ROG Britannia "cabsim" bypass

[swever]

I've boxed my britannia yesterday and after some testing I think I'd like to try to bypass the  "U1b that implements the ultimate toneshaping: a 200 Hz resonance characteristic of a 2x12" cabinet and some high frequency rounding."

Am I right in thinking that I can achieve it this way?

[PRR]

Yes.

[swever]

Finished soldering the switch a moment ago and hmm.. it somewhat does what I expected but not exactly. First of all, surprisingly, switching the mod on lowers the output (as opposed to making it louder as I thought it would). It does indeed remove the low hump and cleans up the highs, making the effect a bit more "transparent". However it still sounds a bit muffled with gain closer to max or when I put another overdrive/fuzz in front of it – must be the other lowpass filters in the circuit.

I'm totally digging this circuit at lower gain settings though and the "mod" allows for a nice variation.

[robthequiet]

It sort of makes sense if you are replacing a feedback resistance with a jumper, as that would make for a unity-gain buffer, although with positive feedback with the 1n cap going back before the 47K resistor. Just for kicks maybe add a resistor in series with your switch, maybe a 47K or thereabouts, and see what that does.

[PRR]

> switching the mod on lowers the output

> "200 Hz resonance characteristic of a 2x12" cabinet..."

"Resonance" is a bump. A lot of things about a 2X12 come together around 200Hz and it bumps-up. This is also where the ear tends to hear "body", the "total size of a sound". Take that out, yes it should be "less".

> still sounds a bit muffled

The ??K, 1n, 47K, 330p are a high-cut filter, emulating part of the natural high-cut of large cones, necessary to cut the inharmonic overtones of strings reproduced too faithfully.

It *may* interact with the 200Hz thing, since the hi-cut uses feedback from the amp, but 200Hz and ~~5KHz(?) are quite far apart so I'd think the opamp is unity-gain where the hi-cut works, either way.

I don't see how the Volume knob can affect it.

[ricothetroll]

Why not cleanly "true bypass" the whole stage between the output of U1a and before the 1uF coupling cap before the volume pot ?

[swever]

Why not cleanly "true bypass" the whole stage between the output of U1a and before the 1uF coupling cap before the volume pot ?

Is that what you mean?

[samhay]

^I would leave the 220n connected and switch the + side of the 1u cap between the output of U1b and the output of the op-amp preceding it (left of the 220n).

[swever]

Would not that also leave out the cap, two leds and a resistor around U1a?.

Btw is it correct to say that these components are in U1a's negative feedback path?

[Ice-9]

It looks like you have drawn the switching incorrectly, when open nothing is connected to the output ?

[swever]

It must be my drawing, haha. The idea is that it bypasses the U1b and all the stuff around it, as ricothetroll suggested.  The output is connected either to U1b or U1a.

[ricothetroll]

Hi,
Like the lower one on this picture :
http://www.geofex.com/article_folders/bypass/bpdiag6.gif

[swever]

ricothetroll, do you think the position of the "first part" of the switch is right?

[ricothetroll]

I think what I would do is this :



Add those two 1Meg resistors to Vcc/2 to fix the potential of the "floating" nodes and avoid pops when switching. Just solder them at the bottom of the PCB or drill some additional holes to fit them in.

[swever]

Finally made the mod and now it does what I wanted it to do. No more "blanket over speaker" effect. I ended up not adding the pull-down resistors - seems to work fine. Actually, I think I will omit the switch as well and just hardwire the circuit without this "cabsim" at all.

Now I decided to check my biasing because I was also getting some nasty gating an notes' tales. Something is apparently not very right with my build.
I am unable to set Q1 drain voltage below 7 volts, and Q2 above 4.3V. Does that mean that my jfets are way off the required specifications or is there something else wrong with my build?

[swever]

BUMP

Sorry for that 

Any ideas why can't I bias it?

[PRR]

> unable to set Q1 drain

Is there a plan which shows these JFETs?

[swever]

Yes, I forgot to post links. Here's the project and the layout I used.

[PRR]

Q1: put 5K or 10K in series with the "5K trim" and see if it pulls into range.

Q2: can't get over 4.3V suggests a build error. The published values should let you jam it over 7V, so something is wrong. Post *all* key voltages on Q2 and Q3.

You may eventually have to replace FETs to find ones that suit these values, or re-value all the resistors to suit the FETs you have.

[swever]

Takes me forever to do anything 

Changed the "trim q1" to 10k and now I can get 6V there.

Q2 voltages:
D: 0.02V ("trim q2" at min)  -  3.82V ("trim q2" at max)
S: 0.00V ("trim q2" at min)  -  1.80V ("trim q2" at max)
G: 0.00V

Q3:
E: 0.86V ("trim q2" at min)  -  3.91V ("trim q2" at max)
B: 1.44V ("trim q2" at min)  -  4.49V ("trim q2" at max)
C: 7.97V ("trim q2" at min)  -  7.90V ("trim q2" at max)