It could be you do need a coupling cap. Do you happen to have a small film cap (>= 15 nF or so) or bipolar electrolytic with at least a 50..63 V rating floating around?
I am reading a Dc at -B1 of 72 mv, and 73 at +B1 with a coupling cap 22nF
Also in any case check whether R671 is OK (270k) and connected to ground, and the DC voltages on both ends in operation.
Yes its ok, voltage is around 80 mv and 0 mv at ground.
Same for R603
Yes its ok, voltage is around 28 mv and 0 mv at ground.
Can you measure on the ends of C605 (otherwise use R607 and R609)?
I am measuring 0,68 volts on both ends (in schematic its 0,6 volts)
A bad connection on Q603's base alone could be enough to cause substantial offset, since base biasing current must flow through R601 and R603 to ground. Between R601 and R607 (ends connecting to Q603) there must be no more than the usual 0.6 V B-E voltage drop.
On R601 I am reading 89 mv and 28 mv
on R607 I am reading 0,65 volts and 0,72 volts
It's certainly not impossible that one of the drivers (Q623/Q625) is toast. How warm do they get in operation?
Q623 has 52,6 degree, Q625 about two degree more
Any suspicious-looking solder joints on them?
No
Also check their emitter reistors R641/R643, 910R 2W.
They are ok
Which voltages do you get when (carefully) measuring on the non-grounded end of R683
1,9 volt
and the non-supply end of R641?
1,28 volt
Same for R681 and R643.
R681 = -358 mv
R643 = 270 mv
For reference, we also need the output DC offset.
240 mv
I am reading a Dc at -B1 of 72 mv, and 73 at +B1 with a coupling cap 22nF
Also in any case check whether R671 is OK (270k) and connected to ground, and the DC voltages on both ends in operation.
Yes its ok, voltage is around 80 mv and 0 mv at ground.
Same for R603
Yes its ok, voltage is around 28 mv and 0 mv at ground.
Can you measure on the ends of C605 (otherwise use R607 and R609)?
I am measuring 0,68 volts on both ends (in schematic its 0,6 volts)
A bad connection on Q603's base alone could be enough to cause substantial offset, since base biasing current must flow through R601 and R603 to ground. Between R601 and R607 (ends connecting to Q603) there must be no more than the usual 0.6 V B-E voltage drop.
On R601 I am reading 89 mv and 28 mv
on R607 I am reading 0,65 volts and 0,72 volts
It's certainly not impossible that one of the drivers (Q623/Q625) is toast. How warm do they get in operation?
Q623 has 52,6 degree, Q625 about two degree more
Any suspicious-looking solder joints on them?
No
Also check their emitter reistors R641/R643, 910R 2W.
They are ok
Which voltages do you get when (carefully) measuring on the non-grounded end of R683
1,9 volt
and the non-supply end of R641?
1,28 volt
Same for R681 and R643.
R681 = -358 mv
R643 = 270 mv
For reference, we also need the output DC offset.
240 mv
I measured the drivers:
Q623:
Base 1,9 volt
Emitter 1,3 volt
Collector 33,63 volt
Q625 :
Base -349 mv
Emitter 266 mv
Collector -33,63 volt
On the other channel I measure the following:
Q624
Base -0,953 volt
Emitter -342 mv
Collector -33,7 volt
Q624
Base 1,4 volt
Emitter 0,9 volt
Collector 33,8 volt
So to me Q625 measurements look weird...
Q623:
Base 1,9 volt
Emitter 1,3 volt
Collector 33,63 volt
Q625 :
Base -349 mv
Emitter 266 mv
Collector -33,63 volt
On the other channel I measure the following:
Q624
Base -0,953 volt
Emitter -342 mv
Collector -33,7 volt
Q624
Base 1,4 volt
Emitter 0,9 volt
Collector 33,8 volt
So to me Q625 measurements look weird...
Last edited:
I do not remember the exact reference of the Rotel amp (maybe RA930) which had its main feedback resistor open circuit but I found that its schematic was very similar to the RA931 power circuit.
The voltage readings on the drivers all look good to me. (Second channel apparently has more quiescent current.) The temperature is not too hot either.
I've been doing a bit of input stage simming in LTspice, using 2SA970 as the closest transistor equivalent I could find (these are a mid-high Vceo, high beta type). After changing R611 to 22k (it looks like 2k2 on the schematic, but I guess that was just some dirt), it seems to work quite well, too. Differential gain is about 12.5, and I can adjust offset just fine.
However, I am quite unable to produce what has to be a +89 mV offset on Q603's base (one end of R601). I'm reading about +5 mV right now, with 1.5 mV on R603 (rather than 28 mV!).
From this I conclude that Q603's current gain hFE must be down to about 22 (it normally is 200..300+). These transistors must have aged and become leaky, which also explains why both channels are similar. Hmm, it's a Sanyo, wouldn't be the first time. Why, however, only Q603 would be affected and not Q605, I'm not sure.
I suggest you look for a few 2SA970/CSA970/KSA970 or similar.
I've been doing a bit of input stage simming in LTspice, using 2SA970 as the closest transistor equivalent I could find (these are a mid-high Vceo, high beta type). After changing R611 to 22k (it looks like 2k2 on the schematic, but I guess that was just some dirt), it seems to work quite well, too. Differential gain is about 12.5, and I can adjust offset just fine.
However, I am quite unable to produce what has to be a +89 mV offset on Q603's base (one end of R601). I'm reading about +5 mV right now, with 1.5 mV on R603 (rather than 28 mV!).
From this I conclude that Q603's current gain hFE must be down to about 22 (it normally is 200..300+). These transistors must have aged and become leaky, which also explains why both channels are similar. Hmm, it's a Sanyo, wouldn't be the first time. Why, however, only Q603 would be affected and not Q605, I'm not sure.
I suggest you look for a few 2SA970/CSA970/KSA970 or similar.
Last edited:
That's interesting. If I modify the input stage circuit so R615 goes directly to ground (i.e. not only AC, but also DC symmetry) and adjust offset trimming again, input stage nonlinearity @10kHz is exactly the same as before, down to a fraction of a dB. Having different DC and AC symmetry via R671 and C613 therefore seems entirely pointless. I'm not surprised nobody else is advocating this.
Differential gain is exactly the same at about 18 for both cases, dropping by about 3 dB (below ~70 Hz, also going from dominant 3rd to dominant 2nd) when R671/C613 are involved.
Differential gain is exactly the same at about 18 for both cases, dropping by about 3 dB (below ~70 Hz, also going from dominant 3rd to dominant 2nd) when R671/C613 are involved.
No, but I measured voltages at the drivers, see here:http://www.diyaudio.com/forums/solid-state/206338-rotel-ra-820bx3-not-working-9.html#post2907856
I ordered SB631 and SD600, I hope they will work
Attachments
Last edited:
oops, I didn't see this post before
I did already order SD600 and SB 631, these ones are in the schematic...
No, but I measured voltages at the drivers, see here:http://www.diyaudio.com/forums/solid-state/206338-rotel-ra-820bx3-not-working-9.html#post2907856
I ordered SB631 and SD600, I hope they will work
Voltages are ok
Since I'm not sure what kind of quality your new drivers would be (which manufacturer, I'd guess ISC or somesuch?), I'd suggest not fixing what doesn't seem to be broken. (That is always a good opportunity for needlessly creating new problems.) They may be of use later, but for now I'd put them aside.
Please measure DC level at R601 / Q603 base, should be approximately 90 mV. As stated, that's way too much. Then you can go ahead and order a few *SA970s (even Reichelt has those), at least 4 but preferably 9 (8 + 1 spare) in case you ever need to replace those in the second differential amplifier stage as well.
Please measure DC level at R601 / Q603 base, should be approximately 90 mV. As stated, that's way too much. Then you can go ahead and order a few *SA970s (even Reichelt has those), at least 4 but preferably 9 (8 + 1 spare) in case you ever need to replace those in the second differential amplifier stage as well.
Last edited:
Aaaah you meant the voltages I measured are ok.
Now I understand!
I don't know so much about electronics, but I thought there must be always a difference about 0,6 volts between base and emitter?
Q625 :
Base -349 mv
Emitter 266 mv
Collector -33,63 volt
-349mV and + 266mV makes difference of above 0.6V and that is ok.
I would check all transistors in the differential stage, VAS stage and the output stage and then if problem still exist i would look for a faulty resistor or capacitor, diode maybe..just check all transistors one by one, it will take u some time but its worth, if they are ok you can go on and look for a problem in some passive elements, if not, just order the faulty transistors.
In one of your previous posts, i think you mentioned that if you are increasing the volume there is a distortion, maybe your output transistors are broken (open circuit) and your amp goes only on drivers.
Last edited:
Since I'm not sure what kind of quality your new drivers would be (which manufacturer, I'd guess ISC or some such?
Yes ISC from Reichelt, is this good?
Please measure DC level at R601 / Q603 base, should be approximately 90 mV.
Yes its 93 mv
As stated, that's way too much. Then you can go ahead and order a few *SA970s (even Reichelt has those), at least 4 but preferably 9 (8 + 1 spare) in case you ever need to replace those in the second differential amplifier stage as well.
Ok, I will go to the local store, they have SA970, the originals on the board are wrapped together with a stripe, Also they look they are glued together or something. What shall I get from my store?
Do I have to measure similar Hfe?
Again: Thanks so much for your help I have been so frustrated in-between, but then an answer from (mostly you) appeared and helped me!
Yes ISC from Reichelt, is this good?
Please measure DC level at R601 / Q603 base, should be approximately 90 mV.
Yes its 93 mv
As stated, that's way too much. Then you can go ahead and order a few *SA970s (even Reichelt has those), at least 4 but preferably 9 (8 + 1 spare) in case you ever need to replace those in the second differential amplifier stage as well.
Ok, I will go to the local store, they have SA970, the originals on the board are wrapped together with a stripe, Also they look they are glued together or something. What shall I get from my store?
Do I have to measure similar Hfe?
Again: Thanks so much for your help
I have been so frustrated in-between, but then an answer from (mostly you) appeared and helped me!Ok I will check all the voltages on the transistors and report back.
Sound from the amp is ok, the distortion was in the test speaker (I think I have reported this before) the bad thing is the dc offset thats far too much....
I see..probably missed your post, my fault. Ill check the schematic again, but DC offset can be a result of not matched output transistors, differential ones...maybe they changed their characteristics durring time. And maybe amp was like that from the beggining, once i had a new Harman Kardon with offset of 0.7V on both channels!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.