Flyback voltages

[Cliff Schecht]

I understand why flyback voltages occur in an output stage, but at what point do they become damaging? I've never had a transformer or set of tubes die on me, perhaps that has been dumb luck up to a certain point. I'd like to hear others thoughts/experiences with this.

I see some pretty big overshoots in my SE KT88 amp when I overdrive the KT88 hard. I'm calculating pulses of almost 200W, albeit for a very brief moment. You can actually hear the flybacks and it sounds pretty cool, it doesn't occur in a nonmusical way as far as I can tell. I'm measuring this on the secondary side mind you, I'm not to keen to blow out the front end of my scope trying to measure super HV spikes. Should I worry about this? Would some series diodes coming off of the plate do anything to protect the amplifier?

[tubeswell]

R.G. Keen gives a good explanation of the flyback protection mechanism here IMO

http://music-electronics-forum.com/t21645/

[Jana]

One of the dangers is arcing. If that happens then a carbon film forms along the path of the arc--usually between pins on the tube socket. This carbon film creates a path for further arcing and also voltage that isn't arcing. This carbon film is hard to remove; it's usually easier just to replace the tube socket.

If the arcing occurs internally in the transformer it punches a hole/creates a path of carbon and from that point on the transformer is compromised.

I would try to avoid flyback even if it does sound cool. The results might not be so cool.

[Cliff Schecht]

I'll take some pictures on the scope. I can't look at the primary side today because it's over the voltage that my scope can handle but I can probably find either a high voltage probe or an entire high voltage setup to test the amp on up at my schools massive pulsed power lab. I just want to make sure that the flybacks aren't actually going below the cathode potential (at which point I would actually worry about it). Looking at the secondary side just doesn't give me enough info, I need DC information.

[Cliff Schecht]

Upon re-reading R.G.'s thoughts on flyback voltages, i've really settled my mind on the whole flyback issue. I'm not working with a push-pull output stage so the diode trick wouldn't prevent anything for me anyways. But, I'm still worried about the transients I'm seeing. They appear on the primary side as clipping on the top of the waveform (when the grid goes way negative and the tube cuts off) and a massive amount of overshoot on the bottom of the waveform. But what I'm seeing on the scope is a flipped view of this because I have to measure everything on the secondary side (annoying!). The huge negative-going overshoot I'm seeing could be caused simply by attaching the scope lead and adding in extra parasitics, but I'm not buying this quite yet. I need to get this on a HV scope soon..

[Cliff Schecht]

I think I figured it out. When I through a purely resistive load on the amp, there was literally no overshoot on the output (secondary side) even at full clipping. I think what I'm seeing is just kickback from the speaker voice coil as it goes from full negative swing to full positive swing with a high powered square wave. I'm not quite sure why it overshoots in only one direction but it seems that because of the assymetrical nature of the overshoot, it's going to add purely even harmonics which is exactly what I want from the power stage. Like I said, it sounds good so I guess I don't need to touch it!