Yamaha RX-A380 Protection Circuit

Hello,

I have a Yamaha RX A830 in protection mode everytime I start It ! Managed to turn It on to self-diagnostic mode, showing DC PRT 2L.

According to the manual, Cause: DC output of the power amplifier is abnormal.
Supplementary information: The protection function worked due to a DC voltage appearing at the speaker terminal. A cause could be a defect in the amplifier.

I checked and I have -5V on one of the Front Right (FR) Channel, only in speakers output (others output are around a few mV, same for preamp outputs)

A friend of mine took It, cut, removed and tested the output transistors marked in green Q1052C and Q1052A. Working correctly, PNP and NPN. We tried to start the amplifier again without the transistors and still -5V at the speaker output, and circuit protection still triggers.

1694554444957.png


My guess is that the first stage of amplification might be faulty (labelled Q1038 and Q1045 if I get It right) ? Would you identify other usual suspects ?

Nothing seem burned/boiled/exploded/.... as far as I can see.

Thank you :)
 
Welcome to the forum!

Q1038 and Q1045 would most commonly be described as driver transistors; they are emitter-follower stages with unity voltage gain, but lots of current gain. The "first stage of amplification" would be typically regarded as the Q1005, Q1006 differential input pair.

I think you can still troubleshoot with the output devices absent--- the amp should be less susceptible to damage. To have the output at -5V is an unusual problem--- more often it shorted to a supply rail by a failed output transistor.

To locate the fault, I suggest beginning by confirming power supply voltages. According to the schematic, there should be +56V at collector of Q1038, -56V at anode of D1008, +61V at collector got Q1006, -56V at collector of Q1045, -61V at collector of Q1045. Measure voltage across R1081; if there's 0.6V, it's likely that the Q1031 current source is working.

To get a first snapshot of amplifier voltage bias, would you measure the following voltages with respect to ground:

Base Q1005
Base Q1006
Junction of R1017, R1031, C1031
Amp output

Base Q1038
Base Q1045

Thanks!
 
Ok I think I managed to measure almost everything except one.


Collector (middle pin) of Q1038 : (+56V ?) : +62V and dropping….


-56V at anode of D1008 ? D1008 is not accessible to measurement, but 0V at base of Q1031 measured at R1073 and -56V at emitter of Q031 measured at R1081.
+61V at collector of Q1006 ? +56V measured
-56V at collector of Q1045 ? 5,6 V measured
-61V at collector of Q1045 ? same question ?

R1081 nothing.

Base Q1005, measured at junction of R1003, C1010, C1017 : 0V
Base Q1006, measured at junction of R1045, C1031, C1038 : 0V

Junction of R1017, R1031, C1031 : 0V

Amp output : -4.7V

Base Q1038 : -4.1V
Base Q1045 : -6.1V

Hope It's helping ! Just remember that Q1052A and Q1052C were cut off.

1695337277038.png


Thank you.
GS.
 
The most glaring at mystery is the voltages at current source Q1031: there should be -56V at anode of D1008 and R1082, and about 0.6V across R1081. [Voltages at bases of Q1038 and Q1045 suggest Q1031 might be working, but the voltage across R1081 doesn't support that notion.]. Would you double check that area?

I would expect at least a few mV at bases Q1005 and Q1006. Would you report them with mV resolution if your meter has that sensitivity? Would you also voltage across R1010 (collector resistor for Q1005)?

Thanks!
 
Ok I think I managed to measure almost everything except one.


Collector (middle pin) of Q1038 : (+56V ?) : +62V and dropping….


-56V at anode of D1008 ? D1008 is not accessible to measurement, but 0V at base of Q1031 measured at R1073 and -56V at emitter of Q031 measured at R1081.
+61V at collector of Q1006 ? +56V measured
-56V at collector of Q1045 ? 5,6 V measured
-61V at collector of Q1045 ? same question ?

R1081 nothing.

Base Q1005, measured at junction of R1003, C1010, C1017 : 0V
Base Q1006, measured at junction of R1045, C1031, C1038 : 0V

Junction of R1017, R1031, C1031 : 0V

Amp output : -4.7V

Base Q1038 : -4.1V
Base Q1045 : -6.1V



l suggest checking resistor values Numbers 1117, 1118,1138, 1139,1150 in circuit with power off to see these are still intact. Some of these have a label IP current protection in theory these expire to avoid damage to semiconductors.

Also while the power is off use the diode test range for the base emitter junctions for the driver devices.For npn types put the black lead on the base and the red on the emitter and reverse the leads for pnp. The forward voltage should be roughly 0.6 volts. As a rough check putting the leads on the emitter and collector there should be conduction only in one direction. If there is conduction in both directions there is a short circuit and if nil open circuit.

The readings you have given are quite odd you have - 6.1volts at base Q1045 and + 5.6 volts at the emitter. If this is a polarity error whereby Q1045 has 0.5volt forward bias on the threshold of conduction then the voltage drop across R1118 at 120R which connects to the output line could reduce the dc offset from -5.6 to -4.7 volts if R1117 is open circuit.

You could measure the voltage drop across R1118 with power on to answer this question.


Hope It's helping ! Just remember that Q1052A and Q1052C were cut off.

View attachment 1215727
Ok I think I managed to measure almost everything except one.


Collector (middle pin) of Q1038 : (+56V ?) : +62V and dropping….


-56V at anode of D1008 ? D1008 is not accessible to measurement, but 0V at base of Q1031 measured at R1073 and -56V at emitter of Q031 measured at R1081.
+61V at collector of Q1006 ? +56V measured
-56V at collector of Q1045 ? 5,6 V measured
-61V at collector of Q1045 ? same question ?

R1081 nothing.

Base Q1005, measured at junction of R1003, C1010, C1017 : 0V
Base Q1006, measured at junction of R1045, C1031, C1038 : 0V

Junction of R1017, R1031, C1031 : 0V

Amp output : -4.7V

Base Q1038 : -4.1V
Base Q1045 : -6.1V

Hope It's helping ! Just remember that Q1052A and Q1052C were cut off.

View attachment 1215727

Thank you.
GS.

Thank you.
GS.
 
Last edited:
I’m in general agreement with mjona’s recommendations. Like him, I wondered if the Q1045 emitter voltage was intended as -5.6V. If there was symmetry in the Q1038/Q1045 pair and if R117 and R118 were equal, output would be -5.1V. Not the -4.7V you reported, but maybe the voltages were drifting.

Are Q1052C and Q1052A still removed? This seems a reasonable precaution.

I had hoped that the measurements would reveal a failed part, but there are too many strange voltages at the Q1005, Q1006 input pair for any conclusions.


I recommend probing both sides of R1045 to confirm that the output voltage appears on right and recheck base voltage on the left. If correct, the base curre is way too large. Further, Q1006 base voltage is higher than Q1005 base, Q1005 should be off; but there’s there’s 1.06V delivered across R1010. Something is very wrong.

Other channel is working, so use it for guidance for reasonable voltages.

Good luck.
 
It is best to have one of the meter probes connected to earth or zero volts with double case insulation allowing the use of the probe other with one steady hand testing the voltages at critcal points for readings with power on. With close transistor lead spacing working with two hands inside an amplifier case can lead to accidental damage to components and to you with high voltages from capacitors and mains supplies.
 
  • Like
Reactions: 1 users
Thanks, Wakh.

The only thing I would add in clarification is that the base voltages at Q1005 and Q1006 will be negative by a few tenths of a volt due the bias currents flowing in R1017 and R1045. Ideally, they would have equal voltage drops, leading to 0V at amp output; there's likely to be some output offset error.
 
  • Like
Reactions: 1 user
Ok, thank you everyone !

It's good advice to compare L/R channels and expect measured values to be identical ! But, since I removed the output transistors, should I expect any differences in measurements ?

I'll give a try to simple resistance measurement to see if anything is dead here.

Thanks for the warning, It's true that some measurement points are not easily reahcable :(

GS.
 
I might be overlooking an issue, but I think most voltages should be very similar even with absent power transistors. First challenge is to discover what's causing the strange DC at the output.

Once that's resolved, exercise caution re the bias when returning output devices, especially if they are replacements. I haven't confirmed, but I suspect there's a service procedure, since there's no pot provided for trimming bias current.

Let us know how things go...

Good luck!
 
If you have a service manual you can download the component list and print this out and make a start on measuring these on your pcb. This is an approach I used to repair a Sansui power amp belonging to my brother earlier this year.

To keep the list in step with the diagram I used high lighter pens of green and yellow to mark good and bad values respectively and marked each item off on this and on the component list as I went. This allows taking a break without missing any clues.

The problem with the Sansui was obviously a faulty output section where an output and driver needed replacing and also a couple of resistors. I checked the semiconductors using the same method. It is logical to work backwards from the output stage.

If the fault lies with resistors and driver output transistors test ok, then after replacing the resistors it will be safe to power up to see if the dc offset returns to normal in comparison to the other channel.

I fit 100 Ohm resistors in voltage supply rails to avoid unforeseen errors causing disaster at power up. There was no convenient place in the Sansui amp and so I made a light bulb limiter from information on the internet and used that instead.
 
Last edited:
l suggest checking resistor values Numbers 1117, 1118,1138, 1139,1150 in circuit with power off to see these are still intact. Some of these have a label IP current protection in theory these expire to avoid damage to semiconductors.
These are OK.

You could measure the voltage drop across R1118 with power on to answer this question.
I have 0.54V across R118
I recommend probing both sides of R1045 to confirm that the output voltage appears on right
I first tried to measure the resistance, but my multimeter show 20 M ohm which is huge compared to the 33 k ohm in the schema. Could the measurement be wrong and including many other elements around R1045 ? When I compare with R1044 (other front channel), multimeter shows something closer around 30 k ohm as expected.

I measure 6.8V now across R1045 measured across C1038 since they are mounted in parallel.

And just a few mV across R1044/C1037 which is the working channel.

Should I cut out R1045 to be sure of the measurement ?

Thanks everyone :)
Once again, I am enjoying trying to save this not-so-old yamaha AV amplifier from the trash since It looks like It's not a very big default...

GS
 
Also while the power is off use the diode test range for the base emitter junctions for the driver devices.For npn types put the black lead on the base and the red on the emitter and reverse the leads for pnp. The forward voltage should be roughly 0.6 volts. As a rough check putting the leads on the emitter and collector there should be conduction only in one direction. If there is conduction in both directions there is a short circuit and if nil open circuit.
Q1059 NPN is OK
Q1052C and Q1052A are OK
Q1038 NPN is ok around 0.6V
Q1045 PNP is ok around 0.7V
Q1024 PNP is ok
Q1031 is ok
Q1017 is ok
Q1005/Q1006 are quite hidden will try to reach them but time to sleep here :)

GS.