Solidstate bias control for tube amp

[richardd1970]

Hello there,
I new to tube amp building, but I have a question. Would it be possible to build a tube amp that had a solid state circuit to monitor and maintain bias on the fly? As this circuit isn't part of the sound production I wouldn't imagine it would impact the tone of the amp (other than maintaining the bias). Is this do-able? Has it been tried in tube amp?

Thanks for any info in advance.
Rich

[Steve Conner]

Hi Richard,

"The Emperor's New Amplifier" by Norman Koren uses servo bias, as it's called:
http://www.normankoren.com/Audio/TENA.html#Bias

This is a hi-fi design, though. In a guitar amp, you have to consider how the bias servo will react to extreme overdriving of the power tubes. You don't want nasty pumping effects, or even radically wrong bias, when it's cranked.

I settled for a compromise: I used ordinary bias adjustment pots but regulated the voltage with a LM337. This doesn't necessarily lead to stability of the bias current, though, unless you regulate the plate and screen voltages too, which I did

[richardd1970]

Yeah, I was checking that out

Thanks for the reply. That was one I noticed in my search. I just came up with the idea while watching a truck tuning program. They had a unit that plugged into the truck in between the computer and everything else. So I got to thinking , if you had a plug and play unit (similar to the compu-bias unit) that plugged into the tube sockets, then you plug the tubes into those. Then you output the info via USB to a PC. The PC would have graphic software to bias the tubes and have different user settings. What do you think?

[Bruce / Mission Amps]

The new Traynor amps use a servo system for self biasing...
it basically uses the small voltage derived from a current sensing resistor on the cathodes of the power tubes to change the bias a set of transistors, that in turn apply more or less bias voltage to the power tube so the current sensing resistors are at the same voltage again... which kinda means idle current.

[Earl]

Has anyone ever tried sticking a reference diode in place of the cathode resistor?

[loudthud]

I was going to mention the Traynor amp but Bruce beat me to it. The YBA200 uses a toriod output transformer. Those are real critical as to the side to side balance. See KOC's book Principles of Power.

As to the reference diode, I've never seen it in a power amp but there was either a Mesa of Acoustic amp that used a diode to bias a 12AX7. The problem with regulated bias in a power amp is that when the line voltage changes and the bias voltage stays the same, you get more variation in bias current than if the bias voltage tracks B+ and heater voltage.

[Enzo]

I think Seymour DUncan used constant current diodes in the cathodes of the input stage 12AX7s as well, back in the days of the COnvertible 100.

When you have like 480v of B+, there is a 4 to 1 ratio of B+ voltage to 120VAC mains. so if the mains rises 3 volts, your B+ rises 12. And that brings us to what Loudthud brought up. What good is regulating the bias without regulating everything else?

I think this whole premise is based upon the notion that bias is some sort of critical setting. Honestly, if you set your tubes for 38ma at idle, do you really think there will be much difference if they drift to 36ma? While it is a common rule of thumb that bias MUST be set at 70% of tube dissipation - or pick your own alternative number - there is in reality a whole wide range of perfectly acceptable bias. And since the current demand on the bais supply is almost nonexistent, the simple voltage divider made of resistors offrs pretty good regulation anyway. And in fact the little changes resulting from mains variation would to a certain extent track the B+ changes. B+ goes up a little, so does the bias.

[Steve Conner]

I put a meter on my first homebuilt amp that showed the power tube cathode current. It means you can see and reset the bias any time. I found that it really wasn't critical, but generally the hotter I turned it up, the better it sounded! So I turned it as hot as I dared without overstressing other parts.

After a couple of jam sessions when the whole back of the amp ended up too hot to touch and smelt a little worrying, I turned it back down some.

I like the cathode current meter enough that I've just painstakingly filed out an odd shaped hole for one on the front of my new build. If I ever get it finished

My day job is mostly computer programming, and I tinker with tube amps to get away from USB interfaces and graphic software.

[Enzo]

I saw a USB phonograph turntable the other day. Plug the damn thing into a USB port and transfer your vinyl music over to the computer.

Oh wow, I just googled it and there are apparently a bunch of them. Silly me.

yes, a lot of guys are into hotter sounds better, and more than one person has said hte best his amp ever sounded was when he turned the bais way up and ran the tubes cherry red until they melted. Sounded great... for not very long.

[richardd1970]

Where this auto-bias idea stemmed from

Just so you have an idea of where this idea came from. A friend of mine who is a musician (and also a paraplegic) prefers tube amps. He has a 1968 Ab165 Fender Bassman amp as well as another (which I don't remember). He generally has to jackass his equipment by himself to gigs that he does and as a result his amps take a beating.
Because of his situation and because he cares for his elderly mom he cannot afford matched tubes and bench time to continually keep his equipment in tip top condition. I thought a method of maintaining the equipment would be great for him, but also for others that want their stuff in tip top condition all the time.
I was thinking that if you had a unit (similar to a compu-bias) that plugged into the sockets of the amp (then you plugged the tubes into those sockets) the unit would then monitor the bias. Another line would come out of that unit and replace the bias pot. When you turn on the amp and let it warm up, you press a button on the unit and it would read and set the bias for a nominal value.
Once set for the nominal value you can select 3 settings (the baseline for one) or down a bit (cold), or up a bit (hot).
If the unit cannot set the bias because a tube is damaged it would sound an audible alarm letting you know a tube need replacing.

I the very near future I plan on building a TWreck Express bases on the A1a schematic. If all goes well I want to incorporate this idea as an integral part of that amp rather than an add on component. I do thin the add on component however can be quite marketable.

What do you think?

[Enzo]

I think it would be an interesting technical exercise. I also think it would be totally unecessary. Fender never sold Bassman or other amps with matched tubes installed in the first place. Mismatched power tubes is no threat to this or any amp. The benefits of matching are subtle tonal qualities, and to me mostly about hum cancellation. Matched tubes ONLY cancel hum generated in the output stage anyway, they have zero effect on hum from the preamp.

That is of course just my opinion.

[Earl]

ENZO sez - "...I think this whole premise is based upon the notion that bias is some sort of critical setting. Honestly, if you set your tubes for 38ma at idle, do you really think there will be much difference if they drift to 36ma?..."

I'm glad to hear someone say that. I can certainly understand the 'need' for a relatively stable operating point in a nice Hi-Hi tube amp...but in a guitar amp? MPO is the same. for the most part, for "matched" output tubes.

I thought the whole point in guitar amps was the most number of non linearities as possible.... :-)

Sorry to drife OT here.....

[Enzo]

Without looking at prior work, I'd find the design specs interesting. the system would have to detect tube current of course. That is simple. It would need to detect plate voltage, also simple. Then it would need to calculate the dissipation from those data and then consult whatever reference standard you program into it. In other words there would need to be a calibrated power level standard to which the auto-bias circuit would refer. Some sort of input control to set the target dissipation.

Now that assumes we are using the old % of maximum dissipation method of determining "proper" bias. it might be even more interesting to have something work off of xover distortion. I'd envision some sort of circuit to compare signals at the grid and plate of the power tubes and generate a difference signal, which would then be used to control the bias servo fashion. Hell, that could even be a dynamic system keeping bias at the desired level in real time relation to the amp performance, since it ignores current and watches for distortion.

[richardd1970]

Exactly

Yes. That was the Idea. Look folks, I know the bias isn't the most important thing but I do find it is often overlooked. I know a lot of folks that keep extra tubes on hand and when their amp acts up they swap out the offending tube. But they never bring the amp in for service to even look at the bias. It happens a lot. Your out at a gig, the amp acts up, you swap out a tube or two to keep you going. Then you never remember to do anything to the amp and might not have the dough for the bench fee to have anything done.

But if you had an amp (or add on equipment) that could allow your amp to automatically adapt to any tubes you put in and possibly let you know that one you put in is already dead or dying. I just thought it would be a handy thing.

And if you look at the piles of useless equipment made for musicians I just thought it would be a viable idea to actually make something that could help.

[Steve Conner]

Well, if I need to swap tubes in a hurry, I just twiddle the bias pot until the meter reads something sensible. And you can tell bad tubes if the meter either starts reading more than it usually does at idle, or doesn't read high enough when you crank it to full volume.

I argue that a cathode current meter and an easily accessible bias pot is all you need. And a pair of oven gloves to make tube swaps even quicker!