Power output load line question

[tpaairman]

In my latest question to learn tube amp design, I'm trying to figure out doing load line calculations for a power output tube. I have a book on tube amp design, but it's not making sense when I look at real world designs, and I'm finding very little on line.

Here's the basics that I'm seeing to do - For single ended, it says to use the supply voltage and half of the impedance of the primary for my resistance value to start a load line the same as I would for a pre amp tube. Then it says to move the line up until it's just at the power dissipation curve, keeping the same gradient. Wherever the end of the line is at the supply voltage mark, that is your bias point.

When i try to verify this with real work schematics, I can't get it to work out. I've looked at a couple Fender amps that use a 6V6, and show a bias of -20 volts on one, -19 on the other. Following the steps above, I couldn't have a supply voltage over 290 volts without pushing my line way above the power dissipation line.

[cerrem]

For academic purposes, it would be best to learn the "book" method first.... Then you can add in real-world numbers later.
Most important is to make sure you have the correct I V curves.... make sure it is for the correct SCREEN voltage you are using...
Keep in mind that the voltages on the schematic are idle voltages.... When doing load-lines you are in affect at full AC voltage swing, thus the B+ will dip a bit.... Not as much for Class A single ended...
I attached a very crude simple example just for basic concept.... It is simplified and does not get into rectification effect or voltage sag of the supply...this is assuming 250V screen voltage...

[tpaairman]

After looking at your example, I realized one place I screwed up when looking at the real thing. I was looking at the load line for a triode connection. The datasheet I used was from JJ, and it only says plate characteristics. But I'm realizing it's a triode one since the grid lines go more or less vertical. There is no chart with horizontal grid lines on this datasheet for some reason.

I found another data sheet with the correct chart, however, it only shows screen voltages of 250 volts. The real deal is 350. So what do I need to do now? Do I need to adjust something in my calculation, or something in the chart?

[cerrem]

Well..it gets a bit tricky if you use the tube outside of the published curves...
I have a Tektronix 570 curve tracer...this way can generate the curves under the given conditions...
You can use the classic book equations of 3/2 power law to derive the family of curves by hand or by computer generated...not very realistic... no pun to Radio Shack..Haha..
The quick and dirty way is to use the published 0 volt Bias curves ...this is where the curves are all Grid 1 @ 0-bias volts and the screen voltage is varied for the family of curves...
The top curve is the 0-bias curve for 250V screen... The increments are in 50V ... so the next curve up would be 300V screen and that would be hand drawn in by eye as I did... The spacing also incrementally increases with increase in screen voltage...this is gm.. SO the 300V curve that I drew by hand also shows a slight increase in spacing....all by eye...keep in mind that tubes have greater variance than this...
Then use this 0-bias curve to draw your load-line... since you typically want to stay right above the knee to stay in the current saturation region..for several reasons..one is maintain the plate resistance from drooping to low...stay linear..and also keeping the screen current from getting to high...However in a guitar amp you going to play loud transients and eventually going to slam the load-line into the triode region and make square waves...its inevitable..
With this family of curves you can roughly fit a load-line and get a rough idea of a fit and power output....
WIth Push-Pull this becomes a lot easier to work with in getting more accurate results... SIngle ended requires more details that his chart doesn't provide..
Keep in mind that the 350V you mention will most likely be around 300V on the screen when the amp is pushing full power... the choke and or the screen resistor with the power supply sag may do this...
What amp are you referring to ???

[tpaairman]

I have a bunch of amp schematics that I had downloaded some time back and was digging through those. Trying to find them on line now to link to them is proving harder than I thought, but here's one (Fender Champ AA764):

[eschertron]

In my latest question to learn tube amp design, I'm trying to figure out doing load line calculations for a power output tube. I have a book on tube amp design, but it's not making sense when I look at real world designs, and I'm finding very little on line.

Here's the basics that I'm seeing to do - For single ended, it says to use the supply voltage and half of the impedance of the primary for my resistance value to start a load line the same as I would for a pre amp tube. Then it says to move the line up until it's just at the power dissipation curve, keeping the same gradient. Wherever the end of the line is at the supply voltage mark, that is your bias point.

When i try to verify this with real work schematics, I can't get it to work out. I've looked at a couple Fender amps that use a 6V6, and show a bias of -20 volts on one, -19 on the other. Following the steps above, I couldn't have a supply voltage over 290 volts without pushing my line way above the power dissipation line.

Can you provide a link? Based on your description, the method seems to be a mish-mash of ideas, not all of which are applicable to your problem.
I've seen a few threads here about this topic (sorry no direct links off the top of my head). The Valve Wizard has some good info, and the Black Magic Amps website has a load line calculator where you can see the resulting load line from a handful of input parameters.
Happy hunting!

[tpaairman]

Looking at Valve Wizard, He shows the same method, however he then goes on to account for the screen voltage, which the book I saw didn't mention, or if it did, I missed it. After looking at Valve Wizard, I'm seeing what needs to be done, but I'm still working on getting my head around this.

http://www.valvewizard.co.uk/se.html

[Dave H]

I'm trying to figure out doing load line calculations for a power output tube. I have a book on tube amp design, but it's not making sense when I look at real world designs

If the real world design is off the chart you could work backwards from the measured data. e.g. For the original (black) voltage readings on the AA764 Champ schematic I calculate a Pd of 12W and a plate current of 36mA (assuming 10% for the screen). The plate is sitting at 350V so the peak plate voltage swing it can have is 350-10-20-50 = 270V (That's subtracting 10V for the OT resistive loss, 20V for the cathode voltage and 50V to ensure the plate is above the knee). To swing 270V at 36mA requires an OT impedance of 270/36 = 7.5k.
The red voltages look far too high. The 6V6 is running at 18W.

[tpaairman]

I probably should have noted this when I posted the schematic. This is one I was using since it does list the voltages, which are the black number ones that Fender put on there. I don't know who put the ones in red on there.