MOSFET FX Loop with LND150

[GainFreak]

Hi,

I need a simple FX loop for a tube preamp with -10dB/0dB switch In some of their latest amp Marshall are using this SS FX Loop:



Since a single LND150 can now be used for recovery stage I was wondering if I can use it as a virtual earth mixer with small amount of gain (overall output level from the tube preamp should be 0dB max so max gain will be -10dB->0dB). I "modified" this schematic in Photoshop to be used with LND150:



The source follower is standard but I was wondering if I would need R27 and R29 in the recovery stage? Also is recovery stage output impedance going to be low enough? I guess I will have to determine experimentally the R33/R34 divider for the gain I need and C21 for flat frequency response.
Any comments would be appreciated.

[GainFreak]

Anyone? R.G.?

[Steve Conner]

As you said yourself, it is a virtual earth mixer. So R29 is needed to set the gain (it's roughly equal to (R33+R34)/R29)

And R27 is needed to provide a DC bias path for the MOSFET gate.

The output impedance can be calculated from feedback theory:

Open-loop gain of the MOSFET stage: A=gm*R31.
Open-loop output impedance: Zout=R31.

Feedback factor B=R29/(R29+R33+R34) (depending on level switch position)

And if I remember right: Zout with feedback = Zout/(1+AB)
(You do the math )

Note that this gives the small-signal output impedance. You also have to ensure that the large-signal conditions are met. For instance, if the MOSFET is biased at 1mA, then you can't expect to drive more than a 10V swing into a 10k load, because it will turn off completely on positive peaks when the load current reaches 1mA.

[GainFreak]

Thanks for your reply Steve.

I read Aiken's article about the virtual earth stage but those schematics don't look quite the same:

Designing Single-Stage Inverting Feedback Amplifier

Sorry my theoretical background is poor and I don't mess with transistors very often. This loop is intended as a quick solution for a tube preamp and the max gain of return stage needs to be around 4 (-10 to 0dB) so it will not swing large signals.
If we assume a gm of 1.5 according to your formula it will give me open loop gain of 150. Open loop Zout - 100k. However I'm a bit confused about the FB factor as I get 10/(10+220k)=0.043 which produces Zout with feedback = 100/(1+150*0.04)=100/7=15k? Isn't that stage supposed to have lower output impedance than a tube based one?
Also I'm not sure which resistor sets the input impedance - R27 or R29?

[tedmich]

Mojotone makes one of these in a kit and JeffW over at AX84 has a good one shown here
Free Cloud Storage - MediaFire
an discussed here
AX84.com - The Cooperative Tube Guitar Amp Project

[Steve Conner]

Your math seems OK. If the output impedance is too high for you, remember that it'll go down as you decrease the recovery gain, and MOSFETs with higher gm than the LND150 are available.

R29 sets the AC input impedance at 10k, because it's connected to a virtual earth. (the DC input impedance will be set by R27 because the feedback is AC coupled)

[GainFreak]

I guess I'll need higher value for R29 , maybe 100k and to make sure it won't load any effect's output and will have to tweak the FB resistors accordingly.
I was hoping for lower output impedance to avoid sound degradation over longer cables (it's a preamp out after all) but it looks like I'll have to add another buffer after that stage which will allow me to add a return level pot between them as well.
Thanks for your help.

[GainFreak]

Hi,

After the theory comes the soldering This is a schematic of my already working tube preamp FX loop:



I would appreciate your comments if you see mistakes and/or room for improvement.

[GainFreak]

There's one thing that puzzles me. With two 47k resistors in the source of the first buffer and an input signal of ~26V from the tonestack I get ~4.5V of output. At the source I get the same input signal of ~26V.
Shouldn't I be getting half of that signal "between" the two 47k resistors?

[imaradiostar]

I'm confused why you need a virtual earth mixer. The lowish input impedance is a bad thing in my book.

I've always tried to build loops so that I could plug a guitar straight into one and I'd have enough gain to drive the output section. To that end why not just wire up a normal common source LND150 circuit? You could get a gain of around 100 which should be plenty and you can reduce it by only partially bypassing the source and having a pot at the output. For that matter I find pots to be more useful for send and return. I'd make the send pot internal and the return pot the "master volume" for the whole amp. Lots of ways to do this, I guess.

I would usually pick R29's value based on the desired rolloff in conjunction with the gate capacitance of the LND150. For example, in guitar amps the value is chosen based on the grid capacitance of a 12ax7.

Jamie

[GainFreak]

I need the VE configuration because this is not a regular amp loop. It's intended for a floor tube preamp with output level of 0dB and the max gain of recovery stage that I need is from -10dB to 0dB. The preamp itself is connected to a separate amp with input sensitivity of 0dB. I need the low impedance buffer to be able drive long cables without sound degradation.

[klingo]

Hi,

After the theory comes the soldering This is a schematic of my already working tube preamp FX loop:

[ATTACH=CONFIG]19146[/ATTACH]

I would appreciate your comments if you see mistakes and/or room for improvement.

Hi,

i dont see the first recovery stage drain resistor on this schematic, i guess you use a 100k?
also, on this stage i'm not sure that the 100n cap between the two 470k resistor is needed, but a zener or two might be necessary to protect the lnd150 from statics.

nice drawing anyway.

[GainFreak]

i dont see the first recovery stage drain resistor on this schematic, i guess you use a 100k?

Thanks for pointing that out.
I have the zeners on the recovery stage, no zeners on the SF.

The corrected schematic:

[Kagliostro]

Hi GainFreak

I'm really VERY interested in your FX Loop

Please have you an image of a PCB for the circuit ? Can you post it ?

Many Many Thanks

You have a PM

K

[GainFreak]

I didn't draw a PCB but used a punchboard instead. I should draw a PCB for it to be available for those that don't want to buy "that" kit.