Tone Filter sim problems

[mth5044]

Hola,

I'm trying to sim the 'Sharpness/Body' control from from Skreddy's Lunar Module as a basic pre-EQ tone shaper for drives and fuzzes -

http://4.bp.blogspot.com/-jtoWoktqNs4/T0AQyiJsOUI/AAAAAAAAA8A/U5vbjO3-zig/s1600/lm_scheme.gif

I don't have BC109 in LTSpice, so I used the generic jelly bean 2N3904. Figured since the gain works, it wouldn't have an impact on the tone control.

I tried simming just up through the C4/R8 node, tried it with R9/C7, tried adding on the buffer stage at the end as in the picture in case it needed some sort of load on the end, still getting the 1db or so different in level while R8 (the Sharpness pot in the schematic) sweeps across 100k. Is there something obvious I'm missing here? The spectrum displayed is from probing the output of C9, but it looks similar at any point after the C4/R8 node.



Thanks for your help.

[ElectricDruid]

What happens if you limit the sim to just the bit you're interested in? (C4,C5, C6, R8, R10)

This doesn't make a complete circuit, but for the sim it doesn't need to. We don't have to worry about input/output buffers and impedances and so on. But if that worked on its own (or even if it didn't) it'd narrow the problem down.
I'd reduce it to those five components with the input and output and see what happens then. Once that's going, you can add buffers/gain stages back in.

HTH,
Tom

PS: Totally agree with your tranny swap. 2N3904 for BC109 is almost certainly not what's going on.

[mth5044]

Thanks Tom!

I've given that a shot and pasted the results below. R10 is supposed to a log pot, so I lessened the intervals that the simulator graphs the various resistances, and it looks like it's responding logarithmically with the spread out lines as the treble gets cut.

BUT, according to Skreddy's user manual, this control is supposed alter the bass frequencies going into the fuzz portion of the circuit.  Not quite how the simulator is portraying that. The chart is graphed from the C4/R8 junction. So strange.

[ElectricDruid]

Well, that looks like something's happening, so it's an improvement, I guess!

Looking at that little bunch for a minute, I was trying to work out how exactly that would control bass. C4/R8 make a variable highpass filter, so that would make sense. Reducing R8 would move the highpass higher and cut more bass. 5n6/100K gives 284Hz as the lowest, so there's always a little bit of bass cut. But then on top of that there's all the interactions with the other stuff.
This would make sense if we were seeing some highpass effect, but we're not. And I'm not good enough at untangling RC networks to figure out what is going on in that mess.

What happened to C16 from the original schematic, btw? Is that for fine-tuning the C4 value?

[merlinb]

You need to include the loading effect of the ring-of-two transistors following the tone control. This has an input impedance of around 100 ohms, unlike your SPICE circuit.

[ElectricDruid]

Nice work, Merlin!

Can you explain to me what's going on there? I mean, I can see from your sim that adding the 100R load changes the response, but I don't understand why.

Ta!

[merlinb]

I can see from your sim that adding the 100R load changes the response, but I don't understand why.

Not exactly sure what you're asking? I know you know that filters are sensitive to loading...
If you squint, you're basically looking at a pair of coupling caps with a 'blend' pot between them. By making the load resistance smaller you are making the time constants smaller, meaning the cut-off frequency rises to where the blend pot actually has some effect. it either puts them both in parallel or it (more or less) removes C2 from circuit.

[mth5044]

You need to include the loading effect of the ring-of-two transistors following the tone control. This has an input impedance of around 100 ohms, unlike your SPICE circuit.

Awesome! Thank you for making that clear. Looks like a pretty lossy filter!

What happened to C16 from the original schematic, btw? Is that for fine-tuning the C4 value?

I've read the lunar module was built off of the Screwdriver and used the same PCB, at least initially. There are a couple other components on the original schematic that say omit or have no value, perhaps this scheme was built on the old screwdriver schematic and was still in transition? In any case, I've built the pedal based on this schematic a few years ago and it's a fabulous circuit. Hasn't left my board since take off.

Thanks guys!