Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

General discussion area for tube amps.

Moderators: pompeiisneaks, Colossal

beasleybodyshop
Posts: 1069
Joined: Tue Jul 02, 2013 12:51 am
Location: East Texas (Yee Yee!)

Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by beasleybodyshop »

So, on this Thanksgiving day in America I am stuffed to the hilt with Turkey, and in my Tryptophan induced haze I have been contemplating tube amp design.

On the Ampeg SVT/Hifi/Dumble SSS, we see the post PI cathode follower arrangement as a way to provide a low impedance signal to the power tubes. Presumably to meet the low grid leak needs of certain 6550/KT88 tubes and to provide a higher current drive to the power amp. This makes sense in the higher power amp designs as they maximize the power in the low end especially.

But does using a setup like this with say, lower power amp designs (50W or less) make any sense? Would it push the power tubes into clipping sooner, or cause the low end to become too excessive?

Would like to know your thoughts on this!
"It's like what Lenin said... you look for the person who will benefit, and, uh, uh..."
pdf64
Posts: 2688
Joined: Sat Mar 12, 2011 9:23 pm
Location: Staffordshire, UK

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by pdf64 »

I think it may increase the instantaneous mid and high freq power output a bit, and may alter the character of the overdriven waveform, as the control grids can be pulled a bit above Vg-k=0 (but not with good linearity).
If there's capacitive coupling to the power tube control grids, then I don't see that proper AB2 operation is achievable.
User avatar
martin manning
Posts: 13079
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by martin manning »

SVT and SSS DBeasley asks about are direct coupled, so there is the possibility of AB2 operation. It would be an interesting experiment to try this with say 6V6 outputs and see what the effect is. A prototype could be rigged up pretty easily on an existing amp to try it out.
tubeswell
Posts: 2337
Joined: Thu Mar 27, 2008 6:42 am
Location: Wellington. NZ

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by tubeswell »

I've seen 'lesser' amps make use of a 12AU7 'buffer/driver'

e.g. attached generic schematic for a 1960s range of 35W to 100W designs from an old NZ manufacturer (albeit the '82' is employed as a common-cathode stage).

Yes its using 100k grid leaks (but these are for 6L6 or 7027A). You could use 100k grid leaks for 6V6s (and some 6V6 datasheets recommend this for fixed bias and I've also seen this in other amps, e.g. Morgan SW22)
You do not have the required permissions to view the files attached to this post.
He who dies with the most tubes... wins
User avatar
martin manning
Posts: 13079
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by martin manning »

In that one the output drivers are capacitor coupled at their plates, so not really in the same vein as the DC CF ones being considered here.
dbeasley
Posts: 144
Joined: Fri Jul 28, 2017 3:07 am

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by dbeasley »

That one is similar to the Dumbleland or winterland I think. Also curious as to how that would sound too!

Sent from my Moto G (4) using Tapatalk

tubeswell
Posts: 2337
Joined: Thu Mar 27, 2008 6:42 am
Location: Wellington. NZ

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by tubeswell »

martin manning wrote: Fri Nov 23, 2018 10:49 pm In that one the output drivers are capacitor coupled at their plates, so not really in the same vein as the DC CF ones being considered here.

Yup I know, that's why I said 'albeit as a common cathode stage'.

Even so, while it doesn't produce as much current as using a CF, an AU7 common cathode stage this way will still drive the output tubes with a bit more current and a bigger output voltage swing than directly from the PI. But its not in the same league as an Ampeg SVT
He who dies with the most tubes... wins
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by R.G. »

I was just thinking about what happens when you drive a grid from a low impedance.

The low impedance can charge and discharge the effective grid capacitance faster. The effective grid capacitance is the actual grid-to-ground capacitance plus the Miller capacitance - the grid to plate capacitance times the voltage gain. For pentodes and power beam tubes (EL34 and its ilk, and 6L6 respectively) the grid to plate capacitance is low - that's the reason there is a screen grid in the first place.But whatever the capacitance is, it has to be charged and discharged before the grid field can change the electron flow from cathode to plate. It's like rolling a rock over a garden hose to squeeze off the flow. Light rocks (low capacitance) are easy to roll, heavy rocks (high capacitance) are harder to roll. Hence, it needs more or less force to overcome the weight (capacitance) to modulate the flow. Using a higher-force rock mover (i.e. lower impedance drive) can speed things up. So using a low impedance follower can drive grids faster and ought to improve high frequency response >if< it's being slowed down by grid capacitance.

That only helps if the grid is operated below Vgk=0. When you make the grid positive with respect to the cathode, electrons are no longer repelled by the grid and some of them start sticking to it as the flow passes. This is grid current. To make the grid go more positive, you have to pull these sticking electrons off the grid by pulling them off with current. The effective grid impedance goes from the grid itself - near infinite impedance - paralleled by the grid leak resistor (which is there to pull off the odd sticking electron in the Vgk negative operations) down to as low as a few K ohms.

Most tubes will continue to increase plate current with the control grid driven modestly positive, as the positive grid sucks the electron flow from the cathode towards the plate. This is something that the screen grid does as well, and for the same reasons. But now even more electrons get sucked into the sticking to the grid itself, and you have to pull those electrons off the grid to keep it positive. That is, you have to supply grid current.

As long as you can supply enough current into the grid to pull the sticking electrons off and so long as the cathode can supply enough electrons, current increases.

What this does is stop the flat-topping distortion that happen when you can't supply enough current to the grid, and the grid stays flat at Vgk=0. When a high impedance driver lets Vgk stop at 0 because it can't supply the current, electron flow stays where it was at Vgk=0 (duuuh) so you get a flat, hard transition on the peaks that hit positive grid voltlages. This is quite different sounding from the compressed distortion of a tube nearly turning off.

This makes a difference in Class A amps, but not in Class AB, because the important parts of peaks in AB always involve the opposite-side tube being off. So supplying low impedance drive can make for a softer distortion in class AB output stages.

And it occurs to me that this is the answer. It's not whether you are in a 10W, 50W or 1000W amp stage. What matters is whether the amp (1) needs the extra high frequency capability that a follower can provide and (2) whether the output tubes clipping at Vgk=0 limits the output or not. If something else limits the output tube plate current before Vgk=0 does, you won't hear it. If you're banging on Vgk=0 with grid drive, a lower impedance drove as in a follower will soften the clipping.

A clearer form of the answer is that if you're driving output tubes with enough voltage on the grids to cause Vgk=0, and it would go more positive with more current drive, then followers will soften grid clipping distortion and provide a bit more power. Amp power doesn't matter so much as the size of the grid voltage drive relative to grid cutoff voltage.
User avatar
roberto
Posts: 1841
Joined: Tue Nov 27, 2007 4:45 pm
Location: Italy

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by roberto »

If you take a look at old VHT (now Fryette) designs, even 60W amps have cathode followers after the PI.

If the purpose is to deliver current, why using tubes for it? I'd go with mosfet set-up to work really linearly (so maybe instead of grounding them, connect them to raw bias voltage.
User avatar
martin manning
Posts: 13079
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by martin manning »

Further to RG’s comments, a low impedance, direct-coupled driver should eliminate blocking distortion (farting out) problems without resorting to reducing low frequency content. That could lead to a punchier low-end in lower wattage amps.
Solid state drivers for fixed and cathode biased power stages are described in MOSFET Follies: http://www.geofex.com/Article_Folders/m ... tfolly.htm
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by R.G. »

roberto wrote: Sun Nov 25, 2018 9:44 am If you take a look at old VHT (now Fryette) designs, even 60W amps have cathode followers after the PI.
If the purpose is to deliver current, why using tubes for it? I'd go with mosfet set-up to work really linearly (so maybe instead of grounding them, connect them to raw bias voltage.
I suspect that it's a cultural bias. The reason I used the words "follies" and "heresies" is that a lot of tube amp fanciers feel that ANY silicon devices in an amp degrades the sound, in a perhaps indefinable, not-measurable ways. It's clearly an irrational bias. The commercial amps don't dare go to anything solid state that close to the audio path for fear of losing the pure-tube fanatics for sales.

Doing it right requires worrying about proper power supplies for the MOSFETs and proper heat removal. A bias voltage supply may not provide enough current capability.

But yes, MOSFET followers offer a great way to drive output grids. I probably ought to go back and look at the MOSFET Follies and see if there's a newer recommended design with better devices since I first did that in 2000, eighteen years ago.
User avatar
didit
Posts: 976
Joined: Thu Sep 25, 2008 3:37 pm
Location: Canada

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by didit »

Not done maths but a FET buffer appears no better than, just different, in a signal path. Ignoring cost, heat, space etc...

Have I missed something key?

Best .. Ian
User avatar
Tony Bones
Posts: 228
Joined: Thu Oct 30, 2014 8:24 pm

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by Tony Bones »

didit wrote: Sun Nov 25, 2018 3:38 pm Not done maths but a FET buffer appears no better than, just different, in a signal path. Ignoring cost, heat, space etc...

Have I missed something key?

Best .. Ian
I don't think you've missed anything. With all due respect to others on this forum, MOSFET followers are certainly a solution, probably even a good solution, but hardly the only good solution.

On paper, followers seem pretty straight forward. So long as the transconductance is high enough, and there's adequate current compliance, any device should work as well as any other and be virtually transparent. But, my own experience with followers in both guitar and playback amps is that they're never as transparent as they should be. That tells me that they might not be as straight forward as they seem.
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by R.G. »

didit wrote: Sun Nov 25, 2018 3:38 pm Not done maths but a FET buffer appears no better than, just different, in a signal path. Ignoring cost, heat, space etc...
Have I missed something key?
Not at all. If you ignore cost of replacement tubes, heat generated by driver tubes, heater current needed by tubes, size of the tubes, chassis space taken up by tubes that could be used for other tube functions, longevity of the devices...

Yep. No better than tubes. Just cheaper, longer lived, and smaller. :lol:
Tony Bones wrote: Sun Nov 25, 2018 6:03 pm I don't think you've missed anything. With all due respect to others on this forum, MOSFET followers are certainly a solution, probably even a good solution, but hardly the only good solution.

I firmly agree - MOSFET followers are not the only solution. But I'm amazed at your post. When I first proposed MOSFETs followers in tube amps back in 2000, I got emails running from skeptical to hate mail. Back then, a great number of people were offended at the very idea of a non-tube follower in a tube amp. Fast forward to today when someone has to note that MOSFETs are not the only solution. Wow! I love that although it's taken 18 years, a once-heretical idea has become accepted enough that people would have to note that they're not the only way. Thank you!
On paper, followers seem pretty straight forward. So long as the transconductance is high enough, and there's adequate current compliance, any device should work as well as any other and be virtually transparent. But, my own experience with followers in both guitar and playback amps is that they're never as transparent as they should be. That tells me that they might not be as straight forward as they seem.
Followers are indeed not as straightforward as they seem in some ways. But that's a valid note for even the simple self biased triode stage, right? Practically nothing in electronics is as simple - nor as complicated! - as it seems at first blush.

The following I say with respect and an attitude of wanting to learn: I've heard "transparency" or the lack of it as a criticism of practically every audio circuit ever used. What makes me a little nuts is that no one can ever say what this "transparency" they refer to actually is, or how it might be measured in the interest of making things more transparent. About the best working definition I've been able to come up with is that transparency means "it sounds like I expected it to sound", and that lack of transparency means "it doesn't sound like I expected it to sound". I'd love to design for "transparency" if I could figure out what it is and when I'm being successful.
dbeasley
Posts: 144
Joined: Fri Jul 28, 2017 3:07 am

Re: Post Phase Inverter Cathode Follower - Pointless in anything under 100W?

Post by dbeasley »

R.G.

Thanks for the follow up details! I reckon I could just use Silicon for this role. I just got really intruiged by these older designs.

Maybe I'll use these cathode follower designs with that UL series vox board I got from you!

All the best,

Dustin

Sent from my Moto G (4) using Tapatalk

Post Reply