Browser-friendly Tone Stack Calculator

[CharvelBlue]

Would it be possible to add the Fender Pro Junior tone stack?

Yes, I can add that stack. If nothing unexpected comes up, I will have it ready at the end of this week (latest). I will let you know when it is ready. RL refers to the input resistance of the following stage, which acts as a load to the tone stack. It will need to exist in the Pro Jr stack model as well.

That's great! Thank you very much! Curious which program you use to do the calculations - is it LT Spice or something else?

[jatalahd]

That's great! Thank you very much! Curious which program you use to do the calculations - is it LT Spice or something else?

The Fender Pro Junior tone stack model is now available at:  https://www.guitarscience.net/tsc/projunior.htm

It can be accessed also from the TSC on the web top menu, by expanding the Fender drop down menu.

For calculations, I write the circuit equations using Kirchhoff's rules and solve them symbolically using Maxima computer algebra system. Then the resulting transfer function is used in the simulation. But I always verify the transfer curves against a SPICE model, to prevent human error in writing the initial equations.

[CharvelBlue]

That's fantastic! Thanks jatalahd! A few surprises for me seeing the curve. Its about 15dB lower signal level than the 5F2A. Also its pretty flat between 80Hz to 10kHz which seems to contradict the very bass shy, middy and trebly sound of my PJ, although there's an odd steeply rising treble past 10kHz due to C2 I think and 13dB higher at 100kHz! Have no idea why Fender would do that. I was expecting to see it look more like the 5F2A curve. Also wondering how you arrive at RL of 1M for the PJ - does this figure need to be calculated? Thanks again - great job!

[jatalahd]

That's fantastic! Thanks jatalahd! A few surprises for me seeing the curve. Its about 15dB lower signal level than the 5F2A. Also its pretty flat between 80Hz to 10kHz which seems to contradict the very bass shy, middy and trebly sound of my PJ, although there's an odd steeply rising treble past 10kHz due to C2 I think and 13dB higher at 100kHz! Have no idea why Fender would do that. I was expecting to see it look more like the 5F2A curve. Also wondering how you arrive at RL of 1M for the PJ - does this figure need to be calculated? Thanks again - great job!

Yes, the curve does seem a bit weird, but SPICE gives the same curves for the circuit ... I have a gut feeling that there might be errors in the (original?) schematic, because who would put a 470k resistor in series with the signal path to attenuate the signal so much. Also 22pF cap? That causes the boost in > 10 kHz, but why? If 220p is used, the curve looks more sane to me. Maybe the Pro Junior refers to a Junior circuit designer ...

The RL is the input impedance of the following stage. I just threw that 1M value out of the hat since tube stages typically have high input impedance. If this tone control is to be used in other context, the loading effects of the following stage can be modeled by adjusting the value of RL.

[CharvelBlue]

Yes it makes you question the designer alright. I've also just finished modelling this in Spice myself (my first attempt at Spice - had to use 2 fixed resistors instead of pots because I haven't worked out how to use the pot models I downloaded yet...) and it checks out with yours. I've checked all the PJ circuit diagrams from 1993 to 2005 and they all show the same values for C2 and R2 although later models have R1 reduced from 56k to 22k which reduces gain a bit more. So it doesn't seem to be an error. Are you sure about changing C2 to 220p? For me that boosts everything from 1kHz instead of 10kHz. Using 0.5 to 1pF flattens the curve for me. Thanks for the info re RL too.

[jatalahd]

My guess for the value of C2 was just an assumption, because it is easy to accidentally leave out one 0. And boosting from 1 kHz to 10 kHz would be audible in the guitar sound opposed to a boost from 10 kHz upwards. But maybe they did not want to make the boost audible, so then it is fine as it is.

I was just thinking if someone would want to make own adaptations of this tone stack, then in my opinion the 470k resistor and the 22pF cap could be changed to more reasonable values.

[PRR]

...Its about 15dB lower signal level than the 5F2A....

It's not a 5F2A. It has an extra stage, no?

The level diagram is more like many Blackface Fenders. One gain stage. A high-loss BMT tonestack -and- volume control. Then a gain stage, then a driver stage.

Look at the level diagram. The mV numbers, probably recorded by someone not looking at the theory, again suggest a big loss. A huge loss! Level at grid of 2nd stage is -less- than signal at the input jack! Also the only place it is likely to overload is at the EL84s. The input overload is near 0.6V.... you "can" get there with a guitar and a strong strumming arm, but it is hard work. It may be a careful balance between dead-clean on naked axe and easy-overload with slight boost (a transistor LPB will give a true tube grid overload).

[CharvelBlue]

Interesting about the mV numbers. I just removed that C2, R1, R2 attenuation circuit in Spice and the gain goes up 20dB or so. Its a few dB down at 80Hz but increasing C1 to 0.05uF brings it up to acceptable levels. Can you see any problems with having an extra 20dB of gain?

[CharvelBlue]

Here's the frequency curve:

[ElectricDruid]

Can you see any problems with having an extra 20dB of gain?

It'll crunch. Whether that's a problem or not is a question of taste, I guess. It may have been what the designer was trying to avoid.

[PRR]

> Can you see any problems with having an extra 20dB of gain?

For any non-insane playing, you would have to turn a "0" to "10" volume knob down to "1/2". Which is a very twitchy point.

Get your overall gain diagram right before too much work on tone-shaping. This is a low-loss TS adapted to an amp designed around a hi-loss TS. Sometimes commercial design is imponderable.

[CharvelBlue]

^^^ Yes I was thinking the same thing. Some have done this mod and are happy with it but the std amp has enough crunch for my tastes.

[ThermionicScott]

That's fantastic! Thanks jatalahd! A few surprises for me seeing the curve. Its about 15dB lower signal level than the 5F2A. Also its pretty flat between 80Hz to 10kHz which seems to contradict the very bass shy, middy and trebly sound of my PJ, although there's an odd steeply rising treble past 10kHz due to C2 I think and 13dB higher at 100kHz! Have no idea why Fender would do that.

...

Yes it makes you question the designer alright. I've also just finished modelling this in Spice myself (my first attempt at Spice - had to use 2 fixed resistors instead of pots because I haven't worked out how to use the pot models I downloaded yet...) and it checks out with yours. I've checked all the PJ circuit diagrams from 1993 to 2005 and they all show the same values for C2 and R2 although later models have R1 reduced from 56k to 22k which reduces gain a bit more. So it doesn't seem to be an error.

I suspect you may be overthinking things a bit, and that there wasn't a plan to be super-clever with the design.

When I look at the Pro Jr schematic, I see a pretty vanilla 2xEL84 amp driven by 12AX7s.  Since it's a one-channel amp, they could have just used one 12AX7 triode at the input, connected that to the LTP phase inverter through the volume/tone control circuit, and just left one triode unused.  Perhaps that wasn't quite enough gain, or they just wanted the sound of running the signal through more stage, or they just would have felt guilty if they didn't use that triode, so we have two gain stages before the LTP.  But EL84s don't need a lot of input signal, so it would be way too much gain if they didn't do something.  So they whacked in a voltage divider between the coupling capacitor and volume control.  But now there's a lot of series resistance between the first two stages.  Hmm, let's add a treble-bleeding cap across that 470k resistor to make sure it doesn't sound too dull.  Yes, at this point they could have pulled out the slide rules or ran Spice simulations to calculate the cap value, but I imagine they just did the time-honored process of trying different caps until it sounded right, and 22pF was enough to do the job.  Done!   

[j_flanders]

Marshall has a cathode follower before the tone stack.
Cathode resistor is 100k.
RIN in the sim app is 1,3k

Vox has a cathode follower before the tone stack.
Cathode resistor is 56k.
RIN in the sim app is 56k

Why so? Should also be around 1k, not?
Same thing in the original Duncan TSC.

[CharvelBlue]

That's fantastic! Thanks jatalahd! A few surprises for me seeing the curve. Its about 15dB lower signal level than the 5F2A. Also its pretty flat between 80Hz to 10kHz which seems to contradict the very bass shy, middy and trebly sound of my PJ, although there's an odd steeply rising treble past 10kHz due to C2 I think and 13dB higher at 100kHz! Have no idea why Fender would do that.

...

Yes it makes you question the designer alright. I've also just finished modelling this in Spice myself (my first attempt at Spice - had to use 2 fixed resistors instead of pots because I haven't worked out how to use the pot models I downloaded yet...) and it checks out with yours. I've checked all the PJ circuit diagrams from 1993 to 2005 and they all show the same values for C2 and R2 although later models have R1 reduced from 56k to 22k which reduces gain a bit more. So it doesn't seem to be an error.

I suspect you may be overthinking things a bit, and that there wasn't a plan to be super-clever with the design.

When I look at the Pro Jr schematic, I see a pretty vanilla 2xEL84 amp driven by 12AX7s.  Since it's a one-channel amp, they could have just used one 12AX7 triode at the input, connected that to the LTP phase inverter through the volume/tone control circuit, and just left one triode unused.  Perhaps that wasn't quite enough gain, or they just wanted the sound of running the signal through more stage, or they just would have felt guilty if they didn't use that triode, so we have two gain stages before the LTP.  But EL84s don't need a lot of input signal, so it would be way too much gain if they didn't do something.  So they whacked in a voltage divider between the coupling capacitor and volume control.  But now there's a lot of series resistance between the first two stages.  Hmm, let's add a treble-bleeding cap across that 470k resistor to make sure it doesn't sound too dull.  Yes, at this point they could have pulled out the slide rules or ran Spice simulations to calculate the cap value, but I imagine they just did the time-honored process of trying different caps until it sounded right, and 22pF was enough to do the job.  Done!   

Ok so maybe that little 2dB kick up at 1200Hz from that 22pF cap sounded so darn good that it didn’t matter what it was doing at 100kHz. Thanks. Good to know how these decisions might be made.

[ThermionicScott]

Ok so maybe that little 2dB kick up at 1200Hz from that 22pF cap sounded so darn good that it didn't matter what it was doing at 100kHz. Thanks. Good to know how these decisions might be made.

In the real world, any content at 100kHz doesn't (shouldn't) make it very far in the amp because the Miller capacitance of the tubes interacts with any series grid resistance to form low-pass filters that roll it off.  That's mostly what grid-stopper resistors are for -- to suppress high-frequency oscillation by filtering it away.

So if your guitar isn't outputting 100kHz, and the tubes aren't picking up 100kHz noise from the air or oscillating at those frequencies, it doesn't really matter what a simulation might show in that range. 

[PRR]

...Same thing in the original Duncan TSC.

Duncan is considering an update.

Meanwhile, of course, users can just manually edit the source value. It's not clear what the "right" value is, 717 or 612 Ohms, but anything 500r to 1k gives pretty-near the same result so it is not fussy.

[duncanamps2]

TSC now updated to use 717 ohms for Vox source impedance, thanks for the heads up!

[mdcmdcmdc]

Just noticed that the SWTC variants have been added - tysm!