Classic Fender 3.3M/10P Reverb Isolation Components

[tdvt]

Recently, I gained (no pun really intended) an appreciation of how hard the regulation Fender pre-amp (2 triode) can hit the next stage if there isn't much/any attenuation.

I was working on a Bassman for a friend, that had previously been partly AA864'd & I ended up changing the voltage divider ratio at the grid of the 3rd stage to clean things up a bit, as he uses pedals for drive & overall lower volumes (added a LarMar as well).

Anyway, I just powered up & I am sorting a recently completed hybrid PR/BLJr-ish amp, built on a Hoffman PR board (AB763 TMB & reverb into LTPI EL84s) & noticed static-y hash appearing at the 3rd gain stage grid.

While totally usable, I would like it quieter, & while thinking about the Bassman, I started wondering about the purpose of that 10P cap paralleled with the 3.3M, as it seem that it would be delivering completely unattenuated high frequency to the doorstep of that next stage.

What is it's purpose. Is it sonic or otherwise?


PS

I DID disconnect the cap to try & it seems better, but waiting on a new tube-set to see where things really are.

[Dave]

Bass frequencies will plow through that resistor like a juggernaut. High frequency content will get stifled. The purpose of the capacitor is to give the high frequency content an easier path than through that resistor.


Dave

[sluckey]

Bass frequencies will plow through that resistor like a juggernaut. High frequency content will get stifled. The purpose of the capacitor is to give the high frequency content an easier path than through that resistor.

A resistor is not frequency sensitive. It will knock down the bass just as hard as it knocks down the treble.

[Dave]

Sorry Sluckey, I'm no expert, but in practice, that's what I am compelled to believe. Take for instance a notch filter (might also be called a T filter) (the kind with series resistors). The only way I understand its function is to see the high frequency stuff go right through the parallel capacitor. The bass frequencies don't need any help, and the mids get shunted at the junction of the resistors.


It's just the way I visualize it I suppose.


Dave

[tubeswell]

… 10P cap paralleled with the 3.3M …

What is its purpose. Is it sonic or otherwise?

The 10pF allows high frequencies to pass. The 3M3 is in series with the Miller capacitance of the following triode, which forms a filter that rolls off HF. The 10pF keeps some high-end clarity.

[tdvt]

Appreciate the replies.

Looking around, I did find some circuits that used just a resistor for reverb send/return isolation, but the 3.3M/10P is definitely the favored design.

I will feel free to tweak... if needed.

[sluckey]

Just keep in mind, the lower the value of that mixing resistor, the stronger the dry signal becomes which makes the reverb seem weaker. Some time back there was a classic example on the forum. The complaint was very weak reverb, almost nonexistent in a new build. The culprit was a 3.3K rather than a 3.3M.

[AlNewman]

I believe the 10pf cap works in conjunction with the 3.3M resistor to allow high frequencies through.  As Tubeswell pointed out, the 3.3M works with the Miller capacitance to knock down high frequencies.

So, (and this is just searching Merlin's page), a 12 ax7 has a miller capacitance of say 200 pF.

3.3M @ 200 pF would be around 250 hertz roll-off.

A 10pF cap bypassed over 3.3M would allow everything above say 5 kH through.

5kH is kinda in the "sparkle" range.

Nothing's stopping you from putting a grid stopper in series with the 3.3M/10pF to work with the miller capacitance to dial out harsher frequencies above 5kH. 

I bet there's all sorts of options to get a frequency you're looking for.  Although dialling out high frequencies to deal with static type distortion may be more like a bandaid, there could be a different issue.

[tdvt]

Plenty of reverb as it sits, just somewhat noisy, but I am hoping it's at least partly tubes. I am using random stuff from the box just to get it going, the EL84s are especially microphonic & mismatched for current.

New set on the way.

Hadn't thought about a grid-stopper there but looks like a few ways to tune that part of the circuit.

That whole reverb recovery/3rd gain/mixer section seems to be where I encounter the most noise in the amps I am fooling with.

[pdf64]

... the 3.3M works with the Miller capacitance to knock down high frequencies.

So, (and this is just searching Merlin's page), a 12 ax7 has a miller capacitance of say 200 pF.

3.3M @ 200 pF would be around 250 hertz roll-off.
 ...

But Miller capacitence interacts with the effective source resistance of the signal at the grid.
That's typically made up of roughly
3M3 // 220k // 470k