Welcome, Guest. Please login or register.

November 30, 2021, 04:49:07 PM

Login with username, password and session length

Recent Posts

 

Author Topic: Fet version of Marshall Superlead Plexi  (Read 33485 times)

J M Fahey

  • SSGuitar Global Mod
  • Legendary
  • ****
  • Posts: 4142
  • Chip Points: 429
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #30 on: February 08, 2011, 12:03:54 AM »
That's the way to go !!! Congratulations.
I am not quite loved in some Forums because I always state that the ROG approach of taking a "Famous" circuit , replacing the tube with a FET (a *very* different thing), feeding it 9V instead of 250V (a *very* different thing), and loading it with "100K" is *not* the way to capture the "old sound"
Far from it.
To boot, that "100k" is not a resistor but a 10 turn potentiometer or variable resistor, which gets adjusted to any value between 1K and 47K (a *very* different thing) whichever is needed by that particular FET.
And they expect to reproduce the original, vintage amp sound.
No way!!!
Quote
I was used to the typical 2k7/680n, 820/680n, stuff like that. Now I don't see any of these values anywhere
And you won't except by chance.
And those values are not "magic" or "sacred".
Fact is, active elements (tubes, transistors, FETs, MOS) have what's called a "working point" which is established by Bias.
You need to set the bias needed *there*, not something used in an absolutely different place.
My Father was a Country Doctor; he was often approached by somebody who asked him to prescribe "the "red pill" which had been so good for a neighbour's ilness". Oh well.
Our friend KMG *does* use individual trimmers (unfortunately Fets and Mosfets still are very variable) but he uses them "the right way": in a mini regulated PSU which feeds the correct bias in a variety of conditions, which does *not* adversely affect tone.
A de-luxe approach. :tu:

THChrist

  • Master
  • ***
  • Posts: 52
  • Chip Points: 2
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #31 on: February 08, 2011, 12:20:43 AM »
Quote
I was used to the typical 2k7/680n, 820/680n, stuff like that. Now I don't see any of these values anywhere
And you won't except by chance.
And those values are not "magic" or "sacred".

what i was trying to say is that, with those "sacred values", i knew where they went and what to change, add some caps and do some mod (caswell, levi, etc). But now i'm lost. I don't recognize anything. What i *DO* know is that KMG is doing a great job at emulating tube sound. I have his website on my bookmarks and I was mesmerized 'bout his results long before I posted my question about his SS poweramps at this forum (today a posted another question at freestompboxes forum).

I know those values aren't sacred, but that was what i was used to (and I don't like ch-ch-ch-ch-changes)

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #32 on: February 08, 2011, 03:49:34 AM »
Quote
what i was trying to say is that, with those "sacred values", i knew where they went and what to change, add some caps and do some mod (caswell, levi, etc). But now i'm lost. I don't recognize anything. What i *DO* know is that KMG is doing a great job at emulating tube sound. I have his website on my bookmarks and I was mesmerized 'bout his results long before I posted my question about his SS poweramps at this forum (today a posted another question at freestompboxes forum).

I know those values aren't sacred, but that was what i was used to (and I don't like ch-ch-ch-ch-changes)
These "sacred values" are suitable for given preamplifier when using the same tubes. For example if you replace tube on Russian 6Н2П-ЕВ these values will be different for prorer sound.
V/A curves of FETs differs significantly from the tube ones.
If you connect a bypass capacitor directly to the source of the FET, increase of gain will be too large.
You`ll get different frequency response of preamp stage.
« Last Edit: February 08, 2011, 03:51:23 AM by KMG »

Steve Conner

  • Regular
  • **
  • Posts: 23
  • Chip Points: 8
  • somebody squished mah trannies!
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #33 on: February 08, 2011, 07:25:12 AM »
Well, that's the issue. I guess some people have this holy grail of a "tube replacement" block that will drop straight into a classic circuit, with the same voltages and component values outside of it. But that is hard, and it's not very forward looking either. KMG's approach is the best I've seen.

I'm used to designing my own tube circuits from scratch (and I don't really care much if other people like the sound of them :) - see http://scopeboy.com/amps for details) so the approach suits me fine too. I would just take KMG's building block as a strange kind of tube with its own characteristics. I'm looking forward to trying them with my transformer driven power amp.

KMG: You used an emitter follower to buffer the bias voltage on the FET source. Is the temperature coefficient of Vbe a problem there? Or does it actually improve stability of the circuit? Likewise for the tempco of the germanium diode.

I understand what the Ge diode is for now! :) I thought it was part of the "scalable diode" circuit that emulates grid current. But you said it was to make the cutoff softer, and that makes perfect sense.

I'm sold on the topology, but I'm determined to use only current production parts in my designs, after all, that is the main reason why I'm uneasy about using tubes in the 21st century. So do you think a Schottky would do instead of a germanium diode?
« Last Edit: February 08, 2011, 07:26:32 AM by Steve Conner »

gbono

  • Elite
  • ****
  • Posts: 308
  • Chip Points: 9
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #34 on: February 09, 2011, 12:44:25 AM »
 KMG - any chance you can repost the schematic and/or pictures? I can't load the files in the original post. Cheers

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #35 on: February 09, 2011, 04:53:23 AM »
Quote
You used an emitter follower to buffer the bias voltage on the FET source. Is the temperature coefficient of Vbe a problem there? Or does it actually improve stability of the circuit? Likewise for the tempco of the germanium diode.
I did not notice any problem with temperature stability.
Quote
So do you think a Schottky would do instead of a germanium diode?
No details of the curves for diodes at currents around 1 mA.
So it needs testing.
Quote
KMG - any chance you can repost the schematic and/or pictures? I can't load the files in the original post.  Cheers
I Think that is a problem of internet service providers.
Do you see my web page?
http://milas.spb.ru/~kmg/index_en.html
Anyway, there you can download project archive:
http://www.megaupload.com/?d=U2E5TY54

J M Fahey

  • SSGuitar Global Mod
  • Legendary
  • ****
  • Posts: 4142
  • Chip Points: 429
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #36 on: February 09, 2011, 08:23:50 AM »
Hey KMG !!
You seem to be involved in pretty complex industrial electronics of the highest level !!
Please tell us something about that.
I'm impressed !!! :tu:

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #37 on: February 09, 2011, 10:22:21 AM »
I work for a leading engineer in the development department of Mega-Milas Ltd. We develop a fire alarm systems, information systems and surveillance systems mainly for railway transport.
For example, one of my designs - PCB (multilayer) for the fire alarm server for railway train.

joecool85

  • SSGuitar Admin
  • Legendary
  • ******
  • Posts: 3436
  • Chip Points: 995
  • SSG Creator
    • View Profile
    • thatraymond.com
Re: Fet version of Marshall Superlead Plexi
« Reply #38 on: February 09, 2011, 02:17:40 PM »
I work for a leading engineer in the development department of Mega-Milas Ltd. We develop a fire alarm systems, information systems and surveillance systems mainly for railway transport.
For example, one of my designs - PCB (multilayer) for the fire alarm server for railway train.
...

Neat!  How'd you happen to come across this site anyway?
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #39 on: February 09, 2011, 03:04:06 PM »
It's very simple, I was looking for links to my site, found at this site discuss one of my projects and decided to join the discussion.

joecool85

  • SSGuitar Admin
  • Legendary
  • ******
  • Posts: 3436
  • Chip Points: 995
  • SSG Creator
    • View Profile
    • thatraymond.com
Re: Fet version of Marshall Superlead Plexi
« Reply #40 on: February 09, 2011, 03:12:25 PM »
It's very simple, I was looking for links to my site, found at this site discuss one of my projects and decided to join the discussion.

That makes sense.  It's good to have you aboard!
Life is what you make it.
Still rockin' the Dean Markley K-20X
thatraymond.com

THChrist

  • Master
  • ***
  • Posts: 52
  • Chip Points: 2
    • View Profile
Re: Fet version of Marshall Superlead Plexi
« Reply #41 on: February 09, 2011, 04:49:59 PM »
It's very simple, I was looking for links to my site, found at this site discuss one of my projects and decided to join the discussion.

The one I posted at freestompboxes? Dude, I knew this will be the right place for your SS Amps.

On a side note, sometimes I imagine the awesomeness that could be if KMG form a team with Teemu... "The SS-Team" (too bad it sounds kinda nazi)
« Last Edit: February 09, 2011, 04:53:41 PM by THChrist »

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #42 on: February 10, 2011, 05:01:23 AM »
The story at freestompboxes begins from this topic:
http://freestompboxes.org/viewtopic.php?f=28&t=8037&p=90747
On ssguitar I come after this discussion:
http://www.ssguitar.com/index.php?topic=1981.0

rowdy_riemer

  • Elite
  • ****
  • Posts: 264
  • Chip Points: 13
    • View Profile
    • www.riemer.us
Re: Fet version of Marshall Superlead Plexi
« Reply #43 on: February 10, 2011, 09:22:52 AM »
I am not quite loved in some Forums because...

I hope you're not confusing disagreement with dislike.  :) Of course, I can only speak for myself in saying I do not dislike you at all. I think you pretty much rock.

Quote
I always state that the ROG approach of taking a "Famous" circuit , ....

I've always agreed about your criticism of ROG and others FET biasing schemes. It just seems to come across as a bit smug when you do so rather than as giving constructive criticism. Of course, I could be totally misreading the attitude in your criticism. You're obviously an electronics bad ass. I think the guys at ROG definitely know a lot more than your average hobbiest, but they're certainly not in the same league as you or KMG or Teemuk, etc. So when you're hard on the ROG guys, it almost sounds like a professional athlete talking trash about how a high school football team sucks. Or maybe a physicist telling an elementary school kid that he sucks at math. Once again, I might be totally misreading the intent behind your criticism.

KMG

  • Elite
  • ****
  • Posts: 147
  • Chip Points: 53
  • completely insane
    • View Profile
    • KMG personal page
Re: Fet version of Marshall Superlead Plexi
« Reply #44 on: February 10, 2011, 10:17:02 AM »
Be patient, guys. All ideas have a right to live.
IMHO, one important thought:
Tuning the stage for different FET  parameters by changing the resistor at the drain will also affect stage gain and its output impedance (which affects the frequency response of schematic).
It impairs the schematic repeatability.
In my version, of course more complex, adjusting does not affect gain and frequency response of the stage.

"Nothing is perfect" (C) Saint-Exupéry. ;)