SVT CL Trips At High Power

[Ted]

Got an SVT CL here. It has a new, matched set of Svetlana 6550Cs in it. It biases up correctly and is stable at idle. You can run it up pretty loud, but if you drive it into clipping the fault detection circuits operate almost immediately - with twinkly flashes from inside the valves. (For those of you concerned about my neighbours, I'm running it into a dummy load).

I've tested the O/P transformer ratio with an AC signal and checked the isolation to 1500V.

I tried biasing completely cold and the fault still occurs. The suppliers have suggested that I return the 6550Cs to them for testing. Before I do this, is there anything else I should check?

[Steve Conner]

Is the dummy load the right impedance?

[jpfamps]

This is a common problem with modern SVTs (and Ampeg won't acknowledge it).

There is a modification to the protection circuit to prevent this happening.

Unfortunately I can't shed any light on the exact nature of the mod, but Mike Hill might be able to help:

Amplifier Valve | Guitar Leads | Speaker Cables | Amplifier Tube

[Ted]

The dummy load is 4 ohms, which matches the output setting.

I don't think this is a problem with an over-sensitive trip circuit. (Unless twinkling and flashing is normal behaviour for 6550Cs. )

I note that one of the op amps in the trip circuit has been slugged with a 0.1uF cap (which I believe is an Ampeg approved mod).

[Ronsonic]

Flashy stuff in the tubes is bad. Most likely just a problem with the tubes themselves.

[Ted]

I'm reasonably sure the problem is with one or more of the valves...

Returned the 6550s to the supplier for testing. They said they were fine.

I asked what voltage they'd tested them at and they said 500V. When i pointed out that SVTs ran at 700 odd volts the guy said that 500V was the "industry standard" for testing and that they would be OK at 700V. (I don't see how one statement follows the other but there you go)


Somehow I don't see me getting my money back or a new set of valves here...
Any suggestions?

[Prairie Dawg]

I had a similar problem with a new set of Winged C 6550s that a guy had bought for his SVT-CL. He couldn't get the green lights. It turned out that three of the six 270 ohm resistors on the tube board were open. Once I got that done it was OK except when the tubes were cold there'd be an occasional pop and an intermittent hard red light meaning that a tube was drawing a lot of current. Once warmed up it was fine.
The overall problem is that SVTs run the snot out of tubes and there's no such thing as a good consistent 6550 any more. I'm in the midst of building a tube matcher that can deliver 550v but anything more might be catastrophic.

[woodyc]

I don't know if this is relevant but one time I had a Marshall that would blow the HT fuse every time it got past clipping.

Long story short, it turned out to be a JJ tube where when the screen got hot it somehow deformed and shorted to the supressor. As soon as the screen cooled down it was back to normal and it tested fine.

Another thing I learned from this is that my tube tester (a B&K 707) can't detect a short like that even if it occurred at low voltage because the short test has the screen and supressor tied together.

[Ted]

Ah shit. Good money after bad. I agreed with the supplier to buy another 6550C matched to the original 6 - on the assumption that it was probably only one bad one one out of the set of 6. Of course it didn't fix it.
I put an old (and not terribly well matched) set of 6550Cs into the CL. And of course it sits there running a solid 350W rms into 4 ohms, dummy load glowing, no problems at all. I put four of the 7 new ones into a V4B. They lit up after about a minute at full power...

Now it seems to me, that the supplier, if he feels that these valves test OK, should be able to accept them back and resell them...
Shouldn't he?

[rf7]

There are a couple of Factory Mods from Ampeg that could apply to this scenario. Do you know if they have been done?

[Gregg]

After all these problems of SVT shutting down we need a fault circuit bypass mod or something. This way you'll be able to check if tubes are OK without guessing which one is bad and retubing the whole power section.

[Prairie Dawg]

That's why I built a tube matcher that can test 6550s at up to 600v.

[Ted]

All the Ampeg Factory mods have been done, so far as I can see.
i) One of the trip circuit op amps is slugged with a 0.1uF cap
ii) There are clamping diodes on R58
iii) The trip circuit diodes are 1N4007

Are there any others I should be aware of?

Most of them are actually concerned with making the trip circuit less sensitive.

BTW I've got 705V on the plates and 406V on the screen grids.

[Billy R aka DynaFreak]

Maybe consider larger value screen resistors. Although the screen voltage is relatively low, 22 ohms is not much current limiting for new production tubes.

[Prairie Dawg]

Makes me wonder how people can match tubes with a maxi matcher that only puts our 400v and say they're matched. Max vP on 6550s is 660v. You've shown Ampeg run's 'em hard. Can you provide more detail on the mods that you have mentioned? I'd like to be able to desensitize the circuit on the next SVT that comes in.