Hi all
I got an Rotel 820BX4 off ebay recently which had one channel not working.
After i got it I had a good look on the parts and for shorts did and quickly found that fuse F601 was blown but couldn't find any other issues.
I've replaced the fuse and adjusted the bias to 2.5mv and dc offset to about +/-5mv (best could get close to 0) and so now the amp works.
However I've discovered another issue. Driver transistors Q623 & Q625 run much hotter on one channel. I can only barely keep my finger on the heatsinks whereas on the other channel the heatsinks feel stone cold.
Any idea what I should be looking at next? I still don't feel I got to the bottom of why this fuse was blown and somehow feel this is related
Cheers Evan
I got an Rotel 820BX4 off ebay recently which had one channel not working.
After i got it I had a good look on the parts and for shorts did and quickly found that fuse F601 was blown but couldn't find any other issues.
I've replaced the fuse and adjusted the bias to 2.5mv and dc offset to about +/-5mv (best could get close to 0) and so now the amp works.
However I've discovered another issue. Driver transistors Q623 & Q625 run much hotter on one channel. I can only barely keep my finger on the heatsinks whereas on the other channel the heatsinks feel stone cold.
Any idea what I should be looking at next? I still don't feel I got to the bottom of why this fuse was blown and somehow feel this is related
Cheers Evan
Attachments
Last edited:
Well if the blown fuse and the hot transistors are on the same channel you may be onto something.
Strange that Q623 and Q625 can run hot without something else glowing with heat. Check if the the base of the transistors are at the +-0.53V as stated in that picture, and the emitters at 0V.
Check the resistance value of R641 and 643 too. If the values in the picture are true the designers used 2W resistors in a location that would emit 1.85W of heat...
If the voltages are correct and the resistance values are correct, then most likely one or more of the output transistors are dead.
Strange that Q623 and Q625 can run hot without something else glowing with heat. Check if the the base of the transistors are at the +-0.53V as stated in that picture, and the emitters at 0V.
Check the resistance value of R641 and 643 too. If the values in the picture are true the designers used 2W resistors in a location that would emit 1.85W of heat...
If the voltages are correct and the resistance values are correct, then most likely one or more of the output transistors are dead.
If you reduce the bias (VR601), do these transistors run cooler?
Is Q621 in thermal contact with the heatsink?
Are Q627, Q629, Q631 & Q633 running hot? Are they good?
Are the voltages across R645, R647, R649 & R651 the same?
Good Luck!
I read +/-1.1V at the bases of the 4 driver transistors, so it would appear there is variation from the schematic, but could be normal... not sure...
The driver transistors are thermally connected to their heatsinks, when I touch the transistors themselves they are hot/cold.
I can't get a good resistance reading for all the 2w resistors, but I do read 35.7V across all 4 of them. Interesting enough I can get a ~0.9kR reading across the resistors in the hot channel but I read in the kilo/mega range for the cold channel. All 4 resistors have slight cracks on them so it could be that they are beginning to go open? I might as well replace them anyway as I don't like the look of them.
All power transistors read well and don't get hot, just lukewarm.
It's difficult to tell but reducing bias via VR601 seems to make the drivers warm up slower, but they seem to reach the same temperature in the end.
I've attached some pics, first one is the cold channel, second is the warm one.
Strange thing is the amp seems to work okay... I wouldn't say it sounds amazing but not bad
The driver transistors are thermally connected to their heatsinks, when I touch the transistors themselves they are hot/cold.
I can't get a good resistance reading for all the 2w resistors, but I do read 35.7V across all 4 of them. Interesting enough I can get a ~0.9kR reading across the resistors in the hot channel but I read in the kilo/mega range for the cold channel. All 4 resistors have slight cracks on them so it could be that they are beginning to go open? I might as well replace them anyway as I don't like the look of them.
All power transistors read well and don't get hot, just lukewarm.
It's difficult to tell but reducing bias via VR601 seems to make the drivers warm up slower, but they seem to reach the same temperature in the end.
I've attached some pics, first one is the cold channel, second is the warm one.
Strange thing is the amp seems to work okay... I wouldn't say it sounds amazing but not bad
Attachments
Last edited:
What voltage do you read at the emitters of Q623 & Q625?
Are the voltages across R645, R647, R649 & R651 the same?
So yea the resistors probably died from the heat. I would suggest changing them to 5W type.
The problem is in the channel that is running cool since the 910R resistors have "gone high".
The driver transistors in Class AB amplifiers switch off at some point on opposite cycle voltage swings. This happens if the transistor emitters drive the load through a resistor or if the emitters are cross-coupled by a resistor.
The driver transistors in this amplifier are resistively connected to the opposite voltage rail which breaks with the normal conventions and hotter running is to be expected.
You are on the right track to replace the 910R resistors.
I've ran into a problem. I was probing the voltages across the emitter resistors when I touched one of the bus bars with my probe. It would appear I have shorted something and now I have one dead channel... school boy error...!!
Now the previously cold channel reads open on the speaker terminals and the main heatsink seems to heat up very rapidly on the side of this channel.
Whoops
Now the previously cold channel reads open on the speaker terminals and the main heatsink seems to heat up very rapidly on the side of this channel.
Whoops
Right then so verdict is not as bad as I thought. Turns out I fried a couple of output and one driver transistor, Q628, Q630 (2SD1047) and Q624 (2SD600K) which coincidentally are all on the cold channel. All remaining transistors appear to read ok at diode test.
Might order a full set of drivers/output transistors for the bad channel, it deserves it after all that trouble!
Emitter resistors seem to be in good knick, all 4 read the same 0.34 ohm resistance but not 0.22 ohm as per schematic. Hmm should I worry about that and change them too?
The 2W910 resistors will also get changed with higher wattage ones.
Anything else I should be changing?
Might order a full set of drivers/output transistors for the bad channel, it deserves it after all that trouble!
Emitter resistors seem to be in good knick, all 4 read the same 0.34 ohm resistance but not 0.22 ohm as per schematic. Hmm should I worry about that and change them too?
The 2W910 resistors will also get changed with higher wattage ones.
Anything else I should be changing?
Last edited:
A multimeter with transistor test is useful, but also
https://www.youtube.com/watch?v=b0jCUsenNsY
As well as temperature measurement, because "hot according to fingers" is seldom accurate when "too hot" can be 50-60 degrees when feeling metal but the silicon is feeling pretty okay. Have you touched the heatsink of a graphics card while playing?
If all 4 of resistors read the same... well for resistors the value is printed on them so...
If they all measure the same and different from what's printed on them, I would question the multimeter. After all, 0.1ohm can easily come from connectors and probe and stuff. I just measured my probes and each (with the wires and stuff) is 0.2ohm.
https://www.youtube.com/watch?v=b0jCUsenNsY
As well as temperature measurement, because "hot according to fingers" is seldom accurate when "too hot" can be 50-60 degrees when feeling metal but the silicon is feeling pretty okay. Have you touched the heatsink of a graphics card while playing?
If all 4 of resistors read the same... well for resistors the value is printed on them so...
If they all measure the same and different from what's printed on them, I would question the multimeter. After all, 0.1ohm can easily come from connectors and probe and stuff. I just measured my probes and each (with the wires and stuff) is 0.2ohm.
Thanks yes my DMM can measure resistance too in diode mode, I usually check mV first then resistances if I get bad readings. I have a uni-t 61d which is quite accurate and a DCA55 for out of circuit checks.
You are right its the resistance of the probe leads they measure 0.12 ohm exactly. Bingo
You are right its the resistance of the probe leads they measure 0.12 ohm exactly. Bingo
Last edited:
Suspicions confirmed. I took the 910R resistors out today and indeed 2 of them measured at about 1.3Meg so they seem to be heat fatigued. So this is why the drivers were cold, not much current going through them...
Also turned out the output devices that I thought were bad tested okay outside the circuit on my tester. It must have been the bad driver that was causing the bad reading then.
The second picture shows transistor Q901 which forms part of a zener based series regulator for the phono stage. I don't have a means of testing phono but I could hear a strange static noise from the speakers when nothing was connected. I replaced it with a 2N3904 which has a different pinout - note the legs crossing. Now the static is gone.
Also turned out the output devices that I thought were bad tested okay outside the circuit on my tester. It must have been the bad driver that was causing the bad reading then.
The second picture shows transistor Q901 which forms part of a zener based series regulator for the phono stage. I don't have a means of testing phono but I could hear a strange static noise from the speakers when nothing was connected. I replaced it with a 2N3904 which has a different pinout - note the legs crossing. Now the static is gone.
Attachments
Last edited:
She lives!
I ended up changing the driver resistors and transistors, a few diodes here and there (for peace of mind), and did a full recap. Upgraded the bridge rectifier which was rated 200v (not 250v...?) to 400v, and the main filter caps from 8200uf to 12000uf too which seems to have made a difference to the bass output.
Sounding great!
I ended up changing the driver resistors and transistors, a few diodes here and there (for peace of mind), and did a full recap. Upgraded the bridge rectifier which was rated 200v (not 250v...?) to 400v, and the main filter caps from 8200uf to 12000uf too which seems to have made a difference to the bass output.
Sounding great!
Attachments
Last edited:
I recently got one of these with the cooked 910 ohm 2w resistors. changed them for 3w metal film and while the amp sounded great these 3w also showed signs of heat damage after a few days use. I've now replaced these with Welwyn vitreous enamel wirewound 7w (1k the nearest I could find) rechecked the mv settings and all seems well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.