Am I driving this P-mosfet right?

Thread Starter

robotDR

Joined Mar 17, 2020
35
Please see schematic below in regards to using Q65 to drive Q64.
VIN_1 will be about 44-57V

Can you please confirm my thoughts.
And can you please critique/review for improvements and/or errors.

I put a 10V zener (d74) between the gate and source of Q64. Is my following understanding correct:
1) with Q65 off, the gate of Q64 will be VIN_1
2) with Q65 on, the gate of Q64 will be -10V with respect to VIN_1 (For example: if VIN_1 is 55V, gate of Q64 will be 45V)
3) the zener 'overrides' any normal voltage divider calculation I would do. The output of just the voltage divider would be 14V with respect to ground. Or -41V with respect to VIN, destroying the gate of Q64.

1634949690268.png

Thank you for taking the time to read this thread.
 

crutschow

Joined Mar 14, 2008
28,170
You voltage calculations appear correct.

But if Vin_1 is 55V, then the current through R2128 and the Zener will be 45V / 340Ω = 132mA when Q65 is on.
The power dissipated in D74 is then 132mA * 10V = 1.32W.
Is the Zener rated for that wattage (at least a 2W Zener)?

Unless you need very fast switching, you could likely increase the value of R2128 to reduce the Zener dissipation.
 

Thread Starter

robotDR

Joined Mar 17, 2020
35
You voltage calculations appear correct.

But if Vin_1 is 55V, then the current through R2128 and the Zener will be 45V / 340Ω = 132mA when Q65 is on.
The power dissipated in D74 is then 132mA * 10V = 1.32W.
Is the Zener rated for that wattage (at least a 2W Zener)?

Unless you need very fast switching, you could likely increase the value of R2128 to reduce the Zener dissipation.
Thank you for your post.
This is a great point.

so the current path will mainly be through the zener (reverse bias) and r2128.
At 132 mA, the resistor will dissipate about 1/2 a watt which is too much.

I can increase r2128. There is an upper limit though right? If I switch q64 too slowly then it can overheat while passing current and not fully on? Is that accurate?
 

Thread Starter

robotDR

Joined Mar 17, 2020
35
lYes.
It depends upon the frequency of the switching.
How frequently will you switch the MOSFET on and off?
It’s not switching on and off with a predetermined frequency. It’s meant to switch once and likely stay that way for hours or days or even weeks.
 

BobTPH

Joined Jun 5, 2013
4,032
Then the frequency of switching is in cycles / month, considered a very low frequency. The switching speeds Crutschow is talking about would be hundreds of Kilohertz.

Bob
 

Thread Starter

robotDR

Joined Mar 17, 2020
35
so using a 10v zener like in the schematic is correct to keep the gate at -10v with respect to source (q64) when q65 is on. And when q65 is off, the gate of q64 will be equal to the source (VIN_1).
I still need to double check my power dissipation but besides that, it’s correct use of q65 q64 and D74?
 

LowQCab

Joined Nov 6, 2012
1,317
The Gate is rated for +/- 20-Volts.
So change the Zener to ~15-Volts.
This may also slightly reduce the Rds of the FET and
will reduce Heat dissipation everywhere.

All this excessive Current is not necessary to drive the Gate,
multiply R-2127, and R-2128, by a factor of 10.
( a factor of 100 would probably be fine,
but might be slightly slower than desirable in Switching-Speed )
.
.
.
 
Top