Need tips on using interstage transformer - phase inverter

[martin manning]

What about using a cathodyne inverter? Balance is inherently better, and it will tend to stay that way driving the power stage through CF’s or SF’s. About the same part count as LTP, assiming a gain stage in front. This is SVT topology, of course. As mentioned, LTP distortion may well be de rigueur for guitar tones by now, or at least be more like the expectation.

[Synchu]

Or a LM317 regulator!

That sounds adjustable. Do tell.

I think it was on DIY Audio...a hi-fi design by Bruce Herran. A push-pull pair biased with an LM317.

I built Bruce EL84 monoblocks for my brother. It’s indeed using CCS, the power supply is royal PITA to build and the problem was sourcing LR8 high voltage reg.

They do sound amazing, though!
Niki

[strelok]

If you don't want to use MOSFET's in your amp a DC cathode follower using a 6SN7 should be more than enough to push the grids well into conduction for most any tube, especially if you run it off of +/- 300v rails. The CCS will do wonders for your balance, but I'm not sure how it would sound. You can do that using a tube as well, if you like. I've never used a CCS in a guitar amp before, its more of a Hifi thing, but if you want something different it's easy to try. You can run it off of ground or a negative supply. The negative supply will give you more headroom and available voltage swing. Of course you'll want to adjust the bias accordingly.

Here's an example of one using just a pair of MOSFET's:



Be aware though that the DN2540 is very sensitive to ESD, so you'll want to include protection zeners from gate to source of each device. I learned that one the hard way last year.

[Lauri]

If you don't want to use MOSFET's in your amp a DC cathode follower using a 6SN7 should be more than enough to push the grids well into conduction for most any tube, especially if you run it off of +/- 300v rails.

Cathode follower won't be able to drive the grids well once grid current starts flowing. Control grid impedance drops very low when it's driven positive.

[strelok]

On the amps we build you can supposedly get about +15Vgk from a 6sn7 using the setup I described. Past that the tube goes into full saturation. This is of course direct coupled to the output section. Now granted, the output tube in question is an extremely low gain triode. Has a mu of about 8 if I remember correctly. So that could mean that it will draw far less grid current than something like an EL34, 6L6 etc. given that its likely it has a much courser pitch to the grid winding. However our amplifiers use a lot of these tubes in parallel, so there's that. I haven't observed this myself, its just what I'm told it can do, so the next time one comes up on the bench for checkout I will verify this is actually the case.

[Lauri]

Years ago I tried 12AU7 cathode follower direct coupled to EL34 and if I remember correctly I was only able to get less than +5Vgk before it would clip hard. 6SN7 cathode follower would probably work a bit better but not much. White cathode follower is a much better solution because it can deliver twice the idle current in to a load. But for a push pull amp adding two double triodes and a complex power supply is probably not worth the effort just to get a few watts more output power.

[strelok]

I just checked its actually a mu of 2. So its really low. Yeah the white follower would be a bit more work to setup and probably not worth it. A more practical approach would be to use the 6bx7 in a normal follower configuration. It has almost 3 times the transconductance of a 6sn7 and about 1/7th the plate resistance. So its probably safe to say it could do the job. Same envelope and pinout as well. They're not made anymore but old stock is available for semi-reasonable prices.

The problem is once you get into positive grid territory is verifying if the driver is clipping or if the tube is actually saturated. I mean actual saturation, as in the cathode is incapable of emitting more electrons. +15Vgk seems to be about the limit in our case, as adding a beefier driver doesn't seem to get you anywhere. That seems to be the only way to verify it, build a progressively bigger driver section until you stop seeing improvements. Its possible you were hitting that point with the EL34 but I have no idea. Of course the datasheets don't go anywhere into that territory so the only way to tell is to run some tests.

[Roe]

The Epiphone EA-300RVT looks interesting, since it has a 12au7 powered interstage transformer driving a quad of 6l6s: http://archive.gibson.com/Files/schemat ... 20trem.jpg

Edit: the gibson GA95 Apollo is similar https://ampgarage.com/forum/download/file.php?id=13241https://ampgarage.com/forum/viewtopic.php?t=12161

[BrownIsound]

Did you ever get around to making this?

I’m planning on a Musicmaster influenced build with some differences. Was wondering about attempting global NFB, but the phase shift of the interstage transformer is a problem with that, right?

[sluckey]

Was wondering about attempting global NFB, but the phase shift of the interstage transformer is a problem with that, right?

Not at all. Very easy to flip the phase if needed.

[Roe]

Did you ever get around to making this?

I’m planning on a Musicmaster influenced build with some differences. Was wondering about attempting global NFB, but the phase shift of the interstage transformer is a problem with that, right?

No, sorry the transformer is still lying here. I ended up building an ac100 with one top boost channel (with switchable coupling cap on v1 for some of the fatter original ac100 tones as well as a switchable 50/150pf cap in the tonestack for the 70s vox 100w sounds). as a second channel, I added a hot rodded 2203 ala soldano x88. amp sounds great and didn't cost much because it was based on leftover parts mostly

[pdf64]

…
Was wondering about attempting global NFB, but the phase shift of the interstage transformer is a problem with that, right?

Yes, there’s a potential that a feedback loop around both an interstage and output transformer might shift from being negative at lower frequencies towards being positive at higher end of the frequency range.

[BrownIsound]

Yes, there’s a potential that a feedback loop around both an interstage and output transformer might shift from being negative at lower frequencies towards being positive at higher end of the frequency range.

So looking at the schematic for the MM, a place to inject the feedback could be at the V1B cathode, perhaps a switchable circuit like Robrobs 5E3, where you switch off the bypass cap when adding feedback.

MM schematic:
https://cdn.shopify.com/s/files/1/2284/ ... 1553745401

5E3 switchable feedback:
https://robrobinette.com/images/Guitar/ ... ematic.jpg

Except if there is positive feedback in high frequencies from the circuit, then a cap of a yet to be determined value could be switched in with the feedback loop to shunt the high frequencies?

[BrownIsound]

…or, possibly also taking some high frequency negative feedback from the IT output/6v6 input via a cap and resistor ala mid 70s silverface twin. Looks like I’ll need to get familiar with my scope again if I’m going to play around with this, haven’t used it in a decade.