610 or possibly Q614 have gone bad.
The service manual shows voltages in various point of the power amplifier. Check those (quickly) and tell us if any are significantly different in your amp from what they should be according to the manual. Check voltages before desoldering anything; you're likely to wreak more destruction if you desolder what's working fine.
Q608 and Q610 form a 6mA current source; the 6mA current from this source flows through Q614 and Q612 to the negative rail. The voltage drop across Q614, which normally should be 2.2V according to the schematic, sets the quiescent current of the output pair. The voltage drop can be adjusted by VR602, but DON'T TOUCH IT for now.
If the amp still works, I would suspect any of Q608, Q610 or possibly Q614 have gone bad.
The service manual shows voltages in various point of the power amplifier. Check those (quickly) and tell us if any are significantly different in your amp from what they should be according to the manual. Check voltages before desoldering anything; you're likely to wreak more destruction if you desolder what's working fine.
Last edited:
The lack of continuity between base and emitter shows that it is time to switch on the soldering iron.
If there was no continuity in Q630, there would be no way for 0.4a flowing through R630, and the amp would not work. Something must be wrong with the continuity test. Also, if indeed R630 (i.e. emitter of Q630) was shorted to the base of Q630, it would not be a disaster for Q630.
Last edited:
Thanks for all the advice.
The meter I used is a good one, or at least it was expensive. I do have another cheaper meter I could use as another test, but it is more likely my technique.
I did the continuity test with the amp off.
Thanks for looking at the schematic Alexcp. I realise I said I grounded Q630with R630, but looking again at the schematic it was actually Q620 (but I think that's what you understood anyway).
I haven't touched VR602, even though I was tempted to!
The recommendation to check the voltages on the board is a good one and I will try that when I get home.
My limited instincts say that the output transistors are ok, and it is driver stage that is making one channel or one half of one channel max out all the time - hence the excessive bias current and heat, but that the amp seems to work.
I only had a tuner to hand when I tested that the amp still functioned, and I had to move the arial so the reception was very sketchy and may not have properly been able to tell if there was a lot of distortion on the bad channel, so will try with a proper source.
Also, I do have an oscilloscope that I presume could be useful, and I did have a signal generator on my laptop that could generate sine waves but the soudcard isn't working right now, but could probably find a way of generating a sine wave from something else otherwise - if it would help the diagnosis.
The meter I used is a good one, or at least it was expensive. I do have another cheaper meter I could use as another test, but it is more likely my technique.
I did the continuity test with the amp off.
Thanks for looking at the schematic Alexcp. I realise I said I grounded Q630with R630, but looking again at the schematic it was actually Q620 (but I think that's what you understood anyway).
I haven't touched VR602, even though I was tempted to!
The recommendation to check the voltages on the board is a good one and I will try that when I get home.
My limited instincts say that the output transistors are ok, and it is driver stage that is making one channel or one half of one channel max out all the time - hence the excessive bias current and heat, but that the amp seems to work.
I only had a tuner to hand when I tested that the amp still functioned, and I had to move the arial so the reception was very sketchy and may not have properly been able to tell if there was a lot of distortion on the bad channel, so will try with a proper source.
Also, I do have an oscilloscope that I presume could be useful, and I did have a signal generator on my laptop that could generate sine waves but the soudcard isn't working right now, but could probably find a way of generating a sine wave from something else otherwise - if it would help the diagnosis.
- Status
- Not open for further replies.