Separating power ground from control ground to exceed NMOS threshold voltage

Question

I'm a novice circuit designer looking to create a high voltage source follower NMOS circuit. I originally planned to use a Teensy digital I/O pin as the gate control signal, so when the Teensy outputted high (3.3V) the source follower output voltage would be ~0V and when the Teensy outputted low (0V) the source follower output voltage would be ~500V (circuit shown below). Unfortunately, however, I didn't realize that the transistor I picked up has a threshold voltage of 4V.

On to my question: I'm well aware that I could solve the problem by (a) getting 40 boost converters or (b) buying another set of transistors (I'm basically just replicating this circuit for 40 of the Teensy's digital I/O pins), but I'm curious about what would happen if I connected the Teensy ground pin and the grounded transistor source pin with a 1-1.5V potential difference (through a power supply, for example), as shown below. Will the new \$V_{gs}\$ value be 4.8V, as I (maybe naively) envision, thus turning on the NMOS and pulling \$V_{out}\$ to the transistor's ground? If this works at all, are there any disadvantages of using this approach compared to buying a boost converter/another transistor?

To clarify, this would mean I have two power supplies running: one to ground the transistor source and for the DC-DC converter that outputs 500V, and one to create the 1.5V potential difference. Thank you very much for your help!

Answer 1

Yes, this will work, but can be cumbersome and somewhat non-ideal.

1. Is the VTH of the FET really 4 V ? How well does it turn on with 4.8 V on the gate ?

2. Now when the Teensy puts out a 0, the FET will get 1.5 V. This may make the FET partially on (leaky).

You would be better off to get a real level shifter device. Perhaps a 74LVC245 would be suitable.

Answer 2

To answer your direct question no, raising the source by 1.5V will not change Vgs to 4.8. Vgs is the voltage between the gate (g) and the source (s) so if you move the source higher you have to move the gate higher yet. (Clarification from OP in comments)

Yes pulling the source of the FET down to increase the Vgs would work but would certainly be unusual. As long as whatever inverted supply you use can handle the current from the MOSFETs when they are all on it should work.

Be careful measuring things when you have a "floating" ground like that though, referencing the wrong voltage and getting something you didn't expect is the least of the possible issues. If you use a scope to measure different parts of the circuit and connect both probe grounds they are usually connected in the scope chassis in which case you'd be shorting your negative supply.

This may be a situation where a BJT is better because your base current can be very low and you only have to be a diode drop or 2 above ground to turn it on. This is very similar to level shifting the gate drive for a MOSFET with a BJT except you can leave the MOSFET out entirely. You could still use a 10V supply with pull-up/down resistors and a set of BJTs to level shift your existing mosfets if you really want to use them and not buy a bunch of high voltage BJTs.