True bypass???

[amz-fx]

Here is a partial schematic of a commercial pedal that I own.



This is what the signal path looks like when the footswitch is bypassing the circuit.  The red strip is where the 3PDT is connected.

Do you consider this to be true bypass as the manufacturer claims?

RG, wouldn't you say that the ferrite should come before the 100pf filter capacitor on the input, as with the output?

There is also a big mistake in how the external power jack is wired up but I have not drawn that yet.

regards, Jack

[slacker]

I suppose technically it's not true bypass.
I guess the important question is whether the components left in the signal path affect the sound at all, if they don't then it's not an issue.

[Mark Hammer]

Is the 200pf of capacitance that the pedal provides enough to detract from tone when summed with the various cables involved?  That's what I would be more concerned with.

In the grand scheme of things, what is never really discussed is the incidental capacitance that might occur between input jack and switch, and output jack and switch, even when the physical switch is nominally a textbook true bypass.  After all, the signal travels along wires to get to the switch; it doesn't get there by magic.   If what you show in the drawing is little different from what normally occurs even in the context of ostensibly "straight wire" bypass then why worry about it.

I guess what I'm getting at is a distinction between the various aspects of TB.  Certainly, such an arrangement absolutely assures signal continuity in bypass mode, even in the event of power/battery failure, which IS part of the definition and rationale for TB.  It also imposes nothing in series with the signal to eat up bottom end.  So, what it does have that doesn't meet the "ideal TB" definition is the potential for some high end loss, or for exacerbating high-end loss stemming from cables.  If the pedal in question is normally preceded by a buffered signal, though, then the circuit should perform as if it met all the criteria for TB, I should think.

Does that sound fair?

[R.G.]

True bypass plus extra tone sucking, maybe?

I like Mark's practical evaluation. If it doesn't affect the signal noticeably in bypass, who's to say?

The 200pf input capacitance is certainly getting up there where it could be noticeable with some pickups, but if it's after a pedal or two, it's unnoticeable.

The cap/ferrite/ferrite/cap makes me wonder what RF they were worried about. If it's an analog opamp/transistor widgie, they were ill informed about how to deal with RF. If there is (a) a switching power converter (b) continuously running digital circuitry or (c) RF speed sampling/modulation techniques in the pedal, then they *may* have been trying to keep RF from leaking out onto the cables where it could radiate.

Some of the most instructive moments in circuit analysis happen when you find things where you have to say "Now, why would they ever do that?"

In terms of the terminology, I really think that "true bypass" is a term which is becoming archaic. Words lose meaning or change meaning based on the cultural context. True bypass was a useful, accurate term back when effects circuits were minimal and when a guitar player had only one effect. In that era, it meant something useful that the single effect the player had could be left true bypassed and the show could go on. But the show could always go on by replugging one cord, so that's not as huge a benefit as it's always been explained. The non-interference with the bypassed signal has always been the real reason for true bypass.

True bypass was the simplest way to get this non-interference with the bypassed signal when an effect only had one to four $2.00 transistors in it. Today, when dual opamps and analog switches are under US$0.50 in hugely inflated dollars ( Cars cost $3K - $6k in the 60's), a really good non-interfering buffered bypass is not only available, it's cheaper and more reliable than the stomp switch itself.

But I'm wandering. The designer was either confused about RF or solving a different problem.

[Mark Hammer]

But I'm wandering. The designer was either confused about RF or solving a different problem.

What role would the ferrite beads play in reducing the likelihood of heterodyning when using multiple digital pedals?

[amz-fx]

Here is the power supply connection.  The ground point is a star ground on the pcb which is directly connected to the in/out jack ground lugs.



-Jack

ps: this is a simple one-transistor booster

[amz-fx]

The only advantage I can see to this arrangement is that the big diode only comes into play when you are using an external supply... i.e. no diode drop when using a battery.

-Jack

[R.G.]

It was probably more like having to decouple the inner circuits from the signal cords. Strictly speaking, any pedal maker that uses circuits which switch at faster than 100kHz or which have harmonics above 100khz is required by law to submit their products for EMI type approval. I suspect that not many actually do this, and just count on not getting caught.

A signal cord forms a great antenna for low-RF signals in some circumstances, and a ferrite bead/cap forms a nice stopper to prevent that.

A stopper for RF coming in from the outside would put the series impedance - the ferrite's inductance - on the outside of the shunt impedance of the cap. The circuit as shown would actually be resonant at some RF frequency dependent on the ferrite's characteristics and could TUNE IN a radio source, providing a nice, low impedance RF source at its resonant frequency to the circuit. What's missing entirely from the picture is any damping resistors in the circuit itself. If the first thing it connects to inside the circuit is a series resistor, that could damp the RF enough to keep it from being a problem. But that's pure speculation on my part.

A star ground on a 1-transistor booster? Wow. That's either slavishly following suggestions or an attempt to correct some RF problem.

That's a bizarre power connection.

The collection of oddities makes me suspect that there's an unknown problem being solved. Given the now three ferrite beads and RF caps, it's maybe an RF problem. But the details are not yet apparent.

If the transistor's a JFET or MOSFET with gain at RF, the "cures" would make more sense than if it were bipolar.

[amz-fx]

The +9v connects to the board at the plus sign in the bottom right.  The power supply ferrite is where I made the black mark on the pcb by the diode in the upper left. The 100pf is next to the power input jack.

The big ground pad in the middle is where the back of the output volume pots rests (now removed)...




.

[snap]

probably a method to get rid of RF influence from those switchmode wallwartz?

[R.G.]

probably a method to get rid of RF influence from those switchmode wallwartz?

Maybe.

The way it's arranged points at keeping RF inside the pedal, not intercepting it from the outside. Unless it's just a mistaken design approach. Or they're solving another problem.

That being said, there are switching wall warts that can cause RF problems. There is one - the 1Spot - that definitely does not. There is another that definitely has the potential to that I have measured. I guess the details always matter. 

[bumblebee]

The +9v connects to the board at the plus sign in the bottom right.  The power supply ferrite is where I made the black mark on the pcb by the diode in the upper left. The 100pf is next to the power input jack.

The big ground pad in the middle is where the back of the output volume pots rests (now removed)...




.

i knew it was an EH effect cause the RI little big muff has ferrite beads in the true bypassed signal as well.