How does the Mesa Boogie SOB P.I. stage work?

[Mark]

I was wondering if anyone has had experience or is familiar with this type of phase inverter circuit. I would appreciate an explanation on how it works.

https://ampgarage.com/forum/viewtopic.p ... 2&start=15

0F2D8B84-BFC9-41FA-8857-148F084A7775.png

[bepone]

this is constant current diode, nothing special, to keep balance of PI, better than tail resistor

[sluckey]

Looks like a typical LTP PI driving a typical push/pull output stage.

[nworbetan]

https://boogieforum.com/threads/mk1-sob ... odes.2674/

I found a sketchy transistor equivalent table that says TCR5305 is comparable to CR200, which is apparently obsolete too.

DSA1206000-139.pdf

But there's at least one datasheet still floating around for the CR200.

https://html.alldatasheet.com/html-pdf/ ... CR200.html

That one looks like a 2ma constant current diode too.

Edit: here's the full SOB schematic.

sob.gif

[Mark]

If someone asked me roughly how does a long tail pair work, I would say the signal comes into the grid of one triode. As the voltage of the signal increases, current flows creating an inverted signal on the plate of the first stage due to the voltage drop across the plate resistor. I would mention that there is a signal present on the shared cathodes of the two valves. I would mention in the second triode is out of phase with the first triode as the voltage on the shared cathodes increase it has an inverse effect on the plate of the second triode as the grid is strapped to earth thus an increase in voltage on this cathode would cause the second valve to conduct less. Thus an output which is 180 degrees out of phase.

This is what I’m surmising about the SOB P.I. driver.
Sure, the signal is on the grid of the first stage and there is the constant current diode on the shared cathodes. I would assume that there must be signal present on the cathodes for the second stage to produce the out of phase output signal. The resistor off to the ground on the grid of the second triode must be enough to bootstrap the grid to ground potential allowing an out of phase signal.

I assume the Limit pot is a type of negative feedback with enough punch to eliminate the signal in the stage altogether.

What is the truth of the matter?

https://www.valvewizard.co.uk/acltp.html

[Mark]

Looks like a typical LTP PI driving a typical push/pull output stage.

What I’m questioning is the amplitude of the signal on the top of the constant current diode, the ability of the grid leak resistor to bootstrap the second and is the negative feedback enough to shut down the entire P. I. stage for this amp?

How does the circuit work?

[nworbetan]

The signal on the cathode is going to be the rail voltage minus the plate load voltage. The only "tricky" thing the constant current diode does is keep the plate load line nearly horizontal. Are you familiar with drawing plate load lines in tube circuits?

Bootstrapping in this context would be a local positive feedback from cathode to grid, where the current through a voltage divider on the cathode creates a bias voltage that's fed back to the control grid.

That's not what's happening here. The grid of the inverted side of this LTP is tied to ground through a resistor that should under normal operation have little or no current flowing through it. This is almost the exact opposite of bootstrapping. This is more like nailing a person's boot to the ground so that the only thing they can do is spin in place.

That schematic doesn't appear to have a "limit" control. It does have a variable resistor labeled "presence" which doesn't work like the vast majority of presence controls. This one is a mislabeled negative feedback control. Presence traditionally only controls the negative feedback at "high" frequencies. If you turn that variable resistor all the way down to 0 ohms as indicated as a possibility in that schematic, the circuit probably won't work very well any more, if it continues to work at all.

As an afterthought, I suppose it could be possible that the input voltage swing the LTP needs is big enough that the output voltage swing at the 8 ohm speaker out wouldn't be so large as to cause problems. In that case it would be a fairly boring negative feedback control. I'd have to run a spice simulation to come up with any numbers though, I don't do that kind of calculations in my head.

[martin manning]

The only "tricky" thing the constant current diode does is keep the plate load line nearly horizontal.

Nah, the two triodes are going to trade current same as they do in the usual LTP circuit, and the load lines will look pretty much the same.

[nworbetan]

Thanks, good catch! My brain isn't quite as good as an actual spice simulation and I only mentally simulated half of the LTP where there wasn't a second triode to share the constant current with.

[mhuss]

A constant current source (CCS) does it's best to keep the current flowing through itself constant (as the name implies ). This means it appears as a very high impedance looking "in" and therefore it will allow the voltage across itself to change almost like an open circuit.

An ideal LTP design would have the tail current held constant no matter what is happening 'upstairs', so this is more 'ideal' (fwiw) than a normal medium-sized tail resistance.

[Mark]

nworbetan thank you you taking the time to explain to me the operation of the SOB P.I. stage. I suspect the presence control was on the face plates, my one has limit on it. I have a LTP in my amp with a LarMar master volume control.

Thanks to everyone else who took time to reply to my question. It was very much appreciated.

My recollection of this circuit is it didn’t work particularly well, nor did it sound very good. I remember salespeople in the store touting it as a built in variac back in the day.

[martin manning]

If there was any magic in this more people would use it.

[FUCHSAUDIO]

@martin Manning, agreed.

Randall tried to avoid direct Fender rips by using CCS diodes on preamps and in some power amps (like this one).
Actually, an LED will do the job of a CCS nicely, but when Mike Bendinelli at Mesa tells you "we rip those dopey things out all the time", that tells you something.
While a CCS allows you to get the maximum gain a circuit can provide and may increase linearity, your not necessarily looking for either or both of those things in a guitar amp.
I've given up trying to figure out the weirdness I see in Boogies honestly...

[dorrisant]

I've given up trying to figure out the weirdness I see in Boogies honestly...

Amen!

[Reeltarded]

Best Boogie ever made for ROCK. Un-Series

Mk0

Seriously great amp if serviced to original specs.