"unstablilized" cathode bias voltage

[EETStudent]

Ive got a pair of 6L6's cathode bias, and both the cathode and plate voltages modulate(?) about 4V each. The cathode rapidly varies between 39-43V, and the plate varies 472-476V. There is no bypass cap and no input signal. Is this common?

On a similar note, according the Weber's bias calculator, 2x6L6 with a Vk of 43 and a Vp of 476 and a Rk of 330 the amp is biased at 62mA per tube and 26W per tube. Well, the amp is rated at 30W Class A. Whats with the super hot bias? Ive already had a power tube temporarily give out( it resurrected though) and of the 3 sets of 6L6GB none of them can handle it....Am I missing something??

[EETStudent]

Meant to add, also in the weber bias calc. the recommend bias setting for class A 6L6s is 40.7mA. Again, why is my amp biased so hot??

[EETStudent]

Ok, I want the previous posts to get a response, but while its fresh here is another concern with the bias. The 'bias feed' resistors in a fixed bias setting are typically grounded in cathode bias. In my amp, I measured the resistance from where the two 220k resistors met to ground, and it read about 100k. The circuit board, mostly handwired with a few traces, shows no traces nor can I think of another connection point for these resistors. Should I fix this or is it doing something different?

[tubeswell]

The cathode bypass cap is there to stop the cathode voltage modulating (which it otherwise will do as you have observed, if there is no bypass cap). Thus the bypass cap diverts any instantaneous rise in cathode current into charging the cap, and when cathode current falls, the cap will supply the 'deficit' from its own charge, helping to hold the cathode voltage constant, and providing more gain to the stage. (The modulating voltage of an unbypassed cap lowers the stage's gain, because it follows the plate voltage in proportion, thus robbing the stage of some of the voltage differential between the cathode and the plate, lowering the amplification).

The 'side effect' of the bypass cap is that, depending on the size of the cap, you end up rolling-off certain frequencies (which is why you need to pick the right size cap to suit your taste).

[Rob Mercure]

EETS,

I'm not sure I understand your posts. OK, this is a cathode biased amp - got that. But your third post seems to read that the grid return (leak) resistors aren't connected to ground - is this true? If so then there really isn't any bias on the tube as the grid is floating free and I'm suprised that the tube hasn't "redplated."

Secondly the lack of a cathode resistor bypass cap shouldn't cause modulation - it causes degenerative current feedback but there are many designs that are unbypassed to force current sharing between output tubes.

But since I'm not sure whether you grid returns are grounded I'm not going to speculate further.

Rob

[Enzo]

With no signal - I believe that is your conditions, right? - the voltages would be steady. Your B+ is bouncing 4 volts. That would correspond to the mains bouncing about 1 volt. Is it? That would explain it all. My mains bounce a little here too.

The cathode bypass cap would prevent the signal from appearing at the cathode, but would not stop this.

The bias current sets the idle dissipation. This has nothing to do with output power deliverd to the speaker. 26w idle dissipation on 30 watt amp does not mean there are only 4 watts left for the speaker.

The idle speed of your car engine does not subtract from how fast the car can drive down the highway.

As always, it would be very helpful to tell us the model amp you are working on or post the schematic you built from. Otherwise how would we know what was wrong or right?

As to the resistors that go to ground and yet are not grounded, where did you ground your meter? Was there any voltage in the circuit to confuse your meter? What was the path for the common point of those resistors to ground?

[EETStudent]

Ok Ill try to sum up all responses in one reply. First, its a new gibson amp and I do not have a schematic, and the layout isnt very straightforward like Im used to in older designs. However, I have visually traced most of the circuit and its pretty straightforward, much like a Fender of Gibson of the old days. So, bear with me.

As for the grid return resistors. With the amp OFF, I connected the black meter probe to the chassis, and I connected the red probe to their common point expecting to get a very small reading(0-1ohm). The meter read 100k. With the black probe, I tried different parts of the circuit to see where the connection point was. I couldnt find one.

Enzo, Im not too clear on what youre saying about the power output. The amp is rated at 30W, but the measurements Im taking show about 52W. I know they dont subtract, but why isnt the amp simply rated at 50W. Is it because its at idle?

[Enzo]

Because the power dissipated just sitting there is not what goes to the speaker.

Consider there are many situations where one thing is measured but has no relation to another value. The output transformer secondary might be rated at 8 ohms. But does it measure 8 ohms on an ohm meter? No. One thing is the resistance of the wire in the transformer, while the other thing refers to how much current can be induced into that wire by the transformer.

How much power a tube uses just sitting there is not the same thing as how much power it can cause to flow through a speaker. Look at the amp, it says 30w into the speaker. I bet it also has another rating over where the mains cord comes in that says more like 60w. Power output and mains draw are not the same.

I have a 1/2 ton pickup truck, yet it weighs two tons. shoudln't it be called a two and a half ton truck then? No, because the 1/2 ton refers only to the cargo.

Your amp might produce 30 watts, but it needs more than that to do so.


Where exactly did that 40.7ma come from? You have about 435v across the tubes. That makes 17 and a bit watts. That sounds about right for your typical class AB push pull 6L6 stage to me. Class A runs at higher currents as a rule.

But let us look at the big picture. Other than that you checked the tube current and it seems higher than yo9u'd expect, what problem are you having? REgardless of what the grid resistors seem to be doing, if they are working, why would we want to rewire them differently? In general, is the amp working and sounding OK?

[mooreamps]

Ive got a pair of 6L6's cathode bias, and both the cathode and plate voltages modulate(?) about 4V each. The cathode rapidly varies between 39-43V, and the plate varies 472-476V. There is no bypass cap and no input signal. Is this common?


....Am I missing something??

I'll help you out here.. What is the variation on the screens?


-g

[stokes]

In your first post you mention 6L6GB's.If they are actually 6L6GB's you have too much plate volts and are drawing too much current for that tube.I know many OS tubes can handle the over voltage,but some cant and the extra 7 watts may be too much for them,combined with the extra plate volts.Maybe the voltage fluctuation is caused by an oscillation?

[EETStudent]

Stokes, my apologies I meant 6L6GC.

So Enzo, youre saying there are other factors making this amp a 30 watter, not 50W? Like what? How did Gibson come up with 30W?

I got 40mA using the Weber bias calculator. At the top of the page is a recommended bias setting feature. Class A, 6L6, 435Vp calculated to 39.3mA. Maybe that was for 2 tubes? Anyway, thats where that came from.

The ULTIMATE problem is that I have a pair of JAN-Phillips ECG 6L6GC and a pair of military spec RCAs that I want to use, but neither of these pairs can seem to handle it. The amp develops a loud hum and begins to crackle with them in too long. I want to set the bias so that I dont kill these, but I cant figure it out. And now, Im very curious...

[Chuck H]

The "Weber" bias calculator is a fairly blunt tool (No flaming please). It doesn't account for screen current, which can be significant. And IMHO the listed "hot" "normal" and "cool" ma indications are lowish. It may be that it is edited for a "safe" zone with currently MFG tubes. But IMHE you can exceed the ratings on that site with good results.

To figure for class A operation, just run the tubes at 100% dissapation (subtract for screen current and subtract the cathode voltage from the plate voltage in cathode biased amps). The amp will produce less wattage under actual playing conditions because in class A the idle is set to the mid point of the tubes operating range. Nowhere to go but down, as it were.

I think you'll find that the static dissapation read from the the current at the cathode doesn't jive with the Weber calculator in the end. Especially if your using cathode bias.

Chuck

[Rob Mercure]

Actually most of these bias calculators overcompensate for the screen current but using the measured cathode current to calculate plate dissipation which included the screen current - the plate current is actually lower. But I'm not sure whether this calculator subtracts the control grid voltage - if not it is further over estimating the plate dissipation which might just be another "safety margin" that is built in to prevent overdissipation.

Still doesn't get back to the fluctuation though.

Rob

[Steve Conner]

Was the tremolo turned on when you measured it?

[EETStudent]

lol you said it Steve, I had the tremolo on. That would explain the fluctuating. Everything is testing out much better now. I adjusted the cathode resistor, and Im getting about 19W per tube at 51mA each. Still nice and hot, but so bad as to burn up tubes. I guess the 30W rating was throwing me off as was the tremolo engaged, but Enzo put the wattage rating in perspective for me.