Bias probes- ma vs. mv ?

[rockon1]

I "think" I remember reading(from a fairly trusted source) that mv type probes are safer to use. The reasoning was that ma probes run the risk of inducing oscillation?in some amps or something to that affect? Maybe I was dreaming...thoughts? Thanks in advance Bob

[defaced]

Pretty much. Milivolt bias probes put a 1ohm resistor in series with the cathode, and we use Ohm's law to calculate bias current. Miliamp bias probe put your DMMs amp circuit in series with the cathode, and since you've got alot of wire between the socket and ground, you can get oscillations, though I've never had a problem with them.

[rockon1]

Pretty much. Milivolt bias probes put a 1ohm resistor in series with the cathode, and we use Ohm's law to calculate bias current. Miliamp bias probe put your DMMs amp circuit in series with the cathode, and since you've got alot of wire between the socket and ground, you can get oscillations, though I've never had a problem with them.

I think I will stick to the mv type. I have a ma type but havent used it since a 6L6 went "runaway" on me while using it. The tube was fine for months previously and I was just checking the bias again for drift. The ma drawn just kept climbing and a resistor inside the amp started to smoke. I quickly shut off the amp and removed the probe. New set of tubes worked fine and I checked the bias with a 1 ohm resistor I installed instead. I then replaced the resistor that got hot(even though it measured fine and still worked).

Might of been a coincidence but I remember reading an old post from John Philips? about ma probes and possible oscillation. Time to get a mv probe. Anybody know a good place to get one? Thanks Bob

[R.G.]

Time to get a mv probe. Anybody know a good place to get one?

Make it easy on yourself. Get one 1 ohm 1% resistor for each tube. Put this between cathode and ground (in the standard circuit at least). Bring both ground and the top of the 1 ohm resistor out to a test point you can reach without removing the chassis.

Now when you bias, connect your meter between ground and the test point. 1mV = 1ma for a 1 ohm resistor. You can use a ten ohm 1%, and get 10mV/ma if you like. Neither will cause any bad effect in your circuit's normal operations. Use a 2W or 3W metal oxide type resistor. The current needed to dissipate 2W in a 1 ohm resistor is 2W/1R = 1A. For a ten ohm, 2W requires 2W/10R = 200ma.

They may get hot at these currents (really only an issue with the 10R) but they will not fail. You could go to 5W 1% wirewound if you use tens if you like extra safe.

Once you put the resistors in, if you make your bias adjusters available from the back panel too, you can (re) bias without taking the amp apart. Might as well put a bias pot per tube and save needing matched pairs as well.

[tubeswell]

FWIW I use 2 x 0.6W 2R 1% in parallel for each cathode. That's plenty (1.2W) and theoretically is 0.5% tolerant.

[R.G.]

... and theoretically is 0.5% tolerant.

Actually, no. It's still 1% tolerance. Both resistors can be 1% high, both can be 1% low, and you have no way of telling that a priori. If the actual distribution of values in the 1% resistors is normal, using two at a time has the same distribution.

None of which is a problem. 1% is a good enough tolerance for setting bias.

Here in the USA, Mouser electronics stocks several 1 ohm 1% 3W resistors for between US$0.50 and $1.00.

[hasserl]

Or just measure the voltage drop across the OT and calculate the current, which gives you plate current which is what you need anyway. I've got a bias probe, but I can't tel you the last ime I used it, the OT method is far simpler, and more accurate.

[R.G.]

Or just measure the voltage drop across the OT and calculate the current, which gives you plate current which is what you need anyway. I've got a bias probe, but I can't tel you the last ime I used it, the OT method is far simpler, and more accurate.

That's true - but you have to first measure and note the resistance of the wires in the OT. This is very frequently not the same on each side, as many amps don't have well constructed OTs which are balanced windings, and even in the ones that do, the balance is often for AC, not for DC. And it's different from amp to amp.

The OT-voltage-drop method is pretty much the same as the inserted-resistor method, except that you're using the OT's resistance, whatever it is, as the measurement resistor. Both work fine, with some caveats.

The OT method measures plate current only, and you get to hook both meter probes up to the B+ voltage potential. An inserted resistor between B+ and the tube plate does the same thing, but now you don't necessarily have to use whatever resistance the winding has. A resistor inserted in the cathode lead reads the plate plus screen current, and both leads are nearly at ground potential.

I'm with you on the "don't have to get out the bias probe" thing. That's why I pop in measurement resistors in the cathode in all my amps. Doesn't help much if you'll never see the amp again, but mine, I can spend the initial time to do the measurement resistors.

[tubeswell]

Actually, no. It's still 1% tolerance. Both resistors can be 1% high, both can be 1% low, and you have no way of telling that a priori. If the actual distribution of values in the 1% resistors is normal, using two at a time has the same distribution.

None of which is a problem. 1% is a good enough tolerance for setting bias.

Here in the USA, Mouser electronics stocks several 1 ohm 1% 3W resistors for between US$0.50 and $1.00.

Well okay - maybe I should've said that with two 1% resistors, you are twice as likely to be within 1% tolerance .

[tedmich]

tip jacks work well for the bias points, insulated ones from Mouser (105-0802-001 Red/105-0803-001 Black) are only 0.66 each and take a DMM probe tip quite well (paperclips too so avoid 5 yr olds!)

locking RV4 pots are nice for bias:

[R.G.]

tip jacks work well for the bias points, insulated ones from Mouser (105-0802-001 Red/105-0803-001 Black) are only 0.66 each and take a DMM probe tip quite well (paperclips too so avoid 5 yr olds!)

That brought back some of my earliest "self-teaching moments"!


It's also a good reason to vote for the cathode resistors. On a fully screaming amplifier, a 1 ohm cathode resistor only gets up to well under a volt on the sense point. It's substantially at ground; won't hurt it to be shorted to chassis, either.

... like I'd have done with the other end of my paperclip.

[tedmich]

That brought back some of my earliest "self-teaching moments"!


It's also a good reason to vote for the cathode resistors. On a fully screaming amplifier, a 1 ohm cathode resistor only gets up to well under a volt on the sense point. It's substantially at ground; won't hurt it to be shorted to chassis, either.

... like I'd have done with the other end of my paperclip.

I clipped an upright lamps cord with a pair of "guillotine" dog nail clippers at ~6yrs old. They are just about perfect to short the two conductors! I can still smell the ozone...

[Bruce / Mission Amps]

1 ohm resistor, with 300ma to 500ma across it, is handling only .09 of a watt to 1/4 watt.
I think a 1/2 watt resistor is OK and a 1 watter is over kill.
I reality, I rarely use anything other then a 1 watt resistor for current sensing on the cathode lead of 6V6s, 6L6s or El34s.
1 ohm@1% resistors are very easy to find and fit the bill with lots of overhead.

Now, 10 ohm resistor changes things.
It would need to be a 1 watter at 300ma, so I'd use a 2 watter.... and to handle 500ma, I'd be looking at a 3 to 5 watter.
However, there aren't many amps that are pulling 300ma to 500ma per single power tube anyhow!!

[soundmasterg]

If you use 1 ohm resistors to sense the cathode current, to figure out your bias percentage of the max, you still need to know the plate voltage right? So you still have to pull the chassis unless you add a tip pin jack with a divider there to bring the plate voltage down to a safe level, or use a meter like the Bias Rite, which can tell you both the plate voltage and the cathode current. For my own amps I sometimes add the tip pin jacks, but I include a divider to get the plate voltage down to a safe level....say 45v instead of 450v for example. For customer amps the Bias Rite is the ticket usually.

Are there any other suggested approaches or are you guys just going by cathode current alone when you set the bias via tip pin jacks?

Greg

[R.G.]

If you use 1 ohm resistors to sense the cathode current, to figure out your bias percentage of the max, you still need to know the plate voltage right?

Right - assuming that either (a) you worry about percentage of max and (b) the B+ voltage ever changes much in the conditions under which you bias.

If you ever bias an amp a second time, you can usually already know (b), so you don't need to measure it every time, unless you're really, really fussy. They're going to drift anyway, so my personal take is that you don't need to measure B+ more than once if you get it right. I personally also view the "percent of max dissipation" as a bit of gilding that can usually be dispensed with; in fact, it's usually contained in the "get it right once" setup. After that, you're looking for a current number.

For my own amps I sometimes add the tip pin jacks, but I include a divider to get the plate voltage down to a safe level....say 45v instead of 450v for example.

Good idea if you have to have high-accuracy percent bias. I view this as a personal preference, not a necessity.

For customer amps the Bias Rite is the ticket usually.

Any amp you'll probably only bias once.

Are there any other suggested approaches or are you guys just going by cathode current alone when you set the bias via tip pin jacks?

Cathode current alone is generally good enough. It may not be perfect, but it's sure good enough for rock and roll.

Hey! How about this? A uC with A-D conversion inside it, where it reads the cathode current and the divided down plate voltage, so it can on-the-fly do the math and tell you if you have it right! Better, it can self adjust the bias to be either at cathode-current priority or percent of max, selectably. Better, it could sense when the amp had been turned on, mute the output, bias the tubes, and have a fresh, correct bias each power-on, or whenever the bias-me-now button was pushed.

I've done exactly that. Worked fine, but was impractical because the tubes heated and drifted slowly enough that the amp was quiet for too long when biased. Not practical for live playing.