Bias range question.

[Raoul Duke]

Tried searching both here and Google, but I think the specificity of the question prevents searching effectively - so on to posting:

Been tweaking my ODS 2nd gen build and really trying to pay attention to the small details that add up to great sound. To that end - I removed the JJ 6L6GCs I built it with (~30 hours on them) and put in some Tung Sol 6L6GCs I had (~60 hours on those) just to see what they might offer. I expected to have to re-bias, so I turned the pot as low as it could go and started the amp. Voltages on power up at every B+ node were in line with expectations, but when I measured bias - it was way hot at the lowest travel of the bias pot. Here’s the numbers:

JJ: ~460 on the plates biased at ~41mV at about 50% travel on the bias pot
Tung Sol: ~455 on the plates at ~62 mV with bias all the way down

Is this normal to have such a difference with the same tube type? I expected there to be a difference - but this seems drastic. Admittedly, I’m new to building amps - so this may be an obvious question for the more knowledgeable folks. Just trying to understand what’s happening here.

Any info is greatly appreciated! Thanks!

[martin manning]

Yes, you will see variation like that, but these two seem pretty extreme at 50% difference in current. You'll have to tweak your bias circuit to get a bit more negative voltage to run those Tung-Sol's

[Stevem]

Tubes pull millamps when at idle, not mv even though your reading mv .

Besides that, something does not seem right for the TS tubes to be pulling 50% more current yet the B+ dropping only 5 volts.

Does your heater voltage change between having the JJs or the TS’s in the amp.

Also note that the TS tubes idling at .062 amps each with 455 volts on the plate is idling them well above the normal max 70% of there 30 watt rating and will not live long like that if left that way.

[GAStan]

What is your actual bias voltage?

[Raoul Duke]

Thanks guys! I appreciate all the help!
All things considered given where the amp is as far as sound - I’ll save the TSs for something else. I don’t think it’s worth the effort to re-work the bias circuit - I’m pretty happy with the amp’s performance as is.

Stevem, I didn’t leave them running more than a few minutes. All the voltages to include the heaters is within the same ranges as the JJs - except the bias.

Glenn, the bias supply stays pretty much consistent between the two sets of tubes. I even subbed in another set of JJs from my 102 and they were pretty darn close to the original (to this amp) JJs. Bias supply was consistent throughout the three sets. It’s the one measurement I didn’t write down - but I recall they were all consistent.

Martin, does that tweaking involve changing the resistor on the bias pot; or the bias supply board, or both? I imagine such a drastic difference might require both? Asking so that I might learn something I didn’t know from this exercise.

Thanks again all!

[martin manning]

Martin, does that tweaking involve changing the resistor on the bias pot; or the bias supply board, or both? I imagine such a drastic difference might require both? Asking so that I might learn something I didn’t know from this exercise.

Still got the 3k3 in front of the bias rectifier? Put another 3k3 across it or try a 1k8 there.

[GAStan]

Raoul,

To be more specific, what is the bias voltage going to the tubes? It can be measured at the wiper of the bias pot.

Also to make sure we are all on the same page, when you say you turned the bias voltage DOWN, do you mean closer to zero (smaller absolute number) or more negative ( bigger absolute number)?

Edit in italics

[martin manning]

So how much variation is there in the idle current at a given grid voltage?
Here's a nice article that speaks to it: https://blog.thetubestore.com/tube-matc ... -together/

According to the text, the data is "simulated," but representative of a sample of 500 6L6GC. Seems like the author is being a little cagy here, probably to keep specific data from a specific manufacturer out of it.

The test voltage at which the plate current is measured is 400V (I'll assume plate and screen are equal). It also says:

"In this test batch we can also make determinations based on the normal distribution discussed earlier. Applying some statistical calculations to the test results tells us that the average idle current is about 24 ma. 68% of the tubes produced will range from 21.4 ma to 26.9 ma. About 96% of the tubes will range from 18.6 ma to 29.7 ma, and 99.97% of the tubes produced will have an idle current between 15.8 ma and 32.5 ma. Only 3 out of 10,000 tubes produced will have an idle current outside of this range."

What I get from that is the sample mean is 24.2 mA, with a standard deviation s of 2.75 mA (11.4%), normally distributed, and the variation in the sample is +/- 3s. At 15.8 mA, the -3s tube tests at just under 50% of the +3s tube's plate current.

[Raoul Duke]

Raoul,

To be more specific, what is the bias voltage going to the tubes? It can be measured at the wiper of the bias pot.

Also to make sure we are all on the same page, when you say you turned the bias voltage DOWN, do you mean closer to zero (smaller number) or more negative ( bigger number)?

I’ll check the bias voltage again tomorrow and report back.

Yes, turned down to a smaller number. I followed Martin’s guide for initial power up and static bias setting - so I turned the bias to the lowest setting anticipating I’d be working my up. The bias swing with the JJs was something like -32 (full CCW) up to -55 (full CW) IIRC.

[Raoul Duke]

Martin, does that tweaking involve changing the resistor on the bias pot; or the bias supply board, or both? I imagine such a drastic difference might require both? Asking so that I might learn something I didn’t know from this exercise.

Still got the 3k3 in front of the bias rectifier? Put another 3k3 across it or try a 1k8 there.

That’s interesting - I would have thought going bigger was the answer; but that’s just me guessing really.

After looking at my notes, I see we ran into this with my 102 as well. Went from the 3k9 in the board plans down to 1k8 and it worked perfectly - so maybe this is just a part of the process and the JJs I have in there now are the odd set. They’re probably at least 10 years old even though they only have a few hours on them.

Reading that tube article next. Thanks as always Martin!

[GAStan]

Raoul,

To be more specific, what is the bias voltage going to the tubes? It can be measured at the wiper of the bias pot.

Also to make sure we are all on the same page, when you say you turned the bias voltage DOWN, do you mean closer to zero (smaller number) or more negative ( bigger number)?

I’ll check the bias voltage again tomorrow and report back.

Yes, turned down to a smaller number. I followed Martin’s guide for initial power up and static bias setting - so I turned the bias to the lowest setting anticipating I’d be working my up. The bias swing with the JJs was something like -32 (full CCW) up to -55 (full CW) IIRC.

You should start with the bias voltage at the most negative setting (largest absolute number). The more negative the bias voltage results in the smallest mA setting of the power tubes. Since you already have powered it on with bias voltage set at -32, try powering it up and slowly adjust the bias until you see a change in the mV reading of the 1 ohm cathode resistors on your power tubes. As the voltage goes more negative the mV reading should decrease.

Edit in italics

[martin manning]

You should start with the bias voltage at the most negative setting (largest number).

Talking about negative voltage is always confusing. The convention I follow is more negative is lower, which is mathematically correct and follows the convention for temperature. I still write it two ways to make myself clear, as in the "most negative" setting is the "lowest voltage."

[martin manning]

What bias voltage range is needed to cover the variation in production tubes?
A typical operating point might be 450V Va and Vg2, at say 63% plate dissipation, 42 mA. From a 6L6GC data sheet, a curve of Ia vs. Vg1 can be found. Scaling that curve to the 450V Va and Vg2, a nominal tube would require Vg1 to be -48.3V to idle at 42 mA. In 1959, anyway ;^)

If there is +/- 3s variation in Ia at Vg1, where s = 11.4% as suggested by the article linked above, then the required Vg1 could be anywhere from -51.6 to -42.8, an 8.8-volt range. This was found by scaling the nominal curve up and down by 3x 11.4 = 34.2%, and finding Vg1 for 42 mA.

I don't know how close the average production tube is to the data sheet nominal (if there is any mean shift), and that may vary from one manufacturer to the next. Individuals will prefer different bias settings, too, so targeting a range of -40V to -60V might be reasonable, especially if you are in the habit of swapping tubes often, and you don't want to open the amp up to get the bias where you want it.

[pdf64]

You should start with the bias voltage at the most negative setting (largest number).

Talking about negative voltage is always confusing. The convention I follow is more negative is lower, which is mathematically correct and follows the convention for temperature. I still write it two ways to make myself clear, as in the "most negative" setting is the "lowest voltage."

Following the temperature analogy, to reduce the potential for confusion, I've come use hotter / colder idle anode dissipation. to describe bias setting.

[sluckey]

-100V is definitely more than -10V. Which will push more current through a resistor? Which would cause more pain if accidentally touched?