Hi smikko,
You have already noticed your DC offset rising, and that is a great reason for the amp to go into protection. I would leave everything else alone until you have explored this issue.
Trimmer controls can oxidize over time just like a volume control. But since these controls are not moved the problem becomes worse. The only reliable cure is to replace them with new parts. To check a trimmer, just monitor the DC offset of the channel and gently move the control a small amount either way. To put the control back where it was you can usually feel a small rut in the control. Sometimes the control can be so bad that tapping it without moving it will cause it to act up. When adjusting this control the DC offset should move steadily in one direction. Sometimes they can be very hard to adjust because their range is too great.
You have two channels, either one which may trip the protection. Both channels should "feel" the same as far as sensitivity of that control.
Once you have set the DC offset, allow the amplifier to warm up by playing some music (not loud). If the protection operates, measure the DC offset on both channels and make a note as to how far off the offset has become. A drifting offset can be caused by a defective capacitor in the feedback network, or a transistor damaged (failing) but not yet open or shorted. I would have to look at the schematic to give you more specifics.
-Chris
You have already noticed your DC offset rising, and that is a great reason for the amp to go into protection. I would leave everything else alone until you have explored this issue.
Trimmer controls can oxidize over time just like a volume control. But since these controls are not moved the problem becomes worse. The only reliable cure is to replace them with new parts. To check a trimmer, just monitor the DC offset of the channel and gently move the control a small amount either way. To put the control back where it was you can usually feel a small rut in the control. Sometimes the control can be so bad that tapping it without moving it will cause it to act up. When adjusting this control the DC offset should move steadily in one direction. Sometimes they can be very hard to adjust because their range is too great.
You have two channels, either one which may trip the protection. Both channels should "feel" the same as far as sensitivity of that control.
Once you have set the DC offset, allow the amplifier to warm up by playing some music (not loud). If the protection operates, measure the DC offset on both channels and make a note as to how far off the offset has become. A drifting offset can be caused by a defective capacitor in the feedback network, or a transistor damaged (failing) but not yet open or shorted. I would have to look at the schematic to give you more specifics.
-Chris
Hi, thanks for your answer. Just to clarify: When you say offset do you mean the idle current at TP1-4 pins or the DC maesured at the speaker terminals?
I will measure once more but from what I can see the idle current at TP1-4 pins is ok and steady even when it trips.
But the DC voltage at the spk terminals starts off very low when the amp is cold, then slowly rises up to about 0.8 something volts and then it trips. Which seems to be around the threshold voltage for the protection circuit? I realize there might be other factors involved here.
Providing the links to uPC1237 datasheet and 315BEE service manual again for convenience.
I will measure once more but from what I can see the idle current at TP1-4 pins is ok and steady even when it trips.
But the DC voltage at the spk terminals starts off very low when the amp is cold, then slowly rises up to about 0.8 something volts and then it trips. Which seems to be around the threshold voltage for the protection circuit? I realize there might be other factors involved here.
Providing the links to uPC1237 datasheet and 315BEE service manual again for convenience.
So, my previous findings are verified. The idle current voltages at TP1-4 seem stable. I measured the pins while manipulating the pots back and forth and I can't see any horseplay from them. The values moves up and down with no jolts. And once warm the pins seem to keep their set values of around 7.5 mV. Measuring the spk out terminals, at first when cold they have near zero DC but then it rises when the amp is getting warm until broken off by the relay at around 0.8 V. Even when the relay opens, the TP1-4 pins are at their stable 7.5 mV. How to proceed? Should I measure for DC somewhere else to track it down? If so please indicate where and how.
Thanks again
Thanks again
Hi smikko,
Now we know for certain that you have an unstable DC offset condition. Any other measurements will simply confirm this. I need to look at a schematic, but it does appear that you have a DC servo there (the op amp). It might be a good idea to measure the DC outputs from the IC (pins 3 and 7) along with the output of that channel before the relay. D105 compensates (sort of) for the base-emitter drop of Q106. The op amp does the fine work of reducing DC offset. So you have a problem that should have been compensated for by the op amp. I do see a muting transistor right on the input, Q105, 2SC2878. Remove that part to see it normal operation is restored. If it is, you must replace the 2SC2878 with another one. These are special transistors with an unusually high reverse emitter-base breakdown voltage. They will test fine with most transistor testers, so don't bother trying. Let me know how this test works. Note that you might hear bangs or pops in your speakers because you will have removed the muting transistor. Pull them both actually.
This amplifier design isn't naturally balanced as it doesn't use a differential pair. DC offset has always been an issue with this design, not to really high levels, but it drifts around a lot compared to a normal amplifier. It doesn't sit naturally near 0 mV.
-Chris
That is called Bias, or bias current. This is what I expected you to find
This is called DC Offset, or DC Offset Voltage. You have confirmed what I expected to see.
Measure it before the relay like you are doing, not on the speaker output terminals.
Now we know for certain that you have an unstable DC offset condition. Any other measurements will simply confirm this. I need to look at a schematic, but it does appear that you have a DC servo there (the op amp). It might be a good idea to measure the DC outputs from the IC (pins 3 and 7) along with the output of that channel before the relay. D105 compensates (sort of) for the base-emitter drop of Q106. The op amp does the fine work of reducing DC offset. So you have a problem that should have been compensated for by the op amp. I do see a muting transistor right on the input, Q105, 2SC2878. Remove that part to see it normal operation is restored. If it is, you must replace the 2SC2878 with another one. These are special transistors with an unusually high reverse emitter-base breakdown voltage. They will test fine with most transistor testers, so don't bother trying. Let me know how this test works. Note that you might hear bangs or pops in your speakers because you will have removed the muting transistor. Pull them both actually.
You may have burned relay contacts by now due to the fact that you are breaking a DC current.
This amplifier design isn't naturally balanced as it doesn't use a differential pair. DC offset has always been an issue with this design, not to really high levels, but it drifts around a lot compared to a normal amplifier. It doesn't sit naturally near 0 mV.
-Chris
OK, so... Let's break this up n00b style:
OK let's see... Uhhm.. The DC servo... An IC. Is that IC11 and 21? TL082CP?
And you want the values of leg 3 and 7 which according to page 1 of this SM is "Non inverting input A" and "Output B". Correct?
OK, Q105 and Q205. I get that. Two questions:
1. Can I hurt my speakers doing this? Or, in general, having them see the nasty DC too many times? During my last tests I have been doing them without the speakers connected though. Guess I don't need them to find out if the protection circuit engages anyway?
2. There are some smaller electrolytic caps around here that I haven't switched out for new ones yet? Should I? And in that case, before or after testing/replacing the transistors? Could those, if faulty be responsible for passing the DC?
Bare with me please! I really appreciate the support and learning a little on the way...
Daniel
OK let's see... Uhhm.. The DC servo... An IC. Is that IC11 and 21? TL082CP?
And you want the values of leg 3 and 7 which according to page 1 of this SM is "Non inverting input A" and "Output B". Correct?
Err need more specifics sorry. I guess you're not talking about the TP1-4 terminals but somewhere else? I understand stuff like "the negative leg of C224" or "the speaker terminals".
OK. Bad op amp.
OK, Q105 and Q205. I get that. Two questions:
1. Can I hurt my speakers doing this? Or, in general, having them see the nasty DC too many times? During my last tests I have been doing them without the speakers connected though. Guess I don't need them to find out if the protection circuit engages anyway?
2. There are some smaller electrolytic caps around here that I haven't switched out for new ones yet? Should I? And in that case, before or after testing/replacing the transistors? Could those, if faulty be responsible for passing the DC?
Bare with me please! I really appreciate the support and learning a little on the way...
Daniel
So if I understood which part you were referring to Chris, the TL082 transistor the pinout is this:
And if the chip orientation on my PCB is this (standing like in the connection diagram), the pins are laid out like this:
And in that case the voltages are:
PIN 3 = 3.84 V
PIN 7 = 1.43 V
This is measuring with the black lead on PIN 3 (V-). If I use speaker ground instead it's different values, lower. But now it tripped so let's provide the values for the protected condition also. I'm not sure if I'm doing this right.
PIN 3 = 3.9 V
PIN 7 = 1.1 V
OK so that was that but I have a feeling it's not only the amp tripping but perhaps me as well... Think I need some guidance.
An externally hosted image should be here but it was not working when we last tested it.
And if the chip orientation on my PCB is this (standing like in the connection diagram), the pins are laid out like this:
An externally hosted image should be here but it was not working when we last tested it.
And in that case the voltages are:
PIN 3 = 3.84 V
PIN 7 = 1.43 V
This is measuring with the black lead on PIN 3 (V-). If I use speaker ground instead it's different values, lower. But now it tripped so let's provide the values for the protected condition also. I'm not sure if I'm doing this right.
PIN 3 = 3.9 V
PIN 7 = 1.1 V
OK so that was that but I have a feeling it's not only the amp tripping but perhaps me as well... Think I need some guidance.
So meanwhile I replaced all the remaining capacitors except for some hard to reach without taking the PCB out of the chassis. The ones I didn't replace are C11, C12, C301, C302, C303 and C501 (all pretty far from heat sources and from what I can gather not in the paths of input/output or the protection circuit). All others are fresh Panasonic or Nichicon caps. But the DC at the speaker terminals is still there, and I switched off the power as it went past 0,4 V figured it's unneccessary to stress the relay referring to:
I will take out the Q105/Q205 transistors next but would love some feedback on my last post before I do. Also, it's time for some sleep here in Sweden.
Look forward to hearing from you when you find the time.
Regards, D
I will take out the Q105/Q205 transistors next but would love some feedback on my last post before I do. Also, it's time for some sleep here in Sweden.
Look forward to hearing from you when you find the time.
Regards, D
Hi "D",
-Chris
Yes, IC11a, a TL082. The "a" indicating section #1, so pin 1 should be a within 200 mV of your common ground (chassis or signal ground). The TL072 has lower DC offset I think, and probably selected for lower noise and THD. It appears they only use 1/2 of each op amp. Make certain the inputs are shorted to ground in that second section (pins 2 & 3). Of course, I didn't see the second section on the schematic, so it is possible they could be used for something else. See if pins 2 & 3 are grounded (signal ground). The pads would be tied together if this is true.
Not unless we prove it is bad. I just listed the possibility and what that part is supposed to do.
The junction of R174 and R177 (0.068R) or 274 and 277. This is where the output is taken.
It could be possible, I was trying to warn you of a loud noise. You don't need the speakers connected for this test.
It is possible of course. Only you know what you have replaced. C124 would be suspect, but no worse a suspect than anything else. Leave it for now.
The dot on the case indicates pin 1. Measure voltages with respect to signal common (ground).
I goofed on the pin numbers. Measure pin #1 wrt (with respect to) signal common. Measure pins 2 & 3 for giggles to see what that can tell us. Ignore the pins 5,6 & 7 as they aren't (I don't think) part of the circuit.
No worries. The capacitors you need to worry about are in the 100's and 200's.
Relay contacts can burn when current is interrupted. Turn the volume down before turning your set on or off. Once you are done with fixing the amp, run a low level, 1 KHz tone through your speakers. Tap on the relay case and listen for the tone becoming distorted or cutting out. 1KHz is suggested because our hearing is very sensitive in this area, so it's easier to hear a problem. You might be able to generate the tone with your computer sound card. That's your homework. Using music can mask the evidence of a problem, that's why I want you to use a constant tone.
-Chris
Hello, I did the measurements:
Note that I drew the wrong pinout on my picture. Looking at the connection diagram in the data sheet I realize the pins are laid out counter clockwise from PIN # 1. So yeah I guess that you're right pins 5-6 are not used. Pin 7 have -2 V idk what that means.
This is referring to the datasheet from Texas Instruments: http://www.ti.com/lit/ds/snosbw5c/snosbw5c.pdf
I used a ground post on the PCB connected to the chassis.
OK, great that's where the TP1-2 and 3-4 connections are conveniently placed right?
So anyway - next is desoldering the tranistors? Thanks for you help!
An externally hosted image should be here but it was not working when we last tested it.
Note that I drew the wrong pinout on my picture. Looking at the connection diagram in the data sheet I realize the pins are laid out counter clockwise from PIN # 1. So yeah I guess that you're right pins 5-6 are not used. Pin 7 have -2 V idk what that means.
This is referring to the datasheet from Texas Instruments: http://www.ti.com/lit/ds/snosbw5c/snosbw5c.pdf
I used a ground post on the PCB connected to the chassis.
OK, great that's where the TP1-2 and 3-4 connections are conveniently placed right?
An externally hosted image should be here but it was not working when we last tested it.
So anyway - next is desoldering the tranistors? Thanks for you help!
Hi "D",
Is that chart measured AC volts? The voltage for the op amp supply is far too low, +1.62 - 1.75V and (-) 3.5 - 3.6 V. These would measure the same on those op amps between L and R channels.
-Chris
Is that chart measured AC volts? The voltage for the op amp supply is far too low, +1.62 - 1.75V and (-) 3.5 - 3.6 V. These would measure the same on those op amps between L and R channels.
Yes, those test points were not on the schematic. You are measuring voltage across both emitter resistors, so your bias current will be x volts/ 1.36 ohms.
Perfect.
It sure is. Let us know how that works out.
-Chris
Hey folks, and Chris!
I now measured again on the legs of the op amp and yup, it's all DC volts. And the values are about the same as before desoldering the transistors.. What does this mean then? And, it trips after about as long so this seems like a dead end. I list the last values of ic11/21 with the q105/205 transistors out:
IC11:
1=0.8V
2~0V
3=0.5V
4=3.4V
5~0V
6=0V
7=-1.99V
8=1.57V
IC21:
1=0.8V
2=0V
3=0.53V
4=-3.4V
5~0V
6=0V
7=-2V
8=1.56V
Any ideas? /Daniel
I now measured again on the legs of the op amp and yup, it's all DC volts. And the values are about the same as before desoldering the transistors.. What does this mean then? And, it trips after about as long so this seems like a dead end. I list the last values of ic11/21 with the q105/205 transistors out:
IC11:
1=0.8V
2~0V
3=0.5V
4=3.4V
5~0V
6=0V
7=-1.99V
8=1.57V
IC21:
1=0.8V
2=0V
3=0.53V
4=-3.4V
5~0V
6=0V
7=-2V
8=1.56V
Any ideas? /Daniel
Hi Daniel,
The voltages on pins 4 and 8 are too low for those chips to operate. They are supposed to be + and - 17 VDC. Because these chips can not operate it could lead to the problem you are having.
Q308 and Q307 are the regulators for your op amps. I would now check in that area. This is not a normal failure mode. Check R317 and R314 (33R 1/2 watt) to see if they are open.
-Chris
The voltages on pins 4 and 8 are too low for those chips to operate. They are supposed to be + and - 17 VDC. Because these chips can not operate it could lead to the problem you are having.
Q308 and Q307 are the regulators for your op amps. I would now check in that area. This is not a normal failure mode. Check R317 and R314 (33R 1/2 watt) to see if they are open.
-Chris
After desoldering the Q105/205 transistors I connected the amp to my speakers. It played well without pops or bangs and played for over 30 minutes without protection mode. I noted that the bias current was up a bit to around 13-14 mV from previously 7.5. The DC at the output was around 0.7-0.8V though. But maybe it keeps going because the speakers transform some of that DC to heat and thus keeping the voltage from rising until it trips? Because I think it breaks around 0.8-0.9 V.
Daniel
Daniel
The resistance metering was made in circuit though so I guess some other component might affect the reading? Are we thinking about for example the Q308 here? Sorry for probably being daft..
OK, so some other random pieces of information:
1: The media player "MP" input at the front isn't working. When a mini tele is inserted here, the input selector automatically switches to this input (like it's supposed to) but the audio is barely audible, sort of like the sound in the headphones when the is in protection mode. (I tried this before, but have since then made all the tests without headphones since I think they break the signal path and perhaps complicates things.
2: I found this amp and a C515bee CD player. I tested the CD with some active speakers and it seems fine. However, when connected to the "DISC" input of this amp, after a while the sound gets distorted and sounds really crappy. This doesn't occur when I connect my smartphone with the volume set at around 80 % to the same input.
Related or not I don't know but I thought I should mention these two anomalies as well.
Best regards, Daniel
OK, so some other random pieces of information:
1: The media player "MP" input at the front isn't working. When a mini tele is inserted here, the input selector automatically switches to this input (like it's supposed to) but the audio is barely audible, sort of like the sound in the headphones when the is in protection mode. (I tried this before, but have since then made all the tests without headphones since I think they break the signal path and perhaps complicates things.
2: I found this amp and a C515bee CD player. I tested the CD with some active speakers and it seems fine. However, when connected to the "DISC" input of this amp, after a while the sound gets distorted and sounds really crappy. This doesn't occur when I connect my smartphone with the volume set at around 80 % to the same input.
Related or not I don't know but I thought I should mention these two anomalies as well.
Best regards, Daniel
And oh yeah sorry I see now those are the very transistors you mentioned as suspects before.
I managed to forgot while measuring the resistors.I get lost with all the numbers hehe... But I almost feel a little bit proud for guessing right then?
So this might be a broken Q308? Does it mean it's shorted inside and that's were the infinite resistance of R317 is coming from? But how to explain that the other channel is shot too? Could Q307 be broken as well but not showing through R314 resistor the same way?
Thanks for your patience!
Daniel
I actually can't find 17V anywhere on the board. I measured on the links that says +17V and -17V that I think are close to where the 5V and 17V supply comes to the board via the CB32 connector. No 17V on those. Then I measured the solder joints of the CB32 connector itself and here are the readings:
PIN#
1=Around 1.2V but really erratic flicks around
2=Around 1.2V but really erratic flicks around
3=0V ground
4=4.38V
1=4.03V
So does this mean that there is no good power coming in from the PSU even? How do I know for sure, is it safe to disconnect the CB32 connector and measure the cable disconnected from the PCB?
Maybe if I was more experienced I would have started here. I realize now.
/D
Hi Daniel,
R317 is open, which would also take down the -17 volt supply as it tracks the positive output. You may have shorted the power supply at some point while replacing capacitors.
You should look at all your previous work under a bright light. Clean the solder flux off with Lacquer thinner and a tooth brush. Don't soak anything. Once the flux is gone you can inspect the work you have done. It only takes a sliver of solder to create a short that would burn out that resistor.
After your inspection, replace R317 with another 33R 1/2 watt resistor and try it again.
-Chris
R317 is open, which would also take down the -17 volt supply as it tracks the positive output. You may have shorted the power supply at some point while replacing capacitors.
You should look at all your previous work under a bright light. Clean the solder flux off with Lacquer thinner and a tooth brush. Don't soak anything. Once the flux is gone you can inspect the work you have done. It only takes a sliver of solder to create a short that would burn out that resistor.
After your inspection, replace R317 with another 33R 1/2 watt resistor and try it again.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.