Fender Deluxe Reverb AB763 -- Normal channel needs more gain

[pkb]

Hello Ampgarage peeps,

In this, my first post, I want to say that the level of competence and helpfulness on this site is amazing. Thank you, everyone, for participating and sharing at such a high level.


Here are two questions:

1. How can I modify the normal channel of a Fender Deluxe Reverb so that it produces more gain? It's wired as a gain stage going into the Vibrato channel and is not giving a satisfying amount of saturation even when turned up to 10. Should I remove the 68k grid stopper resistors? Do I need to change the bias of the tube?

2. Even though the normal channel is not wired to the PI tube (the connection has been lifted after the .047uf cap that sends out the signal), when it's turned up, it manages to reach the output as a flubby soft (think really fat little kid who's not very energetic) treble-less signal. How is this and how can it be stopped?


A bit of background: I was given a '67 Silverface Deluxe Reverb that spent 25 years in a barn without electricity in it. It was likely retired because of many layers of questionable modifications to increase the gain of the amp and ultimately succumbed to a failed solder joint to ground behind the cathode resistor/capacitor of the tremelo 12ax7. When I turned the amp on, it wouldn't do anything but make a full-volume white noise roar.

I've replaced modded caps with original value 715p's (the mods usually had smaller values, for example, the coupling caps between the PI tube and the power tubes had been changed from .1uf to .033uf). I've also replaced some visually damaged 470ohm resistors above the power tubes with beefier 2w resistors. I also replaced the 25uf/25v electrolytics with the modern Sprague versions. The elderly original caps were measuring everywhere from 1uf to 38uf. There was also a solid state tube replacement in the normal channel tube socket. It's a curiouslittle device that did in fact provide more gain, but it took a different bias voltage and required some more messy modifications of the plate voltage that was luckily on a switch.

Part of the modification mess that I inherited was a cascade mod that jumpered the signal from the plate of the first gain stage in the normal chanel and fed it through a .01uf cap to the grid of the first stage in the vibrato channel.

Since the bass knob on the vibrato channel didn't do anything thanks to the .01uf coupling cap, I replaced the cap with a .047uf orange drop and switched the source of the signal from the plate on the first half of the tube to the plate on the second half of the tube AFTER it goes through the tone stack and volume knob. This gave me some additional gain and a some control over the frequency response of the clean signal. I also lifted the wire connecting the normal channel's output to the PI tube so I could turn the normal channel up enough to acheive gain without playing the amp at full volume.

My hope was to end up with something that resembled the saturation of a Boogke Mk1 or Mk2. I'm quite a ways away from that.

Any ideas would be very welcome, even if they don't address my question directly or are suggestions to please stop doing this and just build a different amp.

[DonMoose]

The AB763 is a very nice-sounding circuit. I would suggest you do a full restore - get the thing all the way back to original in terms of values and wiring.

Try on a external speaker - the one that spent all that time in the barn probably isn't healthy.

Now that you know what you're working with, you can try doing something with the channel cascade - your plan of taking NORMAL's output, right where it heads for the PI, into the VIBRATO input should work. I think you may need to pick off that cascade signal from before the 220K into the PI, though.

Hope this helps!

[thejaf]

Are you going to use the vibrato?

I'd keep the 68K grid stopper on the input.

I recently finished a build mostly based on the AB763 circuit, and am 95% happy with it (enough to stop tweaking for a while).

A few things to get some modest gain boost from that channel (what I ended up with):

1. Mid boost cap parallel with the treble cap. Larger value lets more signal through. My boost totals about .004uf, where I found happiness with mid boost in my circuit.

2. Raise the value of the mid "pot" resistor. Stock value is 6.8K, and you could raise this up to about 30K or so until you start defeating the tone circuit altogether.

3. Lowering the value of the .047uf mid tone cap will give a slight gain boost and low end thump. I currently have a .015uf there.

4. Raising the value of the last dropping resistor in the power string will lower voltage to this tube, and give you a softer breakup. I have a 22K as my last resistor value, and I like that a lot (as opposed to my original 10K).

5. Install a post PI master.

6. To tighten some flubbiness, lower one or more of the cathode caps in the first few preamp stages. I have the second stage (post tone circuit) lowered to about 1uf.

[NickC]

I built mine to stock values, except added a rear panel switch to defeat NFB.

I agree with advice above regarding speaker replacement. The storage conditions could not have helped the original speaker any.

[pkb]

Now that you know what you're working with, you can try doing something with the channel cascade - your plan of taking NORMAL's output, right where it heads for the PI, into the VIBRATO input should work. I think you may need to pick off that cascade signal from before the 220K into the PI, though.

Hope this helps!

The mod is set up so that I'm taking the signal directly off of the intersection of the 100k resistor/.047uf cap juncture after pin 6 on the first tube socket (plate). I'm avoiding the 220k resistor and the preexisting .047uf cap and have installed another .047uf cap on the switch that toggles cascade on/off. The switch is above and between the first two tubes. It was there right out of the barn and seems to assist rather than impede this mod. I can post some pictures sometime soon.

Thanks

[pkb]

Are you going to use the vibrato?

.

Very likely not. I doubt I'll use the reverb either, though I don't want to extract that entire circuit in case I decide it's useful.

Here's a picture of the amp with the barn mods. I've since removed and restored a few. The leftmost switch previously from either adding the signal from the first gain stage's plate to the vibrato grid to its current state. The change was to lift the red wire going from the leftmost intersection of 100k and 25uf to the one directly to its right (the output of the second stage).

[DonMoose]

The mod is set up so that I'm taking the signal directly off of the intersection of the 100k resistor/.047uf cap juncture after pin 6 on the first tube socket (plate). I'm avoiding the 220k resistor and the preexisting .047uf cap and have installed another .047uf cap on the switch that toggles cascade on/off. The switch is above and between the first two tubes. It was there right out of the barn and seems to assist rather than impede this mod. I can post some pictures sometime soon.

Thanks

That ought to work - the one thing (I think you covered this) - is to make sure that the cascade switch opens (disables) the non-cascade signal path to the PI.

Sounds like a fun project!

[pkb]

The mod is functional and as you've suggested (no signal travels to PI wen signal travels to PI from normal channel). The amount of saturation, compared to a Mk 1 or Mk2 Boogie circuit, is quite weak.

These Mk circuits use the same 12ax7 and a tone stack with similar options (no mid knob here, though), yet somehow manages to achieve a much higher level of distortion and saturation. Is there some adjustment that can be made in resistor or capacitor values that would change the amount of available gain without too much effecting the tone of the stage?

So far, it's definitely been a fun project. Playing with tube amps (both with a soldering iron and with a guitar) is fantastic stuff.

[DonMoose]

In non-cascade mode, does the NORMAL channel sound normal?

And you're diming that (volume and tones at or near 10) to hit the VIBRATO channel hard?

You may need to play with some known-good tubes.

[pkb]

In non-cascade mode, does the NORMAL channel sound normal?

And you're diming that (volume and tones at or near 10) to hit the VIBRATO channel hard?

You may need to play with some known-good tubes.

In non-cascade mode, the normal channel is mute. The switch, at the moment, only enables the signal to flow between the post-volume knob output of the normal channel to the grid pin of the first half of the vibrato channel (via a .047uf cap). The normal channel is mute because I lifted the connection between the normal channel's built-in .047 cap and the PI input. I can put this on a switch later, but for now, the normal channel doesn't have a voice of its own without soldering something back together.

I hope to (dime to?) hit the vibrato channel's tube hard enough to produce saturated lead tones. =)

Just replaced the preamp tubes in the normal and vibrato channel's sockets with some fresh JJ's.

[Super_Reverb]

It is going to be difficult to get Boogie Mk1/2 saturation and drive in this platform.

You're at a crossroads. IMO, you can't tweak your way there from here.

One good suggestion was to restore this to factory design and enjoy the vinatage tone. If you want to get to a Mesa Boogie sound, you will have to implement some or all of their preamp design techniques, such as hotter biasing, additional gain stages, low frequency filtering, etc.

Another approach might be the SLO 100 preamp or the TW express preamp. Both have very pleasing levels of overdrive way above the AB763 domain.

Agree with Don and earlier posts: first things are to replace preamp/poweramp tubes and speaker.


cheers,

rob

[DonMoose]

In non-cascade mode, the normal channel is mute. The switch, at the moment, only enables the signal to flow between the post-volume knob output of the normal channel to the grid pin of the first half of the vibrato channel (via a .047uf cap). The normal channel is mute because I lifted the connection between the normal channel's built-in .047 cap and the PI input. I can put this on a switch later, but for now, the normal channel doesn't have a voice of its own without soldering something back together.

Strongly recommend you do this - trying to figure out why a preamp not know to be good isn't doing what you want as an overdrive is a dead end.

As mentioned elsewhere, without major revisions to the normal preamp, you won't have a Boogie, but you may wind up with something you like. To get from there to Boogie territory, you will probably delete the tone stack in the normal channel, or modify it to a 5F2 Princeton style for lower insertion loss.

[topbrent]

As stated above, you might like having a different tone available via the 5f2a mod. You can accomplish this without really hurting anything.

Unsolder the slope resistor and the regular tonestack connections and secure the wires out of the way. When you want to go back to the ab763 flavor, it will all be ready to use again.

These pictures tell the tale. Very easy conversion project.

[pkb]

That mod looks fantastic.

I think I've given up on the cascade mod for the moment. I'm going to attempt to replace the vibrato circuit (just the vibrato circuit, not the whole channel) with a Dumble ODS overdrive circuit (copying the #124 OD circuit). I'll have the normal channel on a switch to rough through the OD circuit or not and hope that it sounds really great. The post on that idea lives here: https://tubeamparchive.com/viewtopic.php?t=14983

I did finish the mod related to this thread with a few interesting results. It sounds pretty good but the bass is a bit wooly. The speaker is definitely suspect and I'm planning to replace it with the Weber Michigan at some point.

The strange bassy bleed-through between the normal channel and the vibrato channel was resolved by individually biasing/bypassing the cathodes in the second halves of the preamp tubes. They're bypassed/biasd as one and by adding a second 25uf/25v cap/800ohm resistor that grounds the first cathode separately, the problem was resolved.

I had also lifted the cathode bypass cap on the vibrato tube's 1'st half to promote compression and distortion. This turned out to be not so much to my taste and I re-implemented the cap.

Let me know if any of this is interesting and I can send pictures or detail my experimentations with better clarity.

[rdjones]

The strange bassy bleed-through between the normal channel and the vibrato channel was resolved by individually biasing/bypassing the cathodes in the second halves of the preamp tubes. They're bypassed/biasd as one and by adding a second 25uf/25v cap/800ohm resistor that grounds the first cathode separately, the problem was resolved.

Unless you are treating the cathode value as an additional mod it should be 1.5k each.
2 tubes effectively in parallel (at the cathodes) is ~2x the current through the cathode resistor, hence the 820 Ohm value stock.

rd