Fender reverb in both channels methods

[p2pAmps]

Guys,

I know there are several ways to add reverb and tremolo to both channels of a ab763 circuit.  I have done this a number of ways in the past but I am wondering what your take is on this change...  Curious what else is out there.

Scaggs

[tubenit]

I always insert reverb right before the LTPI.  You can use resistors between 100k -220k and have that work because the reverb pot is 1M with a
470k resistor helping as mixing resistors (sort of).

I've been very happy with this approach on every LTPI amp I've added reverb to.  I like my reverb after the gain vs. reverb that is pushed by a gain stage .

There is an additional advantage in not having a 3.3M resistor in the signal path IF you are wanting more overdrive on a Fender style amp. You can change that 3.3M to a smaller value if one desires.

On Marshall style amps using a 100k mixing resistor is nice because you aren't attenuating much of the signal and the gain sounds about the same.

With respect, Tubenit

[sluckey]

If you are referring to the Dan Torres mod (see attachment) there is one more benefit. The NOR and VIB will also be in phase so you can jumper the two channels together with a short 1/4" phone patch cord. Now you can use every knob on the amp for some different tone combinations.

[p2pAmps]

If you are referring to the Dan Torres mod (see attachment) there is one more benefit. The NOR and VIB will also be in phase so you can jumper the two channels together with a short 1/4" phone patch cord. Now you can use every knob on the amp for some different tone combinations.

There is quite a bit of bad comments concerning the Torres mod.  Look at the attachment Steve on another way...

[tubenit]

With the insertion point just prior to the LTPI, the phase isn't an issue.  Just a thought.

With respect, Tubenit

[sluckey]

There is quite a bit of bad comments concerning the Torres mod.

Like what? Mission and Torres are doing practically the same thing. Mission chose to jumper the channels before the coupling caps, Torres put the jumper after the coupling caps. Torres method is certainly simpler. Mission changed the value of the coupling cap (no biggie) and also used the long wire that's no longer needed, as a very lame shield.

I'd love to here some specific bad comments. Maybe I'm missing something?   

[sluckey]

With the insertion point just prior to the LTPI, the phase isn't an issue.  Just a thought.

Phase is always an issue if you want to jumper the channels together. Doesn't matter where the reverb insertion point is.

In your Bassman/Deluxe example, there are two inverting stages in the Bassman channel, and there are three inverting stages in the Deluxe channel. Both channels will have reverb. But, if you try to jumper the channels together at the input jacks or with an A/B/Both switch box, the phase difference between channels will do a bunch of cancelling when mixed back together, and the resulting sound is usually very thin.

[Mike_J]

Why wouldn't you do what I have shown here?  I don't see the mixer resistors with the Mission method and removing the coupling cap from the normal side would eliminate some of the tuning possibities on the normal channel.


Thanks
Mike

[Fresh_Start]

I built a Super Reverb derivative and just connected the two channels right after the coupling caps.  Can't tell if that's the Torres mod or not because I'm not familiar with the Fender layout.  I had tried something different first which didn't work right - don't put any resistors ahead of the 3.3 meg resistor is all I can remember.

What I liked about this approach is that it let me use different coupling cap values for the two channels.  Note that it puts tremolo on both channels too.  My build used bias vary tremolo anyway.

Chip

[p2pAmps]

With the insertion point just prior to the LTPI, the phase isn't an issue.  Just a thought.

Phase is always an issue if you want to jumper the channels together. Doesn't matter where the reverb insertion point is.

In your Bassman/Deluxe example, there are two inverting stages in the Bassman channel, and there are three inverting stages in the Deluxe channel. Both channels will have reverb. But, if you try to jumper the channels together at the input jacks or with an A/B/Both switch box, the phase difference between channels will do a bunch of cancelling when mixed back together, and the resulting sound is usually very thin.

This is something I read.  I'm not sure which is why I ask the experts here on this forum...

"The Dan Torres method above, while it successfully puts the fx on both channels, creates a capacitive, frequency dependent potential divider at the mixing point, making the normal channel bassier, the trem channel thinner."

and

"I haven't tried this (and I forget where I first read about it) -- but a seemingly more elegant way to get reverb/tremelo on both channels is to jumper the V1B and V2B plates together. You remove both the long blue wire and the .047 cap that feeds it. Essentially, this approach makes those two gain stages act together as a mixer. You can think of them acting like a single triode, except there are two control grids, one for each channel. The cathodes already share a common resistor and bypass cap. The mixed signal is taken off the plates using the .02 coupling cap that feeds the reverb circuit and 3rd gain stage. 
 
The Torres circuit is flawed because it lacks mixing resistors to help isolate the two channels before they hit the reverb circuit. The circuit I described above is nice because you don't even need these resistors. Of course, I wouldn't be surprised if some component value tweaking is necessary to make it work its best. "

[Mike_J]

I built a Super Reverb derivative and just connected the two channels right after the coupling caps.  Can't tell if that's the Torres mod or not because I'm not familiar with the Fender layout.  I had tried something different first which didn't work right - don't put any resistors ahead of the 3.3 meg resistor is all I can remember.

What I liked about this approach is that it let me use different coupling cap values for the two channels.

Chip

What I did above places the mixing resistors in front of the 3.3 meg resistors which I am guessing you probably did initially with poor results.  Where I am confused is where the mixing resistors are using the Torres mod, or either of the mods for that matter.  My guess is the Torres mod is probably the one you probably used.


I agree the Mission method reduces tuning possibilities by sharing the coupling cap.


Thanks
Mike

[sluckey]

This pic should have been attached to my last post...

[Ed_Chambley]

If I want to keep the amp stock, I use the 1 wire. This simply has verb on both since I already have a tremolo channel. Changing the insertion point changes the sound, but you may like it. I have Super Reverb that has a cf tone stack, bias vary and verb inserted at the pi. It does not sound anything like stock. What I am getting at is if you simply want to add verb to the normal channel, I do not know if it is "Torres style has worked on every ad763 I have built. I have always referred to it as a 1 wire reverb.
No additional parts and I guess with vi grid grounded when not using the normal channel it could not mix with the other channel.


Maybe I'm missing something, but I cannot hear any difference except for the normal channel having verb, however I have never tried to add the tremolo.


It is also something I will do on any similar fender circuit that I do work on if the owner wants. Simple and effective. I have never had any complaints, but a lot of happy players.


Really there is only one mod I like on fenders with the roach and that is the switching pot to make use do the gain, but even then you are moving from a true Fender tone.

[mresistor]

From Aspen Pitmans book supposedly keeps gain the same on both channels.   

[tubenit]

if you try to jumper the channels together at the input jacks or with an A/B/Both switch box, the phase difference between channels will do a bunch of cancelling when mixed back together, and the resulting sound is usually very thin.

Sluckey, you are correct.  I've never built a clone of a manufactured amp and my scratch builds always have just one input. I just wasn't thinking thru the idea of jumpering both channels. So, you have made a good point.

With respect, Tubenit

[Ed_Chambley]

http://www.tdpri.com/forum/amp-central-station/261427-fender-twin-reverb-both-channels.html


I cannot get the file to upload, but the link above shows a method that a lot use successfully. I think it is what you already mentioned about tying plates together. I have done this to a twin and changed normal channel to voxish top boost and the guy loves it.

[Mike_J]

Disregarding reverb and vibrato for the moment does that mean you could rewire the normal channel to a cathode follower tone stack using tubenit's fet circuitry on the back end of the cathode follower circuit and end up with two sides that are now in phase?  That would be nice because then you could have a 5f6a Bassman in the normal channel and an AB763 blackfaced amp in one amp and play them both at the same time.


Thanks
Mike

[Fresh_Start]

I built a Super Reverb derivative and just connected the two channels right after the coupling caps.  Can't tell if that's the Torres mod or not because I'm not familiar with the Fender layout.  I had tried something different first which didn't work right - don't put any resistors ahead of the 3.3 meg resistor is all I can remember.

What I liked about this approach is that it let me use different coupling cap values for the two channels.

Chip

What I did above places the mixing resistors in front of the 3.3 meg resistors which I am guessing you probably did initially with poor results.  Where I am confused is where the mixing resistors are using the Torres mod, or either of the mods for that matter.  My guess is the Torres mod is probably the one you probably used.

I agree the Mission method reduces tuning possibilities by sharing the coupling cap.

Thanks
Mike

OK I took the time to chase through the AB763 layout and figure out that I did the "Torres" mod, albeit without ever looking at his work.

Tying the plates of V1b and V2b together the Bruce Collins way (Mission Amps) creates what Kevin O'Connor calls a "Tweed mixer".  Should work fine.  Only downside is not having different coupling caps after the second gain stage but there are plenty of other ways to tweak frequency response.

I came up with a variation of the Aspen Pittman approach on my own and IT DID NOT WORK.  Gain was significantly reduced.  He probably has some mod to offset the impact of the voltage divider that approach changes - a 2.2 meg resistor in place of the 3.3 meg might do the trick.

Note that my friend/customer I built the Uberverb for had no interest in jumpering the channels and I never tested it.  Used separately, the absence of "mixing resistors" didn't seem to hurt anything and the bigger coupling cap on the "Normal" channel fit our design goals better.

Respectfully,
Chip

[Fresh_Start]

Disregarding reverb and vibrato for the moment does that mean you could rewire the normal channel to a cathode follower tone stack using tubenit's fet circuitry on the back end of the cathode follower circuit and end up with two sides that are now in phase?  That would be nice because then you could have a 5f6a Bassman in the normal channel and an AB763 blackfaced amp in one amp and play them both at the same time.


Thanks
Mike

Mike - a DC coupled cathode follower doesn't flip phase at all.  You could use a FET or a tube triode on one channel to reduce tone stack insertion loss.  The tube triode would add the "magic" compression/clipping of any classic Bassman/Marshall circuit and could be really cool.

In a normal AB763 circuit, I don't see a phase cancellation issue.  Both channels have two triode gain stages each, so they should be in phase (except for capacitor frequency effects which I don't completely understand). 

Someone earlier appeared to be discussing an amp where one channel was typical AB763 topology:

gain stage 1 => tone stack => gain stage2

 but the other channel was like a Bassman with

gain stage 1 => volume => gain stage 2 with cathode follower => tone stack => gain stage 3

THAT would create a phase problem because the second channel is out of phase with the first due the third inverting gain stage.

At least that's what I think they were talking about.  If you try to have tremolo only on one channel but reverb on both, you have to design a new circuit AFAICT.

Hope that helps,
Chip

[sluckey]

Disregarding reverb and vibrato for the moment does that mean you could rewire the normal channel to a cathode follower tone stack using tubenit's fet circuitry on the back end of the cathode follower circuit and end up with two sides that are now in phase?

I don't quite understand what you're asking? But here are a few basics that will help you figure it out.

If you put a signal in on the grid and take the signal out from the plate (call this a gain stage if you like) there will be a 180° phase shift.

If you put the signal in on a grid and take the signal out from the cathode (call this a cathode follower if you like) there will be NO phase shift.

Now, as far as being able to jumper two channels together... ie, run the same signal through both channels and recombine the two channels somewhere down the line... The signals need to be in phase at the point where you mix them back together. This means that each separate channel must go through either an even number of phase shifts, or an odd number of phase shifts to be in phase when the two are recombined. IOW, one channel can have two 180° phase shifts and the other channel can have two, four, six, etc. 180° phase shifts, and the signals will be in phase when you recombine them. Or, one channel can have one 180° phase shift and the other channel can have one, three, five, etc. 180° phase shifts and the signals will be in phase when you recombine them. But if one channel has an even number of 180° phase shifts, but the other channel has an odd number of 180° phase shifts, THE SIGNALS WILL BE OUT OF PHASE when recombined and cancellation will occur.

[sluckey]

In a normal AB763 circuit, I don't see a phase cancellation issue.  Both channels have two triode gain stages each, so they should be in phase (except for capacitor frequency effects which I don't completely understand).

This is only true for the non-reverb AB763s like a Deluxe or Showman. But all the AB763s that have reverb have two triode gain stages in the NOR channel but have three triode gain stages in the VIB channel. Jumping the channel inputs on a stock Deluxe Reverb will result in some cancellation when the two channels are mixed together with the 220K resistors. This does not sound good, unless you want a thin, nasal sound, kinda like a wah pedal mashed all the way to the toe position.

[Mike_J]

Disregarding reverb and vibrato for the moment does that mean you could rewire the normal channel to a cathode follower tone stack using tubenit's fet circuitry on the back end of the cathode follower circuit and end up with two sides that are now in phase?

I don't quite understand what you're asking? But here are a few basics that will help you figure it out.

If you put a signal in on the grid and take the signal out from the plate (call this a gain stage if you like) there will be a 180° phase shift.

If you put the signal in on a grid and take the signal out from the cathode (call this a cathode follower if you like) there will be NO phase shift.

Now, as far as being able to jumper two channels together... ie, run the same signal through both channels and recombine the two channels somewhere down the line... The signals need to be in phase at the point where you mix them back together. This means that each separate channel must go through either an even number of phase shifts, or an odd number of phase shifts to be in phase when the two are recombined. IOW, one channel can have two 180° phase shifts and the other channel can have two, four, six, etc. 180° phase shifts, and the signals will be in phase when you recombine them. Or, one channel can have one 180° phase shift and the other channel can have one, three, five, etc. 180° phase shifts and the signals will be in phase when you recombine them. But if one channel has an even number of 180° phase shifts, but the other channel has an odd number of 180° phase shifts, THE SIGNALS WILL BE OUT OF PHASE when recombined and cancellation will occur.

From what you are saying changing the signal from the cathode would change the phase 180 degrees versus the way the normal channel is wired by taking the signal from the plate.  If the AB763 - 2 Channel amps are 180 degrees out of phase as Fender designed them changing the normal channel tone stack from coming off the plate to coming off the cathode would put them in phase.  Therefore an A/B/Y box could be used.  Tubenit's FET design would allow for the use of the same number of tubes that are used in the two channel AB763 amps.  This is not something that someone would likely do to an existing Fender amp but in a new build it could be another option if it does puts both channels in phase.  I realize then that the normal channel would not have reverb.


Sorry if my post wasn't clear I hope it is now clearer as to what I was asking.  Probably should have started a new thread as this does not have anything to do with the reverb discussion but it came to mind after reading the comments about phase.


Thanks
Mike

[sluckey]

If the AB763 - 2 Channel amps are 180 degrees out of phase as Fender designed them changing the normal channel tone stack from coming off the plate to coming off the cathode would put them in phase.

That's correct. But now the normal channel has lost a gain stage. It will not have enough signal to drive the power amp. It will sound very wimpy.

A cathode follower has a gain of slightly less than one.

[Fresh_Start]

Disregarding reverb and vibrato for the moment does that mean you could rewire the normal channel to a cathode follower tone stack using tubenit's fet circuitry on the back end of the cathode follower circuit and end up with two sides that are now in phase?

I don't quite understand what you're asking? But here are a few basics that will help you figure it out.

If you put a signal in on the grid and take the signal out from the plate (call this a gain stage if you like) there will be a 180° phase shift.

If you put the signal in on a grid and take the signal out from the cathode (call this a cathode follower if you like) there will be NO phase shift.

Now, as far as being able to jumper two channels together... ie, run the same signal through both channels and recombine the two channels somewhere down the line... The signals need to be in phase at the point where you mix them back together. This means that each separate channel must go through either an even number of phase shifts, or an odd number of phase shifts to be in phase when the two are recombined. IOW, one channel can have two 180° phase shifts and the other channel can have two, four, six, etc. 180° phase shifts, and the signals will be in phase when you recombine them. Or, one channel can have one 180° phase shift and the other channel can have one, three, five, etc. 180° phase shifts and the signals will be in phase when you recombine them. But if one channel has an even number of 180° phase shifts, but the other channel has an odd number of 180° phase shifts, THE SIGNALS WILL BE OUT OF PHASE when recombined and cancellation will occur.

From what you are saying changing the signal from the cathode would change the phase 180 degrees versus the way the normal channel is wired by taking the signal from the plate.  If the AB763 - 2 Channel amps are 180 degrees out of phase as Fender designed them changing the normal channel tone stack from coming off the plate to coming off the cathode would put them in phase. 

Sorry if my post wasn't clear I hope it is now clearer as to what I was asking.  Probably should have started a new thread as this does not have anything to do with the reverb discussion but it came to mind after reading the comments about phase.

Thanks
Mike

Re-read what Sluckey wrote much better than I did.  Each AB763 channel has two inverting gain stages.  Both even numbers, therefore they are in phase when combined as is.

If you added a cathode follower after the first gain stage of one channel, the CF would take it's signal off that first gain stage's plate and the cathode of the cathode follower would drive the tone stack.  IOW the cathode follower itself does not invert phase because the signal on the cathode is in phase with the signal on the grid.  A FET used in place of a tube triode as a cathode follower works exactly the same way in terms of not inverting the phase.

Personally I think this is a useful detour within Mike's thread because of the confusion about phase mixing related to the original question.

Respectfully,
Chip

[g-man]

Deleted

[Mike_J]

If the AB763 - 2 Channel amps are 180 degrees out of phase as Fender designed them changing the normal channel tone stack from coming off the plate to coming off the cathode would put them in phase.

That's correct. But now the normal channel has lost a gain stage. It will not have enough signal to drive the power amp. It will sound very wimpy.

A cathode follower has a gain of slightly less than one.

I forgot about the first tube in the 5f6a, Plexi, etc schematics.  Just got a little excited because I was going to modify my one channel build but it is still safe because my PT would not support a seventh preamp tube.  Thank you for the explanation.  I am still a little uncertain, for example are the two triodes in the first tube in the aforementioned circuits gain stages and is there an easy way to identify a gain stage?


Thanks
Mike

[shooter]

is there an easy way to identify a gain stage

Until you *master tubes*, if it goes in the grid and out the plate, there's gain

[Mike_J]

is there an easy way to identify a gain stage

Until you *master tubes*, if it goes in the grid and out the plate, there's gain

Very simple explanation.


Thanks
Mike