Hi soulsonic
good to see you here too!
As you probably know I have been playing around with the SHO circuit using a BS170 mosfet and simulating responses with Multisim 9.
I have also built a few of these in the past and had no problems biasing them - I have not tried 2N3700 types though, only BS170. Are you sure you have connected the pins correctly? Looking at the flat face with the writing on it towards you (and legs at the bottom), it is Drain Gate Source.
With a 5k1 (or 5k6 for that matter) connected to the Drain and a 5k source resistor, with two 1Megohm resistors as per the SHO, you should see approximately +8.7volts on the drain (+ 9 volt supply), and +2.1 volts on the gate ( with no signal). Of course you won't get a lot of gain from this (approximately unity gain).
However if you reduce the 5k source resistor to close to zero ohms, you will get at least 30dB of gain from it, and the drain will now be +7.8 volts and the gate bias voltage will drop to +1.89 volts.
The above are simulated examples relating to the SHO gain pot at both minimum and maximum, but in practice they are fairly close to predicted voltages. As you can see they do bias quite differently to Jfets or Bipolar silicon devices!!
A further interesting idea is to increase the drain resistor to 22k and the source resistor to 18k (drain=8.35v, gate = 2.2v)
As you can see, the drain and gate voltages are very similar BUT simulation shows a slight reduction in the high frequency gain - It does not sound as glassy, in fact it sounds a lot more like a triode tube. I have not bothered to try using 100k for a drain load - not yet!! But as you can see the circuit will work with many different values besides the SHO ones!
Changing the 1M bias resistors to higher values has no major effect on the sound of the circuit ( eg: 2.2M, 3.3M, 4.7M or even 10M - as long as they are BOTH the same value - you may notice slightly more bottom end with the higher values, not much though!!).
In the original example though (2 x 1M), if you change the drain to gate OR the gate to ground resistor, you will notice quite a different sounding circuit - in fact you can get very large CLEAN gain with the appropriate changed resistor (not both!!).
The SHO circuit values deliberately exploit the less than ideal bias and allow the mosfet to operate close to its pinch off or cut off voltage though - this is why it generates those lovely 2nd harmonic distortion components as the gain is increased - Marshall have deliberately exploited less than ideal biasing in some (not all) of their triode valve stages way back in the PLEXI era - nothing new eh!
If you want further clarification on how the MOSFET operates just ask Analogguru - he's the MAN
Cheers
Steve
when source resistor is smaller, dc bias is smaller also. at 470 ohms is 4,5v at drain connection
