Web based schematic capture and SPICE-like sim.

[DummyLoad]

look cool! 

https://www.circuitlab.com/

--DL

[spacelabstudio]

look cool! 

https://www.circuitlab.com/

--DL

Overall, I'm impressed.  I banged out a James tone stack as a quick demo.  A few things:

1) It works!  And was fairly easy.
2) I couldn't figure out how to draw wires that crossed without connecting.  Hence the twin 'Out' nodes, as the nodes with the same name have an implicit wire going between them.
3) The simulator doesn't allow you to vary pot rotation in the sim.  You have to run a sim, go back into build mode, click on the pot, assign 'K' a value from 0 to 1, run sim again.  I figured out through trial and error that 'K' must be the amount of rotation.
4) You can't tell it which direction is CW on the pots.  This means that on the treble pot, since it's rotated 180 degrees from the bass pot, K=0 is all the way up and K=1 is all the way down. 

I suspect if this catches on that solutions to 2-4 are probably on their todo list.

Chris

[DummyLoad]

yes you can sweep - only one variable at a time though... :-\

set:

sweep parameter

• (check box)

Parameter: Rx.K   (x is the reference designation of the POT)

Start: .1

End: 1

Step: .1

Outputs:  DB(MAG(V(out)))   (assuming you have a node labeled "out" for output)

click Run...


 

[spacelabstudio]

Weird, I tried that, but for me, the 'Parameter' field is grayed out and I can't change it.  :-(

[HotBluePlates]

That site rocks!!

I don't see a lot of value when modelling common circuits when compared to the Duncan Tonestack Calculator (because the TSC has some easier-to-use features). BUT... when modelling something that's not one of Duncan's typical circuits, it's extremely handy.

It really helped me see the effect of some circuits that would be very tedious to analyze by hand.

[Shrapnel]

Not familiar with that program, although I am a bit more familiar with LTSpice (from Linear Technology.) I had to download tube models for it though. (most PSpice models work)

[HotBluePlates]

Weird, I tried that, but for me, the 'Parameter' field is grayed out and I can't change it.  :-(

I'm sure you've already sorted this out, but in case you hadn't (or if others try):

The grayed-out field is misleading. Click inside the grayed-out box; it may not be obvious the box is selected. Type in the box, and the letters will appear. The program only recognizes capital letters for the parts reference designators.

[spacelabstudio]

Thanks HotBluesPlate!  I hadn't figured that out yet.

Now I'm stumped on this circuit:



It's likely a dumb idea, but I can't get it to simulate for some reason. 

Chris

[HotBluePlates]

It took me a few tries, but I duplicated your circuit. I did not include R11 (I made it a piece of wire).

It's simulating for me, but I suspect wrong values around the T-filter mid circuit, as the frequency range affected was too low.

I'll try to post some suggested values when I get a chance to fiddle some more.

[DummyLoad]

i had no problems. those are some funky values for a james tone-stack - mid control is more like a boost/cut switch.

i made the circuit public so you can download and edit it.  sim settings should be preserved, so all you'll have to do is change the RC values and sweep the appropriate pot.

[spacelabstudio]

Wow, awesome!  Yeah, I just wanted to get some idea whether that was a reasonable approach or not.  Based in initial sims, I'm guessing probably not.  I was hoping for a narrower notch in the middle.  The inspiration for this was basically liking the mid cut network in my Gibson Scout, but wanting to make it adjustable and add some independent bass/treble controls to go along with it.  The challenge, I guess, is finding an arrangement where the responses of the controls are fairly independent, short of just having the mid cut and the James (or something else) just separated by a buffer.  I'll keep monkeying around.

[DummyLoad]

useful for complex PS; something PSU designer has limitations with, e.g. split chains. SSR only though... :-\

i made this schematic public as well - click on schematic image to load it into your own circuitlab workspace. run time domain simulation.

change the value of the transformer by dividing Vin/Vout - click on the part and input your result in the "N" field - that's the turns ratio. for step-up it will be a fractional (<1) value.

--DL
 

[HotBluePlates]

...  I was hoping for a narrower notch in the middle.  ...  The challenge, I guess, is finding an arrangement where the responses of the controls are fairly independent, short of just having the mid cut and the James (or something else) just separated by a buffer.  ...

I promise you the James works as advertised on its own, and the T-filter works as advertised on its own. Why not separate with a buffer?

Place the T-filter between first and second gain stages. Place the James between the 2nd and 3rd gain stages. Then they'll do what you expect... Except that the T is trying to create a mid-dip; it also needs a much smaller resistance value to do that (less than 100k, maybe in the 10-50k range).

[spacelabstudio]

...  I was hoping for a narrower notch in the middle.  ...  The challenge, I guess, is finding an arrangement where the responses of the controls are fairly independent, short of just having the mid cut and the James (or something else) just separated by a buffer.  ...

I promise you the James works as advertised on its own, and the T-filter works as advertised on its own. Why not separate with a buffer?

Place the T-filter after between first and second gain stages. Place the James between the 2nd and 3rd gain stages. Then they'll do what you expect... Except that the T is trying to create a mid-dip; it also needs a much smaller resistance value to do that (less than 100k, maybe in the 10-50k range).

Sure, but what fun is that?   

[HotBluePlates]

Actually, it's a lot of fun...

I've been noodling. Keep the T-filter on its own. Look up typical values from an old Gibson or Valco amp. Use them, at least to start.

The cap to ground may or may not have a series resistor already in place. If you use a ~1M log pot wired as a variable resistor between that cap to ground (or in series with an existing resistor), you'll find the pot controls the amount of mid-dip caused by the filter. Try a decade sweep, from 1k to 1M, 5 points per decade.

If you vary the two resistors forming the top of the T, you'll find that shifts the center frequency of the notch. You're basically looking at a low-pass and a high pass filter, rolled into one circuit. Since you don't want to dink with variable caps, varying the resistors is much easier. I found that using typical values for the T-filter, and shifting the resistors from ~50k to 500k sweeps the notch point from ~100Hz up to 1kHz. Plenty of useable range for a guitar amp.

With this sim, you'll have to manually change the two resistors for each simulation.

[HotBluePlates]

After more playing, I've found you really don't want to combine the bridged-T within the James circuit. It is much better to keep them on their own.

If you think about it, the two circuit fight each other, if you roll them together. The james is trying to control bass and treble, while having indirect control of the mid. The T-filter fights that, because it's trying to form its own mid-dip, and probably not at the center value the James circuit is using.

Another way to look at it, from a tech point of view, is that by inserting the T-filter, you just screwed up all the RC values set within the James which makes it work as intended. The James circuitry is also screwing up the RC values of the Bridged-T that allow it to work as intended.

So the answer is to keep them separate.

And so you know, this has mucked the process of my current build, in a good way. The amp I'm building has a T-filter in it, apart from the regular Treble/Bass tone circuit. I thought I knew how some versions of this amp incorporated a "Contour" control. Turns out I was way wrong, because the Contour alters values in the T-filter, and I wasn't using that approach.

In the end, I'm fixing the implementation of the Contour control in my build, plus adding your idea for mid control, for a much more flexible control of midrange.

[spacelabstudio]

After more playing, I've found you really don't want to combine the bridged-T within the James circuit. It is much better to keep them on their own.

If you think about it, the two circuit fight each other, if you roll them together. The james is trying to control bass and treble, while having indirect control of the mid. The T-filter fights that, because it's trying to form its own mid-dip, and probably not at the center value the James circuit is using.

Another way to look at it, from a tech point of view, is that by inserting the T-filter, you just screwed up all the RC values set within the James which makes it work as intended. The James circuitry is also screwing up the RC values of the Bridged-T that allow it to work as intended.

So the answer is to keep them separate.

And so you know, this has mucked the process of my current build, in a good way. The amp I'm building has a T-filter in it, apart from the regular Treble/Bass tone circuit. I thought I knew how some versions of this amp incorporated a "Contour" control. Turns out I was way wrong, because the Contour alters values in the T-filter, and I wasn't using that approach.

In the end, I'm fixing the implementation of the Contour control in my build, plus adding your idea for mid control, for a much more flexible control of midrange.

Wow!  Cool.  You never know what's going to knock some neurons loose.

[HotBluePlates]

So had you gotten around to the altered sims, or maybe breadboarding the controls?

If so, what did you think?

[spacelabstudio]

So had you gotten around to the altered sims, or maybe breadboarding the controls?

If so, what did you think?

No, sorry.  I've satisfied myself that the particular idea I was trying out in that schem is not going to work out.  I haven't really spent any more time on it.  I've considered maybe having a TMB style tone controls but where the bass is a rotary switch for choosing different coupling caps, the mid is the Gibson style mid controls and the treble is a Vox-style 'Cut' know in the power section.  Or maybe I'll just decide something more traditional will work just fine.  I'll probably breadboard a few things and see if anything grabs me.

Chris

[spacelabstudio]

I finally separated them out.  I'll probably actually build this.

[HotBluePlates]

I finally separated them out.  I'll probably actually build this.

I think you're missing a cap to ground in the middle leg of the T-filter.

You might have checked already via sim, but does the treble/bass range of the T-filter match those in the James circuit?