Low side MOSFET driver with protection?

Status
Not open for further replies.
M

mtwieg

Advanced Member level 6
Joined
Jan 20, 2011
Messages
3,908
Helped
1,311
Reputation
2,628
Reaction score
1,435
Trophy points
1,393
Visit site
Activity points
30,026
For a 15A DC chopper I want some good fault protection. I already have a high side current shunt and sense amplifier for detecting slow overcurrent, but in the event of a hard short circuit this measure would be too slow (several microseconds) to save the FETs from destruction. I was hoping to find a FET driver with desaturation detection, similar to devices like this. However all devices with desat protection have very high trip thresholds (6-8V), which is suitable for IGBTs but not for FETs. Does anyone know of a MOSFET driver with such a feature?
 

What does your overcurrent circuit look like? I'm not sure what topology you're using for your current sense but there you might be able to do something like the LT1161 datasheet suggests (figure 6) and use a diode to speed up the response of your overcurrent circuit under short-circuit conditions.
 

Currently I'm looking at using a current sense amplifier like the LT6108, with a high side sense resistor.

The LT1161 can't switch nearly fast enough to replace the FET in the DC chopper, so it would need to be an additional stage, which would cost a lot and harm efficiency. And I don't think it could have sub microsecond response time, even with an external speed up network.
 

I wasn't suggesting using an LT1161, just thought you might be able to bypass a resistor somewhere.

I can't think of anything of the shelf of the top of my head... the only thing I can think of is to sample the sense resistor directly with a PNP.
 

If you post your circuit, it would certainly help.

Can you use a sense resistor, a fast op-amp configured as differential and a comparator to detect fault?
 

fendertunes said:
If you post your circuit, it would certainly help.
Click to expand...
Here's the current schematic of the relevant section.



So I already use a high side sense resistor and a current shunt monitor LT6105, which goes to a comparator and latches the gate drive circuitry off with some discrete logic. The LT6105 takes >5us to respond, which is fine for motor stall conditions, but isn't fast enough to protect the gates from a dead short (total ESR in the circuit path is around 10-20mOhm, so the FETs would desaturate). I haven't been able to find a decent diff amp or shunt amp with fast enough response, which will work with a single supply.

Desat detection sounds a lot easier than sensing current with a shunt resistor, since it doesn't require an amplifier with high CMRR. I could definitely build such a function out of discrete components, but I'd like to avoid the cost and complexity involved, so I'm looking for a monolithic solution.
 

If all you want is desat detection just put a comparator on the FET drains. You will need to do the same with the gate voltages as well and only allow your desat detection to trigger while the gate voltage is above the threshold voltage for at least one FET, you don't want your desat detection to trigger while your FETs are still turning on.

*EDIT: I suppose it might be nice to catch your gate voltage not being high enough though...as long as your FET turn-on time is significantly faster than your detection circuit you can compare the gate voltage to something just below the threshold voltage.
 
Last edited:

Hi mtwieg,

I had the same doubt some time ago in a 400 V DC bus inverter. The solution we found was placing a zener diode in series with the desaturation diode! This allows a change of the desaturation voltage set value and works pretty well. It also works with more than one fast diode in series so you sum the forward voltage.
Hope this helps.
 

Hi mtwieg,
Here is a way to make things simpler use fully autoprotected Power MOSFET like the one given below and forget about over current, Temperatue etc....
 

Attachments

  • CD00002208-107980.pdf
    467.3 KB · Views: 183

mtwieg said:
However all devices with desat protection have very high trip thresholds (6-8V), which is suitable for IGBTs but not for FETs. Does anyone know of a MOSFET driver with such a feature?
Click to expand...
Hi mtwieg,
I have an Idea for Mosfet Desat protection, How much is the desat voltage for MOSFET?
 

Adding a Zener in series with Desat diode is provably the best solution for this application.
See
" The DESAT detection threshold voltage of 7 V (typical) can be reduced by placing a string of DESAT diodes in series or by placing a a low voltage zener diode in series with the DESAT diode."
 

Hi emontllo,
7v is the desat voltage for IGBT what is the Desat voltage for Mosfet(within which the mosfet is considered safe)
 
Last edited:

The "Desat voltage" would depend on each MOSFET and application.
Check the current - voltage characteristics of the MOSFET for a given gate voltage and operating temperature. Then chose the VDS for the desired short-circuit current. After that, check that for worst case conditions of temperature, desat detection time, zener tolerance, etc, the SOA of the MOSFET will not be exceeded.
 

Hi emontllo,
Thanks for the detailed reply , As per the current voltage chart attached for IRF740 For a current of 8A , the "Desat voltage" is to be taken as 6.5V. Am I right
 

Attachments

  • 740.bmp
    710.6 KB · Views: 111

If you drive the gate at 9 V I would rather set the "Desat voltage" to 4 V.
Therefore (according to the application note I sent on a previous post):
4 = Driver Desat Voltage - Vf - Vz
 

Unfortunately, while the zener diode trick would probably work, I don't think I can use it because I don't have a supply voltage in the range required by IGBT drivers with desat protection (I have 5.9V, but all the drivers I've seen take >9V). So it looks like this time I'll have to role my own circuit if I want desat protection. I'll keep that trick in mind for the future, though...
 

Hi mtwieg,
I have made a circuit (attached) for you please test and reply (you can change the values of resistors as required)
 

Hi mtwieg,
You can add a small value capacitor across R5 to prevent false trigering
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…