Anybody know what the plate-to-catode diodes do?

[anson]

Every once in a while I'll see some schematics showing a diode connecting the plate of the power section to the cathode. Any idea what this is about? Always looking for ways to purposefuly tweak my amp a little further.

Here's an example of an Orange with such a layout:

http://users.telenet.be/orangefg/OFG...Cmainboard.pdf

Anson

[Morris Slutsky]

protection

They're supposed to provide some protection in case the amp is operated with a disconnected load by absorbing the high voltage spikes that might otherwise destroy the output transformer or tubes. They might work in this event. They might just end up destroyed themselves, leaving the amp unprotected. It's sort of a crapshoot there. But diodes are pretty cheap so they figure why not?

[Enzo]

Not quite. When those diodes are destroyed, they almost always fail shorted, which takes out the amp fuse. No better protectiion for an amp than the inability to power it up.

[R.G.]

What many people don't understand is that the diodes catch the plates by clamping to ground, not by breaking over. When you are pulling current from the OT, it's going out of the primary lead to one of the plates. As the driven plate goes down, the un-driven plate goes up, by an equal and opposite amount. If some transient tries to interrupt the current, the transformer inductance tries to make the current continue by changing its voltage. The down-going plate is flipped above B+, the UN driven plate is driven down by transformer action. The first one to let current flow clamps the voltage.

The voltage is bounded by the un-driven plate's clamp diode. When the un driven plate is forced below ground, its diode conducts and clamps the inductive discharge. When this happens, the other plate will be below two times B+. This is not much worse than happens in normal operation where the driven plate bottoms out maybe 50-75V above ground, leaving the un driven plate pushed above B+ by only 50-75V less than B+.

So the diodes protect the plate they're NOT on. But both sides have to work to prevent the problem.

I personally much prefer using MOVs across the plates with a breakover voltage of more than 2*B+. This may break over at a higher voltage than the diode clamp, but MOVs are usually tougher, and eat the excess energy instead of trying to push it back into the power supply.

[trem]

On a similar note......
What is this diode for? The one that is in parallel with the 1 0hm cathode resistor on the power tubes.
Thank You
http://www.schematicheaven.com/fende...rod_deluxe.pdf

[R.G.]

It's an attempt to keep the resistors from opening like fuses when the tube shorts. Sometimes it even might work.

[trem]

It's an attempt to keep the resistors from opening like fuses when the tube shorts. Sometimes it even might work.

You have a wicked sense of humor
So what led Fender down this path? Was it a problem at one time, or something an engineer forsaw.....?
Thanks Again

[Morris Slutsky]

It's an attempt to keep the resistors from opening like fuses when the tube shorts. Sometimes it even might work.

That doesn't seem right to me. First of all the resistor really should open like a fuse, if the alternative is pulling a huge amount of current through the OT and likely frying it. I don't see a B+ fuse in that schematic, and the 5A primary fuse might keep the PT from frying but it won't prevent a huge current surge through the OT. That resistor wouldn't make much of a fuse either, having a 1W rating and being 1 ohm, it'd take an amp or two for long enough for something more expensive to fry. The diode wouldn't even begin to conduct until half an amp was flowing across the resistor, and it would likely make the situation even slightly worse in terms of current.

[Morris Slutsky]

Damn I guess I didn't get the joke.

[R.G.]

That doesn't seem right to me. First of all the resistor really should open like a fuse, if the alternative is pulling a huge amount of current through the OT and likely frying it. I don't see a B+ fuse in that schematic, and the 5A primary fuse might keep the PT from frying but it won't prevent a huge current surge through the OT. That resistor wouldn't make much of a fuse either, having a 1W rating and being 1 ohm, it'd take an amp or two for long enough for something more expensive to fry. The diode wouldn't even begin to conduct until half an amp was flowing across the resistor, and it would likely make the situation even slightly worse in terms of current.

Yup. You're right, all counts.

I didn't say it was a good thing to do, just that I believed that's what it was there for. I personally think it's a silly thing to put in an amp from anything I can see. But maybe there's something there I can't fathom.

What I did was to design a circuit to watch the voltage across a cathode resistor and when it got over X, all properly time delayed and filtered, shut off the B+. It took a couple of opamps and some comparators, and used power MOSFETs to open the ground connection of the output tubes when a fault was detected. Seemed to work well.

But diodes? Not a good idea from all I can tell.

[trem]

Morris,
R.G.
Thanks, I appreciate the info.
The both of you are sharp and humble.
Thank You.....

[anson]

I don't relly have much to reply to the post with--just thank you for the very informative discussion.
Anson

[octal]

I think the diodes might be there for liability reasons. If someone is attempting measuring the bias across a blown (open) 1 ohm resistor assuming that they're going to see millivolts of DC, they've a.) probably got their meter on a really sensitive scale and b.) are probably not taking the same precautions they would take if they were measuring the B+ rail. If the resistor is open, and there's not a protective diode the cathodes could have quite elevated voltages on them.

I bet Fender doesn't want to fry technicians or their meters.

Nathan

[R.G.]

You know, that's a real possibility.

[trem]

I think the diodes might be there for liability reasons. If someone is attempting measuring the bias across a blown (open) 1 ohm resistor assuming that they're going to see millivolts of DC, they've a.) probably got their meter on a really sensitive scale and b.) are probably not taking the same precautions they would take if they were measuring the B+ rail. If the resistor is open, and there's not a protective diode the cathodes could have quite elevated voltages on them.

I bet Fender doesn't want to fry technicians or their meters.

Nathan

Seems like any meter that a tech has would be auto ranging, and not much worry if something measures 1 volt or 500.
But anyway, regardless of that.....Do companies really have to think in those terms? The phrase "Idiot Proof" comes to mind......