Hi-Gain LTP

[Ray Ivers]

I finally got these things scanned - they're from the November 2003 issue of AudioXPress magazine:

[ATTACH]120[/ATTACH]
[ATTACH]121[/ATTACH]

Either of these circuits has enough gain to allow a -10dBm FX-loop return to drive a typical power amp into hard clipping, which was the main reason they appealed to me (I haven't tried either one). Both circuits use 12AX7's, and all feedback - both positive and negative - is local; the 'FEEDBACK FROM VOICE COIL" input in Figure 2 is additional NFB for a greater margin of stability.

Ray

[Satamax]

Ray, just a daft question, why do you call them LTP? Aren't they cascode phase inverters?

[Ray Ivers]

Max,

You're right - they're not LTP's. Technically, they're called 'mu-stage split-load phase inverters'; I thought that 'LTP' sounded a little more familiar, and might reduce the 'WTF? factor', if you know what I mean.

Ray

[Matt T.]

So you could eliminate any dedicated 'Return' stage and just go straight into the PI and still get full power? Is that correct?

What is that line below R1 in the first image? It almost looks like another input but that connection is grounded.

Will this PI will stay much cleaner than a LTP? Like for use in a high gain channel switcher with FX loop where you're not (or don't want to) distorting the PI?

[Ray Ivers]

Matt,

Yes, you're correct - the FX return stage could be eliminated in the vast majority of cases. An 800mV peak-peak signal (this is roughly -10dBm if my calculations are correct - a little hot for pedals but still OK for most IME, and a little low to pull the last dB of S/N from rack stuff but no big deal IMO) will result in a 160V peak-peak swing after passing thru a 200X gain stage - enough to clip even a hi-B+ UL output stage.

The line you referred to is the input ground connection. I thought it was one of those floating deals - requiring two input caps - but then I followed it over to ground like you did and imagine they just put it there for the sake of completeness.

I don't know just how linear this circuit is from a hi-fi standpoint ("reasonably low distortion" is claimed in the text), although for guitar-amp usage it should sound as clean as any other, and should work just fine in the MV design you mentioned. Higher gain is always achieved at a price; you can 'haggle' a bit with meticulous design, but some payment will always be made. S/N ratio, bandwidth, distortion, etc. will be impacted to some extent, and usually in some combination. The note in Figure 1 points out the bandwidth reduction in that circuit when really high gains are used (the article mentions "considerable attenuation of frequencies in the upper end of the audio range", which sounds like above-10kHz to me). BTW, a figure of 1.8% distortion is given for the average 12AX7 LTP PI.

Ray

[martin.k]

I sorry, but I can't really see how this circuit could have that much gain.
Is it possible to get some more explanation?
Thank you,
Martin Kuhn
Sweden

[Ray Ivers]

Martin,

I can't really see how this circuit could have that much gain. Is it possible to get some more explanation?

Sure! The first circuit is very unusual in its operation, the second (Admiral) circuit a bit less so.

In the first circuit, the input signal is applied to pin 2 of V1a, and V1a's output is applied to the grid of V1b through C2. V1a's plate resistor R3 is connected to the bootstrap node at pin 8 of V1b (this node is at a fairly high DC voltage), allowing negative feedback to control overall gain and keep the circuit stable. In addition, positive feedback is applied to V1a's cathode via R5, in a manner similar to Marshall's Silver Jubilee amps. The unbypassed R2's negative feedback counterbalances some of the positive feedback through R5, again for stability and gain limiting. Simple, no?

The Admiral circuit is quite similar to the circuit described in TubeCAD as a "self-biased totem pole", with the addition of positive feedback (via R5), a second polarity-inverted output (via C4), negative feedback from the speaker output added for additional stability and bandwidth enhancement, and additional components for phase-inverter usage.

Ray

[martin.k]

Thank you for your explanation, I guess I have to get a copy of the magazine to get the full article, though, this is interesting...

[Satamax]

Hi everybody!

Ray, do you think i could atack directly a stage like this with guitar, and it would then split the phase and send the signal to a pair of 6V6? I was thinking of using crossline tone controls like some matchless. Well, i mean, would it have enough gain to have power amp distortion.

[Ray Ivers]

Max,

With a hot pickup, you might be able to get the same amount of overdrive as in the intros to AC/DC's "You Shook Me All Night Long" or "Hell's Bells" when driving 6V6's with this PI, but I don't think you'll get much sustain. You could add a JFET or LND150 500V MOSFET between the guitar input and the PI (or use a Menatone Red Snapper), of course.

Ray

[Satamax]

So a preamp 12AX7, one "nuter phase spliter" and two 6V6 I'll see what i can do. I want something in the type of size of the orange tiny terror. I have an old batery charger chassis, 30 wide 20 deep 10cm high but the tranies i have are huge! I'll see what i can do Can i just paste the second one straight into my circuit?

[Satamax]

Hi Ray.

Well, i'd like your opinion. Here's what i got to work with. Big tranies, and small chassis. What do you think? Shall i stick everything inside? I've put an el34 for size comparison, but i'll use 6V6. I realy wanted a box, with nothing outside, exept the handle, but it gonna be hard. I'll certainly work real point to point. Even on the verge of flying leads. Or should i stick tranies and tubes outside?



Thanks a lot.

Max.

[Ray Ivers]

Max,

I think that chassis is a bit too small, unless you want to do a lot of work. OTOH, a preamp w/toroid PT would probably be easy to fit inside.

Ray

[Satamax]

Ray, something completely nuts came across my head! Would it be possible to do this on the power tubes?

[Ray Ivers]

Max,

Pretty much anything is possible with this circuit, if you're willing to go to enough trouble to make it happen. If you're going to use two pentodes, though, IMO you don't really need this circuit.

Ray