raising input impedance of a DS-1

[kissack101]

Ok, so i'm still learning, but I was wondering how to raise the input impedance of a DS-1? I've heard the impedance of the SHO for example makes your guitar sound amazing and I was hoping to replicate something similar in my lowly DS-1.



Would this work? Simply adding a resistor to ground in parallel with the rest of the circuit? Would the input impedance now be ((1/r0)+(1/[original impedance of circuit))^ -1? If so, could you make the impedance as high as you like? Or does it have consequences given that the pedal isn't true bypass?

Any help would be apprecaited.

Adam.

[Roberto]

The DS-1 has an input impedance of 470K (R2 value) and, in most cases, this high enough but  you can try changing R2 from 470K to 680K if you want. I think that higher values, as 1M, will work changing Q1 for another with higher beta. By the way, I'm almost sure that this don't lead to any perceptible tone change.

[nooneknows]

You can also change the input tranny with a fet, a 2n5457 for example, changing the 470K with a 1M (this raises the input impedance a lot). You have to check the pinout. I did this mod on the sd1 and it sounds ok

[no one ever]

Different effects need different impedances. Ask Mr. Vex. He selects his impedances for each of his effects for optimal performance. IE, the Machine's output impedance is designed to work well with other fuzzes and whatnot, and the Fuzz circuits have the usual loooow input impedance for the sound we all know.

[Gus]

the real input R with a simple BJT EF buffer

~=   hfe x emitter R  in parrallel with the input R

But you just can not make the input r as big as you want with BJTs and that simple EF buffer circuit

Now with a simple SF Fet the input R is more just the resistor

With both BJTs and Fets you can increase the input R with a circuit change

[Roberto]

You can also change the input tranny with a fet, a 2n5457 for example, changing the 470K with a 1M (this raises the input impedance a lot). You have to check the pinout. I did this mod on the sd1 and it sounds ok

That's a good idea, very interesting.

Did you check the source voltage?  it will be 4.5V (bias from R2) - Vgs ;  and Vgs with that 10K resistor (very low source current)  will be near Vgs(off), that is from -0.5V to -6V according 2n5457 datasheet, so the source voltage can vary from 5V to 9V (4.5v - (-6v) = 10.5v -> 9V rail), you need to pick a very low Vgs(off) 2n5457 to keep the input dynamic range without distortion.

[Ben N]

Adam:
First, if it hasn't already been pointed out, adding a resistor of any value to ground at the input will decrease input impedance.  Second, where and how do you use this pedal?  If it follows another buffered pedal (Boss, DOD, etc.), then raising the input impedance is just pissing in the wind--it wont make any difference at all.  If you are using it alone, depends on your pickups, but it probably wont make much, if any, audible difference.  It might matter if you run it after other pedals with mechanical non-true-bypass.
Ben

[Roobin]

First, if it hasn't already been pointed out, adding a resistor of any value to ground at the input will decrease input impedance.

Hmm, i've been wondering abou this. There was a thread a bit back about impedance. Have a look at GEO about impedance.http://www.geofex.com/Article_Folders/impednc.htm

I think (although I'm probably wrong) that the input impedance is all the current drawing stuff: so in this case it's: R1, R2, R3, Q1, C1.

I'm doing the maths, ill post it later, see if I'm right.

[nooneknows]

Did you check the source voltage? 

Uhm, no, I didn't. But it works fine, no distorsion at all. Maybe it's just a matter of luck.   I'll check the voltages anyway.
BTW I made this type of mod in front of my crybaby, to mantain the original spdt switch, with the difference of the 1MR, going to ground and not to Vref, still it works fine.
Marcello

[Gilles C]

I used the Boss Fet version of the input buffer a few times, and even adopted it as my standard when I try a new circuit.

But I normally use a 2SK30A type of Fet because I ordered a batch of them a while ago. And I also used a 2SK117 with the same results.




Gilles

[Roobin]

I've been thinking about this. Not really getting this impedance stuff, I read RG's amazing article. Then I tried some of the maths. However, could anyone tell me the Hfe of the transistor?

[Mark Hammer]

As was properly pointed out, if the output impedance of something before your DS-1 in the signal chain is already low enough, then an input impedance of 470k is absolutely more than adequate.  The megs and megs of input impedance on things like the SHO are intended to compensate for the load presented by the guitar itself and that first cable.  If your DS-1 is situated after the 2nd or later cable, then a higher input impedance may not add value unless those first couple pedals are all true bypass.

[Roobin]

Mark: Yeh, good point, if the pedal before is tone-sucking, you can't put it back again.

[Ben N]

Roobin, I think you missed Mark's point, which was that a low impedance buffer before the pedal makes the DS-1's 470k input impedance more than enough, and more wont make any positive difference (although it will very likely add noise).  Actually, high input impedance will help after tone-sucking non-true-bypass mechanical switching pedals, since the input impedances of all of the non-true-bypass pedals mechanicals ahead of the DS-1 or other buffered pedal are in parallel with the input impedance of its buffer.  Thus, for arguments sake, let us say that there is an old MXR pedal ahead of the DS-1 (which for simplicity's sake I will round up to 500k also), also with an input impedance of 500k.  When everything is bypassed, the guitar is seeing an input impedance of then seeing an input impedance of 250k, which is quite likely to affect your tone, depending on your pickups.  In that case, if you increase your input impedance on the DS-1 to 1M, the total impedance facing the guitar is 333.33k -- not outstanding, but better.  Ultimately, though, either true-bypassing the other pedals or buffering them will do you more good.
Ben

[Mark Hammer]

To add to Ben's point, one always should try to make decisions about input/output impedance by considering what is situationally likely to be before the pedal in question and what is equally likely to be after it.  If there is little or no likelihood of the preceding pedal being low output impedance, then one wants to protect against loading by having a high input impedance on the pedal in question.  If you know it's always going to be low output impedance then a medium input can be quite sufficient

One of the issues that was drawn to our collective attention a couple of years ago when people started linking to Pete Cornish's pontifications, was that "true bypass" has a number of faults along with its many virtues.  One of those is that if you have a bunch of TB boxes in a row and they're all "off", then the first thing that is "on" better damn well have a high input impedance because it will have to wrestle with loading from the guitar and all that cable - both before and after.  Conversely, for all their weaknesses, one of the things that electronically-switched boxes like those from Boss, DOD and many others have, is an "always-on" buffer that always poses a high input impedance and low output impedance, whether the effect is bypassed or not.  It may not be God's gift to buffers, but you can guarantee that everything will be buffered appropriately. 

So, if you have your DS-1 after any other pedal that uses typical electronic switching, the output impedance of those other pedals will always be low enough that the input impedance of the DS-1 is quite sufficient.  If it was the case that all those boxes leading up to your DS-1 were straight-wire true bypass, AND you had a long-ish cable from your guitar to the first pedal, the stock input impedance of the DS-1 might not be as good as it could be to combat all that loading and cable capacitance.  Not emphasis on "could".  It may meet the minimum criteria, which is okay, but no more than that.

[Roobin]

To Ben N: Okay, so the 470k isn't that low and is enough, if it is after a pedal that has a low output impedance. Correct? Is that what Mark meant by

if the output impedance of something before your DS-1 in the signal chain is already low enough, then an input impedance of 470k is absolutely more than adequate

?

To Mark: So if say you had a load of Boss Pedals, all with buffers always on, then the input impedance of the say second or third pedal can be say 500k, because you have all the previous buffers.

Conversely, if you had a load of true bypass pedals, and a long guitar cable (btw how long is long?), then you'd need to be careful about the input impedance of the 2nd pedal if it's just coming on by itself, because the 1st pedal is off, no buffer, and the pedal would have to cope with the pickups and cable capacitance.

Is this correct guys? Sorry to be a pain, but I'm still trying to get my head around this stuff, and your knowledge is invaluable to someone like me.

[Mark Hammer]

Not a pain at all.  Clarification is always the right thing to do - that's why so many lawyers make such a good living!

And yes, you are correct.

... and a long guitar cable (btw how long is long?), then you'd need to be careful about the input impedance of the 2nd pedal if it's just coming on by itself, because the 1st pedal is off, no buffer, and the pedal would have to cope with the pickups and cable capacitance.

"Long" will depend on the load and the cable capacitance.  Unless it is absolutely fabulous cable, in most instances once you start going past 12 feet or so, you start to get noticeable effects, though nothing you can't easily compensate for with your treble control.  Once you get out to 20ft+, you start to need either some solid buffering because turning the treble up from 6 to 7 ain't gonna do it.  Take a peek at the Anderton article in DEVICE (http://hammer.ampage.org) called "When is a cable not a cable?" for some excellent discussion and explanation of the issues.  (BTW, anyone know how Albert "Mr. 100ft cable" Collins used to set his amps?)

And while we're in the neighbourhood, one of the best points Pete Cornish made was that when you have buffering in a pedal, you can think in terms of only the guitar-to-pedal cord.  If all the pedals are true bypass without benefit of any buffering anywhere, then the 20ft cable up to the first pedal, and the 20ft cable from pedalboard to amp, through the "magic" of straight-wire bypass, turn into what is essentially a 40ft cable from guitar to amp, and bubbelah that's a LOT of cable capacitance.

[Roobin]

Thanks Marks for all your help - and those DEVICE magazines are really helpful as well.

If you consider that quite a few people here gig with their pedals, and assumingat least a 20ft cord each side of the pedalboard, how come so many people use TB, and not a buffer? I'm thinking about using the Boss switching, but can see a couple of disadvantages:
i) I like the 'click' of a DPDT over a momentary switch (actually, I do't know, both feel right)
ii) Can you get momentary footswitches that look like the DPDTs?
iii) The leaky FET issue - but I have done some research, and seen a couple of mods - esp one that adds another FET switch to bypass the opamp of an SD-1.
iv) Parts - more expense. The switching section and buffer(s).

Also, conversely, could this (TB) be exploited to give some degree of warmth - i.e lack of higher frequencies? Or would this be uncontrollable or too much loss?

[Satch12879]

When it comes to cables I'll throw this in:

1.  "Long" for audio isn't "long" anywhere else in electronics
2.  Two words: "output current"

And I'll +1 on objecting to simply raising the base resistor on the input stage of the DS-1 as a good move.

Oh, and not all bipolars nor JFETs are created equal...

[Connoisseur of Distortion]

you might be wiser to simply place a buffer inside your guitar. i did with a friend's guitar, and there was a very noticeable increase in treble!

we like those 20+ foot cables...