Super Reverb blown fuse and 100 Ohm CT resistors

[osing]

Hi there,


About six months ago, my self wired Super Reverb failed while playing - the two 100 Ohm artificial CT resistors burned up. I had been playing the amp loudly as always (volume at ~5) and was using JJ 6L6GC power tubes. I changed the resistors, and swapped to a new set of tubes (Tung Sol 7581A reissues) and everything seemed fine. I have played the amp several times since then with no issues, but this weekend, while playing at similar volumes and shortly after kicking in a fuzz pedal, the amp started smoking. The fuse blew, and the same two resistors got fried. Some of the cloth wiring near the resistors also burned up because they were close in proximity to the burning resistors.


Today, just to be sure, I checked the multi-tap OT and confirmed that I have the right outputs connected to the right lugs on the switch (i.e. the 2/4/8 Ohm wires are connected correctly to how my switch is labeled). I applied 22.49VAC using a variac to the primaries and measured 0.996V, 0.705V and 0.495V on the 8, 4 and 2 Ohm taps, respectively. If my math is right, these all suggest a primary impedance of ~4.1k Ohm. Are these measurements effectively validating that the OT is good, or could I still have a faulty OT?


I proceeded to replace the burnt wires and resistors, and powered everything up gradually using a lightbulb limiter (no tubes, then rectifier only, followed by preamp tubes). Everything appeared fine. I unfortunately do not remember if I tried the amp through a lightbulb limiter with power tubes and a dummy load attached, but I do know that when I powered the amp with no lightbulb limiter and power tubes and a dummy load installed, the two 100 Ohm resistors immediately went "pooff" again, and the mains fuse blew.


There is no evidence of carbon traces on either side of the power tube sockets between pins 2 & 3, but I am not sure if that necessarily rules out arcing. In all cases, I believe the amp was biased somewhere in the 60-65% of max plate dissipation, but it could perhaps have drifted over time.


I am wondering if one or both of the power tubes are toast and causing the problem, if the OT is blown, or if there is another issue altogether. Before I replace the two resistors and power up again, does anyone have any suggestions about what I should check? Should I assume the power tubes are bad, or is there a way to test for internal shorts without a tube tester?


I don't mind blowing a few more fuses and resistors, but I am concerned about damaging any expensive parts....

[sluckey]

Changing tubes is easy. Check for a short between pins 2 and 3 of the tube sockets with tubes plugged in. Do the same with tubes unplugged.

[pdf64]

A HT fuse, rectifier back up protection diodes, and metal oxide flameproof screen grid and heater balancing resistors would be good improvements that would help to mitigate the risk of valve shorts resulting in collateral damage.

[mresistor]

He's had two sets of power tubes do the same thing..  so check the sockets too ..   but  if this was a recent build then the sockets should be in good shape..   Have you checked any of the small bottle tubes for shorts?

[Willabe]

the two 100 Ohm resistors immediately went "pooff" again, and the mains fuse blew.

..... heater balancing resistors would be good improvements

He has balancing R's, that's whats burning up.

[pdf64]

…He has balancing R's, that's whats burning up.

The wiring damage points towards them not being MO flameproof, perhaps they were ‘vintage correct’ CC?

[PRR]

100Ω at 3.15V is a tenth of a Watt. He does not say(?) what watts were smoked, but <1/10W is rare here. So assume the resistors are/were fine. And I dunno why we are poking the OT, which does not connect to heaters. As others are saying, loose or short connections are real likely. Bright light, magnifier for eyes over 30.

[Beezerboy]

check all the sockets to see if any are shorted.... check them all

and, what voltage do you have?

[pdf64]

I’ve puzzled it over though still can’t hypothesise the exact mechanism, but there’s an output valve failure mode that allows high levels of HT current to flow to OV common via the heater filament.
The fault is related to the valve rather that the socket or wiring, and requires operational conditions to manifest, eg it’s not apparent on a cold valve, or even a hot valve with no HT.
If the heater circuit is balanced and referenced to 0V using a pair if resistors, unless they’re super beefy, they’ll blow.
Often 1W screen grid resistors will blow too.
It’s unclear whether high screen grid current is due to the fault occurring between the screen grid and filament, or due to some other reason.

[Willabe]

If the heater circuit is balanced and referenced to 0V using a pair if resistors, unless they’re super beefy, they’ll blow. Often 1W screen grid resistors will blow too.

Both good, acting like fuses.

[pdf64]

If the heater circuit is balanced and referenced to 0V using a pair if resistors, unless they’re super beefy, they’ll blow. Often 1W screen grid resistors will blow too.

Both good, acting like fuses.

But fuses, in the act of blowing should prevent further damage. Whereas the opening post reports that the heater balancing resistors, which I'm assuming were a carbon based type, in the act of blowing, apparently flamed out and damaged the amp's wiring.
So my point is that a metal oxide flameproof resistor type would be a better choice for this application.

As I see it, there's a big problem for the non faulty valves in the amp when the heater circuit's DC reference to 0V is lost, as it seems likely that the faulty valve will pull the heater circuit up to HT voltage. As that may easily result in damage to their heater to cathode insulation, leaving them liable to complete failure or leaking buzzy hum into the signal path.

Hence I see that as a counter productive application of fusing. Rather it seems preferable for the heater's 0V reference to be more resilient, and to have a HT fuse blow. That way, the good valves should stay good.

[Willabe]

Yes, they should be flameproof in those positions.

Some amp builders fuse all the PT secondaries. 

[osing]

All - thanks for your feedback and suggestions!

It turns out I did have one power tube that had a short between pins 2 & 3. I have been able to successfully start the amp back up again with a pair of 6L6GCs. I believe these are the ones I had in the amp when the 100 Ohm resistors got fried 6 months ago. I have not had a chance to play the amp yet, but will give it a try tonight to see if it holds up.

I was using metal film 1/2W 100 Ohm resistors. If I were to change to 1 W flame proof ones, will they still protect the amp if another power tube shorts out? Are you saying that this change would basically force the mains fuse to blow before the 100 Ohm resistors would fry?

Now the bigger question that probably cannot be answered - what likely caused the 100 Ohm resistors to burn up 6 months ago if there was no internal tube short back then and no signs of arcing on either power tube socket? Is the shorted power tube this time around a likely a result of an underlying issue with the amp, or was this possibly just a poor quality tube?

EDIT - got rid of extra spaces

[sluckey]

The most likely reason for those resistors to burn is arcing between pins 2 and 3. There's not always a visible arc trace on the socket, especially if it is a porcelain socket.

The most common reason for the arc is playing guitar at high volume without a speaker connected. Doesn't matter if you just forgot to plug it in or if the voice coil opened up or if the speaker cable/plug failed or someone stepped on the speaker cord and accidentally unplugged it. All it takes is one strong power cord to create a very high voltage spike in the primary. This spike will usually take the closest path to ground, creating an arc across the space between pins 2 and 3 (may be inside the tube or on the socket) and your 100Ω resistors. Hopefully the only damage will be open resistors. But if the resistors are big enough wattage to survive then you risk burning the tiny wire inside the OT and/or maybe even the PT. For this reason, I prefer 1/2W resistors.

The resistors are not the problem, merely a symptom of the problem. So, try to remember... Was there a time (even a short time) when your speaker was disconnected while you were playing loudly?

[pdf64]

…
I was using metal film 1/2W 100 Ohm resistors. If I were to change to 1 W flame proof ones, will they still protect the amp if another power tube shorts out? Are you saying that this change would basically force the mains fuse to blow before the 100 Ohm resistors would fry?
…

I think it’s a bad idea to rely on the mains fuse to protect a secondary circuit.
My suggestion includes fitting a HT fuse.
eg F500mA in series with the PT HT winding’s CT connection to 0V common.
And back up protection diodes for the rectifier.

[osing]

So, try to remember... Was there a time (even a short time) when your speaker was disconnected while you were playing loudly?

I had connected the Super Reverb to a Leslie 18 (8 Ohm) I recently bought, but I made sure I was using the 8 Ohm tap and also made sure my 4x10 speakers were set to 8 Ohm (I have a switch that I can choose between 2 or 8 Ohm for the speakers). I had already disconnected the Leslie from the Super Reverb and am 99.9% sure I switched back to the 2 Ohm settings by the time the resistors burned. On a side note, the Leslie 18 does not have a cross over but instead just switches between the Leslie speaker and the amp's speaker. I feel like that is a potential risk as there is a moment of no load on the amp when switching (might be ok if you don't play when switching). With the Vibratone and Leslie 16, with the crossover, my impression is that the amp will always see a load, even though it might not be perfectly matched.

I just looked over my amp again, and I was very surprised to see that I have a wire that appears to not be soldered - I am shocked if it has been this way for 12+ years, but it is possible. Basically, the wire from the cathode of V8 (pin 8) is mechanically connected to the turret where the 1 Ohm resistor is connected (which then goes to ground), but there is no solder. If this connection were to be intermittent due to vibration such that at brief moments the cathode is not tied to ground, does anyone know what the potential impact could be? Could this lead to arcing between pins 2 & 3???

I think it’s a bad idea to rely on the mains fuse to protect a secondary circuit.My suggestion includes fitting a HT fuse.
eg F500mA in series with the PT HT winding’s CT connection to 0V common.
And back up protection diodes for the rectifier.

Thanks for the suggestions - I'll need to look into this!

[HotBluePlates]

It turns out I did have one power tube that had a short between pins 2 & 3. ...

The most likely reason for those resistors to burn is arcing between pins 2 and 3. There's not always a visible arc trace on the socket, especially if it is a porcelain socket.

The only tube I've had fail outright was an RCA 6L6GC, and it was for a short from Pin 2 to Pin 3.

That short was only at the tube, and it was internal.  There was no arc path on the outside of the tube or at the socket, and the tube exhibited continuity from Pin 2 to Pin 3 when removed from the amp.  I'd have to look at some other RCAs to see if the heater passes too close to the plate's connection to Pin 3.

Old tube went in the trash.  New tube, new fuse, and the amp never had another problem.

[PRR]

...your speaker was disconnected while you were playing loudly?

A disconnected speaker isn't loud.

That's why it is hard to notice. You have a momentary "Whaaa?", you fix it, you forget it.

[sluckey]

You might be surprised at how many times the first reaction is to turn the guitar volume up, then turn the amp volume up and hit the strings even harder.   

[thetragichero]

^^^^^been there with my jet city picovalve. either the tube or the ot produces the faintest bit of sound. stupid human tricks

[JB]

It turns out I did have one power tube that had a short between pins 2 & 3. ...

The most likely reason for those resistors to burn is arcing between pins 2 and 3. There's not always a visible arc trace on the socket, especially if it is a porcelain socket.

The only tube I've had fail outright was an RCA 6L6GC, and it was for a short from Pin 2 to Pin 3.

That short was only at the tube, and it was internal.  There was no arc path on the outside of the tube or at the socket, and the tube exhibited continuity from Pin 2 to Pin 3 when removed from the amp.  I'd have to look at some other RCAs to see if the heater passes too close to the plate's connection to Pin 3.

Old tube went in the trash.  New tube, new fuse, and the amp never had another problem.

I've had it happen in my 5E3.  Brimar 6V6 died (failed on power up after a car journey).  Took out the 100R resistors, but not the HT fuse.  Also happened to a friends Marshall Superlead, the 100R resistors, which looked like 1/4W max, were completely burnt away.

[pdf64]

…
I've had it happen in my 5E3.  Brimar 6V6 died (failed on power up after a car journey).  Took out the 100R resistors, but not the HT fuse…

What are the details of the HT fusing, eg spec and place in circuit?
1/2W heater balancing resistors may well blow faster than a slow blow HT fuse.

I have a hypothesis that heater circuit DC elevation may reduce the likelihood of that mode of valve failure. But it’s best for the cap and ‘bottom’ resistor of the elevation circuit to be rated to cope with the full HT voltage, in case it does occur.

With a 5E3 the heaters can be elevated by referring the balance node to the output valves’ cathodes, rather than 0V common. If an output valve shorts in that scenario, provided reasonably beefy cathode and heater balancing resistors are used, due to the resulting high HT current, a fuse should blow pretty quickly.

Plenty of 5E3’s competitor amps used that method of heater elevation, I suppose Fender moved straight from single ended heaters with the chassis as a conductor, to fixed bias, so never got around to such tweaks to cathode bias.
https://el34world.com/charts/Schematics/files/Gibson/Gibson_ga20_12ay7_preamp.pdf

[JB]

It's got a T0.2A fuse in each leg of the HT secondary.  So yes, maybe quick blow would be better.  I think the 100R resistors were 1W metal film. 

[pdf64]

Kudos for fitting HT fuses 
They’re doing a great job of protecting the HT winding and other big stuff, but may tend to allow little stuff to fry.
I suggest F200mA, preferably at the CT to 0V return of the HT winding, or between the reservoir cap and the rest of the HT system.

[osing]

I was able to crank the amp and played for an hour with no issues. All seems well. Thanks for everyone’s input along the way!

[pdf64]

… the wire from the cathode of V8 (pin 8) is mechanically connected to the turret where the 1 Ohm resistor is connected (which then goes to ground), but there is no solder. If this connection were to be intermittent due to vibration such that at brief moments the cathode is not tied to ground, does anyone know what the potential impact could be? Could this lead to arcing between pins 2 & 3???…

Yes, the stop / start intermittent current flow in the OT primary, due to the high inductance there, could generate high back emf, leading to such arcing.

Switching the load whilst the amp is attempting to put out a power signal seems a bad idea though.
A make before break switch (as opposed to the more typical break before make) might mitigate for problems.