How much capacitance makes a circuit non TB?

[ExpAnonColin]

"None at all, TB means there is nothing in the path at all whatsoever when bypassed"

OK, OK, but what if you just had .2pf of capacitance when bypassed?  What about 2.7pf?  That doesn't even count, does it?  Can anyone provide any evidence otherwise?  Your guitar cable is going to have more than that (bad cable is what, maybe 2000pf) isn't it?

-Colin

[Gilles C]

In my mind, True Bypass is about what is connected in series with the signal, like a FET for the switching, compared to a dry contact of a switch.

Whatever is in parallel is loading the signal path, and it can be resistive, capacitive or inductive.

It also affects the signal, but it can still be called True Bypass.

For example, take the 1M resistor that almost everybody puts to eliminate popups. It is in parallel with the signal, and everybody still calls that True Bypass. And as you said, it is the same with cables.

Gilles C

[Paul Perry (Frostwave)]

My understandign of True Bypass is that it means the signal doesn't go through any electronic components at all. (switched by a stomp, or a relay).
You are right on the capacitance, personally I would rather have active pickups & then not worry about cable capacitance (plus, really helps for hum & radio interference by lowering the impedance). But, if you have active pickups, you don't have true Bypass (or at least, you might as well not, for a purist.)
Personally, I;m not a True Bypass guy, but, Live & Let Live.

[niftydog]

0.2pF is impossibly tiny... and so is 2.7pF.

I generally think of it in terms of the RC filter equation. Stray capacitance is usually in parallel, so the resistance required to make an effect on the audible frequency range is tremendous.

@ 20khz and 2.7pF you'd require a series resistance of around 2-3Mohms before you'd start noticing an effect. And that's only if your hearing is perfect! Considering the frequency range of a guitar, it's highly unlikely to be a problem.

Besides, if you've got anything near 2-3Mohms series resistance, a little bit of stray capacitance is the least of your worries!

[Torchy]

crp

[ExpAnonColin]

I'm talking about the capacitance being in serires at all times.  Thanks for proving it, niftydog...

.2pf/2.7pf are capacitances of the 4016 bilateral switch, that's why I ask.  Does anyone have any other reason it wouldn't be TB, other than the single digit mV switch bleedthrough and noise?

-Colin

[ExpAnonColin]

An interesting take on True-Bypass from a certain Mr. Pete Cornish ...
http://www.petecornish.co.uk/case_against_true_bypass.html

He is overdoing it, as are the hardcore TBists.  It's good to have one or two buffers in your chain.  One or two decent buffers.

-Colin

[Torchy]

crp

[R.G.]

.2pf/2.7pf are capacitances of the 4016 bilateral switch, that's why I ask. Does anyone have any other reason it wouldn't be TB, other than the single digit mV switch bleedthrough and noise?

Sigh. A question that our dearly departed president, William Jefferson Clinton would love. Is it truly true bypass? I guess it depends what your definition of "is" is.

The commonly accepted meaning of "true bypass" would say, no, it's not true bypass, and can never be because the signal goes into and out of a semiconductor device that depends on turned-off semiconductors for isolation.  Case closed, not true bypass.

True bypass, to purists, holds that when bypassed, there is a metallic conductor path through the box that does not touch any of the effect parts - semiconductors, resistors, caps, diodes, most certainly not transistors.

However, you're sniffing around the idea that  -- well, it's *close* to true bypass, isn't it? What exactly does "true" mean anyway? Why can't we change the definition so it there's a bit of capacitance and some biased-off semiconductors, we can *call* it true bypass? Close counts, right?

Probably not. This is perilously close to writing advertising copy.

See my description of the "Clinton Bypass" in my Technology of Bypasses article. I think it would sound like true bypass, and I named it the Clinton Bypass with the idea that Bill would say "Sure, I don't think you could tell it from true." (Dang, I'm inspired by that man! But I don't think it's true bypass. Nor do I think 4016 setups are TB, nor my bypassing with a CD4051, nor other variants that rely on turned-off semiconductors to keep the signal out. And I believe that most people in the industry that have thought about it a bit would come up with the same definitions.

[ExpAnonColin]

It's understandable, RG.  I guess our views differ in that for me, the fact that your guitar's signal going through a 4016 inside of a black box will sound and act 99.9% the same as your guitar's signal through a switch inside of a black box, it would be OK to call it True Bypass.  If you can't tell, and the other electronic devices in your chain can't tell, then why shouldn't it be?  If this isn't true bypass, then how can the DL4's be true bypass?  This would make a new thread I think...  the switches in the DL4 are without a doubt momentary, and the thing is an EPROM and some components to control it, but it's undoubtedly accepted high and low that the thing is 100% TB.  But as far as I know, there is no tangible way at all to use a momentary SPST (or DPST, as the case may be) to make TB without putting it through something more than the switch.  Or am I wrong?

-Colin

[Eric H]

But as far as I know, there is no tangible way at all to use a momentary SPST (or DPST, as the case may be) to make TB without putting it through something more than the switch.  Or am I wrong?

-Colin

I have no idea what's in a dl-4 --but if that momentary switch controls a latching-relay it's true-bypass.

-Eric

[ExpAnonColin]

But as far as I know, there is no tangible way at all to use a momentary SPST (or DPST, as the case may be) to make TB without putting it through something more than the switch.  Or am I wrong?

-Colin

I have no idea what's in a dl-4 --but if that momentary switch controls a latching-relay it's true-bypass.

-Eric

Edit, I was wrong... is there an relay?

-Colin

[R.G.]

I guess our views differ in that for me, the fact that your guitar's signal going through a 4016 inside of a black box will sound and act 99.9% the same as your guitar's signal through a switch inside of a black box, it would be OK to call it True Bypass.

By the way, I meant no ill wil or censure in that note. Just my understanding of what the words mean.

The term "true bypass" has historically meant that there's a continuous metallic conductive path through the box for the signal when the effect is bypassed, and that the path does not touch any of the effect circuitry - that is, the effect does not touch the signal path at all.

You're free to call anything you want "true bypass". What that means to other people will depend on their own opinions and experiences. If you yell "Nova!!" at an astronomer's convention, you'll get a different response than in a taqueria. The sounds are the same, there's different meanings attached. If you're from the UK, you think it's odd when people in the USA giggle at your asking them to knock you up. If you're from the USA, you will get odd looks from your uses of the words "closet" "fanny" and the term "on xxxx street". I just pointed out that there's a body of people who will either not know what you mean, or not agree with you if you call 4016 switching "true bypass".

If you can't tell, and the other electronic devices in your chain can't tell, then why shouldn't it be?

This is kind of like existentialism arguments, or Von Neumann's definition of intelligence.

One has to distinguish between results and causes or paths to results. I think that you're saying that you believe the term "true bypass" to mean "no audible sound change, and no electrical difference to the signal, except for X percent" and then asking how small X has to be to make that believable.  I'm saying that "no signal change or degradation" can be done by multiple paths. One of those might be hard wired, another might be mechanical switch true bypass, another might be CMOS switches, and yet another might be buffering with careful attention to impedances to keep other devices from either suffering or sounding too good after the buffer. Those are paths. No signal degradation (or, I guess, just the expected type and amount of degradation) is a result. True bypass is one path or cause, not the result.

I think you're saying "I have this result (to within some percent) and I want to know if I can call it one of the paths, because it's almost exactly the same result as that path."

If this isn't true bypass, then how can the DL4's be true bypass?

I don't know the DL4.

This would make a new thread I think... the switches in the DL4 are without a doubt momentary, and the thing is an EPROM and some components to control it, but it's undoubtedly accepted high and low that the thing is 100% TB.

Point me to the circuit and I'l render an opinion. In fact, drag in a few of those people who accept it high and low as TB to tell me why.

But as far as I know, there is no tangible way at all to use a momentary SPST (or DPST, as the case may be) to make TB without putting it through something more than the switch.

Me neither. But I've been wrong before. Show me the circuit.

Or am I wrong?

See above.

[bwanasonic]

It's funny, I use to laugh at "True Bypass Fetishists" , but then I got into this crazy hobby, and started introducing *primitive* effects into to my signal chain that don't always get along with more *advanced* technology. Tone suckage wasn't my concern, as much as it was "will it freak out my Fuzz Face/ Rangemaster, etc.) ? Basically, If I am buying a pedal, I expect that "True Bypass" means the aformentioned:

"there is a metallic conductor path through the box that does not touch any of the effect parts - semiconductors, resistors, caps, diodes, most certainly not transistors.

It's almost like that type of bypass has trademarked the term "True Bypass". If you wan't to market your bypass scheme, the name is taken. I understand your point in theoretical and practical electronics terms, but in *consumer* terms, I prefer the current "standard".

Kerry M

[Paul Perry (Frostwave)]

If any manufacturer claims their unit is 'true bypass', and it doesn't have a continuous metal path from in to out while bypassed, then the customer is entitled to sue. That's because ther eis a consensus that that is what TB means.
My units aren't true bypass, by the way. But if I had my life over again, they probably would be (using a latching relay).

[puretube]

...
For example, take the 1M resistor that almost everybody puts to eliminate popups. It is in parallel with the signal, and everybody still calls that True Bypass. And as you said, it is the same with cables.

Gilles C

where do you put your pulldowns? on the jacks?  :?

[Mark Hammer]

Ton's right.  Pulldowns go on the input of the circuit so that the cap can drain even when you're not using the pedal.

For me, the starting point is not the definition of TB, but a single solitary question:  What do you want an ideal bypass to do for you?

If you can list the criteria of what that bypass arrangement needs to do for you, and you can identify a particular arrangement and check off all those criteria, problem solved.  As it happens, straight wire does pretty much all of it, but it is not the only way to accomplish "all".

Bear in mind that what some folks want from a straight wire bypass is also the capacity to have usable signal even under conditions of battery/local-power failure.  A straight-wire stompswitch does that nicely.  Battery fizzles, just stomp and keep playing.  CMOS and FET switching can't do that, and neither can relays.

That is not strictly the same as the traditional electronic issues regarding what happens to signal when a circuit is never disconnected from the path.

[puretube]

...Battery fizzles, just stomp and keep playing. CMOS and FET switching can't do that, and neither can relays.

got to disagree with that last one: the relay (true!    bypass in the E-H tubepedal series do automatically switch to true bypass @ power failure!
Let me call this: auto-true-bypass .

[Mark Hammer]

Then kudos to you my friend!  I guess I was thinking of latching relays, where there would have to be power to return the relay back to the alternate state.

[Ben N]

Whatever is in parallel is loading the signal path, and it can be resistive, capacitive or inductive.

It also affects the signal, but it can still be called True Bypass.

That doesn't sound quite right.  After all, most old-fashioned mechanical SPDT non-TBP switching (e.g. MXR) leaves the (buffered?) input of the effect connected while switching the output between effect out and the input.  The effect is therefore in PARALLEL with the signal path in bypass mode.  My understanding of TBP was that both the input and output of the effect are isolated from the BP signal path.

Ben