Silvertone 1482 tremolo a no-go

[Apexelectric]

Well I’ve been messing with this one for a few hours with no luck in resolving a dead tremolo. I’ve tried addressing the most obvious components that could be causing the issue and anything that was out of spec has been replaced. I replaced a few resistors and capacitors with no luck. Did the 330K’s that tie to the screen and the plate resistor as well as the one of the oscillator .02 uf film cap, other two are ceramic, the .5 uf wax cap and the .05 molded film cap. I tensioned and put Deoxit the pins on the 6AU6A socket and reflowed some solder joints that might have affected the tremolo. I also removed the footswitch jack from the chassis to eliminate possible grounding. Nada! 

The oscillator seems functional since I get the voltage change at the pot. I also can hear the effect as the amp is powering down. The screen voltage is at 35 instead of 75 volts which is indicated on the schematic.

What else could be limiting the screen voltage? The 130+ voltage drop across the plate resistor seems to indicate that there is current flow through the tube.

Kinda at a loss now. Any ideas?

[Sonny ReVerb]

Any chassis pics? I used that trem circuit in one of my amps and it worked pretty well.

[Apexelectric]

The circuit itself does work well. I have it in another Airline amp and it sounds great. This one just doesn’t want to give up the goods and I can’t find out what’s knocking the screen voltage on the 6AU6A down so low.

[sluckey]

How much voltage on pin 4 of the 6V6s?

[Apexelectric]

I got right around 350 at the plate resistor of the 6au6 and 220 on the other side of it. Haven’t checked it at pin 4 of the 6v6’s but it should be the same.

[sluckey]

I got right around 350 at the plate resistor of the 6au6 and 220 on the other side of it. Haven’t checked it at pin 4 of the 6v6’s but it should be the same.

OK, that's what I really wanted to know.

Is the 6AU6 plate voltage steady or is it wildly fluctuating? Steady would indicate the oscillator is not working.

[Apexelectric]

I’ll have to double check it when I get back but I want to say it was steady. However, when powering the amp down you can hear the oscillating in the output as the volume drops. The speed of the oscillations change with the adjustment of the speed control. With the amp powered on you can’t hear anything oscillating.

[terminalgs]

The screen voltage is at 35 instead of 75 volts which is indicated on the schematic.

seems like the trail to follow...  maybe a bad tube?

[Apexelectric]

I tried two other tubes with no change

[Apexelectric]

Plate voltage is steady at the 6AU6.

[sluckey]

Oscillator is not working. Start by replacing ALL THREE of the .02µF caps.

[Apexelectric]

Will do.

Any reason why two of them would be ceramic and one a film cap?

[sluckey]

I think I see 3 ceramic .02µF caps.

[Apexelectric]

There are but one is off the strength pot.

[Apexelectric]

Well I’ve replaced the three oscillator caps with no change. I’ve left the .02 off the strength pot alone.

[Apexelectric]

Turns out it’s the output tubes. The amp seemed to play fine but the output tubes tested weak. Swapped them out and the tremolo came back to life. Anyone else run into this? I don’t think I saw one mention in any of the info I’ve searched that offered up an output tube swap as a fix for a non-functioning tremolo. I certainly understand the relationship between the two entities but it wasn’t one of my first thoughts as a root cause of the problem.

Won’t make that mistake again. 

[sluckey]

I'm curious. Now that it's working, is the 6AU6 plate voltage still steady?

[Apexelectric]

Yes, the voltage now fluctuates.

Now the tremolo is cutting in and out depending on what I do with the tremolo controls. If I get it to stop by turning the strength pot down it doesn’t want to start back up again. Seeing as the output tubes were the culprit, do you think that the old cathode resistor bypass cap might be playing a part in this as well. It’s an old electrolytic.

[Apexelectric]

Played around some more with a few other component changes including the cathode resistor and bypass cap to try to remedy the tremolo cutting out. Needed the bypass cap done anyway. Nothing has changed. I can get the oscillator working again if I turn of the amp briefly then turn it on again. If you leave the tremolo running it will stay running. If you turn down the strength pot and get it to cut out, it won’t come back unless you shut it down.

The amp seem to be pulling a bit more than I’d expect, 26W. The plate voltage seems in line, 312V plate to cathode, but I’m dropping 23V across the 270 ohm cathode resistor which is shown at more like 16 on the schematic. At 16V across the cathode resistor it brings the dissipation down to around 20W which seems to make more sense.

Can’t help but think that the tremolo and output are related. I know I can change the cathode resistor to reduce the output but it seems like it should be where it need to be with the 270 ohm resistor.

Am I missing something obvious?

[ginger]

Try tacking in a new strength / intensity pot ?

[sluckey]

Put a new/strong 6AU6 in it.

[Platefire]

I would double check that on/off jack to see if it might be shorting out and killing your trem even thought the switch may not be plugged in. I've got two 1482's and those original jacks do fowl up from time to time. Platefire

[Apexelectric]

I previously pulled the jack from the chassis to eliminate that possibility.

I’ve tried 4 different oscillator tubes with no change. I don’t have a new one on hand. But I would thing that one out of the 4 I tried would have worked if that was the issue.

The pot seems to not be the issue either. My resistance readings seem to be consistent and when the oscillator is working the pot seems to do its job with no issues. It’s only when the oscillator stops and has to restart agin, does the issue arise 

[punkykatt]

Did you get the 35v on the screen grid of the 6au6 up to 75v per schematic?

[jjasilli]

For faulty trems, where the tubes are ruled out, Dan Torres recommends just replacing all components in the the whole circuit, rather then the the diagnostic; or the shotgun (educated guess) approaches. 


That said, I'm guessing trem signal is bleeding to ground somewhere intermittently.  Trem pots or trem jack could be faulty.  At the very least I'd replace All the caps in the trem circuit, including the .02 off the intensity pot. 


For diagnostics you can use an anolog meter, preferably a VTVM, to watch the swing of LFO voltage at various points in the trem circuit, to try to isolate the origin of the issue.

[PRR]

While the iron is hot, replace the 0.5uFd 300+V cap on 6AU6 screen. If crapped-out, marginal gain, which fits a symptom. If leaky, low screen voltage which is another symptom.

[Apexelectric]

I’ve replaced all the caps in the tremolo circuit including the .5 uf, the .05 uf, the 680pf, the .02 off the pot as well as some of the resistors. The strength pot too.

When the oscillator is working the screen voltage is at 70V when it is not working it is at 35V

With the jack removed from the chassis there is no way to bleed signal to ground. That did not eliminate the problem.

[jjasilli]

When the oscillator is working the screen voltage is at 70V when it is not working it is at 35V
OK. 70V is about right.

When it drops to 35, try tracing the low voltage back to its source???  Looks like it should be starting out @ 330V at the filter cap supply node. Where is it unduly dropping? (Though sluckey may have covered this above)

[Apexelectric]

That’s correct, when it’s at 35V the voltage at the other end of the resistor remains at 335v regardless of whether it’s oscillating or not

The screen voltage at the oscillator will vary depending on the speed. As high as 70 and as low as 60. Voltage goes up as the speed slows down. That’s when it seems to want to cut out, when the speed is below 3 on the dial.

I’m still confused as to why this amp is dissipating so much wattage. The numbers seem to match up close to ones on schematics but Silvertone stated this as a 15W amp. I’m getting more like 25. No redplatting but I have JJ 6V6S’s rated at 14W ea. I’m inclined to put a larger cathode resistor to settle it down to more like 20W

[ginger]

Maybe I missed seeing it... but I assume you've replaced the filter caps

[Apexelectric]

The original cap can is still there.

[ginger]

Well , if you DO re-cap it... I never fool with replacing the cap can... I leave it there for looks , and use quality caps inside the chassis . Like this Princeton Reverb , that has a cap can.

[Apexelectric]

That would be my plan as well.

[Apexelectric]

Unfortunately I have no way of testing the rectifier tube so I assumed it was working ok but I dug through my stash and found another 6X4 and swapped it out and that solved the tremolo start-up issue.

Can someone chime in on their thoughts on running this amp at 25W dissipation when it was originally billed as a 15W amp. With the worn out tubes I replaced it makes me think that this amp should be dialed back a bit to run at no more than 20W. It would be just a matter of changing the cathode resistor so no biggie.

PT runs pretty warm and the OT not as much but warm nonetheless, all at lower volumes so I’m sure if I cranked this thing up it would heat up pretty good.

[punkykatt]

how did you do the math to get your tube plate dissipation? By the voltages you gave  above the tubes are running at 94.9% max plate dissipation of 14 watts which is 13.28 watts.

[PRR]

> confused as to why this amp is dissipating so much wattage. ...Silvertone stated this as a 15W amp.

Power input is always more than power output. A 15W audio amp is likely sucking >30W of D.

The trem oscillator screen should rise when it is working, that part is OK. It dead-idles at high current for good starting. When oscillation builds-up, G1 is forced negative, current falls. In both plate and screen. Falling to half idle seems reasonable.

[Apexelectric]

I used the plate to cathode voltage, the voltage drop across the shared resistor, and the resistance of the shared resistor to determine the dissipation.

I get the same thing you do, roughly 26W, 13W for each tube. Wouldn’t that fry an output transformer rated for 15W eventually?

[shooter]

Wouldn’t that fry an output transformer rated for 15W eventually?

put a 8ohm 50W R in place of the speaker, inject an input signal ~100mV rms and dime the amp, now measure across the DummyLoad and do Mr. Ohm, I suspect you will get something < 15Wrms

I'm currently using a Kt88 dissipating 42Wdc @ about 18Wrms speaker audio, using a 25W Edcor and she don't get much more than warm

[punkykatt]

Push pull amp with two power tubes only one tube conducts at a time.

[shooter]

Push pull amp

He's running cathode biased, so it is probably idling close to class A?

[Apexelectric]

So don’t equate pdis with output power, is what I’m hearing. Measure output power at the transformer output.

What determines the ratio between pdis and output power?

[shooter]

What determines the ratio between pdis and output power?

sounds like a goodgoogle? 

to many formulas and numbers makes my brain hurt, I've reduced it to this;
in class A you get < 1/2 in audio, the rest is calories
in class AB(something) you get ~~60-80% in audio, the rest is just rest

I still struggle with the majority of correct answers having names like "least common denominator" 

[PRR]

> Wouldn’t that fry an output transformer rated for 15W eventually?

No. How?

The OT rating is about bass distortion, not burn-up. An audio transformer will never "fry" unless something has gone VERY wrong, typically a short through the power supply.

At idle, there's 26 Watts of waste heat in the tubes. The transformer is >95% efficient here, so has maybe 1.25W of heat in it.

When you play loud Sine waves, you can divert up to half the idle heat into the loudspeaker, nearly 12 Watts. This path the transformer is >90% efficient, has another 1.25W heat in it.

(With real tubes in cathode bias, the idle current rises some at full output. The 15W may be a little over the theoretical 13W but may be valid.)

The "50% in class A" comes from the fact we traditionally test with Sine Waves. Being roundy, they only fill-up half the total wave-space. If we tested with Square Waves we could approach 100% conversion of DC to AC. (But square waves are rarely "music"; even gross guitar clipping is not fully square.)

Classic cathode bias gets 40% efficiency clean, a bit more when pushed to the extreme. Fixed-bias tubes do 50%-60%, almost never near the 78% possible with ideal devices.

[Apexelectric]

> Wouldn’t that fry an output transformer rated for 15W eventually?

No. How?

The OT rating is about bass distortion, not burn-up. An audio transformer will never "fry" unless something has gone VERY wrong, typically a short through the power supply.

At idle, there's 26 Watts of waste heat in the tubes. The transformer is >95% efficient here, so has maybe 1.25W of heat in it.

When you play loud Sine waves, you can divert up to half the idle heat into the loudspeaker, nearly 12 Watts. This path the transformer is >90% efficient, has another 1.25W heat in it.

(With real tubes in cathode bias, the idle current rises some at full output. The 15W may be a little over the theoretical 13W but may be valid.)

The "50% in class A" comes from the fact we traditionally test with Sine Waves. Being roundy, they only fill-up half the total wave-space. If we tested with Square Waves we could approach 100% conversion of DC to AC. (But square waves are rarely "music"; even gross guitar clipping is not fully square.)

Classic cathode bias gets 40% efficiency clean, a bit more when pushed to the extreme. Fixed-bias tubes do 50%-60%, almost never near the 78% possible with ideal devices.

Thanks again for all the valuable info and for setting me straight on something I just haven’t been looking at the right way. Makes perfect sense now.

[Apexelectric]

 

sounds like a goodgoogle? 

to many formulas and numbers makes my brain hurt, I've reduced it to this;
in class A you get < 1/2 in audio, the rest is calories
in class AB(something) you get ~~60-80% in audio, the rest is just rest

I still struggle with the majority of correct answers having names like "least common denominator" 

Thanks for boiling it down into simple math. Helps put it in perspective without all the algebra.

{edit- untangled Quote - PRR}

[jjasilli]

Estimating Output Power in Watts:


For PP:  use the Power Formula:  W = V² / R;         W = (Plate voltage)² / the Primary Impedance of the OT


For SE:  W = 40% X Plate Dissipation.  (In theory it's 50% less real world inefficiencies).


(The foregoing assumes that Speaker Load > OT Secondary tap > OT Primary Impedance are in accord with tube specs. )

[Apexelectric]

Estimating Output Power in Watts:


For PP:  use the Power Formula:  W = V² / R;         W = (Plate voltage)² / the Primary Impedance of the OT


For SE:  W = 40% X Plate Dissipation.  (In theory it's 50% less real world inefficiencies).


(The foregoing assumes that Speaker Load > OT Secondary tap > OT Primary Impedance are in accord with tube specs. )

Does cathode biased PP use the same formula as fixed bias PP?

[jjasilli]

Yes.  The PP formula works for cathode or fixed bias.


In fact, I suppose the SE formula would also apply to cathode or fixed bias.  SE is (almost) always cathode biased; but it could be fixed bias.  But SE is always Class A.  Class A, maybe by definition or as a necessary consequence, has the power tube(s) putting out in Watts a theoretical max of 50% of plate dissipation. 


Note, you can bias PP output tubes in Class A.  But that requires a low plate voltage.  So the power formula still works. 

[Platefire]

Apexelectric----Have you got that thing throbbing right yet? I keep watching to see what the solution might be.On my two 1482's the tremolo's have always worked great so far. Those trem circuits work as good or better than anything else I have and I use them quite a bit. So hope you get it up and function right soon. Platefire

[Apexelectric]

Apexelectric----Have you got that thing throbbing right yet? I keep watching to see what the solution might be.On my two 1482's the tremolo's have always worked great so far. Those trem circuits work as good or better than anything else I have and I use them quite a bit. So hope you get it up and function right soon. Platefire

Sure did. The rectifier was weak too. Found a spare in my stash and swapped it out just for grins and it performs like it should now. Didn’t have a way to test the tube so it seemed like it was doing its job but in the end it wasn’t.