Harmonic tremolo - how does it work when bypassed, what is the frequency response?

[itsGiusto]

I'm working on a Revibe build. I was looking at the schematic, and it seems that bypassing the harmonic tremolo (either by foot switch or by intensity rolled to 0) doesn't really "bypass" the harmonic tremolo section, it just turns off the oscillator:

So there's is never really a fully "dry" bypassed signal, there's just a signal that has the effect rolled down to 0. This can be okay, if the frequency response of the system is flat when in bypass (for a reasonable guitar sonic-range, like 100hz to 8000ish hz). I guess I could test this out with my oscilloscope once the build is done, but I was wondering if anyone has a more theoretical explanation. I don't fully understand exactly what is happening when this is in bypass, I only really know that generally when it's on, the circuit has a low pass and a high pass filter modulated at 180 degrees out of phase from each other.

Can anyone explain how the circuit works when it's in bypass, and whether we can tell from looking at the schematic whether the frequency response should be flat?

 

[kleydejong]

As I see it the dry signal on the tremolo side always passes through V1B and then the summing amp in V4A and V4B. V5A appears to be a typical tremolo oscillator with the three caps going from plate to grid inverting phase. I believe V5B and the components feeding into the grid split the signal into two - one high passed and the other low passed.

The J5 footswitch looks to ground the oscillator loop on point T.

I have played a few standalone reverb units and a revibe type unit once. My perception was that they were very much not 'dry' in bypass mode. I felt like it actually was a benefit and it sweetened the base dry signal in a pleasing way. Not so much that it colored the EQ, but it changed the gain slightly and the feel / compression a little.

You could implement true bypass like you'd find in a guitar pedal by putting a switch on the input jack giving a direct line to either the input as normal or to the output jack for a pure mechanical bypass.

 

[itsGiusto]

As I see it the dry signal on the tremolo side always passes through V1B and then the summing amp in V4A and V4B. V5A appears to be a typical tremolo oscillator with the three caps going from plate to grid inverting phase. I believe V5B and the components feeding into the grid split the signal into two - one high passed and the other low passed.

The J5 footswitch looks to ground the oscillator loop on point T.

I have played a few standalone reverb units and a revibe type unit once. My perception was that they were very much not 'dry' in bypass mode. I felt like it actually was a benefit and it sweetened the base dry signal in a pleasing way. Not so much that it colored the EQ, but it changed the gain slightly and the feel / compression a little.

You could implement true bypass like you'd find in a guitar pedal by putting a switch on the input jack giving a direct line to either the input as normal or to the output jack for a pure mechanical bypass.

Yes, that's true, I could put in a true bypass. But if I wanted it to be footswitchable, I'd need to use a relay. And that would mean having another transformer, another filter cap, and a voltage regulator, and finding a place to fit all that in.

Also, I'd only like the harmonic tremolo to be true bypassable. I'd still want the reverb to be in. I think maybe what I'd want to do is have the bypass just directly change the output of V1b after C15 to go directly to the reverb output, connecting to R50

 

[itsGiusto]

I have played a few standalone reverb units and a revibe type unit once. My perception was that they were very much not 'dry' in bypass mode. I felt like it actually was a benefit and it sweetened the base dry signal in a pleasing way. Not so much that it colored the EQ, but it changed the gain slightly and the feel / compression a little.

I am more amenable to it changing the gain, and would expect that, since it's adding at least one more gain stage. I'm just hoping that it doesn't really change the EQ, that's all. If it did, I may want to change my build to be a Torevibe instead, which I could probably expect would change the EQ less when switched off, since the tremolo section doesn't break the signal up into differently EQed parts.

I believe V5B and the components feeding into the grid split the signal into two - one high passed and the other low passed.

@moosie could correct me here, but I think V5b actually inverts the oscillation sinewave, which is what puts the low pass amplification stage out of phase with the high pass amplification stage.

 

[Wyatt]

I can't think of any Fender tremolo that is actually bypassed. There is some insertion loss from the circuit, so unity gain is maintained by just turning off the oscillator. If you true bypass the trem circuit, expect a noticeable jump in signal strength when bypassed. But a resistor in the bypassed pathway could probably negate the difference.

 

[itsGiusto]

I can't think of any Fender tremolo that is actually bypassed. There is some insertion loss from the circuit, so unity gain is maintained by just turning off the oscillator. If you true bypass the trem circuit, expect a noticeable jump in signal strength when bypassed. But a resistor in the bypassed pathway could probably negate the difference.

Yeah, I really am fine with a mild gain difference. I'm just worried, that since this is something I'd keep plugged in more often then I'm actually using the vibrato, that it would alter the tone in some substantial way, especially EQ differences.

I guess I feel like I have some instincts, that how an effect "should" behave in bypass is different if it's a natural part of the amp, vs an additional external effect.

 

[FenderLover]

When the oscillator is killed, Only audio enters V4. That high pass/low pass filter was changed between that version and the -A version Fender schematics, so I wouldn't doubt there is a little mid range character that is part of its sound. Did they improve it or not? It's hard to tell why they changed it in the first place. Notice the Revibe copied the original circuit for some reason.

 

[Wyatt]

Yeah, I really am fine with a mild gain difference. I'm just worried, that since this is something I'd keep plugged in more often then I'm actually using the vibrato, that it would alter the tone in some substantial way, especially EQ differences.

Well, there is nothing "neutral" about the the 6G15 circuit, it's a 5F1 Champ driving a reverb tank. It would be hard to determine how much alteration is the harmonic trem and how much is the 6G15. Personally, I won't fix something until I'm sure it's broken. I would build the Revibe, play the Revibe, then mod it.

 

[Willie Johnson]

Not to high jack the harmonic trem discussion, but--I have a Princeton (no-verb, 5-knobber) whose trem has never been its shining feature. I've read that installment of an intensity pot with an spst on/off detente can be wired to bypass the tremolo circuit entirely. Would that have any significant effect on output and gain? (EDIT: Just reread, "noticeable increase in signal strength"; can this still be accomplished with a switch?)

 

[itsGiusto]

Well, there is nothing "neutral" about the the 6G15 circuit, it's a 5F1 Champ driving a reverb tank. It would be hard to determine how much alteration is the harmonic trem and how much is the 6G15. Personally, I won't fix something until I'm sure it's broken. I would build the Revibe, play the Revibe, then mod it.

Well, that makes sense, but at the same time, I understand it's ridiculously hard to change this thing after it's been made. From Jeff Geheing, the original designer of the Revibe: https://web.archive.org/web/20140410144814/http://home.everestkc.net/jgehring/revibe1-4.html

The backside of the parts board is fairly busy also. It's a good idea to treat the wire on the back of the board like it was going on a space probe to Jupiter--in other words, check it out with the mindset that you'll never be able to see it again. Once the main board is mounted and wired in, you will NOT want to rip it out to fix an underboard wiring mistake!!

I'm attempting to remedy this by moving all rear-side wiring to the front of the board. I hate amps where long-term maintenance and mods are not possible. Still though, I'd be worried about doing major rewiring after the fact, and I wonder if I should figure out my bypass mechanism upfront.

 

[Phrygian77]

You have a network of filters that allows more lows and less highs through one gain stage, and less lows and more highs through the other. The oscillator does nothing more than shift (alternating) the bias of each of stage. The key to that though is that the cathodes are only bypassed above the oscillator frequencies, hence the reason for the small bypass cap on the shared cathode.

 

[Phrygian77]

I guess, in the amp version, you can try to make up for any coloration discrepancy with the tone controls.

There are bandwidth limitations to every gain stage. In this case there are deliberate limitations, especially at the low end. Someone just needs to do some AC analysis on this circuit and publish it.

 

[Bendyha]

Check-out this thread exploring-fenders-harmonic-tremolo, lots of great information to be found there, and the frequency question is also discussed.
And moosie build thread filament-buss-question.

 

[Modman68]

Those two threads were fantastic even for novices!