freestompboxes.org

ppluis0 wrote: ↑09 Nov 2022, 16:59 Hi folks,

In the very first amplifier stage of Dumble amps was an option to employ a fet rather than a half of 12AX7:

I think that the combination of 1Meg and 100pF in parallel are to compensate the mentioned Miller effect...

Cheers,
Jose

It's an interesting idea, reminiscent of oscilloscope probe compensation. There are some observations to make though:

• The JFET is effectively run from an unregulated low voltage supply. This might not be much of a problem since clipping is still unlikely.
• Based on the datasheet transconductance, the gain of that stage is less than half of a bypassed 12ax7. Not much of a problem in an original design, but something to be considered to swap in place of a triode.
• The shelving network does flatten the frequency response (even with the relatively low capacitance of a JFET, considering Miller effect 20pF isn't far off for this one), but does nothing for the dropping input impedance. Honestly, the loading is negligible in this case, making just a 2dB difference, but I suspect that's not the case for a MOSFET.
• The 1M resistor in series with the signal isn't shunted for most of the audio range. This is pretty bad for noise and probably my biggest complaint.

Hi Dylan,

There is another example using a fet input in an amp Made by Seymour Duncan long time ago, named "Convertible Amp" that has a row of edge connectors in their base mainboard to plug and interchange the order of different types of modules at each preamp stage.





I will dig in one of my pile of papers (AKA archive... ) as I'm sure that there are schematic of one of these modules that is a fet stage.

Cheers,
Jose

Or this mixed FET/BJT Preamp of the AMT Electronics - SS-20 topic.
viewtopic.php?t=7694&hilit=amt

I searched my CD archives and found the Seymour Duncan Convertible files.
Here is the JFET module.
B+ = 400VDC

The MPF102 is a replacement type for the ECG312.

Here the socalled hot mod module from the Seymour Duncan Convertible.
The FET symbols in the circuit diagram are drawn incorrectly, both are N-Channel Mosfets.

I think the AMT jFet/BJT cascode circuit probably has potential. Notice that they are overdriving it in the second stage of that SS-20, that means it can overdrive. We all know AMT doesn't mess around with half ass tone.

Edit: fixed the auto-correct "cascade" to cascode.

bmxguitarsbmx wrote: ↑11 Nov 2022, 05:10 I think the AMT jFet/BJT cascade circuit probably has potential. Notice that they are overdriving it in the second stage of that SS-20, that means it can overdrive. We all know AMT doesn't mess around with half ass tone.

Yes, that is true. I ran a simulation and the result was that the positive sine half-wave starts to round out at about 1.1 Vpp and starts flattens out at about 1.35 Vpp input signal level.