SF Twin Reverb no sound

[JAelec]

Sounds like at least one of the old tubes is bad. Or possibly mounting wrong if the guide pins are still broken/missing.

I'll fire the amp up with one power tube installed at a time to see which ones are faulty.

What failure mode would cause a tube to conduct such a large overcurrent that fuse blows?

[Helmholtz]

I see...could you say there has to work done for the voltage to drop?

In a way, yes, as power is V*I = IČR and power is work per time.
But it's best understood using Ohm's law which says voltage drop Vdrop = R*I, so zero current means zero voltage drop.

[Tom Phillips]

I see...could you say there has to work done for the voltage to drop?

I'd say that's a legitimate statement. However, there will always be someone who wants it stated a different way.

Your statement shows that you are getting an understanding of what's going on.
The parameter's follow Ohms law which is Voltage = Current x Resistance. Or V=IR.

With the tubes removed, you have an open circuit. Therefore, no current can flow. When you plug in I=Zero then Voltage will be zero no matter what the value of R is.
I'm sure that there are many YouTube tutorial about Ohms Law that will clarify the concept.

[g1]

What failure mode would cause a tube to conduct such a large overcurrent that fuse blows?

There are a few tube faults that can make them over-conduct and blow the HT fuse.
But usually if there is an HT fuse that blows, the mains fuse will not blow. That's why I wondered if a tube missing it's guide pin got put in the wrong way.

[The Dude]

A good/quick check to see if the tube is inserted incorrectly due to broken guide is to first check if tube filaments are lit with the amp in standby mode. In standby, there is no B+ supplied to the tubes, yet there is filament voltage, so it's usually a safe test. If a tube filament does not light up, there's a good chance it is inserted incorrectly.

[JAelec]

A good/quick check to see if the tube is inserted incorrectly due to broken guide is to first check if tube filaments are lit with the amp in standby mode. In standby, there is no B+ supplied to the tubes, yet there is filament voltage, so it's usually a safe test. If a tube filament does not light up, there's a good chance it is inserted incorrectly.

I don't know how the tubes were inserted originally but I have made sure to insert them correctly during my testing (I can still see where the guide has been and were the "missing" pins 1 and 6 are).

This amp has no standby switch, only a power switch.

One thing I noticed today was that the power switch was glowing faintly while in the off position (it is a power switch with a diode in it). I measured between power switch terminals and got 86V on the Live side of the switch and 176V on the Neutral side. Both are supposed to be 0V in the off position, so there seemed to be some leakage through the electronics of the switch. Replaced the switch with a good one.

I wonder how long it has been like this and what effect it has had on the amp...

After that I tried inserting two old tubes that I got from a Bassman 70 (this amp had a hum problem with these tubes but otherwise they were functioning) so I thought I would give them a try.

I inserted these tubes one on each side and two sockets empty. Tried playing a bit of guitar through them and after a short while the tubes started to red-plate!

So with the new power switch and a pair old tubes... it's not blowing fuses anymore but something is fishy with this amp!

[Tom Phillips]

...I inserted these tubes one on each side and two sockets empty. Tried playing a bit of guitar through them and after a short while the tubes started to red-plate! So with the new power switch and a pair old tubes... it's not blowing fuses anymore but something is fishy with this amp!

This could be a problem with the bias supply or with the tubes themselves. I suggest that you leave the power tubes out and, with the amp powered on, monitor the voltage with respect to ground at pin 5 of each 6L6 socket. Please report what you find.

[JAelec]

This could be a problem with the bias supply or with the tubes themselves. I suggest that you leave the power tubes out and, with the amp powered on, monitor the voltage with respect to ground at pin 5 of each 6L6 socket. Please report what you find.

B+ 511V
All four plate pins: 511V
All four screen pins: 508.2V
All four grids: -50V
Two bias caps: -51 and -62 V

The preamp tubes where glowing and got warm.

Transformer winding resistances:

PT: 4.6 Ohms
HT: 25.6 Ohms
Filament: 0.2 or 0.1 Ohms (jumping between these values)

Quoting my earlier post where voltages were read without any power tubes in.

So the grid voltages seem fine right?

[Tom Phillips]

Quoting my earlier post where voltages were read without any power tubes in.
So the grid voltages seem fine right?

All we can know is that the grid voltages were fine at that time without power tubes installed.
Now it needs to determined why your the last set of tubes you installed were red-plating.
When you said "after a short while the tubes started to red-plate" I don't know the time span.

If the plate dissipation was OK for a while and then started to clime into red-plating territory, I would want to be monitoring the current in real time during and after warm up. I feel that the two most likely causes are a bias supply problem or a tube fault that causes a grid to start drawing current and thus dragging down the bias supply (I.e. dragging it into the less negative voltage direction). If that happens, it turns into a runaway situation.
Maybe consider adding 1 Ohm current monitoring resistors between the cathodes and chassis ground to simplify the monitoring process.

[JAelec]

All we can know is that the grid voltages were fine at that time without power tubes installed.
Now it needs to determined why your the last set of tubes you installed were red-plating.
When you said "after a short while the tubes started to red-plate" I don't know the time span.

If the plate dissipation was OK for a while and then started to clime into red-plating territory, I would want to be monitoring the current in real time during and after warm up. I feel that the two most likely causes are a bias supply problem or a tube fault that causes a grid to start drawing current and thus dragging down the bias supply (I.e. dragging it into the less negative voltage direction). If that happens, it turns into a runaway situation.
Maybe consider adding 1 Ohm current monitoring resistors between the cathodes and chassis ground to simplify the monitoring process.

Progress!

I re-tensioned the socket pins, and also connected an 8ohm speaker (instead of the 4 ohms that I had before) to match the two output tubes. The plates now have an amber glow... not red... can't rember if the plates should glow at all normally...should they?

Checked bias with only two tubes in sockets V7 and V10

Bias for V7 side using transformer method:
CT to plate: 37.5 Ohms
Voltage drop: 1.8 V
Plate current: 1.8/37.5 = 0,048 A
Plate voltage: 485 V
dissipation: 23,28W
% of max: 76%



Bias for V10 side:
CT to plate: 35.1 Ohms
Voltage drop: 1.74 V
Plate current: 1.74/35.1 = 0,050 A
Plate voltage: 485 V
Dissipation: 24,25W
% of max: 80%

So they seem to be on the hot side still. Voltages on the grids is still -50V

I think that the grid pins did not mate properly with the socket. Perhaps I can get an even better connection if I scrape inside the socket pins also, or buy new sockets.

[Tom Phillips]

... The plates now have an amber glow... not red... can't rember if the plates should glow at all normally...should they?...

I would not expect the actual plates to "glow" at all.

... they seem to be on the hot side still ...

I agree. One contributing factor is a higher plate voltage due to the lighter load with only two of the four power tubes installed.

I find that cooler settings still sound great and are often preferred by customers when they are presented with a blind sound test. Many old Fender amps left the factory set as cool as 30% of max. I usually don't set higher than 60%. The 75% target is just, in my opinion, an internet lore thing.

[The Dude]

......I find that cooler settings still sound great and are often preferred by customers when they are presented with a blind sound test. Many old Fender amps left the factory set as cool as 30% of max. I usually don't set higher than 60%. The 75% target is just, in my opinion, an internet lore thing.

^^^^^^ That! Another benefit is extended tube life.

[JAelec]

Bought new quad set of matching tubes and now get these results:

Bias for V7+V8 side using transformer method
CT to plate: 37.5 Ohms
Voltage drop: 3.2 V
Plate current: 3.2/37.5 = 0,085 A
Plate current per tube: 0,043
Plate voltage: 469.5 V
dissipation: 20.2 W
% of max: 67%



Bias for V9+V10 side:
CT to plate: 35.1 Ohms
Voltage drop: 2.97 V
Plate current: 2.97/35.1 = 0,085 A
Plate current per tube: 0,042 A
Plate voltage: 472 V
Dissipation: 19,8W
% of max: 66%

So better then before but perhaps I will lower the bias a bit still...

I also found another issue:

there were some sudden high pitched squeals with the new tubes. One time while measuring voltages, and if I remember correctly one more time when not measuring, although I'm not 100% sure.

I tried to reproduce by rocking the tubes in their sockets, tapping with a chopstick on all caps and at various places in the amp. Also tried measuring again. Was not able to reproduce this fault. But one of the caps is a bit microphonic it seems... a ceramic disk 0.1uF 1kV that is connected to one of the grids (pin2) of the phase inverter. Tapping it makes a sound in the speaker.

Is it normal/common for ceramic disc caps to be a bit microphonic? Could this be the cause of the high pitched squeal somehow? Any ideas?

[Tom Phillips]

...there were some sudden high pitched squeals with the new tubes. One time while measuring voltages, and if I remember correctly one more time when not measuring, although I'm not 100% sure...

It is probable that probing some voltage measuring points caused the high pitched squeals you heard. I have experienced this many times.

...one of the caps is a bit microphonic it seems... a ceramic disk 0.1uF 1kV that is connected to one of the grids (pin2) of the phase inverter. Tapping it makes a sound in the speaker. Is it normal/common for ceramic disc caps to be a bit microphonic? Could this be the cause of the high pitched squeal somehow? Any ideas?

Ceramic disk caps are often microphonic even when they are new and meeting all other performance specs. However, the level of the microphonic sound is not usually a problem in guitar amps. Of course, it's not a "feature" that is a benefit to the overall sound and I say that is not the cause of the high pitched squeal you heard.

[JAelec]

It is probable that probing some voltage measuring points caused the high pitched squeals you heard. I have experienced this many times.


Ceramic disk caps are often microphonic even when they are new and meeting all other performance specs. However, the level of the microphonic sound is not usually a problem in guitar amps. Of course, it's not a "feature" that is a benefit to the overall sound and I say that is not the cause of the high pitched squeal you heard.

Ok, thanks that's good to know.