[Lonely Star] An adaption of the Mesa Boogie Lonestar amp. Seeking comments!

Started by Auke Haarsma, May 05, 2009, 04:28:22 PM

Previous topic - Next topic

Auke Haarsma

I have a thing for the Lonestar amp. Never played it. Just heard it and seen it. Somehow I feel attracted to it. That is why I wanted to try to make a stompbox version of the amp, ala the guru's at AMZ, ROG, GEOFEX, and ofcourse the many examples at this board!

I took the original pre-amp schematic and replaced the tubes with J201's.

I know there is not such an thing as 'FET-for-Tube'-magic. However, I hope it results in a fun circuit and stompbox.

Here is the schematic of the Lonely Star:



I have made a couple of modifications to the original schematic:

  • Tubes have been replaced with FETs (J201)
  • The amp has two tone-stacks. One for the normal chan, one for the lead channel. I combined those. You can have a lead channel->
  • The lead channel has an extra gain-stage. I have opted for a switch that can bypass this stage to have the 'normal' channel
  • SW LEAD A and B should b a DPDT, but for PCB-reasons I have a SPST in the schematic. If you bypass the lead-channel you go from Drain Q1 directly to C9 (the tonestack)
  • I have added 220pF's (C13, C14, C15, C16) to emulate the millar capacitance (uhm, monkey see monkey do ;) )
  • I scaled down the tonestack by a factor of 10
  • R10 and R15 are 'optional'. I have added those because with them you can kill all the sound by turning the drive/gain completely down. With these resistors you always have a (little) bit drive/gain
  • I have breadboarded this schematic and it works. However before putting it to pcb (which I already designed in Eagle) I love to hear your comments and improvements

I hope to receive some feedback to improve this circuit. And hopefully, once verified, I can finalize the  PCB and share it with the board! I want to make a 'modable' version where it is easy to leave pots out (who needs 7 pots???) by e.g. replace the tonestack with a one-knob version.


/edit: corrected SPST to DPDT

Ice-9

Hi, Thats an interesting looking circuit you have there, I am wondering how well the gain pot works from a quick look, it looks like both ends go to gnd. (through resistors). Secondly it might be better to just have the one tone stack in there. Do you have any sound clips of it to listen too.
Good work and i look forward to having a look at the finished layout.
www.stanleyfx.co.uk

Sanity: doing the same thing over and over again and expecting the same result. Mick Taylor

Please at least have 1 forum post before sending me a PM demanding something.

Auke Haarsma

Thanks for your reply.

I do not have any soundclips as of yet. The gain-pot is wired as in the amp, and it does work ;)

The tonestack is just one tonestack. It is the tonestack of the normal channel. In the amp the lead-channel has its own copy of the tonestack. So you would switch from tonestack-normal to tonestack-lead. That would be too much knobs for my liking. ;)

Mark Hammer

When the gain/drive pot is up full, you have a voltage divider comprised of R9 (220k) and the 1M pot in parallel with R11 (220k).  The way the drive pot is wired, as you turn the drive down, you end up with two cascaded voltage dividers.  So, imagine the 1M pot is set halfway (and omit that optional 100k resistor because the Lonestar does too).  You would now have one voltage divider comprised of 220k and 500k to ground, followed by 500k (the other side of the wiper) and 220k to ground.

kurtlives

I dont understand how the lead switch works. Looks like when it is open you get no signal passing at all.
My DIY site:
www.pdfelectronics.com

Ripthorn

I have the same question as Chris, though I think what is happening is that it is an spst switch and he just has the two different poles represented, so the signal leaves the first gain stage and then either goes through the second gain stage and gain pot, or jumps straight to the tonestack.  Someone correct me if I am wrong, though.
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

kurtlives

Quote from: Ripthorn on May 05, 2009, 05:44:38 PM
I have the same question as Chris, though I think what is happening is that it is an spst switch and he just has the two different poles represented, so the signal leaves the first gain stage and then either goes through the second gain stage and gain pot, or jumps straight to the tonestack.  Someone correct me if I am wrong, though.

I thought that too at first.

A DPDT would be needed though. Jumper the outside two lugs so the signal can bypass the extra gain stage though.
My DIY site:
www.pdfelectronics.com

Auke Haarsma

The switch is pretty unclear, I understand. And I made a mistake in the description.

It should be a DPDT.

Position 1 (normal channel): you bypass everything around Q2. From the Drain of Q1 you go the C9, the tonestack.
Position 2 (lead channel): you go from Drain of Q2 to C6, R8/R9. And the Drain of Q2 leads in to C9, the tonestack.

I'll make an updated schematic with a DPDT, just for clarity in the next couple of days.

Auke Haarsma

Quote from: kurtlives on May 05, 2009, 05:53:49 PM
A DPDT would be needed though. Jumper the outside two lugs so the signal can bypass the extra gain stage though.
Spot on!

Transmogrifox

Quote from: kurtlives on May 05, 2009, 05:39:43 PM
I dont understand how the lead switch works. Looks like when it is open you get no signal passing at all.

I think we have to make the assumption the switch is a DPDT switch with parts not shown in the schematic.  It looks like the idea is to bypass one of the gain stages. EDIT: looks like 4 of us posted at the same time :)

Something to keep in mind about this circuit is that it will distort a lot more than the tube amp equal will distort.  The headroom is probably about one tenth of that in the MESA amp.  Remember the rails in the preamp section are often in excess of 100V.  It's possible to have a 50V-p-p signal, whereas in the 9V model your signal will be maybe 5V-p-p.

A way to account for this would be to put a 10:1  voltage divider on the input to scale the signal back so it clips at about the same place as it would in the real thing.   Without the gain reduction on the front end, the input stage will probably clip, when in the real amp it likely does not clip on the input stage.  Clipping the input will make it sound more "fuzzy" and loose.  You may find the thing is more useful just the way it is.
trans·mog·ri·fy
tr.v. trans·mog·ri·fied, trans·mog·ri·fy·ing, trans·mog·ri·fies To change into a different shape or form, especially one that is fantastic or bizarre.

Ice-9

Quote from: Auke Haarsma on May 05, 2009, 05:01:12 PM
Thanks for your reply.

I do not have any soundclips as of yet. The gain-pot is wired as in the amp, and it does work ;)

The tonestack is just one tonestack. It is the tonestack of the normal channel. In the amp the lead-channel has its own copy of the tonestack. So you would switch from tonestack-normal to tonestack-lead. That would be too much knobs for my liking. ;)

Oh yeah sorry ! I didn't look closely enough, i misread the second gain pot after the tone controls, i saw it as a BM tone stack. I'll have to get some glasses.
www.stanleyfx.co.uk

Sanity: doing the same thing over and over again and expecting the same result. Mick Taylor

Please at least have 1 forum post before sending me a PM demanding something.

Auke Haarsma

Quote from: Transmogrifox on May 05, 2009, 06:07:09 PM
Something to keep in mind about this circuit is that it will distort a lot more than the tube amp equal will distort.  The headroom is probably about one tenth of that in the MESA amp.  Remember the rails in the preamp section are often in excess of 100V.  It's possible to have a 50V-p-p signal, whereas in the 9V model your signal will be maybe 5V-p-p.

A way to account for this would be to put a 10:1  voltage divider on the input to scale the signal back so it clips at about the same place as it would in the real thing.   Without the gain reduction on the front end, the input stage will probably clip, when in the real amp it likely does not clip on the input stage.  Clipping the input will make it sound more "fuzzy" and loose.  You may find the thing is more useful just the way it is.

With the trimpot you can set the voltage on the drain to such a setting that Q1 stays clean a long time.

However, would increasing R3 (currently 68k) help in getting closer to the amp?

Ripthorn

I took a look at the original schem and it looks like a project for submini tubes to me, but then again, I'm really into them at the moment.  Looks like a great design.
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

Transmogrifox

Quote from: Auke Haarsma on May 06, 2009, 02:44:40 AM
Quote from: Transmogrifox on May 05, 2009, 06:07:09 PM
Something to keep in mind about this circuit is that it will distort a lot more than the tube amp equal will distort.  The headroom is probably about one tenth of that in the MESA amp.  Remember the rails in the preamp section are often in excess of 100V.  It's possible to have a 50V-p-p signal, whereas in the 9V model your signal will be maybe 5V-p-p.

A way to account for this would be to put a 10:1  voltage divider on the input to scale the signal back so it clips at about the same place as it would in the real thing.   Without the gain reduction on the front end, the input stage will probably clip, when in the real amp it likely does not clip on the input stage.  Clipping the input will make it sound more "fuzzy" and loose.  You may find the thing is more useful just the way it is.

With the trimpot you can set the voltage on the drain to such a setting that Q1 stays clean a long time.

However, would increasing R3 (currently 68k) help in getting closer to the amp?

The trimpot on the drain adjusts the bias.  I agree that you can change the gain by making it much smaller.  However,  this will change the high-pass cut-off frequency through C6 (4n7).  Additionally this will not prevent the first stage from distorting due to cut-off (ie, when the input signal is on the negative swing, it turns off the transistor).

Changing the 68k resistor won't really change the gain.  It will only change the cut-off frequency of the low-pass filter formed by the 220pF grid capacitance emulator (which actually might make it more likely to distort sooner). 

It doesn't seem as though it would be hard to put a divider on the input.
I would probably replace the 2.2M (R1) with two resistors: 511k and 68k.  The 511k would be in series with the input, the 68k goes to ground.  Take the center tap of the two directly to the input capacitor.  This configuration emulates the effect of having the guitar's volume knob turned down to about 9%.  The input impedance will be plenty high.

We're dealing with Alice in Wonderland fitting into two different sized rabbit holes.  One has a high ceiling, the other has a low ceiling.  If Alice is the same size going into the first hole, she's going to bonk her head on the roof of the smaller hole.  When she goes into the rabbit hole entering tube amp, she's just the right size.  Each tube stage she enters has a mushroom so she grows a little bit more.  Each FET stage in the FET amp has the same kind of mushroom, so her size increases by the same ratio as the tube amp...but if the ceiling is shorter, she'll whack her head.

Making a link with the analogy, when you shrink the power supply down from 150V to a 9V supply, you essentially make the entry rabbit hole 10 times smaller than it is for the tube amp.  If you shrink Alice before she goes into the FET amp so the same ratio of her body fills that hole the same as it does for the tube amp, then you have properly adjusted Alice for headroom.  She will grow proportionately in each FET stage thereafter and bonk her head on the roof at about the same bend in her legs as she would in the tube amp.

Without gain reduction on the front end, Alice is going to whack her head on the entry.  Give her a 10:1 mushroom. 
trans·mog·ri·fy
tr.v. trans·mog·ri·fied, trans·mog·ri·fy·ing, trans·mog·ri·fies To change into a different shape or form, especially one that is fantastic or bizarre.

MikeH

The notes say that the channel switching is done with a DPDT, but is represented as an SPST for eagle/pcb reasons,.

They say nobody reads anymore  ;)
"Sounds like a Fab Metal to me." -DougH

Ripthorn

Quote from: MikeH on May 06, 2009, 02:51:52 PM
The notes say that the channel switching is done with a DPDT, but is represented as an SPST for eagle/pcb reasons,.

They say nobody reads anymore  ;)

The edit says that it used to not read dpdt... :icon_biggrin:
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

John G

Hi,
Be aware that MB factory schems have deliberate errors in them, to the casual reader they don't seem apparent.
I believe that the Presence control may be attached to the top side of the master control not the wiper. (see MK1 reissue)
The mid boost is the same connection as for the MKI MB and goes to the top of the treble control, not the wiper??
Why just scale the Tone Stack why not the whole schem.............and are FET impedance's significantly lower than the tube schem ??
Just my observations and 2 cents worth.
John G

MikeH

Quote from: Ripthorn on May 06, 2009, 05:04:47 PM
Quote from: MikeH on May 06, 2009, 02:51:52 PM
The notes say that the channel switching is done with a DPDT, but is represented as an SPST for eagle/pcb reasons,.

They say nobody reads anymore  ;)

The edit says that it used to not read dpdt... :icon_biggrin:

See?  Nobody DOES read anymore  :D
"Sounds like a Fab Metal to me." -DougH

Auke Haarsma

Quote from: Transmogrifox on May 06, 2009, 01:58:48 PM
Without gain reduction on the front end, Alice is going to whack her head on the entry.  Give her a 10:1 mushroom. 
Thanks for the story ;) Would Alice need an additional 1:10 mushroom at the end to bring back the level ?

Quote from: John G on May 06, 2009, 05:22:26 PM
Hi,
Be aware that MB factory schems have deliberate errors in them, to the casual reader they don't seem apparent.
I believe that the Presence control may be attached to the top side of the master control not the wiper. (see MK1 reissue)
The mid boost is the same connection as for the MKI MB and goes to the top of the treble control, not the wiper??
We would need some1 with the actual amp to verify, or a verified schematic. I have just based this schematic on the schematic I found online (at a well known source I guess).

Quote from: John G on May 06, 2009, 05:22:26 PM
Why just scale the Tone Stack why not the whole schem.............
While breadboarding I didn't have the required pots for the original values. Then I decided to down scale the TS as I saw in the Dr Boogey threads. It also helps with reducing noise I understand.

I was not aware that you could downscale more. Could all pots/gains be downscaled without affecting the overall response of the circuit? Could you point to some area' s and values that could be downscaled, and to what the added value is.

Quote from: John G on May 06, 2009, 05:22:26 PM
and are FET impedance's significantly lower than the tube schem ??
Just my observations and 2 cents worth.
John G
Thanks for thinking with me John. I do not understand what you mean with this remark. Could you elaborate?


Quote from: MikeH on May 06, 2009, 05:35:59 PM
See?  Nobody DOES read anymore  :D
Lol! And all this because I couldn' t find a suitable DPDT in Eagle ;) I didn' t want 2x3 pads as a DPDT has, b/c as said above, the bypass is easily done on the switch. I do not need those pads on the PCB. Waste of space I couldn't use.

the_stig

IMO the best way to scale the gain would be to find out how much distortion the triode stages generate. For example a 60V p-p (say at 300hz) signal enters a triode stage, the gain is 60, the output signal should be 60X60=3600V, but the plate voltage is say 300V so we have 20xlog(3600/300)=21db of distortion. Now, we should design the fet stage so that is has the same amount of distortion and do that for every stage. The first stage would be a little more tricky, I think it would need an additional 'input trim' control to make it work with both single coils and humbuckers (assume input voltage 100mV p-p for single coils).