Sansui G-7500 Fried Resistors F2980 Board

[kevzep]

ah ..i missed that bit .. even so knowing each voltage to see if a relative match isnt a bad idea to see ,.

We dont know anything is wrong yet, he's started with the trimmer at 50%, I want to see that trimmed down to make sure we have adjustment.
If he can bring it down to about 3V then that means the bias circuit is working and he then should be able to put the 0.33Ω Emitter resistors back in.

The 100Ω are in there to stop it running away and taking out the output transistors.

 

[Young46]

New readings with VO4 start at full CCW. Wouldn't stabilize.

 

[kevzep]

Yeah okay, so the bias is not adjusting, let me have a look here and see if I can come up with an idea as to why that might be happening.

 

[petehall347]

dc offset might be a clue right now .

 

[kevzep]

dc offset might be a clue right now .

Well, there can only be a few possibilities right?


With 280mA running through the 100Ω resistors, they must be getting a little warm? Can you confirm this @Young46

Output transistors getting warm?

 

[Young46]

Well, there can only be a few possibilities right?


With 280mA running through the 100Ω resistors, they must be getting a little warm? Can you confirm this @Young46

Output transistors getting warm?

Yes, the resistors are getting warm. I’ll have to check the output transistors.

 

[Young46]

No, the output transistors are not getting hot. The driver transistors are as well as the emitter resistors.

 

[Young46]

I meant to say that the emitter resistors and the driver transistors TR16, and 18 are getting hot.

 

[petehall347]

thinking aloud again ..base voltages of drivers should be plus minus 1.2v ,600mv between b and e ..

 

[kevzep]

No, the output transistors are not getting hot. The driver transistors are as well as the emitter resistors.

Okay, that is likely because the emitter resistors are acting the way intended and saving the output transistors.

We need to find out why the Bias is not adjusting down.

Going on past experience with these amplifiers, we need to look at all the traces on the circuit board associated with the trimmer, you need to make sure, the trimmer is making connection via the tracks to where its supposed to joined to.

The tracks on these circuit boards, and the F2980 in particular are very very fragile, they can look fine, but the crack and get hairline cracks around the solder pads. So you need to verify that the trimmer is working in circuit.
You can do this with your multimeter, look at the circuit and then Ω it out...

Also, do you have any freeze spray? You could hit the STV3 Diode on the heatsink and see if the bias current changes, it will go up.
If you have no freeze spray you could hold the soldering on the STV3 and the Bias current will go down.
This will verify the STV, these are not normally a failure point, but there is a first time for everything.

 

[Young46]

Okay, that is likely because the emitter resistors are acting the way intended and saving the output transistors.

We need to find out why the Bias is not adjusting down.

Going on past experience with these amplifiers, we need to look at all the traces on the circuit board associated with the trimmer, you need to make sure, the trimmer is making connection via the tracks to where its supposed to joined to.

The tracks on these circuit boards, and the F2980 in particular are very very fragile, they can look fine, but the crack and get hairline cracks around the solder pads. So you need to verify that the trimmer is working in circuit.
You can do this with your multimeter, look at the circuit and then Ω it out...

Also, do you have any freeze spray? You could hit the STV3 Diode on the heatsink and see if the bias current changes, it will go up.
If you have no freeze spray you could hold the soldering on the STV3 and the Bias current will go down.
This will verify the STV, these are not normally a failure point, but there is a first time for everything.

I don’t have any actual freeze spray but if I can’t find any I have a can of compressed air that will probably work. On the other hand, did you mean to say hold the soldering [iron] on it? If so how hot should should I set the temperature?

I have tested the STV’s using the method that was on a Pioneer forum. I believe it was Mark the fixer, or possibly Echo Wars. I can’t remember the exact method right now but they tested good. I will do the method you suggest to be sure. Things change.

 

[kevzep]

I don’t have any actual freeze spray but if I can’t find any I have a can of compressed air that will probably work. On the other hand, did you mean to say hold the soldering [iron] on it? If so how hot should should I set the temperature?

I have tested the STV’s using the method that was on a Pioneer forum. I believe it was Mark the fixer, or possibly Echo Wars. I can’t remember the exact method right now but they tested good. I will do the method you suggest to be sure. Things change.

Well, we just want to see if heating or cooling them will affect the Bias Current, I am sure they will be okay, but it verify that they are in circuit and functioning.
Normal temp for iron, just hold it very close, as close as you can, and watch the bias current and see if it goes down.

 

[Young46]

[QUOTE="kevzep, post: 11756969, member: Well, we just want to see if heating or cooling them will affect the Bias Current,

Thanks, will do.

 

[rcs16]

Damn those F-2980 boards. it is hard for me to get a handle on what has been done as far as parts replacement, what was used & what has been determined to be faulty. Sometimes it is good idea to summarize, so people like me can get up to speed in this long thread.
It is possible that a sub'd part(s) could make the amp become unstable for which you need to look at the o/p to see if it is oscillating. if it is, this has to be tamed or you will be going in circles. You could over compensate by increasing C17(5pF) to something higher, try a 22pF in parallel with C17, the dominate pole in the comp scheme.
We need to find out why the Bias is not adjusting down.
I agree, suggest to calculate the Current through the Vbe multiplier. I am not sure what channel is misbehaving, so I will refer to the channel schem showing values. measure voltage drop across 150 ohm R35 or R39, both should read about the same, it should be in the ~10mA range, so you should read something like 150 * 0.01= 1.5VDC. if this voltage is changing, It will point to instability.
On my fried F-2980 it had burnt traces when it failed, so you have to verify continuity of the connections.
Another idea is a quick mod to this design and add a high quality ecap in the feedback leg. This will reduce DC gain to unit. It will help a lot in setting offset. Lift one leg of R19(1K) and add a 220-300uF/6.3-16V BP ecap such a Nichicon UES in series.
I do not suggest to use the original o/p bjts ( A1106,C2581) as they are underrated for the supply V. Try to use 150W minimum bjts, 200W preferred, for a +/-71V supply.
I have a Yamaha RX-900U on the bench right now, same sort of issue, under powered ops bjts (A1265/C3182) failed and smoked many parts in the channel.

Good Luck
Rick

 

[kevzep]

Well, at the moment we are trying to verify the connection of components through the PCB....
I know those traces are very fragile.
The idea measuring the current to the driver collectors is a good one, we could do this after we find out of the STV3 is working.
I am thinking this amplifier may indeed be oscillating, but the OP doesn't have a scope so we are flying a little bit blind....

 

[ghazzer]

Is there any "backyard" technique for detecting oscillations? Not to measure the amplitude or duration, but just to get a YES/NO indication, within some boundaries/limitations.

Kevzep is correct about checking the integrity of the tracks and solder connections in the path for the "no bias adjustment" condition. My last experience with this was an intermittent solder connection on one of the emitter resistors - no current flow = no bias. However, the emitter resistors would not be heating up w/o current flow. All paths in the circuit need to be verified. Using mini-grabbers on component leads is a good way to verify connections and tracks, but cannot rule out intermittent faults.

I didn't see anything about VR03 so I am assuming that either the left channel is OK, or it has not been checked yet.

 

[kevzep]

Is there any "backyard" technique for detecting oscillations? Not to measure the amplitude or duration, but just to get a YES/NO indication, within some boundaries/limitations.

Kevzep is correct about checking the integrity of the tracks and solder connections in the path for the "no bias adjustment" condition. My last experience with this was an intermittent solder connection on one of the emitter resistors - no current flow = no bias. However, the emitter resistors would not be heating up w/o current flow. All paths in the circuit need to be verified. Using mini-grabbers on component leads is a good way to verify connections and tracks, but cannot rule out intermittent faults.

I didn't see anything about VR03 so I am assuming that either the left channel is OK, or it has not been checked yet.

This is exactly why I want to know if the STV diode is connected and functioning, and the trimmer and associated components are all correct and working.
We know there is 290mA flowing through those resistors which cannot be adjusted which is consistent with oscillation.
This is why once this step has been done we are going to look at the current flow through the collectors of the Driver transistors.
There is a method to this, and I am waiting for the results of this current step we are at before we move on.

Normally the Oscillation goes like this, you turn the amp on, bias trimmers set to minimum, and everything is normal, you try to set the bias, then at some point close to when the output transistors are turning on enough the bias will suddenly just take off and things get hot very quickly.
The output transistors don't normally oscillate without bias to them. They need a help to get started.
This is, in this case why I am looking at the bias network here.

To answer the yes/no question, experience, its something you just get experience with...

 

[Young46]

This is exactly why I want to know if the STV diode is connected and functioning, and the trimmer and associated components are all correct and working.
We know there is 290mA flowing through those resistors which cannot be adjusted which is consistent with oscillation.
This is why once this step has been done we are going to look at the current flow through the collectors of the Driver transistors.
There is a method to this, and I am waiting for the results of this current step we are at before we move on.

Normally the Oscillation goes like this, you turn the amp on, bias trimmers set to minimum, and everything is normal, you try to set the bias, then at some point close to when the output transistors are turning on enough the bias will suddenly just take off and things get hot very quickly.
The output transistors don't normally oscillate without bias to them. They need a help to get started.
This is, in this case why I am looking at the bias network here.

To answer the yes/no question, experience, its something you just get experience with...

I shot some freeze spray on the STV and nothing...no change. Tried the heat from the soldering iron with same result.

I can buy a Hitachi 223 scope with probes off local Craigslis for $100. Thoughts? Opinions?

 

[kevzep]

I shot some freeze spray on the STV and nothing...no change. Tried the heat from the soldering iron with same result.

I can buy a Hitachi 223 scope with probes off local Craigslis for $100. Thoughts? Opinions?

Hmmm okay, this means a few things.

1, The STV3 is faulty, but unlikely as I have never seen one fail. This needs further investigation.

2, The STV3 is working but not actually connected to its terminations because of cracked tracks on the circuit board.

3, The STV3 is working, but the amplifier is hard in oscillation and so has no effect.

So, can you verify what happens with the Bias when you turn the amplifier on. If the amplifier is oscillating I would expect the bias to be normal for a few seconds, and then suddenly shoot up...
Can you connect the meter, and try this, then report back?

A scope is always handy when doing this sort of work. Its a 20mhz scope so good enough for general audio work.
If you are going to do this sort of thing regularly as a hobby, it would not be a bad investment.

 

[petehall347]

would measuring voltage across the stv and each end reveal some things i wonder ?