Fender Style Reverb Questions

[ChrisM]

Lets refer to this schem for Fender Reverb
http://lib.store.yahoo.net/lib/thetubes ... bab763.gif


1. Why is the input coupling cap so small? 500pF seems very small to me.

2. Why use that 3.3M/10pF filter to mix the reverb and clean signals? Like high resistance means more noise so 3.3M seems huge. Why can't we just use a cap.

3. Never noticed but after the Reverb pot there is a voltage divider. So after you set your Reverb level a bunch more signal gets dumped to ground. Correct? Just seems odd.

4. For switch the reverb on and off I was thinking of breaking the connection going to the 500pF cap and also break the connection after that voltage divider going to the filter. Essentially the input and output would be floating and could not deteriorate tone like the stock switching does. Good idea?


Thanks!

[tubeswell]

Lets refer to this schem for Fender Reverb
http://lib.store.yahoo.net/lib/thetubes ... bab763.gif


1. Why is the input coupling cap so small? 500pF seems very small to me.

It is in front of parallel 12AT7 triodes. A 12AT7 is a high-current pre-amp tube, and there would be a lot more grid current than (say) a 12AX7 (which is possibly what you are thinking of for comparison?). Plus the inter-electrode capacitance is lower than it is for (say) a 12AX7. So these two factors pretty much mean that you ought to:

a) keep bass frequencies to a minimum to avoid blocking distortion (whereas a bigger cap there would have a longer discharge time and therefore be more prone to blocking distortion); and

b) that (even with a 500pF cap), you get the bandwidth you need in that stage (because of the lower inter electrode capacitance). Well that's my take on it anyhow.

2. Why use that 3.3M/10pF filter to mix the reverb and clean signals? Like high resistance means more noise so 3.3M seems huge. Why can't we just use a cap.

The 3M3 is part of a voltage divider, which in conjunction with the 470k in series with the 100k reverb level pot - both of the latter of which are in parallel with a 220k to ground) - form the leg of the divider. This attenuates the dry signal to a point where the wet signal can be heard well. Its all about balancing gain between dry and wet

3. Never noticed but after the Reverb pot there is a voltage divider. So after you set your Reverb level a bunch more signal gets dumped to ground. Correct? Just seems odd.

See above. The 220k is actually a tapering resistor for the 470k in series with the 100k linear pot. The resultant resistance of the level pot is still more than enough to give sufficient juice to the wet signal when dimed, and when you listen to these BF amps with the reverb level dimed, you can well understand that.

4. For switch the reverb on and off I was thinking of breaking the connection going to the 500pF cap and also break the connection after that voltage divider going to the filter. Essentially the input and output would be floating and could not deteriorate tone like the stock switching does. Good idea?

Why bother? The footswitch at the recovery stage's grid works perfectly adequately and is elegant and simple.

[ChrisM]

Cool, thank you for that info sir!


I was actually thinking of using a 12DW7 for reverb. 12AU7 triode for driving the tank cause it has a good power output and 12AX7 triode for recovery. So I guess with a 12AU7 driver I would want even smaller than 500pF?


My Reverb control is going to be 250K(A). I am trying to think how I would adjust those tapering and voltage divider resistors...

[tubeswell]

I was actually thinking of using a 12DW7 for reverb. 12AU7 triode for driving the tank cause it has a good power output and 12AX7 triode for recovery. So I guess with a 12AU7 driver I would want even smaller than 500pF?

A single 12AU7 stage doesn't run as much grid current as 2 x 12AT7 stages in parallel, so I would say if anything you may want to increase the 500pF to get the right sort of bandwidth, But you could start out with 500pF and tweak it from there. Home-brewed amps are 50% rocket science and 50% argy bargy. You might want to change your cathode resistor/bias point tho' - I'd try doubling rk to begin with

My Reverb control is going to be 250K(A). I am trying to think how I would adjust those tapering and voltage divider resistors...

Just chuck the 250kA in there to begin with and then see how you go with the other resistors. If anything I would possibly change the 470k series resistor for 270k. And leave the 220k in there - it'll only increase (and help to smooth the characteristic '2-line') taper that one finds in off-the-shelf audio pots these days.

2CW

[KeithStevens]

2. Why use that 3.3M/10pF filter to mix the reverb and clean signals? Like high resistance means more noise so 3.3M seems huge. Why can't we just use a cap.

The 3M3 is part of a voltage divider, which in conjunction with the 470k in series with the 100k reverb level pot - both of the latter of which are in parallel with a 220k to ground) - form the leg of the divider. This attenuates the dry signal to a point where the wet signal can be heard well. Its all about balancing gain between dry and wet

This is great info. I've read elsewhere that the 3.3M/10pF can be changed to a 2.2M/20pF to maintain the same resistance/capacitance relationship and slightly up the dry gain into the next stage and that it has been done with good results. But I know just enough about all of this to be dangerous. I happened to tack on those values to a order for some other stuff, but I haven't gotten around to trying it.

[Roe]

what should you do if you wanted less reverb? lovering the 3m3, and the other resistors part of the voltage divinder?

[Darkbluemurder]

what should you do if you wanted less reverb? lovering the 3m3, and the other resistors part of the voltage divinder?

Only the 3m3. The other resistors can stay as they are.

[Alexo]

I'm pretty sure that 500pf cap only serves as a high-pass filter to voice the reverb send. You really don't need to worry about grid conduction in your reverb stage unless you are seriously overdriving it, which isn't likely to happen until the rest of your amp is also overdriven. Look at the pi in the same schematic: .01 cap into a 12AT7 with a 2M input impedance (the ltp has a bootstrapped 2X input impedance, iirc). If you read the accutronics datasheets, they talk about the kind of freq response you want to drive their tanks with, but you can play around with that cap to get the tone you like. Warning though - too big and you can get low frequency feedback!

The 12DW7 sounds like a clever idea, but be aware that a 12AU7 takes about double the voltage to drive as a 12AT7.

[Gibsonman63]

I agree with the idea of voicing the reverb. IMHO, you don't want very much reverb on the bottom end because it will just muddy things up. I would think you would want a solid, relatively dry bottom with shimmering reverb on the mids and highs.

[C Moore]

tubeswell.....
I could not find "argy bargy" in the text book.

[David Root]

Old British expression, means "vigorous debate" (polite end of spectrum) to "coarsely expressed disagreement" (impolite end of spectrum).

[C Moore]

Old British expression, means "vigorous debate" (polite end of spectrum) to "coarsely expressed disagreement" (impolite end of spectrum).

Well stated.
That term is definitely a new one for me....
Thanks

[tubeswell]

Of course what I meant by it was 'a mixture of technical discombobulation and bluff and bluster' - but 'rocket science' and 'argy bargy' flowed off the tongue better.

[orrong65]

Great discussion!

I modify the return reverb mixing resistor, be it 3.3M or 470k, to get the right amount of reverb in the blended signal. On some amps the reverb is too passive for my liking even at full on.

An interesting point is that the Fender brown Vibroverb had spectacular reverb that went on forever, and this was due to different circuitry - a fully cathode bypassed 12AX7 driver, although the output blending resistors are both 470k.

[Bob-I]

Great discussion!

I modify the return reverb mixing resistor, be it 3.3M or 470k, to get the right amount of reverb in the blended signal. On some amps the reverb is too passive for my liking even at full on.

Be careful here, changing the mixing resistor also changes the gain structure. I subbed in a 2.2M in place of the 3.3M because I had one on hand. The amp sounded brittle and had a nasty distortion even at low volumes. Changing to the 3.3M cured these problems. I'm sure that modifying the gain structure would've had the same affect.