MOSFET rules, formulas and guidelines

[carpenter]

Hi guys,

I've searched for a concise, consolidated thread of information -- most of what I seek is scattered like precious gold flecks in a creek-bed.

I'm hoping that the brains in our group can use this thread to explain to the budding amateur FET users in this forum how to properly use the MOSFET. How are they similar or different from the bipolar transistor? What formulas can be interchanged, or used when modified?

Examples and associated formula would be most welcome. These examples should be well written and all mathematical procedure laid out. Use and explanations of Ohm's law are paramount.

Then there are the practical "tricks-of-the-trade" -- what rules must be followed to avoid disaster.

What do you think; is this a viable project? Could it be a "permanent" thread?

Don't be shy...

all the best,

John

[lineup]

hi

good idea, carpenter

sure there is knowledge in this forum at absolute highest level
even compared to anywhere else in this world ... !!!

now one thing,
why don't you throw in a practical example of something you have in mind
some kind of amplifier like design
stupid or crazy or more sensible does not matter

because by and by the brainstorming of clever MOSFET posters
will slowly bend your MOSFET creation into something nice
while we go along

who knows
might be one of those everlasting projects topics
that already are among those
www.diyaudio.com Pass Labs forum Classicals


What you say, my friend?
Isn't this a good idea, say?


lineup
a long time member
with several 1.000 posts here around,
even though a long pause was forced upon me
for some reason

[carpenter]

... and I do read your many posts with avid interest.

What I was hoping for was our own version of "PRACTICAL MOSFET SOLID-STATE CIRCUIT DESIGN".

For example: take a single FET and add a voltage divider bias network, then add drain and source resistance. With great care, explain how the voltage and amperage is calculated from known values -- or how to find an unknown. Also, show how to calculate the necessary resistance to set an operating point. Explain the various operating points.

This can be short, sweet and concise.

It could also be a bit like N.P.s Zen articles, but without a specific end product, just the concepts -- well explained concepts.

Basically, I would like to see us design our own handbook that looks and smells like a bipolar transistor handbook, but with MOSFETs and JFETs in mind.

The really neat thing? Any one of us could design anything because the knowledge base is all located in one place. No more skimming and extrapolating N.P.s zen articles, or our hastily scribbled notes. No more wondering where that beautiful little tid-bit of information is located at DIY Audio...

If this were done properly, I can see this becoming the reference base the DIY world craves.

I'd also like to discover a way for us to create donations to assist DIY Audio. Maybe a publication?

This is such a cool forum, I'd like to see it continue to grow and have even more features to benefit the members.

John

[mlloyd1]

for jfets, some good material has already been prepared by an expert:
part 1 is here: http://www.borbelyaudio.com/adobe/ae599bor.pdf
part 2 is here: http://www.borbelyaudio.com/adobe/ae699bor.pdf

mlloyd1

[Nelson Pass]

It's not really a complex subject, and if you go back and
read my articles starting with the Mosfet Citation 12, you'll
see most of the issues covered in the little tutorial ghettos
embedded in the articles, at least for Mosfets and power
JFETs, and you can add those to Borbely's comments.

www.passdiy.com

[lineup]

Originally posted by carpenter

What I was hoping for was our own version of
"PRACTICAL MOSFET SOLID-STATE CIRCUIT DESIGN".
-------------

It could also be a bit like N.P.s Zen articles, but without a specific end product, just the concepts -- well explained concepts.

Basically, I would like to see us design our own handbook that looks and smells like a bipolar transistor handbook, but with MOSFETs and JFETs in mind.
-------------

If this were done properly, I can see this becoming the reference base the DIY world craves.

My suggested idea in my first post,
was that
you throw in a practical, maybe beginners like stupid, real world design example..

Why did I suggest this?

Yes, I know from experience from this and other forums,
that such a general request for theoretical stuff, formulas and good guiding rules
won't attract posters at all.

Most people are not too much bothered with abstract formulas.
They want some real stuff to relate to.
And so eventually get interested and post some good stuff.
For your very good idea to be a success, it is my 100% opinion,
you would need to supply something to get people going.

Could be:

.. a SCHEMATIC of a horribly bad MOSFET circuit - for people to get upset from and so correct you

.. an IMAGE of a lonely MOSFET TO247 N Channel, waiting for some circuit suggestions

.. a LAYOUT of PCB circuit board, with one input resistor to a MOSFET GATE + 2 DRAIN, SOURCE resistors to set Idle Current and Gain

.. a List of DESIGN GOAL SPECIFICATIONS for a power amplifier:
* Two IRF240 / IRFP240 for output stage
* Drive my 4 Ohm speakers
* Minimum 10 Watt output
* Class A
* Input impedance >= 10 kohm
* Simple circuit
* Basic protection function to avoid damage my Speakers
.

carpenter, I think your idea is super!!
This is why I do not want it to die.
So, I give my best advice to try get good activity in this thread.


lineup is certainly not like most people .. but maybe knows what will activate average diyaudio members to post
http://lineup.awardspace.com/

[carpenter]

Nelson, I appreciate being turned on to the "Citation 12" article. In fact, I'd forgotten about the "legacy projects". I'm going to reread all of them.

I'm fascinated by the bipolar publications I've been devouring. I'm hoping the MOSFET isn't any more difficult to calculate for than the bipolar.

mlloyd1, thanks for the lead on the Borbely articles.

Hey lineup, you're terrific. I appreciate your enthusiasm more than you can imagine. Perhaps it would be useful to dissect very simple, popular amplifiers and explain the math that went into their resistor values. Nelson does a wonderful job of suggesting what values to use, but I'm sometimes at a loss with his reasoning; this is NOT his fault, just my ignorance. I'm improving, but its taken years; mostly because this is a hobby , not a necessity. Also, there's a thing called "raw talent". I'm a much better carpenter than an electronics major.

At least this thread has offered some reading samples for everyone interested. Perhaps more links can be added as time goes on.

Once again, thanks everyone,

John

[lumanauw]

I have a strange problem. I use mosfet (IRF9610) for VAS. The schematic is like below.
The differential is Jfet, K30GR. Each is biased about 2.75mA. The differential is cascoded about 25V by Q3-Q4.
The load is current mirror of Q5-Q6.
This differential is driving VAS mosfet Q7=IRF9610 without source degeneration. Q7 has 220ohm base stopper and 20pF miller cap to advoid local oscilation.
The sound of this is not normal (bad). Later I read in here MikeB wrote that if the configuration is like this (differential has current mirror load on top, and driving mosfet VAS), the gain is infinite at DC, since at DC mosfet don't draw any current, while current mirror is high impedance output (collector-collector).

The sound is like have very much intermodulation.

Later I tried to load the VAS drain with R to ground, not much help.

A help comes when I put R1 (// with miller cap). The value is 220k. The sound is getting better (but not like what I expected).

1. What is happening here?
2. Is Mosfet suitable for VAS? Mosfet has big capacitance, while it always have +/-voltage swing (internal cap varies badly) in the VAS position, VDS varies badly without cascoding. I've seen some commercial amps with mosfet VAS without cascoding, how can they do that?
3. If it is suitable, is it bad if the differential has current mirror load for driving mosfet VAS? Do I need a buffer before mosfet VAS gate or something else?

[flg]

With 2.75mA through your mirror source Rs you drop .6V or so, + another .7 or so through Vbe. So your Mosfet has 1.3V at the gate? Don't you need about 4V at the gate?

[lumanauw]

Hi, FLG,

Between Q3's collector and Q5's collector (=Q7's gate drive voltage), the voltage is floating anywhere (collector-collector), not fixed, but surely about the on voltage of the mosfet (3V5-4).