Marshall JCM 800 2204 FX loop advice

[Marco Pancaldi]

Hello,

I need a send-return insert in my amp, searched the net but I haven't found the perfect solution yet.
The classic approach is to insert the loop between the master volume and the .022 cap before the PI.
My idea is use only a MosFET to buffer the send from the master volume pot and rely on the low impedance output from pedals to drive the "power amp" (the .022 cap) return section.

I found several circuits (notably from RG Keen) about using MosFET in tube amps, but I'm not sure about details in this case: if I need a 1/2 V biasing on MosFET input, an input resistance about 220R to control very high frequencies, a 12V zener protection etc.
Any advice and comment welcome.

Thanks
m.p.

[wizard333]

Mojo makes a pre-wired kit for just that purpose, buffered to iirc.

[defaced]

This isn't a complete schematic (you'll need protection zeners most notably), but it is instrument level, unity gain, series loop that can can drive down to 60k loads and is flat across the audio spectrum. This is the schem from the simulation, but in prototyping I changed the following values (from memory, I can't find my notes at the moment). It uses an IRF820 as the send (this isn't the only FET that can do this, others can), and a LND150 for the return (not many other FETs can do this, but there are some, but the stage will need to be reconfigured for an enhancement mode device):
R3 - 100K
R9 - 680R
R1 - 100K
R2 - 270R

[wizard333]

that can can drive down to 60k loads

The problem I foresee there is that many studio devices have a 10k input load, or am I missing something?

[defaced]

This was designed with pedals in mind. At first glance I don't see why one couldn't reconfigure it to work with line level and studio impedances if necessary, but I'd have to play with it before I'd feel comfortable giving a firm yes.

[Marco Pancaldi]

Many thanks! Three questions:
- I will need probably a little RC (about 10K - 10u) to decouple the fx loop powering from other amp stages, right?
- And, the correct biasing is assured from the 1M couple on the IRF820 input and independent from V+, not so for the recovery stage that is biased for gain, right?
- In case I'd wish a different ratio between output attenuation/input recovery gain, how simple is to calculate the recovery stage resistors for a different gain and bias?

This isn't a complete schematic (you'll need protection zeners most notably), but it is instrument level, unity gain, series loop that can can drive down to 60k loads and is flat across the audio spectrum. This is the schem from the simulation, but in prototyping I changed the following values (from memory, I can't find my notes at the moment). It uses an IRF820 as the send (this isn't the only FET that can do this, others can), and a LND150 for the return (not many other FETs can do this, but there are some, but the stage will need to be reconfigured for an enhancement mode device):
R3 - 100K
R9 - 680R
R1 - 100K
R2 - 270R

[defaced]

Yes, this will need it's own filter node. I'd T it off of the PI node so it has as high of a supply voltage as possible. Take voltages before and after so you can adjust the screen to PI node B+ dropping resistor to bring the PI and preamp voltages back up. I tested this circuit at various B+ voltages from 250v to 400v and it performed well at all voltages.

Correct. The the IRF820 is an enhancement mode device, so it needs to be turned on by a positive gate to source voltage. The LND150 is a depletion mode device, so it needs to be "turned off" some by a negative gate to source voltage. So yes, they bias up using different methods.

After I did the simulation, I prototyped the stages and adjusted values as needed (or put another way, with regard to gain, the simulation was incorrect). I'm sure there's a way to calculate the gain, but it's just as easy to sub in different source/drain resistors on the recovery stage to get the gain you want. Just like a tube stage, smaller source resistors and/or larger drain resistors give more gain. Right now it's set to take a 120v P-P incoming signal, divide it down to approximately 0.8v P-P, then bring it back up to 120v P-P. If the stage has too much gain, you can also just divide down the incoming signal (volume knob on effect) to get things in check.