FET circuits - topologies, user experience, etc

[Delta Sigma]

What are people's experiences with different FET circuits? Looking for opinions; not concerned if you use vague non-technical terms like colour; words are hard sometimes.

I'm looking at simple circuits like U87/KM84, 414EB P48, "Schoeps", etc and the many-semi U47FET.

I was comparing the U87/KM84 circuits to see which would be better for a simple "here's my capsule" type circuit and realized they're the same circuit. Seems like only a 3.3p drain to capsule cap, bypass cap value, and transformer size/ratio are different.

The 414EB P48 is pretty simple too. Not sure I get the output stage (is the FET constant current for the BJT?).

My other rabbit hole I'm looking at, is the idea of a cathode or source follower mice, specifically for loud sources like drums. But that's for another day!

--------------------------------------------

Useless background fluff words:

I'm pretty clueless when it comes to semiconductor circuits. A little strange, since I did learn it in an electronics diploma program. Things tend to fade after 20 years though.

I got into mic building out of an interest to learn. I started with tube circuits mostly because of bias from vintage lore. I've built some fantastic tube circuits and love them. Built a 47FET out of leftover components and I really don't care for the result. I used a capsule I'm not a huge fan of, and a toroidal transformer. Is it the capsule, the transformer, or the circuit I'm not big on?

All my winter projects fell apart this year, so I figure it's time to research/plan the direction for 2025/26 winter projects.

 

[untamedfrontier]

Have you played around with the ORS87 at all? That's a pretty simple version of the 87 circuit and can be tuned to preferences, although probably not the "here's my capsule" transparency that you are looking for.

 

[Delta Sigma]

I haven't tried the ORS87.

I guess I should rephrase the question, to what the difference is between some major FET design choices like constant current biasing and BJT output stage before a transformer? I see that the ORS87 project is a good platform for experimenting with some component changes - I haven't read the whole thread though; just some skimming.

U87/KM84 - plain old voltage divider resistor bias
414EB P48 - constant current bias of FET, plus a BJT stage before the transformer
U47FET - constant voltage bias of constant current current for FET, plus a BJT stage before the transformer

I get the feeling that people like the KM84/U87 simplicity for sources like vocals, and constant current biased FETs for more accurate sounding mics, but that's from piecing together various comments and making assumptions.

Also, why the BJT output stage before the transformer on 414/47FET? Simply for the filter on the 414?

 

[micolas]

Hi @Delta Sigma ,
First of all, do you have a personal preference for the sound, the character, of a particular microphone?

 

[Delta Sigma]

Hi @Delta Sigma ,
First of all, do you have a personal preference for the sound, the character, of a particular microphone?

Well, yes and no? I like a lot of microphones for different reasons. When it comes to, say 414 vs U87, etc it can be a hard to compare circuits because of them having very different headbaskets and capsules.

Specifically, I'm looking to learn about FET circuit design. Why bias a FET with resistors and a pot? Why use constant current? Why have an output stage, then a transformer?

My end goal, would be to have built multiple different circuits, then use the same capsule and headbasket and observe differences. The problem is there's never enough money or time. There's a lot of mic builders on the forum, I was hoping to find some feedback from those who have built multiple FET designs.

If I want to design a circuit, where do I start? If I want to put switchable filters in the circuit, is it wise to have an emitter follower before the transformer?

 

[micolas]

Well, yes and no? I like a lot of microphones for different reasons. When it comes to, say 414 vs U87, etc it can be a hard to compare circuits because of them having very different headbaskets and capsules.

Specifically, I'm looking to learn about FET circuit design. Why bias a FET with resistors and a pot? Why use constant current? Why have an output stage, then a transformer?

My end goal, would be to have built multiple different circuits, then use the same capsule and headbasket and observe differences. The problem is there's never enough money or time. There's a lot of mic builders on the forum, I was hoping to find some feedback from those who have built multiple FET designs.

If I want to design a circuit, where do I start? If I want to put switchable filters in the circuit, is it wise to have an emitter follower before the transformer?

These are very good questions.
Try to read carefully, most of the answers are already here, but you have to search patiently.

 

[Delta Sigma]

These are very good questions.
Try to read carefully, most of the answers are already here, but you have to search patiently.

I have found some circuit analysis: what components are meant to do what, etc. What I have not found is people's personal experiences with different circuits. One mic comparing to another is not a measure of what circuit changes make. Capsule and headbasket are likely what people prefer when they hear a mic (disregarding filters). Most threads here are mic build threads based around classic circuit designs. Build a U47, 414, etc and swap a K47 for an M7, then a CK12 and you can hear the differences between capsules in that headbasket & circuit. My experience building some tube mics and swapping capsules between them has revealed some interesting differences in tube circuits. A lot of that translates to FET circuits, but self bias vs heater derived bias vs C12 style B+ derived bias differences likely to not translate to FET circuits.

 

[Khron]

and swapping capsules between them has revealed some interesting differences in tube circuits.

Swapping capsules revealed differences between circuits? Please explain