Peavey Classic VTX

[lowell]

Schematic: http://music-electronics-forum.com/a...lassic-vtx.pdf

This amp has a blown speaker...if not 2 blown speakers. Or voice coil rub to be more specific. I'm not sure if the second speaker is blown or not because it could potentially be a cold biased power amp. I tried a different speaker cab and the amp sounds noticeably better but that speaker sounds darker as well so I'm not sure if it might be just rolling off the "junk." So I checked the bias of this unique cathode -driven hybrid power amp. My readings are in spec with the notes on the schematic:

Vk 80v
Ve (cathode transistor emitters) 17mv

So thats a hair low but not much.

It appears these tubes are only running 3ma of quiescent current. Seems crazy to me....yet it's within spec on the schematic. Thats 1.3watts per 6L6 .

17mv/5.6=3ma
Vak of 6L6 is 450v
3ma*450 = 1.3watts

I even checked the bias using the transformer shunt method just to confirm I wasn't measuring incorrectly.

I just want to be certain this is correct and maybe someone can explain how this works correctly and why it doesn't sound hideous. (or maybe thats why it sounds like crap and not the speaker)

[Tom Phillips]

Lowell,
The amp is designed to run in (or nearly in) class B like the Music Man amps. Therefore, the low idle dissipation is normal. Google Peavey Classic VTX +"Class B" for several past discussions.
Cheers,
Tom

[nashvillebill]

That amp has an excellent clean sound when it is operating normally. It definitely should not sound hideous.

[lowell]

Thanks Tom I'll check it out. I googled and googled...but wasn't using the class b search terms.

[lowell]

Why does the schematic say +540v on the cathode transistor base?? This isn't the case nor should it be.

[nashvillebill]

The decimal point didn't come through. Try 0.560

[lowell]

Theory just for fun/education. And this is if we ignore the Q4/5 limiters.

So lets say Q6/7 Vc is ~85v. Since grid conduction of the 6L6 would start around Vk = +15v, we can subtract 15v from 85v = 70v before grid current flows. So our Peak AC signal applied to the cathodes is 70v peak = 140v peak-peak.

Opamp drivers U8 max output is 15v peak = 30v peak-peak.

We can then assume a voltage gain of 4.7 (70/15) from Q6/7.

Thats what I can gather thus far theoretically from this circuit. The plus side to Peavey designing this way? Seems to me that a 70v peak input before grid conduction is a good 20v or so more than your typical fixed biased common cathode stage. And a good 30-40v more than your average cathode biased output stage. I also read that the cathode has a very low input impedance so it's best to, as in this design, drive a cathode from a low impedance source. Although I'm not sure just how low Zout of a transistor collector is... But I know it's not as low as Zout of the emitter.

I'm quite unfamiliar with Class B operation since it's so rare in guitar amps... Same goes for common grid operation. Gonna do some reading and report back on findings. Obvious question is - how can Class B not cause crossover distortion?

[lowell]

Oh right...thx

[J M Fahey]

Theory just for fun/education.

Ok, I'll play ball ... just let me clear beforehand that what you wrote is correct , if you consider these tubes voltage driven, Class AB1 biased.
Since in this particular case they are neither, understandably it's confusing because it works in a different way.
Let's split this idea by idea:

So lets say Q6/7 Vc is ~85v.

At idle?
Yes.
Which in this case means passing 3mA per tube (what you measured).

Since grid conduction of the 6L6 would start around Vk = +15v,

Correct, because then Vgk would be 0V and, really, what matters is Vgk .

we can subtract 15v from 85v = 70v before grid current flows.

mmmmhhhh:
at idle we have Vk=+85V , Vg=+15V , so we see that to pass 3mA (with proper plate and screen voltage applied, of course), the tube needs negative bias of -70Vgk
Didn't check the datasheet but it must agree, of course.

Even if the circuit is unusual, Physics Laws still apply,and that includes the datasheet .

So our Peak AC signal applied to the cathodes is 70v peak

If you refer to the Vgk peak swing, it's even larger:
* Vg stays fixed at +15V , always
* Vk can go from +85V (idle) to almost 0V , when driver transistor is saturated (in practice 1 or 2V above ground).
In these amps tube drive is *massive* and to boot, forces them to pass current, like it or not, that's why it extracts down to the last ounce of power and tubes work until at least a few electrons can be emitted from the cathode.

= 140v peak-peak.

No peak to peak available since we are not dealing with voltage drive but current drive, which goes from almost 0 (3 mA in fact) to whatever a tube can supply with +15V on grid, and 350V on screens, easily 300 to 400mA instead of the meager 200/250mA available under conventional drive.
Talk about squeezing a dry lemon and filling a glassful of juice

And current always works in the same quadrant, it's always positive, never negative, we have a peak value but not peak to peak (which would imply swinging negative).

Opamp drivers U8 max output is 15v peak = 30v peak-peak.

It actually swings a few hundredths of a Volt, whatever's necessary to drive Driver Transistor's base.

We can then assume a voltage gain of 4.7 (70/15) from Q6/7.

Not the main parameter here, but if for any reason you want to calculate a voltage gain, it's Vk variation / Vb variatiob (at the driver transistor), some 85V/300mV (jut to give it a value) , some 200/300X (just to estimate order of magnitude) .

But the real parameter is that the transistor, sitting at some 3mA idle, is driven to some 300/400mA peak; the slave tube just follows that.

Thats what I can gather thus far theoretically from this circuit. The plus side to Peavey designing this way? Seems to me that a 70v peak input before grid conduction is a good 20v or so more than your typical fixed biased common cathode stage. And a good 30-40v more than your average cathode biased output stage.

The designer was Music Man, and the goal was to get a LOUD amp, which tightly controls tubes, drives them down to the last drop, etc.

If no high power transistors had been invented, this would be the way to go.

Of course they have, so this expensive way to get watts was replace by even louder and definitely much cheaper full SS amps.

It holds its own as an intermediate type though, because speakers are driven from plates, and through an output transformer.

But nowadays people in general goes straight to an SS amp in most cases, and the few who buy tubes go for classic ones.

I'm quite unfamiliar with Class B operation since it's so rare in guitar amps... Same goes for common grid operation. Gonna do some reading and report back on findings.

Cool.
The main point here is not Class B operation, which just is a different bias setting, but common grid/cathode drive.

In a nutshell:

* the tube is FULLY turned on by applying incredible POSITIVE 15V to grid.
Yes, it will pass grid current.
No big deal because it does not come from a current starved PI plate but from a regulated 15V power supply, go figure.

* the tube does not pass all that current all the time, because the real "faucet" is the transistor it has in series with its cathode.

* grounded grid amps have unity current gain, meaning plates will source the exact same current (minus screen current which is relatively tiny) they get at the cathode.

As a side bonus, tube non linearity is irrelevant, the transistor linearizes it.
I bet you can put a 6L6 on one side, an EL34/6550/KT88 on the other and the amp will still work fine, because Vk in that case will be different, whatever necessary to pass 3mA or whatever the transistor sends.
In a way, I consider this an early way of Servo Bias .... think about it.

[lowell]

"If you refer to the Vgk peak swing, it's even larger:
* Vg stays fixed at +15V , always
* Vk can go from +85V (idle) to almost 0V , when driver transistor is saturated (in practice 1 or 2V above ground).
In these amps tube drive is *massive* and to boot, forces them to pass current, like it or not, that's why it extracts down to the last ounce of power and tubes work until at least a few electrons can be emitted from the cathode."

Ahhh so that's the major benefit here. Pushing the tube past grid conduction without worrying about impedance. The question begs - is this any different than using a mosfet cathode follower to drive the grids? Like RG's Mosfet Follies page on Geofex? Is one better than the other? http://www.geofex.com/Article_Folder...osfetfolly.htm

[lowell]

Also, slightly off topic concerning the mosfet source follower. RG, why the need for -Vs? If the source "follows" the gate then the source should just be 5v or so more positive than the gate because of Vgs ON. And since the negative swinging portion of the wave is of no concern in a Class AB power amp why couldn't Rs just be grounded?

[J M Fahey]

"If you refer to the Vgk peak swing, it's even larger:
* Vg stays fixed at +15V , always
* Vk can go from +85V (idle) to almost 0V , when driver transistor is saturated (in practice 1 or 2V above ground).
In these amps tube drive is *massive* and to boot, forces them to pass current, like it or not, that's why it extracts down to the last ounce of power and tubes work until at least a few electrons can be emitted from the cathode."

Ahhh so that's the major benefit here. Pushing the tube past grid conduction without worrying about impedance. The question begs - is this any different than using a mosfet cathode follower to drive the grids? Like RG's Mosfet Follies page on Geofex? Is one better than the other? http://www.geofex.com/Article_Folder...osfetfolly.htm

As Enzo often says, "there's more than one way to skin a cat" , but the ultimate tube squeezer is Music Man, because common base / cathode current drive is absolutely linear (current gain=1 always so output can not deviate from input, by definition) while all other methods get output current controlled by a certain grid voltage, and datasheets show this is not very linear, specially near tube cutoff, so it must be biased AB with a relatively important idle current.
In MM type amps, the *transistors* need to be biased into the linear area,and for that meager 3mA is enough, while the tube would require, what, 30 mA ?
10X larger at least.

On the contrary, the tube loses some of its tubeyness and sits halfway between typical tube and typical SS .

Not bad , but now you see why the Loud and Clean guys (or Bass players) like them, while no Bluesman will be seen close to one.

Or to be more precise, some like Albert King might ... but he's used to SS amps anyway.

BB King's Lab Series amp is actually tubyer than this, go figure, thanks to some clever circuitry.

[lowell]

I have a Lab Series head and just ADORE it. Great amp, very cool compression circuitry as well. Well, for me, every type of amp has its place. Solid state or MM amps are great for attack, punchiness, presence when playing funk rhythm guitar. Wouldn't be surprised if Nile Rodgers funk guitar on Daft Punk hit Get Lucky was recorded with a solid state amp... Or better yet, direct guitar!!! All these tones have their place IMO. When I'm playing roots music I love a legit tube overdrive sound with swirl, sag, compression, all the inherent tube exploits. Anyways... I get off topic.

As you said, this was a design of it's time in the development of solid state devices. Designing a tube amp like this from scratch is pointless these days, unless it's simply to please some ignorant Tube Fanatic who refuses to have an evil SS power amp in his amp, yet knows nothing about audio electronics. I'm not hatin, but we've all fixed/built gear for these folks. It'd make most sense to grab some tubiness from a preamp, then use a SS power amp for wattage if one is going for this loud/clean sound.