Practical Variable input impedance for guitar: 100k to 10Meg

[stm]

Based on another thread there was a discussion about artificially loading your guitar, which is the opposite as using very high impedance buffers. I came up with the following circuit that allows experimenting with your guitar and check for yourself how it responds to different loads (input resitors).  The circuit uses standard components and allows varying the impedance seen by the guitar in a continuous manner between 100k and 10Meg.  When using a 1 Meg audio pot you will have 1Meg at the central position, which is a convenient reference value.

This is more intended as an experiment rather than as stomp box idea per se. However, I have the gut feeling that some people may find this useful and eventually discover a sweet spot.

The key of the circuit lies in bootstrapping the input resistor with a signal 0.9 times the value reaching the other end of this resistor, which effectively multiplies impedance by 10.  Gain is unity (0 dB) independently of the impedance pot position, however you will still notice a gain reduction when loading your guitar with less than 1 Meg.

The two 1 Meg resistors set the biasing of the circuit and provide an equivalent 500k resistor, which in combination with the 56k resistor provide the 0.9 gain factor for the bootstrap opamp.

Regards,

STM

[puretube]

  - yes, that looks about like what I was thinking of in the other thread... ,
if not simulating inductors, strapping the boots...

[Steben]

Ok, cool circuit idea.
Perhaps you could change the range by making that 10k resistor lower for more loading effect and the pot a 500k or so.  100k is still quite in the "high" region and difference between 500k or 1M (resulting in 5M or 10M) isn't audible (by my ears anyway ).

I'm very happy with you dis-grailing the commercial load-concepts.
But again I wouldn't need it  .

[stm]

Ok, cool circuit idea.
Perhaps you could change the range by making that 10k resistor lower for more loading effect and the pot a 500k or so.  100k is still quite in the "high" region and difference between 500k or 1M (resulting in 5M or 10M) isn't audible (by my ears anyway ).

Yes, maybe a 500k audio pot with a 4.7k or 5.1k seres resistor will shift the impedance range to 50k to 5 Meg, with 500k at the center, OR, you may just replace the 56k resistor with 120k and you will have now only 5x impedance boost.

I'm very happy with you dis-grailing the commercial load-concepts.
But again I wouldn't need it  .

I didn't want to dis-grail the product, just tried to understand the real need and purpose of it. My impression seems that despite it *may* add to your tone under certain particular conditions, the product is advertised with the wrong arguments 

Bottom line here is you must try the thing for yourself and see if it works for you before buying. Blindly (or perhaps deafly) believing in advertising is really dangerous for your wallet!

[gez]

Nice idea as it avoids large caps normally associated with bootstrapping. 

I appreciate you'd want to divide down the signal ever so slightly to prevent oscillation, hence the 56k (I presume that's why it's there), but do you really need C2?  Seems a little superfluous. 

[Steben]

I didn't want to dis-grail the product, just tried to understand the real need and purpose of it.

Yes, but you did dis-grail it by posting this, that's all. It's not a bad thing as side-effect.

My impression seems that despite it *may* add to your tone under certain particular conditions, the product is advertised with the wrong arguments  Bottom line here is you must try the thing for yourself and see if it works for you before buying. Blindly (or perhaps deafly) believing in advertising is really dangerous for your wallet!

Ok, I completely agree with all this. Like I said I was a bit hysteric but wanted to say something like you.

[R.G.]

Good idea. Some people might indeed find a sweet spot for their particular pickups and playing technique.

Bootstrapping is a quite old technology, from the days when you simply couldn't get high enough input impedance devices. It's been fairly well plowed. The various stages of the Univibe, for instance, are all bootstrapped for higher input impedance. Bootstrapping any follower is simple - you capacitively couple the output back to the input through a large value resistor. As long as the input voltage has a lower impedance to work through than the large value resistor, it does not oscillate.

In this case, you don't really need the second opamp. The output of the first opamp is quite low enough to boostrap its own input up to many megohms. Then you can just adjust load with a simple loading resistor on the input. So for this circuit, I'd put the loading pot outside the input capacitor. Then I'd change the 10K to 100K, and the two biasing resistors to 220K. I'd connect the now-100K input bias resistor to the junction of the 220Ks and then connect a 0.1uf film cap from the output of the opamp to the junction of the two 220K's and the 100k bias resistor. Now the opamp is bootstrapped to within an eyelash of whatever it can do, and the loading is all set by that variable 1M resistor load on the outside of the input cap.

The size of the bootstrapping cap is determined by the frequency range you want in the output. For 220K/220K/100K, the 0.1uF film cap gives a bootstrapping rolloff of 31 Hz. It's possible to raise the values of these resistors and use smaller caps as well.

It's not widely known, but you can also increase the input impedance of an INVERTING stage by either some feedback tricks or a variant of bootstrapping with a second opamp.

[stm]

Nice idea as it avoids large caps normally associated with bootstrapping. 

I appreciate you'd want to divide down the signal ever so slightly to prevent oscillation, hence the 56k (I presume that's why it's there), but do you really need C2?  Seems a little superfluous. 

The 56k is not there to prevent oscillation. As explained before, its purpose is to set an exact attenuation factor whose inverse plus one multiplies the input impedance.

I rechecked the schematic and you are absolutely right: C2 is superfluous and should be removed.
This is an updated schematic with a better impedance range.

[Gus]

It is good to see people experimenting with this.  For some of my petals I change the input R and the cap for a different Highpass freq and load.  Some distortions sound better at different loading and highpass turnover.

I did about 3 months of testing a few years ago with input loading of effects for different effect types.  You can see hints of it in my schematics.  You also need to think about the volume of sound an effect will often sound good at low vol and not so good at high vol or just the opposite.  FWIW a friend of mine has a few TS that I worked on for Live or home vol.

IMO most of the change is from 100K to 1meg however a SHO or SD are interesting.

I would guess the circuit might be a SF buffer or even a bootstrapped BJT EF.

[zachary vex]

excellent experiment, stm.  bravo. 

you might change the input resistance range so it's variable down to 10k rather than 100k, as was mentioned before, sheerly for the purposes of experimenting with low impedance inputs and what they do with your particular pickups.  the rangemaster, for example, has particularly low input impedance (below 100k) which helps shape the frequency response of that "treble" booster.  8^)

have at it!

[MartyMart]

Nice work stm, very interesting idea !
( I'm an official "stm" fan now )

Marty.

[Phorhas]

Yeah - that's a cool tweek. I once built a simpler input impedance thingy with just a JFET buffer with no cap and a 2M resitor to 1/2V in series with a 100k or so. and it did help to find the sweet spots of differant pickups quickly and easily.

I really recommend incorporating a veriable input impedance option if you can.

[CS Jones]

Yes, thank you. I hope this thing gets more airplay. I missed it the first go around. Seems like a great tool to have around.

[bwanasonic]

I'm very happy with you dis-grailing the commercial load-concepts.
But again I wouldn't need it  .

I still don't understand your problem with the *commercial* product. You have no idea what topology the circuit is, and you claim you don't need it, yet you seem to have an issue with Radial's product. There are no claims of *holy-grail* for the Dragster circuitry, just a practical solution to an issue for at least some wireless users. Radial is known to make some of the best direct boxes available, endorsed by DIY icon Craig Anderton. You might want to wait until you know more about a product, and how well it functions before you take issue with it.

Kerry M

[no one ever]

i feel uneducated compared to the great minds that have posted before me, but still i must ask if i plan to learn... where is the input on the schematic?   

[A.S.P.]

V1 (~) "is" the P.U.

[Steben]

I'm very happy with you dis-grailing the commercial load-concepts.
But again I wouldn't need it  .

I still don't understand your problem with the *commercial* product. You have no idea what topology the circuit is, and you claim you don't need it, yet you seem to have an issue with Radial's product. There are no claims of *holy-grail* for the Dragster circuitry, just a practical solution to an issue for at least some wireless users. Radial is known to make some of the best direct boxes available, endorsed by DIY icon Craig Anderton. You might want to wait until you know more about a product, and how well it functions before you take issue with it.

Kerry M

I guess I got all started with the (JRC)4558 hype ... sigh. Maybe my pills aren't working anymore?

stm: It's very much appreciated you changed the range to 50k-5M. 

[bufferz]

Not sure why this thread ended without addressing some of the possible adjustments mentioned. I apologize if this post is a zombie, it is not intended in distaste.

I am a newb but I went ahead and tried to draw out what R.G was suggesting. I suspect I did it wrong, so please let me know how to ammedn it:



Also, i was thinking since we are only using one side of the opamp maybe the other half could be used as in the same manner as a variable output impeadence in order to allow for placement before a fuzz pedal....does this make sense.

My last though for now is, if we add a pot to the feedback loop could we add a boost to this rather than unity gain? I suspect it would not be that simple though

[PRR]

> I suspect I did it wrong

Both inputs of opamp *must* have a DC path to ground/bias.

You have no DC path to "+" input, which means it can "float" to "any voltage", always a voltage which causes crap-out. (We had a case of that here recently: circuit would run for a minute then quit, it floated very slowly to a no-good point.)