What's happening in this plate circuit?

[Groover]

Yes, but I still don't know how to calculate the load line for the stage driving the primary. What figure do I use for the load on the anode?

And why is it the "Po' man's" phase inverter? I read somewhere (Jack Darr's book? Gerald Weber's book?) that transformers are more expensive than the components of a tube PI?

[Wilder Amplification]

Yes, but I still don't know how to calculate the load line for the stage driving the primary. What figure do I use for the load on the anode?

You use the value of the resistor that's in parallel with with the x-former primary. The primary load is so high that most of your signal current would pass through the resistor, while the x-former primary would see the signal voltage across the resistor.

Now to answer your question about how inductors react to AC...DC is a constant current. Therefore, the magnetic field surrounding an inductor with DC applied to it is also constant. The two don't see each other (the current and its resulting mag field).

However, with AC, you have an alternating current flowing, which also generates an alternating mag field around the inductor. When the AC rises positive, the field density builds. Once the voltage starts falling back toward zero, the field density collapses around the primary and secondary. It does the exact same thing on the negative swing, but with an opposite magnetic polarity.

This building/collapsing action induces a counter-voltage into the secondary. The turns ratio between primary and secondary determines how high or low this voltage will be.

In the case of a choke, the DC ripple generates the building/collapsing mag field, which induces the counter voltage into itself, which serves to remove the DC ripple.

[Groover]

So if I have 253V at the B+ node and divide it by the 15k which is the value of the resistor in the anode circuit I get a max current (when Va = Vk) of 16.8mA and change.

It looks like the load line for the 12ax7 will be very steep, > 60 degrees. Is this right?

[Wilder Amplification]

So if I have 253V at the B+ node and divide it by the 15k which is the value of the resistor in the anode circuit I get a max current (when Va = Vk) of 16.8mA and change.

It looks like the load line for the 12ax7 will be very steep, > 60 degrees. Is this right?

Actually, you add the value of the cathode resistor to the value of the plate resistor, then divide your node voltage by that value to get your max plate current (current draw of the entire branch is limited by both the cathode and plate resistors since they're both effectively in series with the tube). Then you'd multiply the plate current by the 15K to figure out what your max peak voltage swing would be. Works very similar to a voltage divider.

[Groover]

If one needs to consider the cathode resistor to draw the load line and one needs to consider the load line to choose the bias point, and therefore the cathode resistor, one is in a circular trap.

In Merlin's book the Rk is not considered when drawing the load line, according to my reading of Ch 1.

Ultimately I am trying to set the optimum bias point for center-biasing this valve.

Also perhaps I would like to increase gain. Can I increase the R which is in parallel with the IT primary?

[Wilder Amplification]

If one needs to consider the cathode resistor to draw the load line and needs to consider the load line to choose the bias point, and therefore the cathode resistor, one is in a circular reasoning trap.

In Merlin's book the Rk is not considered when drawing the load line.

Ultimately I am trying to set the optimum bias point for center-biasing this valve.

Rk is not considered when Rk is much lower value than the plate resistor. But I can't imagine a stage with a 15K plate resistor using a lower value...which means that it needs to be considered. What is the value of your cathode resistor?

Also perhaps I would like to increase gain. Can I increase the R which is in parallel with the IT primary?

Yes...that is where you SHOULD increase gain. But you'll have to drop the value of the cathode resistor to maintain the same bias point.

[Groover]

FYI what I have going is the front end (with some tone stack tweaks) and vibrato of a Vibro-Champ with the power supply, PI and output section of a Musicmaster Bass amp.

Currently my cathode resistor in this stage, V1b, is 1k5. The original Musicmaster Bass amp, from which I appropriated this circuit from V1b plate to the output, including OT and PT, used a 470 ohm cathode resistor for V1b. I did not change the cathode resistor at the time I did this conversion because I was less informed than I am now about all this stuff. I kept the 15k Ra||primary because I didn't know how to analyze the plate circuit and the relationship between the Ra and the primary.

It is also worth noting that the B+ voltages are 100~150 volts lower than the original VC power supply provided.

The amp is working and sounds pretty good, but is lacking in gain/volume compared to the MMB amp when it had a 5f2a circuit from the input up to the PI. Incidentally when I had that circuit working it also had a 1k5 Rk in the V1b circuit. I attributed the comparative lack of gain/volume to the increased loss from the TMB tone stack compared to the tweed one-knob tone control.

I am also considering yanking the PI trans, and using the PI and vibrato circuits from a BF Princeton (non-reverb), but that might necessitate a new eyelet board. And I hear that that amp circuit is low in gain itself.

I have several threads going on this conversion. Especially amusing is this one http://music-electronics-forum.com/t14218/ in which catnine keeps trying to talk me out of doing it at all!

[Wilder Amplification]

FYI what I have going is the front end (with some tone stack tweaks) and vibrato of a Vibro-Champ with the power supply, PI and output section of a Musicmaster Bass amp.

Currently my cathode resistor in this stage, V1b, is 1k5. The original Musicmaster Bass amp, from which I appropriated this circuit from V1b plate to the output, including OT and PT, used a 470 ohm cathode resistor for V1b. I did not change the cathode resistor at the time I did this conversion because I was less informed than I am now about all this stuff. I kept the 15k Ra||primary because I didn't know how to analyze the plate circuit and the relationship between the Ra and the primary.

It is also worth noting that the B+ voltages are 100~150 volts lower than the original VC power supply provided.

The amp is working and sounds pretty good, but is lacking in gain/volume compared to the MMB amp when it had a 5f2a circuit from the input up to the PI. Incidentally when I had that circuit working it also had a 1k5 Rk in the V1b circuit. I attributed the comparative lack of gain/volume to the increased loss from the TMB tone stack compared to the tweed one-knob tone control.

I am also considering yanking the PI trans, and using the PI and vibrato circuits from a BF Princeton (non-reverb), but that might necessitate a new eyelet board. And I hear that that amp circuit is low in gain itself.

I have several threads going on this conversion. Especially amusing is this one http://music-electronics-forum.com/t14218/ in which catnine keeps trying to talk me out of doing it at all!

It appears our load line here is in fact quite steep.

I pulled up the scheme for the Music Master and used their voltages on the schematic to work it backwards and draw out the load line of that stage.

1.2 volts at the cathode on a 470 ohm resistor translates to a bias current of -

1.2V / 470R = 2.5mA

Now...since the grid is referenced to ground this places it at 0 volts. The 1.2 volt voltage drop at the cathode resistor places the cathode 1.2V more positive than the grid, which makes the grid -1.2V relative to the cathode. This puts our bias dot at the -1.2 volt grid line.

Then, I drew a line from that dot to the 250V node voltage point, and drew the line extending through the dot as you can see.

The curved line in the upper right corner is your maximum plate dissipation line.

I also drew a horizontal line showing the 2.5mA bias current. According to the load line, this should place our plate voltage at 230V although the schematic shows 246 at the plate...I suspect the low DCR of the x-former primary has something to do with this.

So this is what your load line should look like...not a very high gain phase inverter stage (showing a gain of about 25) -



From the load line shown, it appears that your 15K/x-former combined equivalent load is about 8.5K based on the fact that voltage can swing up 30 volts from the shown bias point, which will cause a 3.5mA increase in plate current -

30V / 3.5mA = 8,571 ohms

This would mean that the reflected primary Z of that transformer is about 20K (20K in parallel with 15K = 8.5K).

Now if you look at how the load line hits the 0 volt grid line at 210V and hits cutoff at 250, with the bias point at 230V this allows for a 40Vpk-pk swing centered on the dot, which means that the stage is perfectly at center bias as shown in the schematic. If you were to bias it colder so that it's center biased as per the grid curves, you'd have a grossly asymmetrical output signal. It's a trade off that exists due to the non-linearity of the grid curves ( they bunch up at colder bias settings).

[hasserl]

It's been a while, but I was experimenting with an MMB and I did play with increasing the value of the 15k plate load resistor and found this will increase the gain of that stage and will result in a higher AC level on the grids of the power tubes driving them harder. I don't recall what values I used, but I did not go extreme, probably around 25k.

BTW, that is a great sounding amp. It's currently sitting next to my bench, plugged in and ready to go. When I need a break from whatever I'm working on, or a good song comes on the radio and I want to play along, I plug in a guitar (if ones not already plugged in) and wail away until I've wasted enough time, then I set it back down and go back to work. I have a 2nd MMB over at my bands practice pad and that is what I use to practice with the band, great sounding amp. Would love one with about 35 - 40 watts, maybe a 15" alnico speaker, for gigs and jams. It's on my "to do list", one of htese days I'll build one.

[Groover]

So Wilder, or anyone else, what would you suggest to better optimize this stage? I would prefer to do it using theory and calculation rather than trial and error, simply because that would help my learning process right now, but perhaps t&e would ultimately be easier, seeing how the PIT primary makes the load calculations more complex.

Wherefore art thou, Merlin?

[Wilder Amplification]

So Wilder, or anyone else, what would you suggest to better optimize this stage? I would prefer to do it using theory and calculation rather than trial and error, simply because that would help my learning process right now, but perhaps t&e would ultimately be easier, seeing how the PIT primary makes the load calculations more complex.

Wherefore art thou, Merlin?

A lot of times trial and error is best to get the "best tone", then reverse engineer it to figure out what's going on that makes it sound so good.

[Groover]

Time to buy (or build) some substitution boxes I guess!

[hasserl]

Time to buy (or build) some substitution boxes I guess!

Just try a few different values in place of it, it's a pretty simple task. If you've got a signal generator and a scope hook em up and watch what happens. The signal strength will increase as resistance goes up.

I see this plate load resistor functioning the same as any other plate load resistor. In order to get a voltage gain out of a tube you have to provide a resistance. Voltage is a function of current x resistance, if you increase the resistance while current remains the same you'll get an increase in voltage. The tranny is parallel to it to perform the job of splitting the signal for the push-pull power tubes. I don't know what the ratio is between the two sides, but that would determine the step up value. It's a simple system but it works quite well. The beauty of it is the direct coupling with no RC network, which protects against blocking distortion and the common Fender Flatulence problems. If you raise the AC voltage on the primary side (which you will do if you increase the resistance) then you will also increase the voltage on the secondary side, and that will drive the power tubes farther into the non-linear range. But it already should be driving the tubes pretty good, so you don't need to got apeshit here, try something around 25k and see. Or, hook up a 100k pot as a variable resistor and play with it, when you get it set so it sounds good, measure what you've got and replace with the closest available resistor.

[Steve Conner]

The resistor is for damping resonances in the transformer, like I said in that other thread. If you want to "optimize" this circuit, get a better interstage transformer! It won't resonate so badly, so you can use a larger resistor and get a shallower load line. Of course that's no guarantee that it'll sound better. And if it's already driving the power tubes fully, does it need optimizing? You could get lower distortion, but do you want that for a guitar amp? That "bad" load line is giving you lots of fat-sounding even harmonics.

I have nothing against bodging a MM and a Vibro Champ together, for the record. I've done plenty worse: I'm currently busy bodging a hybrid output stage into a Selmer Corvette.

Just let us hear what it sounds like when you're done!

A lot of times trial and error is best to get the "best tone", then reverse engineer it to figure out what's going on that makes it sound so good.

This pretty much sums up the history of guitar amplification, I'd say.

[Groover]

Well for the record I don't know if it's driving the power tubes fully or not. The amp as a whole is pretty clean up to volume settings of at least 7.

I would probably rewire it with a tube PI before I would get a better PI transformer, but then I'm sure a transformer PI imparts different things to the sound than a tube PI would.

Just trying to come up with tools that serve my needs and learn something as I go.

Thanks to all who are along for the ride!