MOSFET Drain Source resistance

Irving

Joined Jan 30, 2016
2,769
Drain-source resistance, Rds, is lowest when MOSFET fully turned on, ie gate-source voltage >> gate threshold voltage Vgs(th).

MOSFET is never a current source, but can be configured as a constant-current sink and is often used as such for active loads. A typical circuit being this one I am currently testing...

1655402162747.png
 

BobTPH

Joined Jun 5, 2013
5,227
A MOSFET does act somewhat like a current source when it is in saturation mode. For any given gate voltage, there is a max current after which the current increases very little with increase in Vds.

Bob
 

Irving

Joined Jan 30, 2016
2,769
A MOSFET does act somewhat like a current source when it is in saturation mode. For any given gate voltage, there is a max current after which the current increases very little with increase in Vds.

Bob
Hmmm - when configured in a constant current configuration I suppose you could say its a current source, but that's not the MOSFET, that's the circuit configuration. It isn't a current source in the sense that a battery is a current source.
 

Audioguru again

Joined Oct 21, 2019
4,793
Is the load actually 0.75milli-ohms which is only 0.00075 of an ohm?

The Mosfet needs a Vgs of 10V to fully turn on and the minimum opamp output voltage will be 11.2V then the output current will be (11.2V - 10V)/0.00075= 1600A (!) and explode it and the battery if the input voltage is at the maximum input for the opamp of 12V - 1.5V= 10.5V.

If the load is actually 0.75 of an ohm then the output current will be 667mA.
 

crutschow

Joined Mar 14, 2008
30,102
when configured in a constant current configuration I suppose you could say its a current source, but that's not the MOSFET, that's the circuit configuration.
Well, the added circuitry controlling the MOSFET does help in establishing and stabilizing the current-limit, but the MOSFET makes a pretty good current-source by itself if operated in the saturation region (below), as can be seen by the near constant current for a change in the drain-source voltage.

1655422767228.png
 

Irving

Joined Jan 30, 2016
2,769
Is the load actually 0.75milli-ohms which is only 0.00075 of an ohm?

The Mosfet needs a Vgs of 10V to fully turn on and the minimum opamp output voltage will be 11.2V then the output current will be (11.2V - 10V)/0.00075= 1600A (!) and explode it and the battery if the input voltage is at the maximum input for the opamp of 12V - 1.5V= 10.5V.

If the load is actually 0.75 of an ohm then the output current will be 667mA.
OK, that might have been a bad choice of example, it was a WIP on another project and I picked the first diagram with a single MOSFET as an example confi. For completeness here are the LTSpice generated curves for that device for Vgs = 4.15 - 4.85V:

1655475341108.png

I don't have the actual real device curves to hand but they were remarkably similar - better than I expected.

Here is the actual circuit I'm testing, though right now the heatsink isn't big enough to go to 40A @ 28.8v design spec at 10m/s air flow. I'll probably end up using two of the heatsinks I have now and 6 MOSFETs in total (at £28 each I'd rather not fry them too often). The sense resistors are 50A/75mV ammeter shunts - a lot cheaper and easier to use than PCB mounted devices - I was originally using 100A/75mV ones as I had a couple to hand but they are a little insensitive at low currents.

1655473848190.png

This works pretty much as per the simulation though it drifts off a little as the devices heat up - only to be expected. Right now I'm playing with ideas to select 1, 2 or 4 MOSFETs and changing loop gain to make it a little more flexible at lower currents, say <20A and <10A,the approx limits for 2 and a single MOSFET.

Even so, a 16bit DAC on 3.3v full range gives a (noisy) resolution of around 20mA though realistically I'd say 200mA steps are more practical for all 4 MOSFETs which is fine for my immediate needs.
 
Top