Champ 12 overhaul... Output stage question

[RomanMtz]

I decided to revamp my Champ 12 completely… staring with the power supply. I used the same PCB and components… just added a couple of diodes for a full bridge rectifier. I measured the open circuit voltage at the transformer’s secondary and got 220V AC which rectified came out to around 300V DC at the first filter cap.

I have 2 9 pin sockets and 1 8 pin socket available. With that B+, what output tubes do you recommend? The OT’s primary is around 5K…

Even though I decided to revamp the power supply, I’m still wondering what the difference in tone is provided by running the output tube with 500V plates and 250V screen grids (like it was set up originally with a voltage doubler) compared to a more standard 300V plate, 295V or so screen grid like I’m planning to do.

Does the plate voltage affect the output tube’s output impedance?

Thanks in advance!

[J M Fahey]

What are you trying to get?

[redelephant]

Does the plate voltage affect the output tube’s output impedance?

Not really. But u have lower voltage over the tube and can run higher current thru it. That means u can load it with a lower impedance to get max symmetrical signal swing, at the expence of distortion. But for guitar that never a bad thing

All in all u'll probably just end up with less power, great for a recording/bedroom amp. But the increase in standing current will increase ripple and hum a little, so u may need some more capacitance in the supply.

5k OT, guess that leaves u with a 5881. 6V6 will like the voltage, and work fine, but really wants a little higher impedance.

[RomanMtz]

Not really. But u have lower voltage over the tube and can run higher current thru it. That means u can load it with a lower impedance to get max symmetrical signal swing, at the expence of distortion. But for guitar that never a bad thing

All in all u'll probably just end up with less power, great for a recording/bedroom amp. But the increase in standing current will increase ripple and hum a little, so u may need some more capacitance in the supply.

5k OT, guess that leaves u with a 5881. 6V6 will like the voltage, and work fine, but really wants a little higher impedance.

Hey guys... thanks for the responses.

By "lower voltage over the tube" you mean the screen grid voltage, correct? If that's the case, why would they need to double the plate voltage to close to 500V?

My original plan has been and still is to simplify the whole circuit... gone is the line ins, headphone out, OT based reverb and all their related components. I decided to go for something similar to a AX84 High Octane with 3 gain stages, a cathode follower, Bone Ray tone stack (on AX84 forums) and a 6L6 at the output. After checking the PT output and modding the PS to an acceptable 300V at the first filter cap, I'm gessing this is a viable option, correct? Is there anything I should look out for that I'm not taking into account?

Even after deciding this, trying to learn as much theory as I can, I still wondered why the original design was made this way with a close to 500V plate and 250V screen grid and how this affected the tone of the output section. And that's why I posted the question here.

[loudthud]

In general if you keep the same load impedance and cut the B+ in half, you get one forth the output power or about 3 watts. With a 6V6 you may get closer to 4 watts but that's about the limit for a 5K primary and B+ around 300V.

"Output Impedance" and "Load Impedance" are not quite the same thing although they are frequently used interchangably. In this case we are more concerned with the load impedance. Any tube can only pass so much current reliably. This is usually a function of the cathode design. There is also a plate voltage limit that is a function of electrode spacing inside the tube and at the base. Plate dissipation also has limits and is the product of current and voltage.

As a general rule, if you reduce the B+ voltage, you must lower the load impedance and increase the current to keep the power delivered to the load the same. But with most tubes used in guitar amps, you quickly reach a point where the current rating of the tube becomes the limiting factor. There are also voltage losses within the tube that become a larger percentage as B+ is reduced.

The net result of all these factors is that with 300V B+ you could maybe get six or seven watts out of a 6L6 if you optimized the load impedance (primary impedance), probably around 1K. How will it sound is anyones guess. The 500V B+ allows the output to easily get to 12 watts and the lower screen voltage reduces stress on the tube. It's probably brighter and cleaner than what you could get from a lower B+.

[RomanMtz]

The net result of all these factors is that with 300V B+ you could maybe get six or seven watts out of a 6L6 if you optimized the load impedance (primary impedance), probably around 1K. How will it sound is anyones guess. The 500V B+ allows the output to easily get to 12 watts and the lower screen voltage reduces stress on the tube. It's probably brighter and cleaner than what you could get from a lower B+.

Thanks loudthud, I get what you're saying... basically this is probably why the amp has very decent cleans and what seems to be a poorly designed preamp based distortion gain channel (I'm guessing by what you said, you probably don't get much output tube distortion).

So as long as you have a Screen Grid voltage lower than the plate voltage, you're OK? Regardless of the value? Or is there a limiting factor with this voltage as well?

I ask this because I get what you say about a tube's plate voltage limit, plate to cathode current limit and (taking both of these) power limit... but I haven't seen any parameters for the screen grid voltage besides not getting over a certain screen grid current (which is another point I don't get when no screen grid resistor is used) and not getting over the plate voltage.

Most single ended designs I've seen have a screen grid voltage a couple of volts under the plate voltage... With the chamo 12, it's the first time I've seen that much of a difference.

[loudthud]

The screen voltage does not have to be lower than the plate voltage. If you look at the curves for a 6L6 you'll see that if screen voltage is 300V, plate voltage can change from 200V to 500V and the current through plate and screen doesn't change very much. When plate voltage gets down around 100V, then the screen current can increase dramatically and even exceed plate current.

There are three reasons most consumer audio gear runs such high screen voltages. Cost, cost, and cost. Making a separate screen voltage supply was expensive in the days before high voltage transistors and solid state diodes. You had to use a regulator tube or another rectifier tube, filter caps and more windings on the transformer. The voltage doubler is the lowest cost solution these days. So the tubes that became popular were those that could be adapted for use with screen voltages near the plate voltage. It does take more to drive a tube with a high screen voltage but this doesn't raise the overall cost very much.

Screen voltage has a huge influence on how much current a tube can draw in any given circuit. In an SE amp, the maximum instantaneous plate current is only roughly two times the idle current. Using the 6L6 as an example, at 500V a plate current of 60mA would take the plate dissipation to 30W. (Too hot for todays 6L6s) The maximum instantaneous plate current might be 140mA. In a push pull amp with 500V B+, a 6L6 might idle at 45mA but the maximum plate current could be 400mA. So the push pull amp needs the higher screen voltage to perform well.

There are push pull amps that don't run the screens at near B+ levels, the so called split rail amps. These usually have a B+ over 600V. The SVT and Music Man amps are good examples. The 6550, EL34, KT88 and 8417 run well in split rail amps. IIRC Paul Rivera designed the Champ 12. He was smart to use a split rail topology. I wonder if he'll ever write a book on amp design....

[RomanMtz]

The screen voltage does not have to be lower than the plate voltage. If you look at the curves for a 6L6 you'll see that if screen voltage is 300V, plate voltage can change from 200V to 500V and the current through plate and screen doesn't change very much. When plate voltage gets down around 100V, then the screen current can increase dramatically and even exceed plate current.

Screen voltage has a huge influence on how much current a tube can draw in any given circuit. In an SE amp, the maximum instantaneous plate current is only roughly two times the idle current. Using the 6L6 as an example, at 500V a plate current of 60mA would take the plate dissipation to 30W. (Too hot for todays 6L6s) The maximum instantaneous plate current might be 140mA. In a push pull amp with 500V B+, a 6L6 might idle at 45mA but the maximum plate current could be 400mA. So the push pull amp needs the higher screen voltage to perform well.

OK but those voltage swings you talk about that are described in the curves happen during the amp's operation, correct? From a design standpoint, most designs I've seen have a slightly lower SG voltage than the plate at idle... when the Champ 12 has half the plate voltage. I guess from what you're saying that this lower screen voltage at idle regulates how much current goes through the plate? I'm guessing less current would flow though the Champ 12 wouldn't it?

Also, secondary emissions would be higher (and would therefore affect the amp's performance) with this lower SG voltage, wouldn't it?

Last but not least... how can you get away with not using a SG stopper resistor if the tube has a limit for SG current?

Sorry for all the questions... just trying to understand as much as I can.

[loudthud]

The plate voltage changes I was refering to could occur for any reason but they have to be made in isolation. If you just twist the knob on your Variac, you are changing all the voltages on the tube, not just the plate voltage.

If you looked at plate and screen voltages with an oscilloscope, you would see something interesting. The plate supply in a guitar amp typically has 20V pk to pk of ripple on it at idle. The Screen voltage has comparatively very little ripple. So while your DVM reads 495V on the screen and 500V on the plate, the scope shows the plate is really going between 490V and 510V many times a second. The point is, don't worry if the screen voltage is slightly higher than the plate voltage at idle.

Secondary emissions are usually lower with lower screen voltages aren't they? Seconday emission is rarely a problem in guitar amps.

Screen supplys usually have poor voltage regulation so they sag quite a bit when an amp is overdriven. Even if an amp does have stopper resistors, the screens are operated well above their ratings. That's probably the reason power tubes don't last very long in guitar amps. Stopper resistors are sometimes left off as a cost cutting measure, but it should be considered tube abuse.