Tone Bender Mk II - Impedance sanity check question

[iainpunk]

you can add a Jfet buffer to the input, and use a 8k resistor between the first gain stage and the buffer to not make it sound super harsh.

get well and cheers

[Electric Warrior]

Due to the biasing from leakage current, this is a design that is only suitable for room temperature or a narrow temperature range, even more so than the Fuzz Face from which it is derived.

Not at all. The MK1.5/Fuzz Face quickly farts out at higher temperatures, especially when biased as hot as many vintage units.
The MKII fixes this issue. I think the additional gain stage keeps the signal over the threshold for longer. When it starts gating, it's doing it in a nice way. These can still sound good at over 30°C.

In my limited experience, cleanup, like that of a Fuzz Face, is not really possible in this circuit. What you get is varying levels of dirt but never the "cleans" of a Fuzz Face rolled down.

Totally depends on your transistor selction or bias setup. They can clean up well. If yours doesn't, try less leakage on Q2 or tweak its collector resistor a little. It can be used to dial in the sweet spot between gating and clean up.

[Big Monk]

Due to the biasing from leakage current, this is a design that is only suitable for room temperature or a narrow temperature range, even more so than the Fuzz Face from which it is derived.

Not at all. The MK1.5/Fuzz Face quickly farts out at higher temperatures, especially when biased as hot as many vintage units.
The MKII fixes this issue. I think the additional gain stage keeps the signal over the threshold for longer. When it starts gating, it's doing it in a nice way. These can still sound good at over 30°C.

In my limited experience, cleanup, like that of a Fuzz Face, is not really possible in this circuit. What you get is varying levels of dirt but never the "cleans" of a Fuzz Face rolled down.

Totally depends on your transistor selction or bias setup. They can clean up well. If yours doesn't, try less leakage on Q2 or tweak its collector resistor a little. It can be used to dial in the sweet spot between gating and clean up.

I'll have t9 check my voltages on Q2. I've got some OC75s coming in the mail and I'll be testing those prior to boxing up.

[pacealot]

(Edited to note that most of what I wrote below is off-topic, as the question was about the effect on input impedance, and I didn't address that at all...)

My completely non-EE, technically-ignorant, thoroughly subjective experience with several of these circuits has been that increasing the value of the Q1B resistor to ground from 10K to 100K has resulted fairly consistently in increased gain/sustain/saturation of the fuzz, but with no specific gating or other misbiasing artefacts. This has worked to my benefit thus far, as my main guitar is a Rickenbacker which benefits immensely from as much additional sustain enhancement as it can get. 

It does mean, however, that the slightly less blown-out fuzz at the minimum attack settings is still pretty well blown-out. I agree with Derek that pre-gain is a more effective method for controlling the over-the-topness of the thing than rolling off the guitar's volume, but I generally use mine for the full whammy and rely on other fuzzes for the more in-betweeny gain stages (or else bypass Q1 entirely and use it as a modified FF/Vox TB-type circuit, but that's a whole other kettle of tarantulas).

But I've only built or tweaked these circuits with either OC75s or 2N281s, so I can't predict what the results would be working with actual OC81Ds. Just my two cents....

And also get well soon!

[Rob Strand]

LT spice will work.   You just plot Vin/Iin for an AC analysis.     The impedance is the value at small signals where the circuit isn't clipping.   Sometimes it's convenient to add a small resistor in series with the input say 1 ohm or 1 milli ohm (1m, not 1MEG) so you can type Vin/(R10) to get the current Iin, where R10 is the added resistor.

Aside from a simulator, what else is a good way to solidly measure impedance, as in, I have the circuit board in front of me and I have a soldering iron and breadboard?

There's some easy to understand methods but I wouldn't say they are solid.    There's so many things that can stuff-up the measurement.

So one method is to *add* resistor in series with the input.

You can set-up experiments in LTspice to check the answers and the limitations of the method.

Start with this voltage divider and say a DC source,   


R2 is the "unknown" input impedance.  R1 is the added resistor.   Vin is a convenient source voltage.  Vout is the output of the divider.   Be careful in that Vout for the divider is actually the input voltage for the circuit.

If you measure Vin then adjust Vout until you get Vout = Vin / 2 then you know your R1 value.  So you measure R1 and that's the answer.

In order to not have to adjust R1 by hand you just measure Vin and Vout and use calculations,

R2  = R1 / ( (Vin/Vout) - 1)

You can see when Vin/Vout = 2  you get R2 = R1 which is a special case of the hand adjustment method.

For accuracy, it's best to have Vout less than Vin / 2.   In fact in some cases it is better to choose a large R1 so Vout is Vin/10.


The above method works for AC signals.  All the same calculations.  The difference is you need to choose a test frequency.

So here's where the limitations come out:

Suppose the input impedance of the circuit looks like a capacitor.



Here R is the added R1 resistor and the unknown impedance magnitude |Z2| is the impedance of the capacitor is |Z2| = 1/(2*pi*f*C).   The cap has a phase shift and it messes with things a bit.  The method of adjusting R1 to get Vout = Vin/2 doesn't work, since when R1 = |Z2|   Vout is actually Vin / sqrt(2) not Vin/2.     You will get a ball-park answer which is often fine.  In this case making R1 larger so Vout is say Vin / 10 helps.  It makes the calculations work when the unknown impedance |Z2| is not resistive.

A third simple scheme is to follow the method you use in LTspice.  You need a larger R1 than LTspice to get a good measurement.   Measure voltage across R1 then calculate the input current Iin = "V across R1" / R1.   Then measure Vout and calculate |Z2| from |Z2| = Vout / Iin.    That seems easy enough but connecting across R1 has not grounds and can inject a lot of noise.   Also when you measure "V across R1" and Vout the meter input resistance (1M or 10M depending on you meter) upsets the measurements in a way that is hard to compensate for with simple calculations.

In the previous tests when the unknown input impedance R2 is large and in the same order as the meter input resistance we can correct our answers by removing the effect of the meter using the parallel resistance formula    R2_actual = 1/ (1/R2_raw - 1/Rmeter).   The correction isn't quite accurate for a capacitive input impedance.

As a side note, if you have a circuit with a very small input capacitor.   The input impedance will look like the capacitor in series with the input resistance of the following circuit.   If the cap is very small the input impedance can look more like a capacitor.     The other thing to realize is in cases like this the input impedance is varying with frequency.  Sometime you might want to know the impedance at difference frequencies, or find the frequency where the input impedance is least - say on a Baxandall bass/treble control with both bass and treble boosted.

Thanks for the well-wishes!

Yeah, good luck.

[PRR]

If you think you want to "measure" input impedance, note there is a linear (undistorted) impedance which transitions to a distorted impedance.

If you just put big signal in and do not listen/watch the output, 9 out of 10 you will be in the distorted zone. Which is not only different than clean, but different everywhere in the cycle, and usually varies with frequency.

Music will almost always decay to the "clean" (or clean-ish) zone, so you do want to know this. "Distortion" etc effects will also run into the unclean zone.

It's not a simple single answer.