PDA

View Full Version : Help understanding Current Sense MOSFET



Tobias
- 5th August 2012, 03:18
I am trying to understand how to make this MOSFET work. Ultimately I want to use the current sense output to go to a comparator. If the voltage output that correlates with a max current I want the MOSFET to see, the comparator will pull the enable pin on the MOSFET driver low and shut down the MOSFET.

I am attaching three PDF files. The first is my try at a schematic without the comparator side of the schematic included. It is just how I am interpreting the AN10322 PDF, which is an app note for the MOSFET. The last PDF is the spec sheet on the MOSFET

My first question is do I have the schematic right?

Second question is about the RSense resistor value. What is n? I want to scale it so 50 amps is 5v.

Third question, I understand the first op-amp is converting amps to voltage (OA3b), whats the purpose of the other op-amp (OA4b)?

Finally is a OPA227 Op-amp a good part to use for this application?

661566166617

SteveB
- 5th August 2012, 06:16
here are some app notes that may be good reading and have some answers:

Current sensing power MOSFETS (http://www.nxp.com/documents/application_note/AN10322.pdf)

EDIT: Nevermind, I see you have the datasheet in the OP. ;)

HenrikOlsson
- 5th August 2012, 08:43
Hi,
I've never used these devices so don't take what I'm about to say for granted - in fact, don't ever do that ;-)

1) Schematic looks like the one in the app note so it's probably allright (possibly with the exception of [3]). Just remember that you must have a negative supply for the OP-amp due to the inverting nature of the op amp configuration.

2) n is probably the Id/Isense ratio, which according the datasheet for your MOSFET is 500. For 5V output you want a 500 ohm resistor -50A * 500ohm / 500 = -5V (again, note the negative output)

3) I believe the purpose of second OP-amp is to invert the signal so you get it on the positive side of ground in order to measure it with the ADC. It does look like it's missing its feedback though so I'm not sure... I'd probably add a 100k resistor in series with the negative input and another 100k from the output back to the negative input.

4) The OPA227 looks like a nice device but it's not rail-to-rail so you need atleast +/-7V to get a 5V output from it.

Since you're using this on the low side the source (and kelvin source pin) is "basically" at ground all the time but if it was at high side, in for a example push-pull stage or H-bridge, then the source (and kelvin source pin) wouldn't be at ground. That's why they are using the virtual ground setup (and to get higher accuarcy).

/Henrik.

Tobias
- 6th August 2012, 00:51
Henrik
Thanks for the input. One question, how do I get negative supply to the op-amp? What part number do you suggest?
Thanks again

Tobias
- 6th August 2012, 01:57
I think I found a good one





DCP010507DBP-UE4

HenrikOlsson
- 6th August 2012, 07:31
Hi,
That DC/DC converter should work but remember that it's isolated so you'll need to connect its negative input to it's positive output. Here's an alternative aproach that I found in an ON Semiconductor application note on current sensing MOSFETs:
6618

/Henrik.