WARNING! Risk of Death with Mismatching an SVT's Speaker Load!!!

[jazbo8]

Follow on to frus and R.G's point on Class AB load line, I made the following graph to show what happens when you double the load, as bob and I mentioned above, it is not the "double plate voltage" (clearly wrong) that kills the tubes but rather the significant increase in screen grid dissipation at full output power. The operating conditions for the output stage is taken from the Ampeg SVT schematic found here. I rounded up the plate voltage to 700V, so it is easier to see. The output transformer has a Ra-a of 5k with 8 Ohm load and 10k with 16 Ohm on the secondary. Under normal operation with the 5k load, the screen dissipation is around 6W, well within the 8W limit; but the dissipation really shoots up to 22W with 10k load, can you say ka-boom!

Class A - solid, Class B - dotted. Per RDH4, Ig2 diss = Eg2 x (Ig2 max/4 + Ig2 idle/2).

[bob p]

At some point, the mains fuse will blow, as the steady state power of 700V and 6 ohms is 2.94 megawatts, and neither the rectifier, the wires from the rectifier, the OT, nor the speaker load will sustain that. Nor will the input power. If they could, the incoming power line would need to supply 24,500 amperes. The mains fuse would likely pop under those circumstances...

I think you meant to say 1.21 gigawatts. Great Scott! You'd need a nuclear reaction to deliver that kind of power!

[bob p]

Thanks to everyone for their help. I was thinking about drawing up a set of composite load lines so that we could get the real answer about voltage swings. I'm having a problem though, in that the only decent set of plate characteristics that I have for a 6550 are from the GE-6550A data sheet, and it's plate characteristics aren't drawn to represent the SVT's 660V plate voltages. Do any of you guys with modelling software have the ability to produce 6550 plate characteristics at 650 VDC?

If any of you guys have modelling software that will draw composite load lines, that would be even better. Being an old-school kind of guy, I'm limited to xeroxing, cutting and flipping to make my composite charts, and I need a decent 660 VDC plot to get started.

One reason to look at the composite load lines, rather than the single load lines that everyone has been pointing to, is the fact that the actual operating points move in a push-pull setup. Here's an example of a "model tube" (not a 6550), courtesy of Steve Bench's tube pages:

http://diyaudioprojects.com/mirror/m.../composit.html



As you can see, the composite curve for idle voltages falls in between the independent curves for idle voltage.

[bob p]

How does this electrocute the user? Well, to start with, it doesn't in any kind of normal operation. The OT has to be able to withstand at least 1500V normal operation, so you probably have to design it to 3-4kV to withstand "normal" transients. Changing the load line doesn't change this unless changing the load line makes BOTH SETS of output tubes turn off at the same time, letting stored energy in the primary go into flyback. Even then the secondary voltage is limited by transformer action unless the spike punches through the primary-secondary insulation to the actual secondary copper.

The SVT actually has THREE SETS of paralleled output tubes, driven by cathode followers. I don't know if you've ever taken a look at an SVT's OT, but the thing is mammoth in size, and extremely well made. Ampeg has been making them for 44 years now, so they've had plenty of time to learn from any mistakes that they might have made originally regarding insulation in the windings.

If that happens, things get more uncertain. An "8 ohm load" is probably about 6 ohms resistive, so the loading on the primary goes from open circuit to 6 ohms plus some inductance. The spike energy, although substantial, is probably not deadly to a human holding the speaker wire, much as a 20kV electric fence pulse won't kill you - but it will make you sorry you touche the fence. If and only if the punch-through melts wires and leaves a solid copper trail from primary to secondary wires will the situation turn possibly deadly. If and only if a conductive trail is established from the B+ side of the OT to the secondary wires, then this could put B+ on the secondary wires.

This is a mechanism by which an OT failure could take speakers with it, but I don't see it as a way of harming the musician, unless he's got the speaker wire clenched between his teeth at the precise moment that a voltage spike causes the OT insulation to fail.

The net is that changing to an 8 ohm load will make things unpleasant for the amp because of the high voltages and currents it handles, and demand some sincere high voltage design in the power transformer, but to be dangerous to a human on the secondary side, a whole chain of thing have to go bad just the right way and stay that way long enough to kill the human, which is tough for the setup to sustain for any time. People are more abused by Tasers.

Earlier you had mentioned that you're not familiar with the SVT's internals. Just so you know, it does have some of the "immortalizing" mods designed into the circuit. Of particular interest to note are the fusible-link screen resistors and screen diodes, and diodes across the OT primary.

example: SVT-2 Pro & SVT-CL:

29992_0_41941h3_Sch.pdf

An open secondary might be a worse problem, but it is possible it also might kill the amp internally before a human had a chance to die. Things get hard to predict when you're dealing with multi-kV arcs that melt copper and iron.

Yes, an open secondary is a completely different situation. Back to the subject of multi-KV arcs -- can anyone tell me about the breakdown voltage spec for a 6550? I can't find it anywhere.

[jazbo8]

Do any of you guys with modelling software have the ability to produce 6550 plate characteristics at 650 VDC?

The one I posted above goes to 1000V, unless you mean the screen voltage?

One reason to look at the composite load lines, rather than the single load lines that everyone has been pointing to, is the fact that the actual operating points move in a push-pull setup.

I can do one if you really want to see it, but most of the important information are already contained in the single-ended graph above, the only real difference you will see on the composite chart is the load line gradually changing from Class A to Class B, but the end points are still the same as far as the single tube is concerned.

[bob p]

I'd really like to have a set of composite push-pull load lines for the SVT application -- not for this discussion, but for other personal uses. so if you can do it, i'd be one happy camper. optimally, i'd like to print it out at 11x17. (old eyes)

[jazbo8]

Sure thing, I am on the road now, so give me a day to come up with something legible.

[Jazz P Bass]

What I like about these 650 volt amps (Old Univox U-1061 comes to mind) is the fact that they hold the screens at 345 volts.

[J M Fahey]

Regardless of flyback voltages (which are another can of worms), plate voltages reach 2X +B voltage on every single cycle they reach saturation, meaning all of the time in a live stage situation.
So we are talking about reaching >1100V peaks, with sinewaves and "easy" resistive properly matched loads.
Anything weirder (such as any real world speaker, overdrive, etc.) can only push those voltages even higher.
Just sayin'.

[bob p]

Yes, the normal type of swing voltages than can reach 2x B+ aren't the thing that had me scratching my head. What I was scratching my head about are the voltage changes in the system that are directly attributable to the mismatched load, and how those voltages can cause an OT to fail, when we know that the ratings on the insulation in the windings is on the order of kilovolts, and these amps have a real world reputation for being able to handle a 2:1 mismatch even when they are being pushed to their limits. (Ampeg SVT are incredibly well built!)

I was thinking that there had to be something going on in terms of flyback, etc., that I was missing. I don't have the kind of expertise in magnetics that other people here have, so I was hoping that they'd chime in and explain things for me. You're right -- it's definitely a can of worms...

[jazbo8]

Still working on the composite load lines, but in the meantime, I found one of R.G. old posts that's relevant to the discussion in addition to what he posted earlier. So the tubes do see nearly double the B+ under normal conditions (more than what the SE graphs suggest), and perhaps even more if the OPT windings have significant parasitic inductance. But even with twice the B+ on the plate, it should not kill the tubes or cause damage to the OPT. So unless there are actual field data that suggest otherwise, I am still going with the "screen dissipation kills" thesis.

[cbarrow7625]

And what's a rail? Sounds like transistor-talk to me

Ha, ha. My balls hurt, stop please. I don't know why I used rails instead of B+ yesterday, it was early for me. That's my excuse and I'm sticking to it.

Seriously though, thanks for the explanations and the links. Very helpful stuff. Thanks for the explanation of why/when the screen could go into meltdown from these conditions; that clearly makes sense. I'm going to have to review my load line "knowledge" now to refresh. I don't often have to think about them very often. Cheers!

[nickb]

Real life is always more complicated than theory. Consider this possible scenario.
(1) The tubes are not quite matched (almost certainly true)
(2) Therefore one tube screen melts before the other
(3) Therefore, at some point only one tube is working so now you have large d(i)/d(t).
(4) High d(i)/d(t) causes large voltages and things start breaking down and, as I think RG already said in one way or another, it's gets really hard to predict, but none of it is good.


Oh.. and then there's instability caused by the higher than designed-for reactance in the load and it's toxic effect on the negative feedback.

[frus]

I learned a lot from the other guys too! Thanks all!

[Pedro Vecino]

Earlier you had mentioned that you're not familiar with the SVT's internals. Just so you know, it does have some of the "immortalizing" mods designed into the circuit. Of particular interest to note are the fusible-link screen resistors and screen diodes, and diodes across the OT primary.

example: SVT-2 Pro & SVT-CL:

[ATTACH]25892[/ATTACH]

The most effective of the "immortalizing" mods in my experience is the 220 ohms resistor in each cathode of the 6550īs, although is not represented in the diagram.