Groove Tubes Soul-o 45 Rev B 1997 8 18 blowing main fuse

[shooter]

If I install the power tubes, it GLOWS.

Voltages with the Power Tubes installed is now around 250 VDC

what's the current across the 1-ohm R's????

[wsscott]

With the current limiter connected, and the bulb glowing, input 120VAC, the left 6L6GC is 83 mv and the right one is 55 mv!  I guess that's .83ma/.55ma respectively.

WOW, the bias is way off.

[wsscott]

Sorry, 83ma/55ma respectively.

[wsscott]

If I switch the tubes in the sockets, the voltages follow the tube.

But even if these tubes, which are replacements for what was in the amp when the fuse got blown, are not well matched, it doesn't seem that would cause the fuse to blow or cause what may be a short.  The current limiter shows the same problem even with a different set of tubes.

[sluckey]

A long time ago I suggested you should measure voltage on pins 3, 4, and 5 of the output tubes. Now's the time to do that.

Pull the 6L6s and plug the amp power cord straight into the wall. Measure the voltage on pin 5 of each output tube. What have you?

[wsscott]

I pulled the output tubes, set meter to DC Volts, connected - probe to chassis, and + probe to Pin 5 of each tube.

I could not get any stable voltage reading on either tube.

[sluckey]

I pulled the output tubes, set meter to DC Volts, connected - probe to chassis, and + probe to Pin 5 of each tube.

I could not get any stable voltage reading on either tube.

This is the rabbit you should chase! Don't put the 6L6s back in until you have a steady dc voltage of about -50VDC on pin 5 of both 6L6s. Chase the rabbit back to the bias supply diode to find out where it is missing. You may even need to check the AC voltage on the other side of the diode.

Before chasing the bias voltage, turn the power off and measure resistance from the bias diode anode (non banded end) to chassis. This will check for a short on the bias circuit and bias cap. What resistance do you measure?

[wsscott]

I don't have a schematic for this amp, so I'm having to guess a lot.

I think the diode is the one that connects to the bias adjustment pot. The Banded end has continuity and 0 ohms to chassis ground.  The anode end jumps all around when connected to ground.  That end also connects to a 1.5K resistor and I think it also connects to a 10uf cap.  This is on a PCB so I'm limited in what I can see the connections are.

With meter in diode mode, it reads OL in one direction, and with probes reversed .593 in the other direction.

[wsscott]

So if this is the correct diode to check for AC voltage for the bias, I set my meter to VAC, connected black probe to chassis ground, and red probe to anode side of that diode where it meets with the resistor, and it reads 0 VAC.

[sluckey]

Just check the negative dc voltage on the non banded end of the bias diode. See pic...

[wsscott]

Its fluctuating a little but between 17-18 mVDC.  Black probe to chassis, and red probe to diode.

[wsscott]

Can you see the PCB's trace going from that point on the diode up to the area below that 10uf cap?
Is that the source from where the diode should be getting its voltage that we're looking for?

[shooter]

17-18 mVDC.

can you check the same point only measure vAC, not DC

[wsscott]

Shooter I did that, see #58 above, and it read 0-1mVDC.

So it seems no voltage is getting to that point.  Maybe the 10uf cap is bad, but there are no bulges or anything that looks like a failure.

[sluckey]

Now with power off check the resistance between the bias diode anode and chassis. If it's very low ohms or even zero ohms, the most likely suspect is the bias caps. Since they are radials, you will have to get to the other side of the board to remove them.

[wsscott]

It reads 139.2K ohms.

[wsscott]

Here's the problem with removing the PCB from the chassis.  Only 1 tube socket, V1, has this shield base attached to it, and as a result the PCB's movement is restricted such that the pots on the front of the chassis can't be moved back to the rear of the chassis far enough to exit their holes because the shield base is above the chassis' surface and restricts the movement of the PCB.  Attached is a photo

Anyone know a way to skin this cat?  Without removing the socket from the PCB?  I don't know how it can be removed from the socket.

[sluckey]

GT put that pcb in the amp with the pots and tube sockets installed. Surely you can figure out how to remove the pcb. Worst case scenario is to remove the shield from the socket using diagonal cutters and needle nose pliers.

[shooter]

it read 0-1mVDC.

I ask for AC, when a diode that BLOCKS AC shorts, guess what gets through????
EDIT, it's sorta mute point now that you must face the PCB on it's terms n defeat it 

[wsscott]

Sluckey-In the photo you posted in #59, I checked the AC voltage since this is where the Secondary voltage comes out of the PT.  I get 190VAC going into that red .047uf cap, and out of it, then into the Cathode/Banded side of the Diode.  No voltage coming out of the Diode.

Then the only other place where I read AC Voltage is 190VAC going into the 62K resistor, BUT nothing coming out of it!  That's the resistor that has the red wire connecting to a GND solder point on the PCB.

No voltage at any other point in the bias circuit

I just checked that resistor and its Open.

I'll see if I've got another resistor that I can jumper across that one and see if it is the problem.

[wsscott]

Now, I've got -90VDC on both of the Grids!  There are no tubes installed so this may drop when installed?

I'm now reading about 350VAC going into the cap vs. 190VAC. So only 1/2 of the voltage from the secondary was being delivered to the circuit. And after it comes through the .047 cap and at the Cathode of the diode, it's down to about 208VAC.

What else should I check.

[sluckey]

See if the bias pot will change the voltage on pin 5 of the 6L6s. If so, can you set it to -50V?

[wsscott]

Done.  Got it set at -50 with red probe on Pin 5 connection, and black probe on chassis ground.

No power tubes installed.

[sluckey]

Put the power tubes in and monitor the mV across your 1Ω cathode resistors. Also watch for red plating. Depending on how long those 6L6s have been run in the amp with no voltage on pin 5, there's a good chance they are now bad. That resistor may have cost you the price of a set of 6L6s.

BTW, -50V is just a ballpark figure that should be safe for 6L6 but you will need to change that a bit to get the tubes biased properly.

[wsscott]

It's running through the current limiter.  Just a low glow at 120VAC.

The bias is reading 22.5ma and very steady.  I guess that's a little cool for these tubes, but the limiter might be affecting that reading?

[sluckey]

Put the limiter under the bench and plug the power cord straight into the wall (or a variac set to 120VAC). Recheck the bias.

Your amp uses a FWB rectifier for B+ and the bias circuit is cap coupled to the high voltage winding of the PT.  Your bias circuit should look a lot like this circuit...

[wsscott]

It's been running straight to the wall now but through the Variac at 120VAC for about 30 minutes, and it looks like the bias is reading a pretty stable 46.5ma.  That may be a little cool for 6L6's, but a new set of tubes is going to be ordered anyway.  So probably will just leave it as is.

Thanks for everyone's help.  Getting directed by you guys to check out the bias circuit was really what I needed.  Thank goodness I didn't have to pull the PCB.

Its really frustrating not to have a schematic, particularly when dealing with a PCB!

Best.

[dude]

Looks like you're in business, 46 mA is ok, if a 6L6 red plates, you'll know it the long plates inside tube will glow very bright yellow/red, if they do, turn off get a new set, 46ma should not red plate if the tubes are still good.
Sometimes you can replace a radial cap without getting under the bd. Cut the component out leaving the old leads as long as you can, you can solder the new radial cap or resistor using the two leads but be easy on the heat, just a little solder, should do it, like two seconds.
 

[wsscott]

Turns out there was another issue that I was able to diagnose and fix.

There is a 62K resistor in the bias circuit that I found was blown.  So I replaced that.

In addition there is a 100 ohm dropping resistor between A+ node and the C+ node that was reading 10K!!  So I replaced that.

And finally the C+, 100uf/450V, filter cap was shorted.  So I replaced that.

I think all of these things were related to what I believe was a blow out when one of the output tubes with the broken locator pin was probably inserted in the wrong position.

Now I've got it running.

Here are some numbers for you. It's running 6L6GC's. The numbers sound high, but maybe that's just because it's a higher powered amp that I'm used to dealing with.

Plates 494VDC
Screens 491VDC
Grids -50.1 VDC
Bias 43.1/40 ma
A+ 497VDC
B+ 491VDC
C+ 492VDC
D+ 420VDC

I guess that's about it for this one.  Thanks for everyone's help.

[sluckey]

In addition there is a 100 ohm dropping resistor between A+ node and the C+ node that was reading 10K!!  So I replaced that.

I'm pretty sure that resistor is supposed to be 10K. The very high voltages at node C and D seem to support that guess.

[wsscott]

Sluckey, it's bands appear to be Brown, Black, Brown with a Gold band for tolerance. So that's 100 ohms.  It's pretty big, physically, and on one of the schematics, not this particular amp because there isn't one that I've been able to find, it's shown as a 3W resistor.

But in looking at the Voltages, I'm concerned about the C+ since it's a 450V cap.  Both the A+ and the B+ are 2 caps wired in series.  The A+ is two 350V caps in series for 700, but I can't read the B+ voltage because its installed face down to the PCB.

D+ is a 22uf/450V cap and its reading 420VDC.  So it may be okay.

The PT's secondary reads correctly at around 380VAC, so when rectified its around 535VDC.

When I checked the voltages with the current limiter in-line, the voltages read about 100VDC lower.  I know they would be lower in a restricted mode, but I don't know if 100VDC indicates anything.

So there's a lot of voltage going into this amp's circuit.  An amp built in 1997-if that indicates anything.

[wsscott]

If you look at Reply #14 of mine above, in the first image posted there, the resistor in question is the big one directly below the 2, Black, B+ node filter caps.

[shooter]

Brown, Black, Brown

when you HEAT up RED or ORANGE, they tend to turn brown 1st, then they smell, then they......

[wsscott]

If you look at the other large resistor above those 2 caps, it appears to read 470 ohms, and that’s what it also measures.  But it may be serving a different purpose.  It’s connecting to the B+ node. 

[astronomicum]

No, it only lights up if the power tubes are installed.  If I remove the 2 of them, and even if I leave all the other tubes in, then no light up on the current limiter.

…verify voltages are good on pins 3, 4, and 5. Check voltages with tubes pulled…

I apologize if I missed it but what were your voltages on pin 5?

[wsscott]

Sure. In my Reply 78 above, the grids which are pin 5 of the 6L6’s read -50VDC.

[wsscott]

If I replace that 100 ohm resistor with a 10K, what wattage should I use, and what might the resulting voltages look like?

That resistor links to the C+ 100uf cap that I had to replace because it was shorted to ground.

[sluckey]

The original looks like a 3 watt that Hoffman sells. Put the 10K in then measure the voltage ***ACROSS*** it. Use the power formula P=E²/R to calculate actual wattage dissipated. Double the calculated wattage for safety.

EDIT... Many times when a resistor gets hot, the color bands will change a bit. The colors don't have to change much to make orange look like brown.

[wsscott]

I’ll give it a try.  Is E the figure I measure for the voltage drop?

[sluckey]

Hmm. What do you think?      

[wsscott]

Ha Ha--I was testing you.  The Voltage drop is 140VDC, so with the formula, and a R=10K, that's Power of 1.96 watts.  So I guess a 3W would be fine.  And I think that's what this is, given its size.  Maybe even a 5W.

I checked the voltages and they are the same for A+ and B+, which is ok because of the caps being in series giving a higher Voltage.

C+ dropped to 356VDC
D+ dropped to 306VDC

Much better than before, ie. 492 and 420VDC

Plus the Grids stayed at the same -50.1VDC
and the Bias stayed at the same 43.1/39.8ma

So I think this feels much better to me, and obviously it looks like the colors on the resistor had faded.  A 10K/3W seems to be the one to use.

Thanks again for your help.