Hi , I was making experiment of the schematics below in the picture with Vdd = 12v , R=8.2k and couple of IRFZ44N. The problem there is ,
since the Vth is not as predictable as Vbe in BJTs , current mirror did not worked well. I have tried different IRFZ44Ns and the output current was varying.
I do wonder , what should I expect , what should I care more when desing with mosfets. Can you give me some tips ?
I also was wondering , since Vth is not constant and datasheets dont give the value of (unCoxW/L) = k but gives the minimum gm value , how can I calculate the current in a simple mosfet circuit for example biasing IRFZ44N with 3K load and 1mA current biased from 12v VDD ?

 

As you note, the mismatch between two discrete MOSFETs' Vgs(th) can make a current-mirror with widely varying currents, making doing that problematic.
They can use MOSFETs for a current-mirror on an IC since they are much better matched when on the same chip.

If you want to make a discrete device current-mirror than use BJTs,
You can buy matched BJTs in a single package for an even better current match.

 

As you note, the mismatch between two discrete MOSFETs' Vgs(th) can make a current-mirror with widely varying currents, making doing that problematic.
They can use MOSFETs for a current-mirror on an IC since they are much better matched when on the same chip.

If you want to make a discrete device current-mirror than use BJTs,
You can buy matched BJTs in a single package for an even better current match.

Thank you so much , I like learning practical paths

 

I was making experiment of the schematics below

Don't see a schematic.

Can you give me some tips ?

This method isn't suitable if you'll be making more than a few of whatever you're planning to make.

Sort through enough devices to find 2 that have the same threshold voltage. When you measure threshold voltage, make sure you short the gate to drain as the manufacturers suggest. I did it both ways and the threshold voltages actually varied in the opposite direction I was lead to expect. I had previously binned a couple hundred devices by threshold voltage spreads of a few hundredths of a volt. Some went down to the next bin when I re-measured "properly".

 

Don't see a schematic.
This method isn't suitable if you'll be making more than a few of whatever you're planning to make.

Sort through enough devices to find 2 that have the same threshold voltage. When you measure threshold voltage, make sure you short the gate to drain as the manufacturers suggest. I did it both ways and the threshold voltages actually varied in the opposite direction I was lead to expect. I had previously binned a couple hundred devices by threshold voltage spreads of a few hundredths of a volt. Some went down to the next bin when I re-measured "properly".

 

You can make a current mirror with power MOSFETs, but why would you want to? Just use a BCV61 like normal people do!

You might be better off with a dual MOSFET as the chips are likely to be from the same batch, but in a BCV61 they are part of the same chip.
You might improve things by putting some resistors between the source terminals and ground.