As BSST pointed out in post #17: This amp is unusual in that it has flat gain to DC, and only the input has AC coupling.
Normally there's a cap in the lower divider to bring the DC gain down to unity. One of the first things I look for in amps that have a lot of offset is that these caps are shorted or leaky. Add in the fact that this amp has a lot of gain - over 40 - and it's very possibly you have an input transistor mismatch.
Try heating or cooling one of them gently - not a blast of freeze spray, more like just touching it - and see if the offset goes crazy.
Normally there's a cap in the lower divider to bring the DC gain down to unity. One of the first things I look for in amps that have a lot of offset is that these caps are shorted or leaky. Add in the fact that this amp has a lot of gain - over 40 - and it's very possibly you have an input transistor mismatch.
Try heating or cooling one of them gently - not a blast of freeze spray, more like just touching it - and see if the offset goes crazy.
Assuming Mute control line is enabling amps, Q417 and Q418 collectors should both be about -0.1V, voltages at Q420 look reasonable, but those at Q419 look wrong. If the amps are biasing, the voltage at the junction of R436 and R438 should be very roughly 0V, splitting the supply rails. Similarly for R435, R437.
Last edited:
I agree with you more and more, I have the feeling that something is wrong around the muting circuit.
Seems to be time to isolate q417 from the steering circuit and reach board stability in both situations - mute on and mute off - by driving its base directly.
From troubleshooting experience, reverse bias in any BE junction means some other transistor has catastrophically failed (fully open or shorted) so the rest of the fully DC coupled circuit is desperately trying to compensate.
Failing at that.
I have often seen much worse, say reverse 5 V or higher, but in any case negative few mV already shows a problem.
For sake of completeness, a few forward mV also shows a problem, because no junction can be turned on by such a low voltage.
Bonus: measuring voltages is usually better (and definitely faster) than pulling transistors and measuring outside.
And you don't butcher PCBs
Failing at that.
I have often seen much worse, say reverse 5 V or higher, but in any case negative few mV already shows a problem.
For sake of completeness, a few forward mV also shows a problem, because no junction can be turned on by such a low voltage.
Bonus: measuring voltages is usually better (and definitely faster) than pulling transistors and measuring outside.
And you don't butcher PCBs
can be reasonable so if q417 dont have with the problem to do. maby its in the preamp section ?
With no signal voltage coming in W505 L and W505 R should be nearly zero.
i have measure q417 with a ransistor tester and it looks okay
hfe = 511
ie = 2.7ma
ube = 651mv
2sa992
Last edited:
If Q417 is back in circuit, skip this paragraph. But if Q417 is still removed, measure voltage at Q417 collector pad. It should be ~~ -45V
Measure voltage at Q417 base pad. It should be some negative voltage, assuming the mute control is asking for the amp to be enabled.
Use ohmmeter to confirm emitter pad of Q417 has continuity to ground. If all is correct, reinstall Q417.
Confirm collector of Q417 is conducting and reads about ~ -50mV. Double check for correct insertion of Q419 and remeasure its voltages. Those reported in post 59 look wrong, and suggest the device is damaged. Q419 should be biased into full saturation, i.e. base should be about 0.6V more positive than emitter, and collector should be pulling R437 to -45V. Replace if failed, but be alert to possible broken traces or failed resistors; probe directly on resistor leads near the body of the resistor if any doubt of trace integrity. Junction of R435 and R437 should be only a few volts re ground if is all is correct and they are splitting the rail supply voltages.
Keep us posted. Good luck.
Measure voltage at Q417 base pad. It should be some negative voltage, assuming the mute control is asking for the amp to be enabled.
Use ohmmeter to confirm emitter pad of Q417 has continuity to ground. If all is correct, reinstall Q417.
Confirm collector of Q417 is conducting and reads about ~ -50mV. Double check for correct insertion of Q419 and remeasure its voltages. Those reported in post 59 look wrong, and suggest the device is damaged. Q419 should be biased into full saturation, i.e. base should be about 0.6V more positive than emitter, and collector should be pulling R437 to -45V. Replace if failed, but be alert to possible broken traces or failed resistors; probe directly on resistor leads near the body of the resistor if any doubt of trace integrity. Junction of R435 and R437 should be only a few volts re ground if is all is correct and they are splitting the rail supply voltages.
Keep us posted. Good luck.
Read and heed post 66. Transistor testers test at 2 volts or less. Your power sections operates at +-25 v or higher. Junctions that can withstand 2 v can sometimes not withstand dozens of volts. You need to put a tungsten lightbulb series the AC input, say 60 to 75 watts. Then attach negative of DVM or scope to center of power supply with alligator clips so you do not use 2 hands probing for voltages. >25 v across your heart can stop it. Tale all jewelry off hands wrists & neck. Then probe for voltages. I would start at q445 q407 and q409 and then go on to q405 and q407.
Problem could be a bad solder joint or leaky electrolytic capacitor in stead of transistor. Look for b to e <0.6 volts npn -0.6 volts pnp. look for Vce <2 v npn or <=2 v pnp. Or Vce opposite polarity, which means something ahead of that is wrong. Check the voltages differences by subtraction of notations on paper, not by using two hands to probe the power section with two probes.
Problem could be a bad solder joint or leaky electrolytic capacitor in stead of transistor. Look for b to e <0.6 volts npn -0.6 volts pnp. look for Vce <2 v npn or <=2 v pnp. Or Vce opposite polarity, which means something ahead of that is wrong. Check the voltages differences by subtraction of notations on paper, not by using two hands to probe the power section with two probes.
Last edited:
My point is and many share it, that only removing parts to be tested outside is not enough.
Broken tracks, cracked solder, poor contacts, lack of supply voltages, no or poor bias, etc. all play a part.
That's why turning amp on, with a bulb limiter if necessary, and measuring DC voltages, with no signal applied, no load, can lead to the suspect area.
THEN you remove a couple parts as needed if suspect.
This week a Musician in The Gear Page is repairing a Marshall guitar amplifier, he replaced everything, every single component, except jacks, pots, power and output transformers.
It still does not work 🙁
EDIT:
In post #70 you tested the transistor. Cool, it works.
Then you have not found the problem ... yet.
Post #71 gives you a set of circuit, not parts tests .
Follow them and report back.
Broken tracks, cracked solder, poor contacts, lack of supply voltages, no or poor bias, etc. all play a part.
That's why turning amp on, with a bulb limiter if necessary, and measuring DC voltages, with no signal applied, no load, can lead to the suspect area.
THEN you remove a couple parts as needed if suspect.
This week a Musician in The Gear Page is repairing a Marshall guitar amplifier, he replaced everything, every single component, except jacks, pots, power and output transformers.
It still does not work 🙁
EDIT:
In post #70 you tested the transistor. Cool, it works.
Then you have not found the problem ... yet.
Post #71 gives you a set of circuit, not parts tests .
Follow them and report back.
two ******* transistor q421,q425 is going away reading 600 and 650 hfe when max is 300 so I ordered some new once.
scheduled delivery in the beginning of February !!
scheduled delivery in the beginning of February !!
Last edited:
I’m reminded of a cartoon in Popular Electronics, circa 1965:
A man is deep into the innards of his radio with parts on his bench, and other dangling parts still semi-attached. His wife is nearby and says “I forgot to tell you, I had to unplug your ham radio when I vacuumed.”
I reiterate, with Q419 not conducting, the amp is not biased into operation. As a consequence, some voltage readings may be meaningless due to off-state leakage. This may be an explanation for some of the drifty behavior. I suggest a 10k resistor across the output to provide a very modest load, but heavier than a 10M DVM. A bias work-around is to install a short across Q419 collector-emitter. This should establish bias on the amp and defer uncertainty about Mute issues. Confirm expected voltage at junction of R435 and R437. Then you can examine any suspicious readings in output devices.
A man is deep into the innards of his radio with parts on his bench, and other dangling parts still semi-attached. His wife is nearby and says “I forgot to tell you, I had to unplug your ham radio when I vacuumed.”
I reiterate, with Q419 not conducting, the amp is not biased into operation. As a consequence, some voltage readings may be meaningless due to off-state leakage. This may be an explanation for some of the drifty behavior. I suggest a 10k resistor across the output to provide a very modest load, but heavier than a 10M DVM. A bias work-around is to install a short across Q419 collector-emitter. This should establish bias on the amp and defer uncertainty about Mute issues. Confirm expected voltage at junction of R435 and R437. Then you can examine any suspicious readings in output devices.
Last edited:
Pity the like function does not have an option for "I'm laughing my *** off!". O.P. will be weeping when the new ones are installed.
Class AB transistor amps have negative feedback to linearize the gain of intrinsically exponential gain bipolar junction transistors. Which read about in a textbook or wikipedia. Gain of transistor is different at different test currents and temperatures. Did your tester exactly duplicate the test conditions listed in the transistor datasheet?
Probably not.
The two transistors that need to have the same gain at the same operating point are the long tailed pair, Q403 & Q405 on the schematic in the service manual post #1.
Post 75 suggestion to put a 10 k load, or even a 50 ohm load, on the speaker terminals is a good idea. Then take a piece of paper, ruled, put q # (numbers) down left column with e , b, c, on separate lines. Fill up the paper with Q#. Check rail voltages (V+, v-) are symmetrical with light bulb installed in AC line. If so, measure voltages at all those points with the DVM minus at power supply center (ground). Make algebraic subtraction & judgements as explained in post 52. Violations of the rules of class AB amp behavoir, report back. I would start at the preamp input and work right towards the speaker, since this amp did not suffer the typical damage scenario of playing into a shorted speaker or wire until the output transistors blew up and took a lot of parts left of the output with them.
Last edited:
You ask for directions but DO NOT follow them 😱
Those transistors are perfectly fine (as tests prove), and NOT your problem.
Just another manifestation of:
https://en.m.wikipedia.org/wiki/Streetlight_effect
I first heard the Streetlight effect from a colleague. Another of his anecdotes:
As a train conductor was punching tickets, he encountered a passenger frantically searching for his ticket. "Have you checked your shirt pocket? Passengers often put them there" advised the conductor. Replied the passenger "No, I don't dare look there! If I looked and it wasn't there, all hope would be lost!"
😉
Best,
Steve
I don't know if I would dare to start the amplifier with some components removed. the components depend on each other in a circuit..
Collector of q417 conducts and reads -11mV
Yeah the reading in post 59 looks wrong,i will check it again 🙂
I'll check q419,R425,R437 and get back to you. continuity is checked at regular intervals
TNX have a nice day