Could somebody please explain to me how this works? (Tube amplifier content)

[zachomega]

http://books.google.com/books?id=bmeE_HNkAV0C&pg=PA55&lpg=PA55&dq=diode+spike+protection+tube+amplifier&source=web&ots=bgQaamOFqg&sig=JQw4rhBwkQoPUB0i4o2t7ceMJwQ

I just don't get how this works as spike protection.  To me it looks like the diodes would never conduct unless the ground all of the sudden became positive?  I've been staring at this for too long and I'd like an explaination of how this does what it does. 

Any help would be appreciated!

Thanks in advance,
-Zach

[R.G.]

It helps because it's attached to a transformer.

The center of a pushpull OT is attached to B+, and power is transferred by the output tubes pulling their side towards ground. That means the "off" side rises above B+ by transformer action. But not by over 3000V worth, and that's what has you confused.

There is always some magnetic field that leaks; all windings are not perfectly coupled. This leakage flux acts just like an inductor in series with the OT leads. When there has been some current flowing in the primaries of the OT and the tubes try to stop it instantly, this leakage inductance does what inductances always do - it PUSHES on the plate of the tube that had been pulling it. And by transformer action, the other plate gets a spike in the opposite direction. One of them will be pulled below ground by the spike, and it's then that the diodes conduct.

This is an OK-ish method of coping with spikes. I much prefer my method -  a pair of high voltage MOVs across the OT primaries. The MOV directly clamps the leakage spikes and is designed to do just that. rectifier diodes like the 1n4007's, are not designed for it.

[zachomega]

So if i am reading this correctly...

The tube that is "off" has an increase in voltage.  The diodes do nothing to protect against this.

The tube that is "on" has a reduction in voltage  due to current draw and the bucking action of the primary inductance causing the plate to reduce below ground - which is where the diodes protect. 

Is that correct?  Or am I missing something.  Because that at least makes some sort of sense in my mind. 

-Zach

[Ronsonic]

Here's the deal on the flyback diodes. They protect against the extreme high voltage spikes that can occur if you run the amp cranked without a load connected, or if it oscillates without a load or if a speaker blows in mid-power chord.

Under normal operation those diodes never conduct. But with the 1,000 volt ratings of the three they will break down and turn into a permanent short circuit if the plate voltage ever spikes up to 3KV. That is lower than the breakdown voltage of the output tranny and tubes, sockets and bases and higher than the normal operating voltages. So if the voltage spikes you'll kill the diodes and have to replace them and the HT or mains fuse instead of having an arc through the OPT or tube or socket. Diodes and fuses are cheap and real easy to diagnose - outputs and arced tube sockets not easy or cheap.

The failure mode they protect against is called flyback, coils store energy and transformers work both ways. If you slam a signal into an unterminated OPT, the spike generated in the secondary will collapse causing an extremely high voltage on the primary that could otherwise kill parts. Ampeg used these in all their big stuff and it's a pretty reasonable safety measure for gear that's subject to heavy use and abuse.

We also see flyback phenomena in huge solid state power amps driving lots of big woofers in clubs. On a small scale, it is why we put diodes across the coils in signal relays, so that they won't dump an overvoltage into the control circuits.

I am sure a more clear explanation is possible, but I'll need more coffee.

Ron

[zachomega]

Between the two explainations...and my reading in the Art of Electronics, I totally understand.  Thank you guys!

-Zach