mosfet CF bootstrapped?

[tubenit]

In this article, Merlin discusses using a boot strapped CF to increase gain and eliminate some of the fizziness in overdrive:

This will increase the gain of the previous stage so it becomes almost equal to the mu of the valve! An added bonus is that if you don't use a cathode-bypass capacitor, the gain is still very hgh (about 85-90 with an ECC83). Leaving the cathode unbypassed reduces blocking distortion and reduces 'fizzy' sounding overdrive, giving better crunch.

http://www.valvewizard.co.uk/dccf.html

Has anyone tried this?  Did it sound smooth or more "crunchy" (using Merlin's term).

Would this work using a mosfet CF?      (or do you need a tube triode CF?)

Would one still use the 220R/1w resistor and the diode with the mosfet CF if boot strapping it?

This TBM with 5879 OD doesn't need more gain, but I am simply curious if this would work for more gain?

With respect, Tubenit

[HotBluePlates]

"Leaving the cathode unbypassed reduces blocking distortion and reduces 'fizzy' sounding overdrive... "

When Merlin says this, I read it as local negative feedback due to the unbypassed cathode resistor keeping the plate-loaded stage cleaner.

Has anyone tried this?  Did it sound smooth or more "crunchy" (using Merlin's term).

I think it might be an excellent thing for you to try! It seems like only a few folks have MOSFET followers in their amps.

Would one still use the 220R/1w resistor and the diode with the mosfet CF if boot strapping it?

Dunno... try it first with the 220Ω. 

This TBM with 5879 OD doesn't need more gain, but I am simply curious if this would work for more gain?

Hopefully Merlin sees this and answers.

I'm not familiar with this particular circuit connection. It seems like it attempts to get "more gain" by making the effective plate load resistor larger (R2 in the article). That makes a smaller "voltage divider of mu" when working against the internal plate resistance of the triode.

But pentodes operate on a different principle. Internal plate resistance is so high that pentode gain is figured mainly by Gm * Rl, or transconductance times plate load resistance. So to increase the gain of your pentode stage, you make the plate load larger than the existing 56kΩ. The catch is that reduces plate current (more series resistance reduces current), and reduced plate current also generally means reduced Gm, so you often have to re-jigger the screen resistor and/or the cathode bias resistor (both contribute to the pentode bias).

All this said, if the bootstrapped effective plate load resistance becomes huge then it also helps boost pentode gain. I think you trying the circuit is the best approach to answer the question.

[foghornleghorn]

I too have been pondering bootstrapping a pentode with a CF triode recently, both as a first stage and a driver stage.

Has anyone considered a 6U8 or similar "TV tube" for this? Three possible  benefits, one unknown.

Benefit #1  only one tube socket needed.

Benefit #2  cheaper than 5879 + 12AX7.

Benefit #3   6U8 pinout (9AE) is common to a few of these old TV triode/pentode tubes so tube rolling is facilitated.
Many more seem to share the 9DX pinout, so those are possible candidates too. 6EB8, 6GN8, 6JT8, 6LY8, 6MV8, and these are just the ones with 12AX7 like triodes. 6EB8 triode is more like a 12BZ7 though.

Unknown    tone quality. I have it on good authority that a 6U8 sounds good in a guitar amp cathodyne PI.

I'm also interested in switching the feedback in and out with a relay as a clean/overdrive FSW. Trick is how to do this without pops/clicks when switching. Haven't looked at this aspect in detail yet so I'm not yet equipped to make any suggestions, just want to throw this idea out there.

[kagliostro]

I would like to know one other thing

(with respect the question if a penthode may be used in a bootstraped circuit)

The triode that follows and has its output connected to the cathode

gives the usual feature of a low impedance output of a standard CF also here ?

I think so, but I'm not sure about

K

[tubenit]

An advanced search on the Hoffman forum pulled up some threads regarding the 6U8 tube.

quote from leoman May31,2011

I tried the 6U8A and IMO it really has no tone.

 

Reply #4  http://www.el34world.com/Forum/index.php?topic=11684.0

&

In this thread reply #21 there is a schematic for a 6U8 using a triode as a CF .  Don't know if this was actually built into an amp or not?  

http://www.el34world.com/Forum/index.php?topic=14545.0

&  

Reply #3 here shows a schematic for a working amp by zendragon (who has great designs) using a 6U8 in preamp.  And again in
reply #10.   Very interesting design!

http://www.el34world.com/Forum/index.php?topic=8253.0

With respect, Tubenit

[tubenit]

From this thread  reply # 6 suggests that the 6U8 is not a good choice for the overdrive?

http://www.el34world.com/Forum/index.php?topic=9654.0

went back tonight to see what I could do with a 6U8A. Did a load of calculations stepping through it and dialed it in pretty close right off the bat but I no matter what I adjusted, I just cannot get a decent OD sound with it. I am again surrendering to the adage that I am not yet able make a audio silk purse out of a TV local oscillator/mixer combo

Reply #7 of the same thread expresses some success as a preamp tube but that 6U8's tend to be microphonic

I did find that about 70% of the 6U8As I bought were too microphonic to use.

From this thread, some concern that the pentode should follow the triode in the 6U8  Reply #4

http://www.el34world.com/Forum/index.php?topic=8965.0

So far from all my experiments the pentode sounds best as a driver with the triode first. You can really tune the amp that way. You can either use the tone of the triode or pentode by changing values and gain, lower the gain of the pentode and the sound of the triode comes through more. Add gain and tone shaping from the pentode and it's sonic qualities become more apparent

I think I'll stick with the 5879 using a mosfet as a CF.  I'm really happy with it and think it sounds fantastic!

With respect, Tubenit

[TIMBO]

Hi T, Merry Christmas, I've read your post a few times and I couldn't quite work out in what direction you were heading in ( too many beers and prawns has had effect on the brain cells)

Are you looking to split the plate resistor of the 5879 to feed the mosfet as per Merlins info

Or are you looking to replace the 5879 with something else.

I strapped a mosfet to 6AU6 and sounded fine 

[spacelabstudio]

In this article, Merlin discusses using a boot strapped CF to increase gain and eliminate some of the fizziness in overdrive:

This will increase the gain of the previous stage so it becomes almost equal to the mu of the valve! An added bonus is that if you don't use a cathode-bypass capacitor, the gain is still very hgh (about 85-90 with an ECC83). Leaving the cathode unbypassed reduces blocking distortion and reduces 'fizzy' sounding overdrive, giving better crunch.

http://www.valvewizard.co.uk/dccf.html

Has anyone tried this?  Did it sound smooth or more "crunchy" (using Merlin's term).

I haven't, but I'm curious.

Would this work using a mosfet CF?      (or do you need a tube triode CF?)

It should, as far as I can tell.

Would one still use the 220R/1w resistor and the diode with the mosfet CF if boot strapping it?

Bootstrapping, as far as I can tell, doesn't impact the potential for a reverse bias condition on the Mosfet, so I'd think you'd still need that.

This TBM with 5879 OD doesn't need more gain, but I am simply curious if this would work for more gain?

I don't know, but probably, yes.  Easy to try.  Replace your 56k load resistor with two 27k and add that cap.

Chris

[tubenit]

are you looking to replace the 5879 with something else

.

No, I like the 5879 ALOT!  Have no desire to replace it.  Question is about bootstrapping the mosfet CF.

I may try it out of curiousity and see what happens & then just reverse it IF I don't like it.

with respect, Tubenit

[HotBluePlates]

Question is about bootstrapping the mosfet CF.

You won't be bootstrapping the MOSFET. You're bootstrapping the previous stage's plate load.

It dawned on me how this circuit works: the same as the large apparent input resistance of a bootstrapped CF (with a grid reference returned to the junction of cathode bias and cathode load resistors).

Bootstrapping Explained
Refer to the cathode follower diagram below. The output of the cathode follower is at the top of the 1kΩ resistor, and the 1MΩ input resistor is returned to the top of the 56kΩ cathode load.

Assume the cathode follower is a 12AX7 with a gain after the CF's feedback of 0.98. If a 1v input signal is applied to the grid, 0.98v appears at the cathode. The junction of the 1kΩ and 56kΩ will also rise about 0.98v * [56k / (56k + 1k)] = 0.96v.

The current drawn from the source though the 1MΩ will be that due to the difference between the input signal and resulting output, because they are of the same phase. 1v - 0.96v = 0.04v  ->  0.04v / 1MΩ = 0.04μA (micro-amp).

Pretend you are the previous stage, and the circuit is a black box. You have applied a 1v signal and have a resulting current draw of 0.02μA, which means the input resistance of the driven stage looks like 1v/0.04μA = 25MΩ. This apparent resistance due to the circuit connection is the bootstrapping effect. The presence of a portion of the in-phase follower output makes the circuit act as though a much larger resistance is present at the bootstrapped point.

Now realistically, gain may be slightly lower than 0.98 and largely depends on how much gain the device is throwing away in feedback to make the follower (the more gain, the closer to a final gain of 1, and the higher the bootstrapped resistance).

The Subject Circuit
Instead of a grid reference resistor, a portion of the plate load resistor is being used as the bootstrapped resistance.

It seems the bootstrapped point (the part of the plate load connected to the follower's output) cannot be the B+ end of the plate load, because that's at AC Ground so output signal would be Zero.

The bootstrapped point cannot be at the plate of the previous stage, because there would be no difference in voltage between the input and output of the follower, so it would do nothing.

So it looks to me like tapping the plate load at the midpoint allows this circuit to function properly, and makes the portion of the plate load between the previous stage's plate and the follower output (R2 in Merlin's diagram) causes the bootstrapped resistance to both look like a very large plate load for the 1st stage and like a very large grid reference resistor for the follower.

Having talked through all this, the pentode's gain may get very large, but will be limited by its output voltage capability (does it run into cutoff or saturation). When you try it, it may be profitable to use a 50kΩ pot in place of the plate load, and adjust to find a suitable tapping point... I'm thinking you may run it towards less-than-full-effect.

This looks promising, but maybe even more so in an amp with a minimal preamp of just a pentode (and the follower) before the phase inverter.

[spacelabstudio]

The way I'd look at is this: you know the gain of the CF is a little under one, so the signal voltage at the output is almost exactly the input voltage, which is the voltage at the R2/Plate node.  The capacitor allows signal (but not DC) to be returned to the R1/R2 junction.  So for signals, the voltage at R1/R2 is almost exactly the same as the voltage at R2/Plate.  Almost no voltage change for a significant change in current, means effective resistance looks very high.  So it makes R2 look like a giant load resistor for signals. 

[tubeswell]

I did a bootstrapped CF with a LFO stage and it worked absolutely fine to wiggle some 6L6s

[tubenit]

OK, it works just fine doing the 5879 bootstrapped mosfet idea!   LOTS of gain/boost from it! 

As Merlin described,  ....... it does have some real "crunch" overdrive to it with the amp cranked.  I thought it had a dash of chime also with the crunch. Pretty intense overdrive.

What surprised me is I expected more of a compressed slightly distorted tone and it was not anything like that.  Very touch sensitive response and very intense overdrive cleanish/crunchy tone  (as best as I can describe it). 

My guess is that guys that like a JCM 800 cranked wide open might like this idea. Or maybe the Soldano guys might like it?  Lots of crunch, harmonics and chime.  I would describe it as more crunch than creamy smoothness for sure.

Having said all of that,  it had so much gain that it was difficult to match the clean to OD switching at the same or similar volume. I had to dial back the trim and OD level so much that there was no real advantage for me to have the bootstrapping 5879/mosfet idea.  I put it back to the original  5879 with CF. 

However,  IF someone did NOT want or need clean to OD switching, then I think this idea of bootstrapping is an excellent one! 

SILVERGUN, ....... if you read this, I have a gut feeling that you might really like the tone and touch sensitivity of this. Hope you'll try it out in an amp design at some point and report back.

With respect, Tubenit

[spacelabstudio]

Thanks for the report!  Good to know!

[tubenit]

SILVERGUN,

I'm thinking this might be a great way of applying this idea to a super super touch sensitive crunchy harmonically rich tone ..........

Just an idea .......................   

With respect, Tubenit

[foghornleghorn]

+1! Thanx mr. t.

I am looking at doing something with a 6U8 first stage and a 6GH8A driver/DC coupled CF and I'll post it once I think it through a bit. Just chicken scratch on paper right now. These TV tubes have got me all worked up! I hope they can be made to sound good.

Key is how to footswitch the bootstrapping CF feedback in and out, and also allow preset balancing of Clean and OD levels and tone.

[tubenit]

Key is how to footswitch the bootstrapping CF feedback in and out

To give you a starting place to think about this.  The original 5879 & mosfet CF had the OD level set at 7 and the trim at maybe 5.5
AFTER the bootstrapping,  the trim was at around 3.5 and the OD level around 3 to have the same volume. 

There was a HUGE jump in volume with it bootstrapped.   

[foghornleghorn]

I believe it. Merlin says in his book that a 12AX7 gain went from 60 to almost 90, so with a pentode I would expect even more.

[SILVERGUN]

SILVERGUN, ....... if you read this, I have a gut feeling that you might really like the tone and touch sensitivity of this. Hope you'll try it out in an amp design at some point and report back.

Oh yeah,,,I'm right with you.....I originally posted here as a "bookmark" to remind me to come back and check on your findings.
I had also read that paragraph in Merlin's book and was waiting for the right opportunity to try it out.

A couple of weeks before you posted this idea, I had begun a "Son of a Breadboard" to be able to take home as an experimental platform for small PP and SE amp designs   (before now, everything has been at the shop)

This will be the first design that hits that board, and I CAN'T WAIT to try it! ,,,but I'm still gathering parts  
I have been giving my ears a rest and planning my next step....thanks to you, I don't have to waste any time deciding where to start.
I've got a free Saturday coming up next week and I'm hoping to have it up and running in a couple weeks

Thanks again for all of your work, and for helping to keep me focused.....that seems to be the hardest part of this, for me, because there are so many things to try  
I appreciate the fact that you specifically directed my attention to this, and I'm going to trust your ears and experience, and go with it........I was thinking of starting with the original TBM with 5879 and then bootstrap the CF to hear the difference
I was also thinking of doing it as a tube CF......hopefully I'll have the time to try it both ways and see if there's any difference between the sound with the mosfet and/or tube CFs, when bootstrapped.

I needed something to get me going forward again, and start the new year off with a bang  

[kagliostro]

Many Thanks for sharing your experience Tubenit 

Very interesting tread !

K