Which bias? Chuck?

[daz]

I decided to go ahead and redo my bias circuit properly and was going to use the one chuck posted for me. But i was looking at the popular mark huss 6V6 plexi schematic and the values are very different and one change in the diode location. Any thoughts as to which one to use? I realize my pot being 35k if i use the one that show a 10k other values might need changing. I read a old thread discussing the bias circuit and how it can affect the low end and i want to make sure the amp is at it's best so i thought i'd see what y'all think.

[Enzo]

They are really the same circuit.

The large resistor and the diode? Well, imagine you had a 2" brick and a 4" brick on top of it. The stack is 6" high. Now put the 4" brick on the bottom, and the 2" brick on top. Now how tall is the pile? Still 6". The order doesn;t matter much. The rest is the same, just different values. One guy used 10uf, and with two stages of filtration, maybe that makes clean. I've seen amps with less. The other one used 47uf caps. OK, sounds good too. I usually use 100uf caps for bias. I bought a big wad of them for that purpose, and have little reason to vary the value from that.

Larger caps will take a little longer to charge up at power on, but a few seconds don't worry me. I was going through my used tubes the other day, and I would plug a power tube into the socket, hold the test button until the meter woke up. It usually took 20-30 seconds before the power tube started conducting. More than enough time for bias to charge, in my view. So in my humble opinion, you need caps large enough to kill the ripple, and extra won't hurt up to any reasonable level. In other words, 100uf is fine, while 10,000uf maybe not.

That leaves the voltage divider resistances. The top one is 10k over 10k-35k. Starting with the voltage at the first cap, I'll call it Vc, the range at the bias is from 10k/20k or 50% of Vc, up to 35k/45k for 78% of Vc.

The bottom one. 15k over 47k-57k. So with Vc (the 54v in the print), we get 47k/62k for 76%, then the other extreme is 57k/62k, for 92%.

So the dividers vary from a range of 28% on top example, to 16% on bottom example. Not terribly different, just one has a more limited range. On the other hand, most bias controls don't use their whole range anyway. I should say "percentage points" rather than just "percentage." If you weant more range in the bottom one, then change the ratio of pot and guard resistor. In other words instead of 47k and 10k pot, you could use a 47k pot and a 10k resistor. Or a 25k pot and a 30k resistor.

Then the value of the large resistor? Did both circuits start with the same AC voltage? Did they both intend to result in the same bias voltage? A larger resistor will drop more voltage. It is just another value you can fiddle with to set the level and range of your bias supply.

If I assume Vc is -54v in both, then the bias voltage in the top circuit has a range of 27v up to 42v. The bottom circuit then ranges from 41v up to 50v. Are these for EL34 amps? The values of those divider resistors isn't critical, it is the ratios that really matter.

[daz]

It's 6V6. Let me ask you this....whats your take on larger caps and tone. I read a thread where chuck and others discussed it and my take on what they said was that larger caps can affect low end tone. I assume the rest of the circuit is mainly nothing more than getting the voltage range within spec, but if thge caps can affect tone in a good way by going larger i;d do that if only for piece of mind. Heres a post on it....

Unless you change your bias filter cap to 100uF, which seems to be popular.
It causes the low frequencies to distort...and boom...

But as I said, people who don't understand (or don't care) ignore it.
People who are trying to preserve sound quality pay attention to it.

Further, your test with O scope is not making you appreciate it.
Test must be a guitar.

So yeah, you actually do have some control over low frequency with the bias filter cap.
And this seems to happen more often in Fender amps.
But it would apply to Marshall.

Ever wonder why Marshall is using 8uF in Plexi bias circuit? That may be why.
Why does fender use 70 instead of 100uF?
You can really hear a difference, if you can hear...

They are really the same circuit.

The large resistor and the diode? Well, imagine you had a 2" brick and a 4" brick on top of it. The stack is 6" high. Now put the 4" brick on the bottom, and the 2" brick on top. Now how tall is the pile? Still 6". The order doesn;t matter much. The rest is the same, just different values. One guy used 10uf, and with two stages of filtration, maybe that makes clean. I've seen amps with less. The other one used 47uf caps. OK, sounds good too. I usually use 100uf caps for bias. I bought a big wad of them for that purpose, and have little reason to vary the value from that.

Larger caps will take a little longer to charge up at power on, but a few seconds don't worry me. I was going through my used tubes the other day, and I would plug a power tube into the socket, hold the test button until the meter woke up. It usually took 20-30 seconds before the power tube started conducting. More than enough time for bias to charge, in my view. So in my humble opinion, you need caps large enough to kill the ripple, and extra won't hurt up to any reasonable level. In other words, 100uf is fine, while 10,000uf maybe not.

That leaves the voltage divider resistances. The top one is 10k over 10k-35k. Starting with the voltage at the first cap, I'll call it Vc, the range at the bias is from 10k/20k or 50% of Vc, up to 35k/45k for 78% of Vc.

The bottom one. 15k over 47k-57k. So with Vc (the 54v in the print), we get 47k/62k for 76%, then the other extreme is 57k/62k, for 92%.

So the dividers vary from a range of 28% on top example, to 16% on bottom example. Not terribly different, just one has a more limited range. On the other hand, most bias controls don't use their whole range anyway. I should say "percentage points" rather than just "percentage." If you weant more range in the bottom one, then change the ratio of pot and guard resistor. In other words instead of 47k and 10k pot, you could use a 47k pot and a 10k resistor. Or a 25k pot and a 30k resistor.

Then the value of the large resistor? Did both circuits start with the same AC voltage? Did they both intend to result in the same bias voltage? A larger resistor will drop more voltage. It is just another value you can fiddle with to set the level and range of your bias supply.

If I assume Vc is -54v in both, then the bias voltage in the top circuit has a range of 27v up to 42v. The bottom circuit then ranges from 41v up to 50v. Are these for EL34 amps? The values of those divider resistors isn't critical, it is the ratios that really matter.

[eschertron]

...People who are trying to preserve sound quality pay attention to it...

Certain sounds from the history of Rock'n'Roll are a result of bare-minimum engineering for operating envelopes that are routinely exceeded by players. A bias supply that's theoretically sufficient for an amp - for power tube grids that are effectively infinite impedance - is designed as a voltage reference and is not expected to be loaded at all. That bias supply starts to show artifacts when the grids conduct, though. If that affects the sound in a way we've come to expect from the vintage designs, then we MUST exactly clone the design to get the exact same sonic artifacts. In short, if you like the sound of a bias supply being punished, then don't make it any beefier than the historical standard, in regard to the cap values under discussion.

Merlin's advice on tweaking and building is to install a cap that's as small as you'd need; clip in additional caps in parallel until the sound is what you you want. Designing by ear, rather than theory.

[Chuck H]

...whats your take on larger caps and tone. I read a thread where chuck and others discussed it and my take on what they said was that larger caps can affect low end tone.

First... This implies that "I" said larger caps to improve low end. I MIGHT have said to another member that it's been speculated. I MIGHT even have discussed a mechanism where it may. But I don't believe it and I don't think I said definitively that it was valid (link?). If I did, I've since reasoned better. Here's why...

That cap bypasses perhaps 5% to 20% of the power tube grid path to 0V. So any improvement to that portion of the load MIGHT reduce some local NFB by an amount dictated by the improvement @ frequency of the larger cap and the actual ratio of 0V path set by the bias pot. The difference would be VERY small. Also, there's no reason to think that a tiny bit less local NFB at some very low frequency would be an "improvement". It may make the bottom end looser or muddy. But I don't even think this is the case because the circuit, even with the smaller cap values, already filters to below something like 15Hz. So, in the end, any frequency improvement isn't even in a useful range for anything but finding ships under water.

Larger caps will offer more hum rejection, if it makes a real difference I don't know. Consider that any improvement is limited to that amount it improves as observed through any lack of rejection in the push/pull power amp circuit. So, not a lot Is the increase in charge time a problem for power tube safety? Probably not. Does any of it matter beyond some acceptable min/max (like 8uf to 100uf)? Probably not WRT guitar amps. I suppose if you were using the worlds most sensitive bench gear to test some high end audiophool hi fi you could probably find some insignificant improvement claim in the tenths of percent.

What Enzo says about ratios is the most important part. You need to recognize your present ratio WRT any new/improved circuit you use. That is, the new supply circuit should, ideally, be revalued to recognize your particular requirements.

[daz]

Not sure how much of that didn't go whoosh, but am i safe to assume the short answer is doesn't matter in any way shape or form which i use?

[eschertron]

Not sure how much of that didn't go whoosh, but am i safe to assume the short answer is doesn't matter in any way shape or form which i use?

The way I read it, there are two important points to consider:
1) the resistor and pot values for the voltage divider which form the bias adjustment per se. They need to be selected to provide the range of control desired. The 'math' side of the brain does this, through design or experimentation.
2) the bias reservoir cap. From the other thread, the idea that stiffer = muddy is what I get out of it. So the 'music' side of the brain selects the value for this, provided it's above some minimum usable value as Enzo mentioned.
That's all there is, as long as the bias supply or output tubes don't burst into flames. The two schematics are otherwise equivalent

[daz]

Well, i wired up the bias as in the top diagram with the two 10uf caps and it doesn't work. I assume it has to do with the resistor values VS the supply V. I went over and over it and i'm as sure as i can be it;s wired correctly. The issue is i turn it one and the bias voltage starts to read after about 5 seconds, goes up to about 5mV, then goes back down to 0 mV. I tried 2 different values of the feed resistor....220k which was there to begin with and i put another 220k in parallel to get about 100k as per the schematic. Also tried lowering and raising the one that feeds the pot. I have no idea beyond randomly changing the resistors to get within range but i don't know which way to go and which resistor to target. I would assume the pot feed resistor but going from 2.2k up to 47k made little difference and i don't wanna keep trying for fear of hurting the 6v6's.

[Tom Phillips]

... i don't wanna keep trying for fear of hurting the 6v6's.

Just don't plug in the power tubes until you have the bias voltage correct to pin 5 of the power tube sockets.

[daz]

Just don't plug in the power tubes until you have the bias voltage correct to pin 5 of the power tube sockets.

Ahhh, ok. I figured that would not allow you to see the correct bias voltage as you would with tubes in. Not i just gotta figure out which way to go ! Although what gets me is it always goes back to zero which makes me think ti's more than just wrong values.

Hmmm....i can adjust the resistors any way i want even to the point the voltage at the 220k grid R;s is way high. At one point 100v ! Had it as low as 20 but no matter, the tube pin shows 0. It read fine b4 i changed the bias circuit.

[daz]

Ok, i *think* I see what you meant when removing the tubes. Check voltage at the grid 220k junction, but that didn't seem to matter, it still read 0 at the tube socket when i put tubes in no matter what voltage was at the 220k junction. But just now i halved the 10k that feeds the bias put and while the range wasn't right, it did finally read current with tubes in. 25mV min with pot all the way. so i could adjust that feed resistance to get it right, (and would i need to reduce both 10ks to a balanced value?) but theres now another problem. No sound. Checked connections and guitar/amp volumes up, just no sound even with 25mV at the tube.

EDIT: turns out he lack of sound was a unrelated thing, broken lead on a plate R in the gain stages. So it works, and I think i know why. I believe it;s the order in which things happened, but rather than try and think about it, it's done so thats that.

[Dave H]

The issue is i turn it one and the bias voltage starts to read after about 5 seconds, goes up to about 5mV, then goes back down to 0 mV.

It would be in meltdown with 5mV bias voltage. I think the 5mV you are measuring to here is actually the voltage across the 1ohm resistor between the 6V6 cathode and ground so it is a measure of the bias current (not voltage) i.e the bias current is 5mA (because the resistor is 1ohm). Referring to this as the bias voltage is confusing. The bias voltage is the negative voltage on pin 5 of the 6V6s which can be measured with the 6V6s removed (unlike the bias current). With the 6V6s removed measure the bias voltage on 6V6 pin 5 and sweep the pot from end to end and report the max and min values then we should be able to tell you what to change to generate a bias voltage that will give you the bias current required.

[daz]

It would be in meltdown with 5mV bias voltage. I think the 5mV you are measuring to here is actually the voltage across the 1ohm resistor between the 6V6 cathode and ground so it is a measure of the bias current (not voltage) i.e the bias current is 5mA (because the resistor is 1ohm). Referring to this as the bias voltage is confusing. The bias voltage is the negative voltage on pin 5 of the 6V6s which can be measured with the 6V6s removed (unlike the bias current). With the 6V6s removed measure the bias voltage on 6V6 pin 5 and sweep the pot from end to end and report the max and min values then we should be able to tell you what to change to generate a bias voltage that will give you the bias current required.

I get the whole voltage across the 1 ohms and how that works. But i was getting voltage at the grid resistors and none on the pin, but it was a goof.....I realized i had a probe in the probe jack i have on back thats connected to the tube pin when i thought i had it sticking in the chassis and grounded. Thats why 0 reading on the pin. Doh ! It works fine now. When it was doing that up then down voltage thing was when i first wired it up and not sure what that changed. The timeline of what i did is confusing....can't recal the order. No matter, it works.