Class AB2: $5, .5 hr

[Shea]

Great idea - "Zener scaling"! Maybe you could even get away with just using a MOSFET "gain stage" set to a gain of 2, with the adjustable bias voltage fed to both the MOSFET gate and the grid resistors, the source thru an Rs to ground, the drain thru an Rd to the existing raw bias voltage (if it's not high enough, circuits tweaks or a doubler could be used), the threshold diode to the drain, and AC bypass caps? Maybe I'm missing something...

Ray

That sounds simpler, but, I'm embarrassed to say, I'm not sure I know how to create a MOSFET gain stage with a gain of 2. Do you just make Rd twice the value of Rs?

Shea

[Ray Ivers]

Shea,

Yes, I think it actually is just that simple with most MOSFET's, unless super-low resistor values are used. An Rs of 3.3K and an Rd of 6.8K should pull maybe 6mA or so from a -70V raw bias supply (hopefully this current is available, IME bias supplies are notoriously non-beefy) and should stay pretty solid during electron dumps when the diode conducts.

Ray

[Slobrain]

Ray,
That mod looks great, Man, you have sparked my interest in amp mods just reading this post.
Thanks...
SLO....

[Ray Ivers]

SB,

Thanks - I'm psyched about it too! I started out thinking it was only a non-master-amp kinda thing, but not any more - it should make just about any amp sound better when pushed, and at lower levels it just sleeps.

Ray

[Shane]

cool mod! what about cathode bias?

Nice preliminary review Ray! sounds like a good Mod well worth further research.
What kind of changes, if any would be needed for cathode biased amps or self splits?

looking at Paul Rubys original article he seemed to set the zener voltage at a bit more than the cathode bias voltage.
Cheers
Shane

[Ray Ivers]

Shane,

Thanks! I just re-read Paul's article, to try and get a little better handle on the cathode-bias thing.

If the cathode resistor is unbypassed, I would think the most precise way to figure out a ballpark Zener value would be to use a use a dual-trace 'scope, with one probe at the cathode, and one probe at the grid. Turn up the input signal until grid conduction occurs, then subtract the Vk idle voltage from the Vg-Vk peak voltage (possibly using the 'scope's A-B feature if available); the resulting value times 2 would be your Zener voltage, at least to start. I realize nobody's going to go to all this trouble - it's just for example's sake.

With a large-value Ck, I would say double the idle Vk would be a good starting point, and with a small-value Ck I guess you'd have to try different Zener values out to see what sounded best to you.

A typical self-split stage should be pretty similar to a standard cathode-bias situation as far as grid drive goes, I'd think.

Ray

[BobW]

Ray,

thanks for the feedback, I'm tdy w/work and was unable to try out the zener mods on the preamp section. I bought a few zener values to try out, so I'll report back once I get back to try it out.
Also was wondering if anyone has had success using a digital cam to take a snapshot of an analog display Oscope? Using your suggestion to measure grid conduction and then making a pdf would help understand, to me at least, at what point should the zenering take effect vs. tone? Wish I was back in my shop!

[BobW]

You'd have to make sure you don't use an N channel MOSFET with the built in freewheeling diode, (anode to source, cathode to drain) otherwise it defeats the purpose of the ckt.

[Ray Ivers]

BobW,

Here's a link to Paul Ruby's page describing his experiences with the original circuit:

www.paulamps.com/18watterbuzz.html

About halfway down the page there's a section called 'Cross-Over Distortion Root Cause', and below the title there's two 'scope photos showing the downward average bias shift, with the waveform tops clipped around 0Vg and the bottoms free to descend as far as they can go, with no Zeners to limit them. You're looking to get this same waveform on top, but with a matching clipped bottom - i.e., a symmetrically-clipped waveform with its center at the idle bias voltage. This is for starters - after you've played with the circuit a while you can obviously go up and down with the Zener voltage, but I would recommend starting with the double-bias-voltage Zener value as sort of a 'home base' you can always return to and have it sound acceptable.

I really haven't taken any 'scope photos with my digital camera, although I took some a long time ago with my Nikon that came out OK.

As a side note - most discussions I've seen on the Web regarding 'scopes seem to mainly focus on bandwidth (IMO the least important feature of an audio oscilloscope since about 1955 ), and tend to ignore the many cool features some of today's digital 'scopes have - one of which is taking waveform snapshots and outputting them on disk, USB, etc. And IMO it's hard to beat that rock-solid 60Hz display!

Ray

[BobW]

Ray,
Thanks for the Paul Ruby article, it looks like he's already captured the particular waveforms so no sense in me 're-inventing' them. Although his waveforms are not crystal clear, his discussion makes sense, can't wait to try it. cheers

[Arthur B.]

Sounds cool, Ray, thanks for the report.

I've been thinking that if you could generate a negative voltage rail that can supply at least double the expected bias voltage, then you could set up a bias circuit consisting of a pot, a 1:1 voltage divider, and maybe a couple of source followers, so that one bias pot would set both the actual bias voltage and another voltage of twice that amount. Then instead of using a zener, you'd just need one silicon diode per side to chop off the signal when it goes more negative than the higher (i.e., most negative) voltage. That way you wouldn't need to change zeners when you change tubes and rebias.

Shea

If you had a decent bias supply (a.k.a separate from the main PS) you wouldn't need the diodes in the first place.

Shea,

Just when I thought everyone had lost interest in this thing... It was great to get back to the bench again, even if only for a few days.

Great idea - "Zener scaling"! Maybe you could even get away with just using a MOSFET "gain stage" set to a gain of 2, with the adjustable bias voltage fed to both the MOSFET gate and the grid resistors, the source thru an Rs to ground, the drain thru an Rd to the existing raw bias voltage (if it's not high enough, circuits tweaks or a doubler could be used), the threshold diode to the drain, and AC bypass caps? Maybe I'm missing something...

Ray

What you're describing is called a rubber diode.

[Ray Ivers]

Arthur,

If you had a decent bias supply (a.k.a separate from the main PS) you wouldn't need the diodes in the first place.

What you're describing is called a rubber diode.

Can you elaborate on these statements?

Ray

[Arthur B.]

Arthur,



Can you elaborate on these statements?

Ray

Oops, I think some explaination is in order.

If you had a decent bias supply (a.k.a separate from the main PS) you wouldn't need the diodes in the first place.

From the first post of this thread:

According to the article, the component values can be adjusted to allow positive grid shift during heavy drive conditions. This would allow the output stage to be idle-biased a bit cooler for a crisp and dynamic clean sound with the guitar turned down, with an automatic shift to a hotter bias point kicking in during overdrive for a more compressed/"woodier" dirty sound if desired; might sound pretty good!

If you had a robust enough bias supply, it could deliver the current that the grids need when they conduct. It will also cause the positive bias shift because the cathode capacitor will charge on the positive half of the waveform.

What you're describing is called a rubber diode.

Here is a Wikipedia article that details the operation of a rubber diode. Look for the heading "VBE multiplier voltage source"

[Ray Ivers]

Arthur,

If you had a robust enough bias supply, it could deliver the current that the grids need when they conduct. It will also cause the positive bias shift because the cathode capacitor will charge on the positive half of the waveform.

When the control grids conduct, the electron flow is out of the grids into the 'downstream' cap plate; it is then the job of the device driving the 'upstream' plate to pull that plate positive, pulling that current thru the bias supply and into the positive driver power supply. Merely beefing up the bias supply isn't sufficient; it must be accompanied by an increase in driver-stage power capabilities, or the current will never flow in the first place, and electrons will merely accumulate on the downstream cap plate and cause bias shift. It is important to have a low-impedance bias supply in an AB2 driver to maintain a constant bias voltage in the face of drastic changes in grid impedance during grid conduction, but not to sink grid current per se.

Here is a Wikipedia article that details the operation of a rubber diode. Look for the heading "VBE multiplier voltage source"

Thanks for that link - I've seen this circuit before, in SS power amp drive stages. I've never heard it called a 'rubber diode' or 'rubber Zener', though.

Ray

[Arthur B.]

Arthur,



When the control grids conduct, the electron flow is out of the grids into the 'downstream' cap plate; it is then the job of the device driving the 'upstream' plate to pull that plate positive, pulling that current thru the bias supply and into the positive driver power supply. Merely beefing up the bias supply isn't sufficient; it must be accompanied by an increase in driver-stage power capabilities, or the current will never flow in the first place, and electrons will merely accumulate on the downstream cap plate and cause bias shift. It is important to have a low-impedance bias supply in an AB2 driver to maintain a constant bias voltage in the face of drastic changes in grid impedance during grid conduction, but not to sink grid current per se.

Ok, so a stiff bias supply is important, but not the determining factor. Would an isodyne work as a good driver in this situation?

Thanks for that link - I've seen this circuit before, in SS power amp drive stages. I've never heard it called a 'rubber diode' or 'rubber Zener', though.

Ray

Yeah, that's what it's called.