Millenium Bypass Application - controllable pull down resistor

[winston_smith]

Hi everyone,

This is my first post here and I should start by apologizing for my poor English. Also, I write in a hurry (it's 5:00 AM here) after a successfull experiment that I want to share with you. This is the short story. I'm a beginner pedal builder and I must admit that I usually avoided using pull down resistors in order to get rid of the dreaded "pop". As I was fascinated by the millenium bypass circuit, I realized that I can use it in order to connect a pull down resistor only when the effect is bypassed. I tested my little circuit this evening and it works great. Am I missing something? I would greatly appreciate your comments.

W.S.

http://www.aronnelson.com/gallery/main.php/v/diyuser/millenium_bypass.gif.html

[R.G.]

Actually, if what you were trying to do was to apply a pulldown resistor to the input of an effect only when it was bypassed, you could do it with the Millenium. That should work fine if implemented properly.

However, a couple of things come to mind.

First - you can only do this with the input pulldown, as you have to have an output pulldown to run the Millenium.

Second - why is it you want to avoid pulldown resistors? I can't come up with a good technical reason to do that, or certainly a reason good enough to use a Millenium just for that purpose and no other.

[winston_smith]

Hi,

I was hoping for your answer. I don't really have a good answer to your questions. I'm trying to preserve - as much as possible - the signal of the guitar and not to lower the impedance, especially because I like using several pedals engaged at the same time. Could you please take a look at my layout? I already tested it and it works flawlessly, but I'm afraid I may miss something from the bigger picture.

Thank you,
W.S.

PS Please excuse my horrible English. Not only it is poor, but my technical vocabulary is even more limited.

[R.G.]

I was hoping for your answer. I don't really have a good answer to your questions. I'm trying to preserve - as much as possible - the signal of the guitar and not to lower the impedance, especially because I like using several pedals engaged at the same time. Could you please take a look at my layout? I already tested it and it works flawlessly, but I'm afraid I may miss something from the bigger picture.

In general, preserving the signal of the guitar and not lowering the impedance of the signal are diametrically opposed. The impedance of a guitar is mainly inductive, modified by the resistance of the guitar controls. That means the impedance is low at low frequencies, perhaps as low as a few K, and rises to 100K or more at perhaps 7KHz. That variation of source impedance means that any capacitance or resistance loading the signal will selectively cut out treble to the extent that it's significant compared to 100K. This leads to the use of 1M pulldowns. 1M causes a generally insignificant loss of treble to a raw guitar signal. 100K resistive loading leads to a noticeable but perhaps acceptable dulling of the treble. And cable capacitance causes a bad dulling of treble.

The way to preserve the guitar signal is to use a buffer with a high input impedance to not load the signal, and put that buffer as close to the guitar as possible to prevent cabling losses. The buffer converts the variable low-to-high impedance of the pickup to a constant and very low impedance so loading on the output of the buffer will make no difference to the frequency distribution of the signal, no dulling of treble.

I *think* you may be deciding that an open circuit is much better than even a 1M, 10M or 100M resistor pulldown. And you are correct, kind of. However, the law of diminishing returns sets in early. You get maybe 95% of the total possible signal preservation with a 1M, 99% with a 10M, and 99.9 with a 100M, just for illustration. The extra not-loss of higher resistances adds less and less benefit.

I'm not saying that what you're doing is not good. It's a good attempt to get zero pulldown loading. But unless you are doing something about using very low capacitance cables, the cables themselves will cause more treble losses than any 1M pulldown resistor. And the number of effects makes much less difference than you think. If you use true-bypass switching, then the bypassed pedals don't load the signal in bypass no matter how many they are, and in a series-effect setup, only the first pedal the signal hits has any signal loading effect. In a buffered-bypass setup, there is some signal loading all the time, but the use of the buffer makes any following pedals immaterial. This is in fact how I design all my pedals - buffer first, and do it well.

For best signal preservation, a buffer inside the guitar itself will make a big difference, as this will prevent the cable capacitance from loading the signal no matter what the cables are.

PS Please excuse my horrible English. Not only it is poor, but my technical vocabulary is even more limited.

Well, first of all let me state that from what I've read of your English, I would not have picked up that English is not your native language. I know many people here in the USA whose English is worse than you've typed in here. And second, I respect your ability to converse in a non-native language very much. I know bits of three other languages, but very poorly indeed compared to your command of English. No apology needed. You're doing that very well.

[winston_smith]

Hi again and thank you for your patience and great explanations. Perhaps I was over-enthuziastic about my adaptation of the millenium bypass, but I still think that 99% is better than 95%. One example comes to my mind. Zvex SHO is what it is because its very high input impedance. My SHO pops a little. With a "switchable pulldown resistor", I can preserve most of the input impedance of the SHO (when effect is engaged) and still get rid of the pops. However, I have complete trust in your wisdom and experience (the millenium bypass is the greatest invention of all times in pedal building), so I'll be more tolerant to pull down resistors and more open to your advice.

Thank you very much,
W.S.

[BadIdeas]

For best signal preservation, a buffer inside the guitar itself will make a big difference, as this will prevent the cable capacitance from loading the signal no matter what the cables are.

I have read about effects that actually interact with the guitar's impedance, like fuzz effects and the Blackstone Appliances pedal. I'm still not sure about the reasoning, but if you are like me, and like to control your sound with the volume pedal and dynamics, you may prefer to put such an effect first with a buffer only used in bypass.
I may be wrong about this, so hopefully RG will correct me, but if an opamp was running with +9v and -9v power supplies (perhaps with the MAX1044), there shouldn't be a need for an input capacitor on the buffer, therefore only the effect would need to be pulled down.
That's the sort of thing I've been planning to do if Smallbear gets the PCB kit back in stock, or else I'll just get the stuff separately. I just want to let you know I have not tried this, and make an excuse.

[R.G.]

I have read about effects that actually interact with the guitar's impedance, like fuzz effects and the Blackstone Appliances pedal. I'm still not sure about the reasoning, but if you are like me, and like to control your sound with the volume pedal and dynamics, you may prefer to put such an effect first with a buffer only used in bypass.

There are a very few effects where the guitar's impedance makes a difference. These are almost entirely the original fuzz face and it's many clones and mods. In this case, the pickup impedance actually forms part of the circuit. These pedals form a very small special case, albeit an important one, in the number of pedals which exist. I've written about this interaction many times before.

There are two approaches to this. First is to use these few pedals with only an unbuffered guitar, and second is to put some kind of resistor or inductor right at the front of the pedal to "fake" the pickup's impedance. Both of these work well enough for most people, but if you are very fussy about your sound, you may prefer to, as you say, put such an effect first and not buffer that one pedal. Do what sounds best to you, but keep an open mind. Many people don't mind the differences; in fact, many guitarists don't use this small class of pedals, so the issue does not matter to them.

There is a more detailed set of technical reasoning behind this if you want to hear it.

Bipolar power supply opamp circuits do not have to have input capacitors if their inputs are not driven with DC, so a capacitor is not needed always. There are a lot of special cases that go with that generality.

[brett]

Hi
a couple quick notes:
RE: no input cap if running at +/- 9V
While a good idea in theory, it quickly fails in practice because amplifying a small DC offset (say 0.01V) by a hundred or more (eg typical distortion pedal) gives an intolerably large DC offset.

RE: but I still think that 99% is better than 95%
That also sounds good, but those extra little details from the pickup that are being transmitted are no different to background noise, which will be increasingly transmitted down the line.  So a 1M (or less) pull-down resistor is also a handy noise-limiter.
cheers

[BadIdeas]

While a good idea in theory, it quickly fails in practice because amplifying a small DC offset (say 0.01V) by a hundred or more (eg typical distortion pedal) gives an intolerably large DC offset.

I see what you mean, but in this application, the opamp would be a unity gain buffer, not 100+ gain. Looking at the tl071, the maximum DC offset is about 10mV, like you said. The extra pins are used as something called an offset null. I can't tell what is going on from the schematic diagram, but it sounds like it can be adjusted to reduce this offset.

http://www.datasheetcatalog.com/datasheets_pdf/T/L/0/7/TL071.shtml