1 ohm bias resistors. Needed?

[turbofeedus]

These are the resistors I use. (Mouser link - Vishay CPF.)

[angelodp]

Ah those are the little ones I was looking for, thanks

[roberto]

I use them too. You can measure the dc resistance of the two half of the primary and measure the voltage drop to calculate the bias, but those resistors are safer. You can even use 10 Ohm to have more accuracy.

[Roe]

I'd add, get 1W or 2W rated. You'd probably be fine with 1/4W rated, but the higher wattage ones have thicker leads, which makes things a whole lot better if you are taking measurements on an open chassis. If you run tip jacks into the chassis, then it probably doesn't matter at all.

0.6w works well - it acts as a fuse if the tube shorts.

[roberto]

Does it really works in reality as a fuse?
A 600 mW 1 Ohm resistor can constantly let 600 mA pass through with no issues.
To use it like a fuse (so with a prompt response) you need to exceed dramatically that value, and this means the current to flow through the primary of the OT, that can have up to 100 times that resistance, so it will dissipate 100 times the wattage of the resistor. I'm not so sure the resistor will blow first.

[pompeiisneaks]

Well a 6L6 in push pull can get to 200mA at max in normal operating mode. If you have 4, then it would be 400mA. I would think if it goes into runaway that 600mA would happen pretty fast and the resistor would smoke like a fuse pretty quickly too.

BUT that's a guess, I've never had it happen myself.

~Phil

[roberto]

A resistor won't burn at its rated power.
But even so, primary has a way higher resistance, so higher power to be dissipated.
What will blow first?

[pompeiisneaks]

You're correct that the current will go through the primary windings as well. I don't know all tranformers, but generally they are over specced and also can take some extra current/heat. I would hope the 1 dollar part(1 ohm resistor) is going to fail quicker than the hundreds of dollars part (OT).

Again, this still is from theory, as I've not seen either myself. I don't know the gauge of the actual windings, but I have a transformer here in front of me and measuring the leads size it looks like the primary is about 20 gauge which should handle 3 amps that's a lot less than 600mA. Maybe they oversize the leads from the taps inside, not sure, but i'd think for cost reasons they'd match them to not have to pay for larger leads in the connectors?

~Phil

[roberto]

You can find 100 W OPT specs I posted here years ago: AWG26 for primaries.
I have to say that I have substitued many OPT and no cathode resistors, since I started working on tube amps.

[Roe]

Well a 6L6 in push pull can get to 200mA at max in normal operating mode. If you have 4, then it would be 400mA. I would think if it goes into runaway that 600mA would happen pretty fast and the resistor would smoke like a fuse pretty quickly too.

BUT that's a guess, I've never had it happen myself.

~Phil

I have tested it and blown at least one of these resistors. But .25w might also work

[roberto]

Well, a standard guitar amp with 6L6GC, 500 V B+ and 4k4 load, reached g1=0 at around 380 mA: a 250 mW 1 Ohm resistor could blow before the amp reaches full power.
It is very case dependant, but personally I would still keep it not as fuse.

[R.G.]

The stuff from the start of this old thread still applies. A small resistor might open on a tube short, but it's really just wishful thinking to rely on it to be a fuse and protect things.

It is hard enough to choose real, no-fooling, designed-for-the-purpose fuses to blow at just the right time. Deciding that you can second guess when a resistor of unknown materials and unspecified overload characteristics might open and protect something is ... a lot harder.

If you manage to get one variant of a resistor to work the way you want it to, a resistor of the same resistance and power rating from another manufacturer may not perform the same way.

In my opinion, if you want current protection, use something designed for the purpose.

Sorry... end of rant.

[Littlewyan]

Agreed. If it's protection you want you could put an actual fuse on the cathode. Marshall did something like this on the 6100LM, but with an LED and a 100K resistor in parallel to the fuse. When the fuse blew the LED would light up. They used a T500mA fuse per pair of EL34s in the 100W versions. Pretty cool as it meant only one pair would potentially shut off, leaving you with an amp that was operational enough to finish a gig. I read that they didn't include this feature in future amps as the quality of EL34s improved, which I thought was a bit bizarre.....

R.G. Have you got any extra information on the biasing circuits you mentioned on the 1st page of this thread? I'm quite intrigued by what you described!

[Phil_S]

<snipped for brevity> In my opinion ... use something designed for the purpose.

IMHO, FWIW, hardly a rant. This is first rate advice and probably can't be said often enough.

[R.G.]

@Phil: Yeah, but you didn't read what my brain was trying to get my fingers to type...

@LittleWyan: I'll go dig some of it out. In general, it relied on small cathode resistors to ground for each tube. The resistors' voltages were ... um, IIRC, amplified and then lowpass filtered with a long filter, a second or two, and that was applied to a window comparator. The window comparator outputs were used in the logical sense to get too low (i.e., below the target window) too high (above the target window) and just right (inside the target window). The target "that's good!" window was small, a few millivolts. So the lights driven by the comparators were blue for too little current, green for just right, and red for too hot. At the time, I just implemented it in discretes and analog ICs. Today, I'd probably use a PIC. I might even use a PIC to read the cathode resistor voltages by A-D and do the math for indicating hot/cold/just right with a single IC.
The actual bias adjustment was the old manual way, with a human turning a pot. It was one pot per tube, and one set of indicator LEDs per tube. Today, I'd use RGB LEDs, one per tube. I was surprised at how fast a human can set even a touchy "green light" per tube.

I'll go dig it out.