The wires being connected to wrong channel on a working amp will only reverse what you hear as L and R but it also makes bias adjustment via the method in the manual meaningless because the method relies on using the speaker terminal as one of the measurement points.
There just seems to be so many man made errors with this 🙂 that it makes it difficult to follow. Measuring voltage across any of the 0.47 ohms is the sure fire way to get the correct result.
so ignore this, what you have on those 2 i think are corrct.when princomes off it normaly shows a pattern of what was solder print, but i dont think it is in this instance as this is where those 2 transistors are located, and the print that has come off is both soler points and non solder areas.the diagrams are not particularly clear.
That circled in red solder pad had lifted but looked still intact and was soldered to the component. But to be certain I soldered in a bridge. I re-flowed the circled in blue connection and it is not in contact with that other track - no continuity.
Ain't that the truth.
To quote Monty Python...."My brain hurts."
Measuring again across all of the 0.47 ohm resistors, each on the left side group read a few millivolts. A little different than in post #196 but maybe that is because of varying lengths of time that the amp has been idling?
The right side group read 0.0mV across each one. And also as in post #196, bias is still 4.5 volts on both channels.
To avoid confusion, mostly mine, should I reverse that red and white wire on the speaker switch to what they were when I got the amp? The balance control will operate backwards but the power amp channels will be hooked to the correct speaker terminals.
That sounds good. Yes, bias current does drift with temperature. All normal.
Bias is the voltage across the 0.47 ohm resistors 🙂 Offset (which we are not measuring at this point) is the voltage at the speaker output as measured from ground. This should always be close to zero. If the voltage on the 0.47 ohm is 4.5 volts as measured from ground then you have DC offset problem which is unrelated to bias current.
It sounds to me like you are measuring a DC offset on the faulty channel and that is what would happen if the speaker connections were reversed.
Given there is so much confusion over all this its best if you look at each speaker terminal feed and make sure it goes to the correct place. The output of each amplifier (the 0.47 ohm resistors) should ultimately go to the PLUS speaker terminal for that channel. The speaker negative terminal should go to ground.
Over 20 years ago (at the time when a friend became a member of eBay Germany, which had just been founded at the time), I received an integrated amplifier, and the owner told me that I too would not be able to get it back into working order (I think it was a NAD 3130).
In previous repair attempts, the output transistors were replaced 5 times without success (as were numerous other components on the main board around the power amplifier).
Since the service-friendly design of this model made it possible to quickly get an overview of the condition of the solder side of the main board, I decided not to start the repair until I had an other device of the same model with a power amp unit in working condition (solder site of main board around the power amp parts looks very similar to those from image in post #203.
I found a cheap model with various errors in the pre-amplifier unit (like crackled volume control) but with power amplifier in perfect working condition.
I then desoldered all components in areas where there was a soldering condition on the underside of the board as image of post #203 show.
Then I removed the solder and then the green solder mask (with very fine sandpaper). As next step I tinned all the conductor tracks and recreated the missing conductor tracks with pre-tinned copper wires. The second device with power amplifier in working condition served as a template for this.
After re-soldering the parts on the main PCB -,partly replaced with new ones - the procedure as described in post #113 (120) was carried out (additional introducing 0,47R emitter resistors on the power transistors, which were missing in the original state) and this integrated amplifier still works today without any problems.
Because a second device of this SONY-Model TA1150 isn't very hard to find, I think that the same approach leads to the goal most quickly here.
P.S.: Maybe one of the moderators can add the brand and model naming "TA-1150" in the headline of the thread.
In previous repair attempts, the output transistors were replaced 5 times without success (as were numerous other components on the main board around the power amplifier).
Since the service-friendly design of this model made it possible to quickly get an overview of the condition of the solder side of the main board, I decided not to start the repair until I had an other device of the same model with a power amp unit in working condition (solder site of main board around the power amp parts looks very similar to those from image in post #203.
I found a cheap model with various errors in the pre-amplifier unit (like crackled volume control) but with power amplifier in perfect working condition.
I then desoldered all components in areas where there was a soldering condition on the underside of the board as image of post #203 show.
Then I removed the solder and then the green solder mask (with very fine sandpaper). As next step I tinned all the conductor tracks and recreated the missing conductor tracks with pre-tinned copper wires. The second device with power amplifier in working condition served as a template for this.
After re-soldering the parts on the main PCB -,partly replaced with new ones - the procedure as described in post #113 (120) was carried out (additional introducing 0,47R emitter resistors on the power transistors, which were missing in the original state) and this integrated amplifier still works today without any problems.
Because a second device of this SONY-Model TA1150 isn't very hard to find, I think that the same approach leads to the goal most quickly here.
P.S.: Maybe one of the moderators can add the brand and model naming "TA-1150" in the headline of the thread.
Last edited:
It was an unexpectedly busy week. It is near midnight, July 3 and the soonest opportumity I've had to spend more time with this Sony. There were other things that needed attention. Bathroom flooring buckled, lawn mower crapped out and two vet visits with my dog.
I described switching two wires on the speaker switch 3 years ago so the balance control worked properly. I put them back to where they were. So now left channel is left channel and right channel is right channel. I think I found where the previous repairer goofed. I believe that the balance control will now function correctly.
I changed those wires around last night and the amplifier has been ideling for about 24 hours. It's a good thing that I had no wild expectations of bias and offset both magically falling into place.
What I see now is:
Left bias adjustible from minimum 0.0mV to maximum 23mV and right channel bias adjustible from minimum 0.0mV to maximum 5mV.
Left channel offset 4.2 volts and right channel offset 32mV.
EDIT: I again measured voltages across each 0.47 ohm resistor. Left channel resistors showed 2.4mV to 3.5mV and right channel resistors were all 0.0mV.
There is continuity between all of the left channel 0.47 ohm resistors and left speaker + and all of the right channel 0.47 ohm resistors and right speaker +.
And yes, I am seeing the same voltages between 0.47 ohm resistors and ground as the offset voltage measurements. 4.5 volts at left channel resistors and 32mV at right channel resistors.
So, yet another problem then?
I have seen numerous Sony TA-1150 amplifiers for sale on ebay but with asking prices of several hundreds of dollars. An intact TA1150 power amp board would be far more usefull and affordable. So far, unfortunatly, the only such board I've seen offered for sale is in bad condition. Perhaps worse than the one I have.
I described switching two wires on the speaker switch 3 years ago so the balance control worked properly. I put them back to where they were. So now left channel is left channel and right channel is right channel. I think I found where the previous repairer goofed. I believe that the balance control will now function correctly.
I changed those wires around last night and the amplifier has been ideling for about 24 hours. It's a good thing that I had no wild expectations of bias and offset both magically falling into place.
What I see now is:
Left bias adjustible from minimum 0.0mV to maximum 23mV and right channel bias adjustible from minimum 0.0mV to maximum 5mV.
Left channel offset 4.2 volts and right channel offset 32mV.
EDIT: I again measured voltages across each 0.47 ohm resistor. Left channel resistors showed 2.4mV to 3.5mV and right channel resistors were all 0.0mV.
There is continuity between all of the left channel 0.47 ohm resistors and left speaker + and all of the right channel 0.47 ohm resistors and right speaker +.
And yes, I am seeing the same voltages between 0.47 ohm resistors and ground as the offset voltage measurements. 4.5 volts at left channel resistors and 32mV at right channel resistors.
So, yet another problem then?
I have seen numerous Sony TA-1150 amplifiers for sale on ebay but with asking prices of several hundreds of dollars. An intact TA1150 power amp board would be far more usefull and affordable. So far, unfortunatly, the only such board I've seen offered for sale is in bad condition. Perhaps worse than the one I have.
Those are fine at this stage in the game and at least it is not drawing excess current now. I would probably set both channels to minimum bias now while we carry on fault finding.
The 4.5 volts is a problem for sure. And just to be sure, that is plus 4.5 volts from ground?
You need to check the voltage on the base of Q301 (the first transistor at the input). It should be slightly negative, perhaps as much as -200mv or so. What voltage do you see?
Now check the voltage on the base of Q302, the other input transistor. In a working amp that should be equal to the other transistor base voltage. What voltage do you see?
The emitter of both those transistors are joined and should be at around -0.7 volts give or take. Maybe a little more negative than that. What do you see?
Now place your meter negative lead anywhere on the CB1, L301, R325 line (any of those parts and any end of those parts will do as we are now going to measure from this point to elsewhere in the circuit) and measure the voltage (using the red meter lead) to the base of Q307 and Q309.
You should see about +0.5 volts.
Now move the red meter lead to R313. You should should see about +1.1 volts on either end of that resistor.
Finally if you put your meter on AC volts and now measure the voltage across R324 (the 10 ohm in series with a cap at the amp output) you should see zero volts. 0.000 volts.
It is minus 4.5 volts from ground.
It is -145.5mV.
It is -4.2 volts
It is -.697 volts at base of both Q301 and Q302
It is +0.594 volts
It is +1.18 volts
It is 0.005 volts AC
Thanks, that's all useful results. We need to be very methodical now and work through this really slowly and carefully.
Typo 🙂 you mean emitter of both. The emitters are joined.
Try and follow this and lets see where it breaks down by you measuring along the way...
Working with -4.2v on the base of Q302 and -0.145v on the base of Q301 should totally unbalance those two transistors and that should cause Q301 to turn fully on. This means Q301 pulls base current through Q303 turning on Q303 fully. With Q303 fully on you should have a very high positive voltage on the Q303 collector and that in turn should switch on fully the upper NPN driver and output and give you an offset of +30 volts or more.
Now that is not happening because you have a -4.5 volt offset so we need to see where in that chain it all goes wrong. Remember we are measuring this in the faulty state and so these are the voltage I would expect in this state.
So things to check.
1/ Measure the voltage across R307. There should be at least 2v here.
2/ Measure the voltage across R310. There should be 2v or more across this resistor.
3/ Measure the collector voltage on Q303. That is the lead going to R313 and R309. You have already measured the voltage here as +1.18 volt so this double checks that result.
Q303 is actually a prime suspect at this point. Make sure you have it fitted correctly. Lets try and prove it all first though
4/ Finally one careful measurement (don't short anything 🙂) between Base and Emitter of Q303. You should see about 0.7 volts across this junction and the emitter should be the more positive.
Typo 🙂 you mean emitter of both. The emitters are joined.
Try and follow this and lets see where it breaks down by you measuring along the way...
Working with -4.2v on the base of Q302 and -0.145v on the base of Q301 should totally unbalance those two transistors and that should cause Q301 to turn fully on. This means Q301 pulls base current through Q303 turning on Q303 fully. With Q303 fully on you should have a very high positive voltage on the Q303 collector and that in turn should switch on fully the upper NPN driver and output and give you an offset of +30 volts or more.
Now that is not happening because you have a -4.5 volt offset so we need to see where in that chain it all goes wrong. Remember we are measuring this in the faulty state and so these are the voltage I would expect in this state.
So things to check.
1/ Measure the voltage across R307. There should be at least 2v here.
2/ Measure the voltage across R310. There should be 2v or more across this resistor.
3/ Measure the collector voltage on Q303. That is the lead going to R313 and R309. You have already measured the voltage here as +1.18 volt so this double checks that result.
Q303 is actually a prime suspect at this point. Make sure you have it fitted correctly. Lets try and prove it all first though
4/ Finally one careful measurement (don't short anything 🙂) between Base and Emitter of Q303. You should see about 0.7 volts across this junction and the emitter should be the more positive.
I measured 37.5 volts
I measured 2.9 volts
I measured (-)3.1 volts ground to collector of Q303. I re-checked the voltage at R313 (black probe to R325 and red probe to R313 and verified that it is +1.18 volts.
I measured 0.57 volts
Last edited:
This seems wrong.
You should measure the voltages (against ground) at resistor endpoints throughout the whole vertical chain (R306-R307-R308-VR301-R309-R310 and for double checking the voltage across each one of them.
Resistors in that chain can be open (e.g. R307) or short (e.g. R306) even without visual indication.
Last edited:
1/ I'm sorry, I've given a wrong resistor reference for that. It should have been R304 (2k7) but no matter at this point because the next step shows it is probably correct and around 2v
2/ 2.9 volts is getting a little on the high side if the 270 ohm is OK. 2.9 v across 270 would over 10ma but where is that current going? Something amiss here I think.
3/ Hmm. -3.1v on the collector definitely doesn't compute. It should be sky high and positive in this faulty state.
Switch off and check that R310 (270 ohm) is OK and not high in value. You should isolate one end of it to get an accurate reading if there is any doubt. That resistor going high in value would give all the symptoms you have. A high negative offset (how high depends how much in value the resistor has increased), it would also give the -4.5 volts on the base of Q302 that you have and also give the -3.1 or so on the collector of Q303.
Also check R312 (the 100 ohm). You can probably check that in circuit. If it reads 100 its fine.
4/ That is fine and would still be expected to be so in this faulty state.
2/ 2.9 volts is getting a little on the high side if the 270 ohm is OK. 2.9 v across 270 would over 10ma but where is that current going? Something amiss here I think.
3/ Hmm. -3.1v on the collector definitely doesn't compute. It should be sky high and positive in this faulty state.
Switch off and check that R310 (270 ohm) is OK and not high in value. You should isolate one end of it to get an accurate reading if there is any doubt. That resistor going high in value would give all the symptoms you have. A high negative offset (how high depends how much in value the resistor has increased), it would also give the -4.5 volts on the base of Q302 that you have and also give the -3.1 or so on the collector of Q303.
Also check R312 (the 100 ohm). You can probably check that in circuit. If it reads 100 its fine.
4/ That is fine and would still be expected to be so in this faulty state.
🙂 I gave the wrong resistor reference. Hopefully I've got a handle on this now. The seemingly high VAS current (voltage across R310) and the negative collector volts on the VAS transistor is all wrong. If that resistor is high the VAS current could be far to low and unable to pull the output stage drive voltage high enough.
At this moment it all fits into place... hopefully 🙂
Voltages at resistor end points measured to ground:
R306 (-)38.7 volts and (-)41.0 volts
R307 (-)6.0 volts and (-)41,6 volts
R308 (-)5.3 volts and (-)5.9 volts
VR301 (-)5.4 volts and (-)5.4 volts
R309 (-)3.2 volts and (-)5.0 volts
R310 (+)37.4 volts and (+)41.0 volts
Voltages across resistors:
R306 2.8 volts
R307 35.7 volts
R308 0.65 volts
VR301 36.0 volts
R309 1.89 volts
R310 2.96 volts
I detatched one end of R310 and it measured 267 ohms.
In curciut check for R312 is 99.3 ohms.
I was really hoping that 270 was high...
OK, we persist with the same line of reasoning now because something doesn't add up around Q303.
With 2.9 volts across R310 means there has to be about 11 ma flowing in that resistor. Look at the circuit and see if this makes sense:
The only valid path is via the vbe multiplier and the R306/307 chain to the negative rail. Base current into the driver stages should be miniscule so we can discount that. 11ma flowing in that R306/307 chain would bring the voltage up to near the positive rail value and that is not happening. You would get 11ma if Q303 was faulty (near short) but that would put supply voltage onto the collector and yet you have just -3.1 volts here.
The theory (2.9v across 270 ohm) is absolute. There is 11ma there. There doesn't seem to be any viable path.
Lets try and prove this. Kirchhoff's law... the sum of currents leaving a node must equal the sum of currents entering it.
Look at Q303. The current enters via the emitter and this is the 11ma we calculate. (Might be worth rechecking that 2.9 volt measurement across that resistor R310). The only place that 11ma can go is out via the collector and out via the base.
The collector current can only flow into these paths and we can calculate the total easily by just measuring voltages across the resistors, calculate the current and add them together.
The current in R306 should normally equal the current in R310 but that would mean around 36 volts across R306.
Non of this is making sense now for there to be just a single fault or problem causing this. Not that I can just see anyway... more than one problem could it but it becomes ever so much harder at a distance to diagnose this I'm afraid.
I say a single fault, I can see only one other possible single fault that might do this and it would Q311 but it would have to leaky from base to collector and open on the emitter/collector path.
All I can report at this time is that there is 2.9 volts across R310. Also that I pulled Q311 and it checks as good. I looked closely at the solder pads fir Q311 and they are good.
But I did notice something, perhaps.
That is the solder pad and trace for emitter of Q306. It could be that there is missing pad/trace material. I cannot see well enough tonight to examine it. I need better light.
Maybe it's my aging eysight but with this board I find it hard to tell what is trace and what is mask between traces. There is little colour contrast.
If the emitter of Q306 was not connected, would that factor in to this problem? There was significant damage to solder pads of Q306 and was one of the areas I had to create jumpers. I may have missed one.
But I did notice something, perhaps.
That is the solder pad and trace for emitter of Q306. It could be that there is missing pad/trace material. I cannot see well enough tonight to examine it. I need better light.
Maybe it's my aging eysight but with this board I find it hard to tell what is trace and what is mask between traces. There is little colour contrast.
If the emitter of Q306 was not connected, would that factor in to this problem? There was significant damage to solder pads of Q306 and was one of the areas I had to create jumpers. I may have missed one.
That is high. Lets just go back on something I missed in your measurements on Sunday.
R306 and R307. These should have equal volt drop across them as they are in series and the same value.
(R311 that connects to the junction of these resistors is AC coupled by cap C308 and so should not play any part in causing an error here. As a check the voltage across R311 should always be zero)
Can you measure the value of these two 3k3 resistors and make sure that both are 3k3 as per the circuit. Make sure there are no shorts or solder blobs causing a problem around them. Have these been replaced before and maybe a wrong value fitted? Perhaps measure the value in circuit first and see if the one with only a couple of volts across it reads low in value in circuit.
I can't believe I missed not seeing that earlier. The voltage across those two has to be equal at all times.
---------------------------------------------------------------------
Q306 emitter on its own if not connected I can't really see causing this. Obviously check it carefully though if there is doubt because having more than one problem is something I do suspect as possible. So it could be one part of the puzzle.
Voltage drop across R306 and R307 are not equal.
R306 2.9 volts
R307 31.3 volts
Voltage across R311 is basically 0,0 volts. It flickers between 0.0 volts and 1.0mV. Or perhaps it was between 0.0 volts and 0.1mV
The in circuit measurments for R306 and R307 are not equal. Only R307 measures correctly.
R306 2.2K ohms
R307 3.3K ohms.
There are no shorts or blobs around them.
Neither has been replaced, they are both original. And both are marked with correct banding for 3.3K ohms - ORN/ORN/RED.
I shall hold off on scrrutinizing the Q306 emitter trace for the momant.