Choosing central bias

[snowballfight]

Guys,

Posted earlier about my silly ideas to build an amp... got schooled, went away and did it all again properly. Just a query about the biasing.

I understand that you can run the tube in central, hot, cold and so on to get the drive tones/types of drive you want, but I wonder what is the actual value of central bias? Is this found under typical characteristics on the datasheets or is this an agreed upon value?

My first stage is using a 12AU7 hopefully running on center biasing. So what kind of number should I start by using? Of course, I can fiddle with it afterward. I was also wondering about using diodes (probably more than one) to regulate whilst achieving full gain, is this worth it?

Thanks!

[jmaf]

There is a range where anything between the max(cutoff) and minimum bias(plate bbq) voltage is OK, then you choose somewhere in between that range according to your needs. Center would be somewhere in the middle, I guess.

How do you find the approximate middle? For guitar amps, techs normally shoot for 70% of full tube dissipation. So, if you have +B voltage and you know the tube's wattage, bob's your uncle: you know the 100% current. Take 70% from that and you've got the current to shoot for.

This normally isn't done for 12AU7 tubes though, this is power tube stuff. For preamp tubes their maximum dissipation is 1 to 3 watts max, so you're not really worried about quiescent current, but about tone shaping, frequency response, etc, instead.

Good luck with the project.

[Wilder Amplification]

"Center bias" on a single ended preamp stage would be the bias point that gives 100% symmetry at the output (i.e. the point in the positive swing where the tube hits cutoff is exactly the same as where it hits saturation on the negative swing). However there's really no way to achieve "center bias" on a SE preamp stage due to the fact that the grid curves are not linearly spaced. You can get close but not 100% symmetry.

If you draw up a load line it becomes pretty easy to see.

[mooreamps]

Guys,





My first stage is using a 12AU7 hopefully running on center biasing. So what kind of number should I start by using? Of course, I can fiddle with it afterward. I was also wondering about using diodes (probably more than one) to regulate whilst achieving full gain, is this worth it?

Thanks!

Ya, it is worth it. To bias the first gain stage, start with 2 1N4007 in series for a total forward bias of approx. 1 volt. The remaining preamp gain stages, use a single red ultra bright led for a total forward bias of approx 1.7 volts.

good luck !!!


-g

[snowballfight]

Thanks, I was thinking of doing something similar already. Is there any way of putting in the idea of red LEDs into the valve power circuits without actually affecting values? What I mean is, I was trying to use LEDs to quickly show that the valves are recieving power (ie - wired correctly and on = LIGHT) but without changing the characteristics of the B+ or cathode voltages. I could put them in the heater circuits but that wouldn't be a direct measurement on voltage travelling down B+.

[snowballfight]

Yeah, this is what I gathered - do you know of any threads that describe that process of mapping audio/AC curves onto load lines around here? I've got Merlin Blencowe's preamp book infront of me as reference, but this is the one section of the biasing I really don't get. I understand how to change it, and what affects this would encur but not how to plot it to get the p-2-p voltage swing....

Hope you can help?

[Wilder Amplification]

Here is an example load line for the front end of a Marshall Super Lead Normal and Bright channels. You can see how the different cathode resistors give each stage a different bias point -



You can see on this load line, the 820R cathode resistor appears to provide the closest to "center bias" as the grid curves that exist between 0 and -2V seem to have a "linear" spacing. Of course this only holds true with a 2Vpk-pk input signal (1Vpk or 0.707Vrms). Once the input signal swings higher than that, it will push the tube into saturation LONG before it pushes it into cutoff, and you end up with the asymmetrical output. For a front end 1st input stage with a typical passive guitar pickup though, this input voltage is more than likely never reached.

Now for push pull output sections, you have to do what is called a "composite load line". This is where you draw the load line for 1 tube, mark the bias point, then take that same load line and "inverse mirror" it (i.e. copy and flip it), then line the bias points on the two load lines up. One of them gets flipped to reflect the fact that it operates out of phase relative to the other tube.

Here is one I did for a KT88 amp using a 5K plate to plate Z with a 600V B+ and a 300V screen voltage -



You can see here how the load line makes the two tubes look like 1 big tube that can swing between zero and double the B+.You can also see that the ends of the graphs line up with 1200V, which in this case is double the B+. In reality though, it can only swing to the voltage at which it crosses the "Vg1=0" line, which in this case would be down to 125 volts on each side of the swing. This would amount to about a 950Vpk-pk swing.

If we increase the screen voltage by a factor of 50 volts or so, we can shift the curves up so that the load line is crossing the Vg1=0 curve just below the knee rather than right through the knee as it is doing in this example. Crossing the load line just below the knee will yield maximum possible voltage and current swing and therefore maximum power transfer.

This load line also applies for a 4 tube amp...but you double the current scale on the vertical axis.

[snowballfight]

Ah ok I see, this has cleared up alot! I'm taking the area with linear-ish spacings as the area to head for, and the center of that is about central bias. Looking at my JJ tube datasheets for an 12AU7, first stage tube, I should be heading for 0-Vg-8V and centering at -4V.

Only problem here is that this gives a voltage of around 130V - is this not a little low?

[MWJB]

Snowball fight wrote: "Only problem here is that this gives a voltage of around 130V - is this not a little low?" A 12AU7 draws a lot more current than a 12AX7, so runs at far lower plate voltage in a similar circuit, e.g. in a middle of the road preamp triode with 1.5K cathode resistor & 100K plate resistor (not that I am suggesting those values, they're just for illustration purposes) that 130v on the plate would equate to around 435vdc at the B+ rail feeding the plate resistor. In the same circuit we around would see around 295vdc on a 12AX7...so in the grander scheme of things 130vdc is really quite a high plate voltage for a 12AU7. I would perhaps build in a little leeway so that you can also try 12AY7 & 12AT7 and go by ear.

[snowballfight]

Yeah I have been toying in my mind between the 12AU7 and 12AT7 and I don't really know which to choose. This is for a preamp for bass so I'm not so concerned with the high overdrive sound of lots of 12AX7s or whatever, I heard this tubes are nicer for clean tone. If I've only got a B+ of around 250VDC to play with then you mean I should be down lower? Sub 100? In terms of leeway, this is still on paper, so anything is up for change :S ha

[MWJB]

"If I've only got a B+ of around 250VDC to play with then you mean I should be down lower? Sub 100?" Well, you won't really be able to avoid going lower unless you use a very high value cathode resistor (>2.7K?) & a low value plate resistor (I'd start out aiming around 100K as the smallest value at the plate).

Why only 250vdc available at the B+ rail?

[Merlinb]

Yeah I have been toying in my mind between the 12AU7 and 12AT7 and I don't really know which to choose.

I would recommend 12AU7 as it has a much more pleasant grid-current (overdrive) characteristic than the 12AT7. 12AT7 tends to get kinda fuzzy if you get too close to overdriving it, because it has way higher grid current, and it starts flowing sooner too.

Is there any way of putting in the idea of red LEDs into the valve power circuits without actually affecting values?

You can put LEDs in series with the anode resistors of the preamp valves quite safely, which has negligible effect on operation. It's trickier with power valves because the currents are higher.

See here for some LED ideas: http://www.freewebs.com/valvewizard2/ampwithLEDs.jpg

Also, you might like to try my load-line plotter program. Just plug in values for HT, anode resistor, cathode resistor, and view the load lines instantly (preamp valves only). Where they cross is the bias point. See bottom of the page: The Valve Wizard

[snowballfight]

Well since this is a stand alone preamp, I was trying to minimise the size/weight of the power transformer. Rectified, the one i'm looking at states that it should give 250VDC. Think its a 190-0-190 @ 80mA winding. Is that too small?

[MWJB]

No, that's pretty substantial for a preamp, in terms of current capacity anyway, how many preamp tubes are you anticipating? If you aren't drawing significant current from the PT, then your B+ may end up higher than you anticipate, not necessarily a bad thing.

[snowballfight]

Erm. 1x12AU7 and 2x12AX7s. Might even bypass a tone stack for now at least if I've got the power amps EQ bands, if that makes a difference to the loading. A lighter load means its ratings will go up? How can I practically measure that? I say this because I'm not really looking for high end gain, more a full tone and little amount of tube attack. I'm not a great fan of bass overdrive atm.