SA-8800 Trim Pots
- Thread starter spark1
- Start date
it will happen, but I;m finally ambulatory and am trying to catch up with everything I neglected the last 4 weeks while I did leg elevated bed rest for my knee and swollen leg. baker cyst and BAD swelling, as well as avoiding pressure sores from light contacts - that skin was fragile!
Are the trimpots for the idle current adjust the same model and resistance and what are their part number - i (mis?) remember a pix with strange trimmers.
we might have to resort to swapping driver transistors to move the symptoms to the other channel, and if that doesn't do it, moving the outputs.
Are the trimpots for the idle current adjust the same model and resistance and what are their part number - i (mis?) remember a pix with strange trimmers.
we might have to resort to swapping driver transistors to move the symptoms to the other channel, and if that doesn't do it, moving the outputs.
spark1
Lunatic Member
Thanks for jumping back in...and sorry to hear about problems with your leg. Getting older is not for the weak!
New bias trimmers:
3329P-1-101LF = 100ohm
3329P-1-104LF = 100Kohm (to replace 68Kohm originals)
https://www.mouser.com/ProductDetail/652-3329P-1-104LF
I don't know about them being strange...although they definitely look different than the originals. They are single-turn, Bourns. I tested all before installing, and after installing. They were installed at minimum resistance setting, and all work smoothly, and are almost perfectly on spec. The good (right) channel came easily to 56mV (at full power), and shows about 14mV on the DBT (60 watt bulb). The left channel was around 270-280mV at full power, zero resistance, and is around 50mV on the DBT. The "secondary" bias trimmers are currently at zero resistance.
The new trimmers work the same as the originals from the standpoint of CCW = less resistance, CW = more resistance.
New bias trimmers:
3329P-1-101LF = 100ohm
3329P-1-104LF = 100Kohm (to replace 68Kohm originals)
https://www.mouser.com/ProductDetail/652-3329P-1-104LF
I don't know about them being strange...although they definitely look different than the originals. They are single-turn, Bourns. I tested all before installing, and after installing. They were installed at minimum resistance setting, and all work smoothly, and are almost perfectly on spec. The good (right) channel came easily to 56mV (at full power), and shows about 14mV on the DBT (60 watt bulb). The left channel was around 270-280mV at full power, zero resistance, and is around 50mV on the DBT. The "secondary" bias trimmers are currently at zero resistance.
The new trimmers work the same as the originals from the standpoint of CCW = less resistance, CW = more resistance.
spark1
Lunatic Member
I'm doing more swapping today....I assume this is the only option left. I was hoping the testing Tarior had me doing would lead us to a diagnosis, but since he dropped out of sight and Mark is busy (and with a bum leg to boot) the only thing I know to do now is swap parts.
Last edited:
spark1
Lunatic Member
Yes, all resistors have been tested and compared, good to bad channel, with no differences that seemed significant.
Edit: If by shotgun you mean replace each part on the bad channel, it seems like it will be easier to swap likely suspects from one side to the other and test after each swap...but I guess the work is about the same either way. And with the swapping approach, I have the needed parts on hand!
I know wholesale replacement is not good repair protocol, but I don't know what else to do at this point.
Edit: If by shotgun you mean replace each part on the bad channel, it seems like it will be easier to swap likely suspects from one side to the other and test after each swap...but I guess the work is about the same either way. And with the swapping approach, I have the needed parts on hand!
I know wholesale replacement is not good repair protocol, but I don't know what else to do at this point.
Last edited:
The parts I was talking about swapping are the drivers (Q39, Q40, Q41 & Q42 ) and output transistors (Q1, Q2, Q3, Q4, Q5, Q6, Q7 & Q8).
BUT FIRST, (it just occurred to me) measure the voltages across the emitter resistors for the eight output transistors, that will show how well the output transistors are sharing the idle current on each channel. GWH-115 R131 through R138 inclusive.
Then very very carefully measure the emitter to base voltage at EACH of the output transistors, (Q1, Q2, Q3, Q4, Q5, Q6, Q7 & Q8). RIGHT AT the leads..
Then do the same emitter to base voltage readings on the drivers Q39, Q40, Q41 & Q42 .
That may tell us if the extra idle current is the fault of extra base drive on the outputs, or damaged outputs.
I will look back in periodically for when the results are posted, to try to catch them faster, so you can start on the swapping.
DON'T start swapping the transistors until I give you a green light to do so, please.
You see, things are NOT that radically off in the readings taken so far, i.e. no smoking gun,
so when the usual suspects are ruled out, the improbable MUST be considered.
BUT FIRST, (it just occurred to me) measure the voltages across the emitter resistors for the eight output transistors, that will show how well the output transistors are sharing the idle current on each channel. GWH-115 R131 through R138 inclusive.
Then very very carefully measure the emitter to base voltage at EACH of the output transistors, (Q1, Q2, Q3, Q4, Q5, Q6, Q7 & Q8). RIGHT AT the leads..
Then do the same emitter to base voltage readings on the drivers Q39, Q40, Q41 & Q42 .
That may tell us if the extra idle current is the fault of extra base drive on the outputs, or damaged outputs.
I will look back in periodically for when the results are posted, to try to catch them faster, so you can start on the swapping.
DON'T start swapping the transistors until I give you a green light to do so, please.
You see, things are NOT that radically off in the readings taken so far, i.e. no smoking gun,
so when the usual suspects are ruled out, the improbable MUST be considered.
spark1
Lunatic Member
OK, finally got my Sansui2000X cleared off the workbench and got the measurements done on the SA-8800. All measurements were done with the unit on DBT with a 60 watt bulb.
The left channel (high bias channel) emitter resistors:
.017V / .017V / .018V / .017V
The right channel emitter resistors:
.010V / .009V / .010V / .010V
The driver transistors, E to B voltage:
Q39 = .560
Q41 = .567
Q42 = .573
Q40 = .558
The left channel output transistors, E to B voltage:
Q1 (C2525) = .605
Q2 (C2525) = .607
Q6 (A1075) = .608
Q5 (A1075) = .608
The right channel output transistors, E to B voltage:
All were at .595
Looking forward to your thoughts on these results. Thanks so much for the help.
The left channel (high bias channel) emitter resistors:
.017V / .017V / .018V / .017V
The right channel emitter resistors:
.010V / .009V / .010V / .010V
The driver transistors, E to B voltage:
Q39 = .560
Q41 = .567
Q42 = .573
Q40 = .558
The left channel output transistors, E to B voltage:
Q1 (C2525) = .605
Q2 (C2525) = .607
Q6 (A1075) = .608
Q5 (A1075) = .608
The right channel output transistors, E to B voltage:
All were at .595
Looking forward to your thoughts on these results. Thanks so much for the help.
rather rushed now, will be up all night BUT I DID look!!!
( I will be up for the next 24 hours "fighting fires" )
they are doing as commanded. 0.017v on one channel vs 0.010 v on the other.
needed to have left / right distinguished in drivers -
Now try ALL the same readings referenced to the Channel's OUTPUT before the filters ( which is between the 0.5 ohm emitter resistors)
instead of using the emitters as the black lead, still probe the base voltages of all 12 transistors. Pay attention to and post polarity
(it was appropriate to NOT post polarity on that first set of measurements)
this might be better at revealing what I search for.
The high channel has the 100 ohm pot set at zero?
Set the OK channel at zero too.
When checking the idles on these settings, also slowly and carefully watch for any effect
when adjusting the SECOND idle current pot.
( I will be up for the next 24 hours "fighting fires" )
they are doing as commanded. 0.017v on one channel vs 0.010 v on the other.
needed to have left / right distinguished in drivers -
Now try ALL the same readings referenced to the Channel's OUTPUT before the filters ( which is between the 0.5 ohm emitter resistors)
instead of using the emitters as the black lead, still probe the base voltages of all 12 transistors. Pay attention to and post polarity
(it was appropriate to NOT post polarity on that first set of measurements)
this might be better at revealing what I search for.
The high channel has the 100 ohm pot set at zero?
Set the OK channel at zero too.
When checking the idles on these settings, also slowly and carefully watch for any effect
when adjusting the SECOND idle current pot.
spark1
Lunatic Member
Q39 and Q41 are left channel. Q40 and Q42 are right channel.
Yes. left channel was set to zero resistance on the 100ohm pot this yields a bias of 53mV (with the unit on DBT). The right was not; it was set to yield the specified bias voltage while off the DBT (early on in the process, prior to me realizing I had a problem on the left channel). It is allowing a bias of 14mV on the DBT, and 56mV off of it. I will turn it to zero for the next round of measurements. The second set of pots has been at zero throughout all testing.
To be clear...what do I connect to as reference for the base voltages in this next round? And, still on the DBT I assume?
Sorry if I'm being dumb here, but your comments about checking and adjusting the idle voltages are confusing to me. Are you saying that you want me to put everything to zero and check them again, and then while monitoring the voltage turn the second set of bias pots to see if this has any effect?
THANKS!
EDIT: After looking at the schematic, I'm thinking/guessing that test points 17 and 19 = output before the filters...so that's where I would connect the black probe, while using the red at transistor bases to get voltages referenced to the output. Can anyone help me out with a confirmation?
Yes. left channel was set to zero resistance on the 100ohm pot this yields a bias of 53mV (with the unit on DBT). The right was not; it was set to yield the specified bias voltage while off the DBT (early on in the process, prior to me realizing I had a problem on the left channel). It is allowing a bias of 14mV on the DBT, and 56mV off of it. I will turn it to zero for the next round of measurements. The second set of pots has been at zero throughout all testing.
To be clear...what do I connect to as reference for the base voltages in this next round? And, still on the DBT I assume?
Sorry if I'm being dumb here, but your comments about checking and adjusting the idle voltages are confusing to me. Are you saying that you want me to put everything to zero and check them again, and then while monitoring the voltage turn the second set of bias pots to see if this has any effect?
THANKS!
EDIT: After looking at the schematic, I'm thinking/guessing that test points 17 and 19 = output before the filters...so that's where I would connect the black probe, while using the red at transistor bases to get voltages referenced to the output. Can anyone help me out with a confirmation?
Last edited:
1. not on dbt
2. pin 15 for the left, pin 20 for the right as the reference - filters are actually on another board AWR-182 power supply.
3. yes BOTH with zero ohms - otherwise we are comparing apples and oranges.In fact please redo the base to EMITTER readings at both zero ohms..
4. watch that idle current like a HAWK!!! Especially when trying adjusts of the second set of idle current pots (50k? 100k?)
2. pin 15 for the left, pin 20 for the right as the reference - filters are actually on another board AWR-182 power supply.
3. yes BOTH with zero ohms - otherwise we are comparing apples and oranges.In fact please redo the base to EMITTER readings at both zero ohms..
4. watch that idle current like a HAWK!!! Especially when trying adjusts of the second set of idle current pots (50k? 100k?)
spark1
Lunatic Member
Mark - I can certainly re-do the earlier measurements with all bias pots turned to zero resistance (specifically, voltages across the emitter resistors and voltages across base to emitter on the driver and output transistors). Earlier you had told me to do this while on the DBT. When I re-do these measurements, should it be with or without the DBT?
After I re-do those measurements, my understanding is that you then want me to do a different set of measurements; specifically, between the output pins (TP 15 for left, TP 20 for right) and the bases of the driver and output transistors (black on test point, red on transistor bases). And, you want me to do this OFF the DBT (and with all pots set to zero resistance). Correct?
To refresh you on the current situation; when off the DBT and with pots at zero resistance (which I have double and triple checked), the left channel idle voltage = approximately 270mV. Turning the 100ohm pot moves this voltage even higher, as you would expect.
As I mentioned earlier in the thread, the right channel was at the correct initial bias setting (56mV off the DBT, 14mV on the DBT) for all the testing done to this point. The left channel bias pots were at zero for all testing, and the secondary bias pot on the right channel was also at zero for all testing.
When you say "watch the idle current like a hawk", I still don't understand what you want me to do. We already know that even with the 100ohm pot at zero, the left channel bias is way above normal. We also know that adding resistance with the primary bias pots makes the bias voltage rise...I've already done it. Are you saying that you want me to see what happens when I add resistance with the secondary bias pot, while leaving the first pot at zero? FYI, the new secondary bias pots are 100Kohm.
I must admit that I'm a bit nervous about running the unit off the DBT and then pushing the left channel bias up even further past its current 270mV baseline.
After I re-do those measurements, my understanding is that you then want me to do a different set of measurements; specifically, between the output pins (TP 15 for left, TP 20 for right) and the bases of the driver and output transistors (black on test point, red on transistor bases). And, you want me to do this OFF the DBT (and with all pots set to zero resistance). Correct?
To refresh you on the current situation; when off the DBT and with pots at zero resistance (which I have double and triple checked), the left channel idle voltage = approximately 270mV. Turning the 100ohm pot moves this voltage even higher, as you would expect.
As I mentioned earlier in the thread, the right channel was at the correct initial bias setting (56mV off the DBT, 14mV on the DBT) for all the testing done to this point. The left channel bias pots were at zero for all testing, and the secondary bias pot on the right channel was also at zero for all testing.
When you say "watch the idle current like a hawk", I still don't understand what you want me to do. We already know that even with the 100ohm pot at zero, the left channel bias is way above normal. We also know that adding resistance with the primary bias pots makes the bias voltage rise...I've already done it. Are you saying that you want me to see what happens when I add resistance with the secondary bias pot, while leaving the first pot at zero? FYI, the new secondary bias pots are 100Kohm.
I must admit that I'm a bit nervous about running the unit off the DBT and then pushing the left channel bias up even further past its current 270mV baseline.
Last edited:
Somehow, I completely overlooked that Q21-24 are 2SA750, a known troublemaker. Replace them with KSA1015, or 2SA1015 and replace Q19, 20 and 25,26 with KSC1815/2SC1815.
Also, replace Q13,14 with KSA1015. For now, only do the bad channel.
We're not even going to go any further with this until the known junk transistors are out of there.
Your measurement of Q19, 23 got me thinking "hmm...somethin's leaky here". It was only then that I checked the parts list.
Also, replace Q13,14 with KSA1015. For now, only do the bad channel.
We're not even going to go any further with this until the known junk transistors are out of there.
Your measurement of Q19, 23 got me thinking "hmm...somethin's leaky here". It was only then that I checked the parts list.
Last edited:
spark1
Lunatic Member
Interesting. So the first thing you do when servicing an SA-8800 is replace all those transistors, and the bias pots? Or would this only be the course of action if you were seeing a problem?
I don't have any 1015s on hand, so will have to add some to my next Mouser or Digikey order. That will give Mark time to provide his thoughts on this course of action as well. I don't want to follow two different paths...it's been confusing enough with just one!
Weird that you were not getting notifications. I sent you a PM after you went AWOL, asking if you were dropping out or just taking a break.
I don't have any 1015s on hand, so will have to add some to my next Mouser or Digikey order. That will give Mark time to provide his thoughts on this course of action as well. I don't want to follow two different paths...it's been confusing enough with just one!
Weird that you were not getting notifications. I sent you a PM after you went AWOL, asking if you were dropping out or just taking a break.
I always replace those transistors (2SA750) whenever/wherever I see them, whether they are acting up or not. I ascribe a fairly high probability that one of them is leaky, and replacing them will resolve your issue. KSA992 is probably OK there as a sustitute as well if that is what you have. Like I said before, do the bad channel first, that way if the problem isn't solved, we still have the good one for a reference.
As for bias trimmers, yes, I always replace them immediately on any amp that uses this type of biasing scheme. It's just cheap insurance.
AK has been a little weird since they went to the new format. Sometimes I get notifications, sometimes I don't.
Oldsansui441
Super Member
Good call tarior, l just did the trim pots on an SX-3800 as when one of the bias trim pots was set to min it would suddenly go to 1k instead of 1ohm
.
