Tube Amps / Music Electronics
For current discussions, please visit Music Electronics Forum. The sunn still shines online!

ampage archive

Vintage threads from the first ten years

Search for:  Mode:  

Re: Spike protection? Diodes?

6/21/2005 11:25 AM
Re: Spike protection? Diodes?
Hi loudthud,  
I get the idea of a "flyback" effect when the speaker blows / is disconnected, but don't quite follow why the spiking would occur when the amp is overdriven hard...  
With the speaker load in place, it seems that the loading would prevent spiking, even in quasi-square-wave mode..... no?  
6/22/2005 3:43 PM

What I think happends goes like this. When you are outputting a positive current into the speakers inductance and then suddenly try to go negative there is a short span of time when both output devices are off. Durring this time the inductance tries to keep the current flowing by making the voltage more positive. The device that was conducting is now off and the device that is starting to conduct is getting a negative plate voltage.  
I used the term device because the same thing happends in solid state amps. They have diodes across the output transistors to catch those spikes. I designed a FET amp that operated without the diodes and just used 200V parts and an MOV for protection. For evaluation I used a Groove Tubes speaker simulator. It made the spikes just like a speaker.  
I'm currently working on a 5E3 type amp. It makes spikes of about 60V. They'd probably be bigger but right now the bias is too hot cutting down on the crossover region. I think the BlackFace Fender's feedback resistor (820) limits spikes in those amps but in a Marshall, without such a resistor, you are much more vulnerable.  
Traynor YBA-1A and YBA-3 used strange looking MOV's across the transformer primary. I put one of these on a curve tracer and it had a strange characteristic that once it started to breakover, it would start clamping at a lower and lower voltage as you turned the current up.  
Anyway, the spikes are there, you just need a third hand to adjust you scope while you are playing to see them. Sorry I didn't see your post right away.  
Thanks for the old thread. I really wish there was a search function for those old threads. A lot of good info there.
6/20/2005 9:20 PM

They didn't use three "in case one blows," they use threee so none of them DO blow. The voltage spikes that can form from an unloaded transformer secondary are substantial. Three 1N4007 in series is cheaper than one high voltage diode. But for example Peavey uses the 2000 volt MR2502 for their flyback diodes.  
When the diode shorts and takes out a fuse, wouldn't that be preferable to losing the output transformer? Either way the amp is down for the night, but it is a hell of a lot easier to fix a bad diode than it is a bad transformer.  
As to making a tonal difference, you cannot compare different amps because they are not the same, even if they use the same tubes. The only meaningful comparison is to try the same amp with and without them.  
They are reverse biased, so they don't conduct until the spike exceeds B+ in a negative direction. When they are not conducting, they are not in the circuit. In the positive direction they won't blow until the PRV rating is exceeded.
6/21/2005 8:04 AM
Dave H

"They are reverse biased, so they don't conduct until the spike exceeds B+ in a negative direction."
That should clamp the other end of the primary to 2 x B+ in the positive direction. If it can blow three series 1N4007s itís not very effective spike suppression. Maybe a few hundred ohms across the primary would be better?  
6/21/2005 8:09 AM
Dave H

Duh! That should read ďa few hundred ohms across the secondaryĒ
6/21/2005 11:15 AM

I like the power resistor across the secondary.  
I can hear the difference after adding these diodes to an experimental amp.  
I can hear the resistor too but it seems not to be a tone compromise compared to the diodes.  
6/21/2005 11:51 AM
Rob Mercure
Hey Folks,  
Haven't posted in a while - haven't had the shop set and have been on the road mostly doing electrical contracting. But I still check out the board from time to time and it's interesting and perhaps sad, how much "lore" and "wisdom" get lost over the years. I've been on/off Ampage for about 8 years I believe (perhaps 9) and the topic in question was well researched/discussed sometime around 1999-2000 (wish I had copied those posts). In any case there was experimentation using a test chassis and a storage scope to capture the flyback spike/damped waveform from suddenly unloading an OT secondary (I remember the Doc Horner was heavily involved in this - perhaps Steve Ahola captured some of the info)and different menthods of spike suppression were tried. Overall having a resistor 10 or 20 times the output tranny speaker impedance tied across this secondary provided the best damping/protection - and the cost/hassle factor was minimal - I use this on my builds and have also installed said resistor on all of my own amps.  
On a personal note - the first major manufacturer that used the diodes that I remember was Ampeg after Phillips/Magnavox bought them. The diodes shorted frequently and most musicians opted for risking their OT's by having the diodes clipped out before risking a downed amplifier at a gig.  
PS: There seemed to my more "meta experimentation" to define parameters of electronic circuits back in Ampage's "old days" - more "taste tests" of tubes (Mook still around?) and components. Unfortunately I suspect that the boon of computer modeling has also deprived us of much of the "sit up all night swapping parts" experimentation that provided the "holy grail" circuits that we use/copy.

<<First Page<PrevPage 2 of 9 Next> Last Page>>