I read this interesting interview with Bob Gallien and he talking about the "5-string" switch on newer amps that alters the longstanding GK EQ tone stack to lower the roll-off frequency of the bass from 40-ish to 30-ish Hz.
Do any of you have any ideas how this might have been accomplished in changing a resistor/capacitor or two in the GK tone stack, found in the attached schematic? Would love to mod mine to this effect if I can. Does anyone know of an online tone stack calculator that I could punch this into ? The TSC online works for Fender / Marshall amps, but how about this type? Any feedback welcome.
Do any of you have any ideas how this might have been accomplished in changing a resistor/capacitor or two in the GK tone stack, found in the attached schematic? Would love to mod mine to this effect if I can. Does anyone know of an online tone stack calculator that I could punch this into ? The TSC online works for Fender / Marshall amps, but how about this type? Any feedback welcome.
...or is it maybe that the Bass roll-off switch involves a high-pass filter at a certain frequency by default, and then a higher frequency with the switch engaged? Is that what's going on at the front end?
If someone could tell me exactly what kind of circuit that EQ is, that would be very helpful. Using the Texas Instruments Filter Design tool I can reconstruct something pretty close.
If someone could tell me exactly what kind of circuit that EQ is, that would be very helpful. Using the Texas Instruments Filter Design tool I can reconstruct something pretty close.
The R45 stage has the biggest caps and is the low-bass stage.
Change the caps in proportion to frequency.
My lookup says we shift from 41Hz to 31Hz. That would be from 0.1uFd to 0.132uFd. 0.12 is close, two 0.068 parallel is well within a semitone and also within typical part-tolerance on a mild-slope filter like this.
Change the caps in proportion to frequency.
My lookup says we shift from 41Hz to 31Hz. That would be from 0.1uFd to 0.132uFd. 0.12 is close, two 0.068 parallel is well within a semitone and also within typical part-tolerance on a mild-slope filter like this.
@PRR Wow, thank you so much. That's fantastic. Would you be able to share the calculator that you used for this or one I could look up?
I was looking at this calc for another bandpass filter and thought about changing the R44 + R46 resistors, but I like your solution to get bang on 30hz.
https://www.changpuak.ch/electronics/Narrow_Bandpass_2.php
Why is it that it has the two 3.3k and two 220k resistors on this circuit instead of one? Would like to learn how this works.
I was looking at this calc for another bandpass filter and thought about changing the R44 + R46 resistors, but I like your solution to get bang on 30hz.
https://www.changpuak.ch/electronics/Narrow_Bandpass_2.php
Why is it that it has the two 3.3k and two 220k resistors on this circuit instead of one? Would like to learn how this works.
I am confused about the Baxandall filter, however, because I am not seeing any roll off of low frequencies. It looks like a low shelf on the low end.
LTSpice is a good tool for that sort of analysis. I'd suggest you start with a newer schematic than something from 1981 though! 😉
@Passinwind My amp is from 1984 so this is actually the right schematic - I love that it's hand-drawn. Later models are surface mount, which I would not have attempted to mod. As it stands, I can alligator-clip alternative caps in, which is great.
Thanks very much for the LTspice tip - looks like that would involve some learning curve to get an answer, is there a library of pre-made circuits, where I could open a Baxandall filter?
The approach mentioned above looks very successful. It is just a matter of tuning it with the right capacitance. I can drop caps in in parallel, which is easier than replacing them. The bass EQ center frequency - notwithstanding Gallien's comment about roll-of below 40hz - is 60hz as labeled on the front panel.
I tried out some other caps and looked at the scope; So in stock form, 30hz is maybe 9db below 60hz. By doubling the capacitance to 0.2uf, it brings 30hz within about 1db of 60hz, at the expense of response around 120-250hz, which is exactly what I want - to reduce the second harmonic. This is awesome as far as I'm concerned. Now if only I can make a speaker cabinet that really produces 30hz...
I'm going to figure out where to put a "deep" switch on the chassis. Would be perfect if I can re-purpose the low-cut switch. Cutting lows is something I would never do.
Good times.
Thanks very much for the LTspice tip - looks like that would involve some learning curve to get an answer, is there a library of pre-made circuits, where I could open a Baxandall filter?
The approach mentioned above looks very successful. It is just a matter of tuning it with the right capacitance. I can drop caps in in parallel, which is easier than replacing them. The bass EQ center frequency - notwithstanding Gallien's comment about roll-of below 40hz - is 60hz as labeled on the front panel.
I tried out some other caps and looked at the scope; So in stock form, 30hz is maybe 9db below 60hz. By doubling the capacitance to 0.2uf, it brings 30hz within about 1db of 60hz, at the expense of response around 120-250hz, which is exactly what I want - to reduce the second harmonic. This is awesome as far as I'm concerned. Now if only I can make a speaker cabinet that really produces 30hz...
I'm going to figure out where to put a "deep" switch on the chassis. Would be perfect if I can re-purpose the low-cut switch. Cutting lows is something I would never do.
Good times.
Just out of interest I looked up how GK does this on the newer (1001RB) amps. The tone stack is a little different from mine although I think it is to the same net effect as a series equalizer at the same points. Not sure how I would do this in my circuit. This may be of interest to future travelers who might want to modify older 400RB/800RB amps.
I was just suggesting that to reverse engineer something, then working off the schematic for that something is a pretty good place to start, assuming it's avaialable. ;^}
You can find many LTSpice example files here: https://groups.io/g/LTspice . There is definitely a learning curve involved, but IMO it's well worth the effort.
Once you get your Spice models for your existing circuit and your target one up and running to get a good solution you ideally need to look at the big picture of cumulative rolloffs throughout the whole preamp and preferably the power amp section as well. For instance, the section you highlighted with the red arrow does something like this when modeled and just considered in a vacuum:
Green is with R6 and C7 grounded, blue is with ground connection lifted. We can can pretty much count on there being a reciprocal curve generated somewhere downstream, yielding a classic treble pre-emphasis/de-emphasis response format, which can enhance signal/noise ratio. I didn't model past the contour circuit...that's your mission if you choose to accept it!
Wow, that's great. I will take a look at the Spice models. I think the best way to learn is just mucking around with existing stuff. I don't think the GK tone stack is a straight Baxandall because the low is not a shelf (?)
Pardon the noob attempt to get this, but it looks like the 5-string switch on the amp in Post #11 puts C29 (.1uf) in parallel with C40 (.1uf) yielding a similar kind of effect as discussed in #3, which is what I ultimately did ??
But my circuit is different from later 400RB amps and this 1001RB in post #11. Do you think these two solutions are doing the same thing, or is there a better way to do it on my amp?
I like the effect implemented so far. But I assume Bob Gallien did it in the best possible way.
@bremen nacht Okay thanks. Is there such a high pass filter on my schematic, in post #2?
That's what I mistakenly thought might be R6 and C7 as discussed above.
That's what I mistakenly thought might be R6 and C7 as discussed above.
PM me when you get up and running with LTSpice and I'll see if I can scare up a few starter files related to bass applications. You can also open up test fixture examples for many components right from within LTS which will give you some hints on command syntax. I know there's a complete SVT preamp model out there, and complete Fender and Mesa tube guitar amps, not sure how readily available whole bass amps are though.
Will do - thanks man, appreciate it. I think you're sending me down a rabbit hole I will occupy for a while 🤣
Quick google and I see that the model of a Fender ab763 available through Robinette's site. I have a Super Reverb I've been learning to work on the past couple years, so this is handy. I guess I'm underway. So this simulates the frequency response of the circuit, eh?
Meanwhile on this GK, I'm looking for what would be the highpass filter on mine - if there is one - like a 33k resistor and .1uf cap or 150/.022? Seems weird there would be a highpass filter on a bass amp, but maybe it was to protect against speaker pops, thinking the amp would never have to see anything below 40hz.. And then this switch would defeat it.
Modern bass amps are using more involved high pass filters these days pretty often, often variable 4th order ones with steep cutoff slopes. There's alot of junk under 40Hz in a bass guitar signal, including enharmonic transients on the attack of the note. This creates a lot of cone movement that just wastes power and wears out the soft parts of the speaker, and downtuning and/or 5 string basses just make that worse.
And yeah, LTSpice can simulate tone control sweeps showing frequency and amplitude, transient response, harmonic distortion content, signal/noise, and even allow you to play wav or mp3 files through your model. So you can record your own bass played with your own technique and get at least some rudimentary idea of how things might work for you.
I'll dig into your specific request over the weekend if I can make a little time. I have to get ready for my regular Friday night jam and need to prepare a few DIY widgets for the upright player over there to try out.
And yeah, LTSpice can simulate tone control sweeps showing frequency and amplitude, transient response, harmonic distortion content, signal/noise, and even allow you to play wav or mp3 files through your model. So you can record your own bass played with your own technique and get at least some rudimentary idea of how things might work for you.
I'll dig into your specific request over the weekend if I can make a little time. I have to get ready for my regular Friday night jam and need to prepare a few DIY widgets for the upright player over there to try out.