Hot Rod Deluxe Bias Ramping up ... all by itself.

[christarak]

There is no one on this earth who doesn't find a question he is unable to answer - yes, within his expertise.

Elsewhere here, I told the tale of a little Laney amp I had been poking at off and on for ten years before I finally found the simple mistake they made when assembling the amp. How it took me that long to see it, I have no idea.

Every home run king in baseball strikes out. Every star NFL quarterback throws interceptions. Ask the engineers on this board if they ever had a circuit not work. You can't expect the sun to shine every day. Learn to have the confidence you have a valid systematic approach to repairs, and that in most cases that will lead you to a successful repair. And the times you strike out? Shake them off and get up for your next at bat.

Thank you, my friend.

[christarak]

Yes, it could be tubes with excessive g1 current; tubes that have previously red plated may become prone to that.

It's worth checking the values of the 220k and 1k5 resistors in the grid circuit, maybe even try warming them up (eg with a hairdryer) whilst monitoring them.

Note that the 6L6 g1 circuit limiting resistance in fixed bias is 100k http://www.mif.pg.gda.pl/homepages/f...93/6/6L6GC.pdf
Your problem HRD is >3x that, so issues due to excessive g1 current may be exacerbated!

I hadn't checked the 220K and 1K5. Good idea PDF64! I need to check with the customer on his options for this repair. I want to solve it for my own sanity, but I have other repairs that are now overdue. "I'll be back"!

[christarak]

Were the capacitors in the bias circuit changed yet?

Hi Mozz. Thanks for your suggestions. I haven't changed them. They scoped with only small ripple. Can a filter cap cause a ramping bias level? I suppose its possible if it leaks to ground... I am waiting on an directive from the customer. If I continue with the repair, I will try that swap out.

[christarak]

That is a key question. It's normal for bias current to rise as tubes warmup and you should of course continue to monitor and adjust until things stabilize. Is it possible we are trying to fix something that's normal? Or, maybe it never stabilizes. It might help to know.

That's a good question, Dude. I couldn't see it stabilising. I set the bias at 30mA (Vplate ~ 435v) and I would pull the pin when it ramps up to 44mA (after typically 8-10 minutes). I have biased quite a few amps. I normally bias them, then play them, then come back and bias them again. But, man, this amp had an evil mind of its own.

[Leo_Gnardo]

This problem reminds me a lot of the 2000 series Marshalls where bias voltage creeps upwards as the circuit board heats. Could be a case of the same? In these cases either replace the entire circuit board with a new one that doesn't have the same circuit board conductivity as temerature rises problem, or cut a hole in the board around output tubes' pin 5. Christarak you may want to have a look at this:

The Marshall TSL122 TSL100 thermal bias drift repair page

[Mick Bailey]

Just my own thought on this;

The bias only creeps up with either one or both tubes installed and is stable with the the tubes removed. To me that doesn't suggest something amiss with either DC leakage on the coupling caps or the bias circuit. The voltage on pin 4 is also present whether the tubes are inserted or not, so I'm coming back to either tubes or sockets. Or, is it just the presence of something inserted into the socket that's the problem?

You need an octal plug. If one is not to hand, what I do is get an unwanted tube, wrap it in a rag and break it open, then remove the electrode assembly and glass. I've done this many times to obtain an octal plug. I would then solder an insulated lead to pin 5 (I remove the solder from the tubular pin first). Insert the plug and test the bias to see if it rises or not.

[Jazz P Bass]

I have seen leaky coupling capacitors from the PI do some strange things.

The best one was a leaky cap that would send the current draw through the roof but not when I was monitoring it with my Fluke meter.

Heisenberg's uncertainty principle in play.

[mozz]

The reason i asked if the bias caps were changed is maybe under slight load they are leaking. Try without the tubes but with a resistor to ground on pin 5, see if it creeps then, maybe 1 meg to start, maybe 100k, not sure how tiny bias current is.

[Enzo]

He did lift the coupling caps from the PI.

We asked him to clean the board, but that doesn't mean the board was not becoming conductive anyway. I don't recall anyone able to clean the problem out of a MArshall.

Looking for a tubeless tube base? If he has a bias probe socket, that would do.

[Mick Bailey]

I recently experimented with bias creep on DSL Marshalls and ended up making a tubular milling cutter to take out the solder/pad/PCB around the pin 5 socket pin (the inside diameter is just a snug fit over the pin, so self-centres), then connect a 5K6 resistor directly off the pin to the bias supply, having removed the existing grid stoppers.

6 hours continuous running (upside-down, too) and no bias movement, even with additional heat blowing on the board. YMMV. Beforehand it red-plated on one tube after about 20 minutes.

The thing with Marshalls is the bias voltage shifts even with no tubes installed once the board starts to break down. Maybe there's a threshold where inserting a tube could make the difference between OK and not. I've not seen that, but wouldn't rule it out. The possibilities are endless.

Team America;

Gary Johnston: OK, a limousine that can fly. Now I have seen everything.
Spottswoode: Really? Have you seen a man eat his own head?
Gary Johnston: No.
Spottswoode: So then, you haven't seen everything.

[christarak]

This problem reminds me a lot of the 2000 series Marshalls where bias voltage creeps upwards as the circuit board heats. Could be a case of the same? In these cases either replace the entire circuit board with a new one that doesn't have the same circuit board conductivity as temerature rises problem, or cut a hole in the board around output tubes' pin 5. Christarak you may want to have a look at this:

The Marshall TSL122 TSL100 thermal bias drift repair page

Thanks Leo. I will check it out.

[christarak]

I recently experimented with bias creep on DSL Marshalls and ended up making a tubular milling cutter to take out the solder/pad/PCB around the pin 5 socket pin (the inside diameter is just a snug fit over the pin, so self-centres), then connect a 5K6 resistor directly off the pin to the bias supply, having removed the existing grid stoppers.

6 hours continuous running (upside-down, too) and no bias movement, even with additional heat blowing on the board. YMMV. Beforehand it red-plated on one tube after about 20 minutes.

The thing with Marshalls is the bias voltage shifts even with no tubes installed once the board starts to break down. Maybe there's a threshold where inserting a tube could make the difference between OK and not. I've not seen that, but wouldn't rule it out. The possibilities are endless.

Team America;

Gary Johnston: OK, a limousine that can fly. Now I have seen everything.
Spottswoode: Really? Have you seen a man eat his own head?
Gary Johnston: No.
Spottswoode: So then, you haven't seen everything.

OMG Mick! I would never have thought of that. Brilliant. I don't think I'm game to drill holes in a customer's PCB.

[christarak]

I have seen leaky coupling capacitors from the PI do some strange things.

The best one was a leaky cap that would send the current draw through the roof but not when I was monitoring it with my Fluke meter.

Heisenberg's uncertainty principle in play.

I appreciate that suggestion. I had lifted a leg on each of the coupling caps and still the bias ramping took place. I will definitely store this gem of your experience. Thanks

[g1]

I'm not sure if anyone has raised the possibility of oscillation yet?

[Enzo]

I don't think I'm game to drill holes in a customer's PCB.

It is all in how you look at it.

I might enjoy a lovely piece of salmon with maybe some teriyaki glaze. Someone else might say "Ewg, someone tossed a dead fish on a plate and put goo on it.

When you have a cancerous growth, you cut it out. if your board has a defect that makes it conductive, you remove that material. If that is the problem, what is the alternative, buy a whole new board?