If the parts are good and connected properly, there must be an error elsewhere telling the amp to turn off the pull up...corgan4321 said:
I'll give you some more things to measure tomorrow.
JR
Troubleshooting Wurlitzer 200a amp board for bias and crossover notch distortion
- Thread starter corgan4321
- Start date
Help Support GroupDIY Audio Forum:
I'll give you some more things to measure tomorrow.
JR
Gene Pink
Well-known member
A test that may help:
Let's throw some negative current at the output, to see if the topside has the ability to compensate for it.
Get a 2-watt resistor in the 270-470 ohm range, and while measuring the output voltage with no load, jump this resistor from the negative rail (-22V) to the output, to pull it south with dozens of milli-amps. 1/2 watt resistor works too, but it will get hot fast, but it only takes a few seconds to read the voltage at the output.
It will do one of two things:
1) the output will only drop negative a small amount, less than a half a volt, meaning the top side tr-11 and 12 do actually function as a block, but with some leakage there or elsewhere back to the long tail pair that prevents them from turning on at idle, meaning something is causing an offset and pulling north.
2) the output voltage will drop a lot, meaning that the topside has a major problem, most likely in the TR-11,12 block. But we can't discount something in the long tail pair preventing it from pulling north.
Neither of these address your original problem of crossover distortion, but are you sure that is what it is, and not one side only drive?
Sure wish you had a 'scope right now.
Gene
Let's throw some negative current at the output, to see if the topside has the ability to compensate for it.
Get a 2-watt resistor in the 270-470 ohm range, and while measuring the output voltage with no load, jump this resistor from the negative rail (-22V) to the output, to pull it south with dozens of milli-amps. 1/2 watt resistor works too, but it will get hot fast, but it only takes a few seconds to read the voltage at the output.
It will do one of two things:
1) the output will only drop negative a small amount, less than a half a volt, meaning the top side tr-11 and 12 do actually function as a block, but with some leakage there or elsewhere back to the long tail pair that prevents them from turning on at idle, meaning something is causing an offset and pulling north.
2) the output voltage will drop a lot, meaning that the topside has a major problem, most likely in the TR-11,12 block. But we can't discount something in the long tail pair preventing it from pulling north.
Neither of these address your original problem of crossover distortion, but are you sure that is what it is, and not one side only drive?
Sure wish you had a 'scope right now.
Gene
I fear I got distracted by low accuracy voltage measurements (not your fault) and my fixation on assumptions that tr10 is turned on.
The original complaint is crossover distortion which points to low class A bias in the output devices.
Ignoring for the moment that the output is not exactly 0V. lets measure the current in the emitter degeneration resistors (R37 and R 38 0.47 ohm. ). 10mA of class A bias should generate 5 mV across each resistor.
If not we need to tweak up the voltage in the Vbe multiplier TR9 . The schematic shows an optional 1k across R34... that will reduce voltage and class A bias, so if 1k is in there, remove it.
Very carefully reduce the resistance of R33... you can blow up the amp from too much bias so do this carefully maybe starting with a 5k in parallel and slowly drop lower, until you get 5mV (each) across R37 and R38.
Be careful messing with this. After getting the class A bias current right, we can worry about output voltage not being exactly 0V.
JR
PS: Sorry about getting multiple suggestions, maybe listen to both, Gene knows what he is talking about.
The original complaint is crossover distortion which points to low class A bias in the output devices.
Ignoring for the moment that the output is not exactly 0V. lets measure the current in the emitter degeneration resistors (R37 and R 38 0.47 ohm. ). 10mA of class A bias should generate 5 mV across each resistor.
If not we need to tweak up the voltage in the Vbe multiplier TR9 . The schematic shows an optional 1k across R34... that will reduce voltage and class A bias, so if 1k is in there, remove it.
Very carefully reduce the resistance of R33... you can blow up the amp from too much bias so do this carefully maybe starting with a 5k in parallel and slowly drop lower, until you get 5mV (each) across R37 and R38.
Be careful messing with this. After getting the class A bias current right, we can worry about output voltage not being exactly 0V.
JR
PS: Sorry about getting multiple suggestions, maybe listen to both, Gene knows what he is talking about.
corgan4321
Well-known member
I measured the output with a 420 ohm resistor across the output and the negative supply and watched the output drop 4-5 mV to -0.016mV
Unfortunately, it also blew up ome other transistor, I get very distorted sound out of the output along with a dial-up squeal, and the aux out pit is just a loud buzz as if disconnected. Ill have to figure out what blew later this weekend...
Unfortunately, it also blew up ome other transistor, I get very distorted sound out of the output along with a dial-up squeal, and the aux out pit is just a loud buzz as if disconnected. Ill have to figure out what blew later this weekend...
The good news is that proves the pull up is working.corgan4321 said:
Bummer... an extra watt or two of dissipation should not distress a power output stage. Maybe some part was "fixin to blow up".
Good luck, we'll get it right eventually.
JR
Gene Pink
Well-known member
corgan4321 said:
Sorry, I feel really bad, and responsible that this happened.
But I have to ask, not to be condescending or trying to absolve myself, are you sure that you put that resistor from the minus 22V rail to the output? The resistor is 420 ohm as in yellow-red-brown?
The only thing that this test could have possibly put past it's design maximums would be TR-10, but this could only happen if TR-11 wasn't functioning and doing what TR-10 was telling it to do, and the TR-10 driver was doing all the outputting for that rail. The MPSA06 has a max collector current of 0.5 amp, which would see the 60ma, but a max dissipation of 0.625 watts, and this would have been more than double that. Shouldn't be a problem for just a few seconds, but JR would know more about that than I.
In normal operation with a speaker load, this amp would routinely see much more forward, and back-emf current from the drivers changing velocity (cone/air inertia, box tuning), than my suggested "disturbance current" test.
The Aux out humm is puzzling as that signal comes from before the power amp section, perhaps something happened with the power supply? Perhaps a now-shorted TR-10 is dropping other parts of the power supply out of regulation?
Apologies again, looking forward to finding out just what went wrong. I'd start with TR-10, if it blew, then it is because TR-11 is non-functional. If TR-11 still shows "two of six" in the classic diode voltage drop test of all combinations, it is no guarantee that it actually works
What common part number did you replace TR-11 with?.
Gene
Amps blowing up is always a hazard when troubleshooting.
The extra current should not have broken the amp.
What broke, may reveal what happened.
What doesn't kill us makes us stronger (smarter).
JR
The extra current should not have broken the amp.
What broke, may reveal what happened.
What doesn't kill us makes us stronger (smarter).
JR
...as time passes by, Wurlitzer 200a amps seem to develop this problem on a regular basis.
I have had this exact crossover distortion (after a lot of trouble with hum which led to a partial redesign of the PSU part, but that's a different story), resulting in some noise appearing whenever you played a note.
To bias the amp I had a 1k trim pot installed in place of R58. It turned out that turning the pot improved the sound while the TR9 emitter voltage slowly moved towards -0.58V, increasing the resistance got me closer. So I decided to raise the value of R34 and replaced it with a 220Ω resistor first, which knocked distortion completely out, and a 270Ω resistor eventually since that gave me -0.63V on the emitter of TR9. Actually, I screwed down a bit again to get a proper Vbe drop at TR11, that had the same symptom as OP's in the beginning.
I'm not sure if this is a lasting solution, because the output voltage is not 0V as it should be, but about -10 to -11mV, so I assume there is something else that needs to be investigated. Collector voltage of TR13 is ~26mV, and 3mV of TR11, respectively. While trimming the bias pot does change these measurements it does not have any influence on the resulting Output voltage of -11mV.
Any ideas?
Too sad that neither the OP corgan4321 nor JohnRoberts or Gene Pink went on solving the riddle but maybe one of them could pick up on this? It was really fun reading through their investigation. And yes, I have a scope.
I have had this exact crossover distortion (after a lot of trouble with hum which led to a partial redesign of the PSU part, but that's a different story), resulting in some noise appearing whenever you played a note.
To bias the amp I had a 1k trim pot installed in place of R58. It turned out that turning the pot improved the sound while the TR9 emitter voltage slowly moved towards -0.58V, increasing the resistance got me closer. So I decided to raise the value of R34 and replaced it with a 220Ω resistor first, which knocked distortion completely out, and a 270Ω resistor eventually since that gave me -0.63V on the emitter of TR9. Actually, I screwed down a bit again to get a proper Vbe drop at TR11, that had the same symptom as OP's in the beginning.
I'm not sure if this is a lasting solution, because the output voltage is not 0V as it should be, but about -10 to -11mV, so I assume there is something else that needs to be investigated. Collector voltage of TR13 is ~26mV, and 3mV of TR11, respectively. While trimming the bias pot does change these measurements it does not have any influence on the resulting Output voltage of -11mV.
Any ideas?
Too sad that neither the OP corgan4321 nor JohnRoberts or Gene Pink went on solving the riddle but maybe one of them could pick up on this? It was really fun reading through their investigation. And yes, I have a scope.
Its been 7 years so this is not fresh in mind...
JR
JR
corgan4321
Well-known member
I never bothered spending more time on it because it worked just fine through an amp. The issue is in the output stage. Plus, it’s my brother’s problem now - I have an older 100 series as my own.
If I were to run into this issue again, or any 200a amp needing service, I would STRONGLY consider replacing the electronics with one of the reproduction options. The electronics of the 200 series is, in my opinion, not up to par with the quality of the rest of the instrument.
If I were to run into this issue again, or any 200a amp needing service, I would STRONGLY consider replacing the electronics with one of the reproduction options. The electronics of the 200 series is, in my opinion, not up to par with the quality of the rest of the instrument.
