Comments requested on MOSFET high-side driver

[nickgammon]

Have you tried using a pot to vary the resistance to see if you can find a "sweet spot" for the resistance?

I am trying to have a mathematical basis for choosing values, so that if someone has a different size load (eg. 500 mA or 5A) they have a way of calculating what ought to work, without just using trial and error.

[zoomkat]

If I had an o-scope I'd use pots so I could make adjustments and see what is going on real time. I'd try driving the 2N3904 base with maybe a 250 ohm resistor to ensure that the voltage drop across the 2N3904 collector/emitter is as low as possible. This would help keep the PNP MOSFET gate at its lowest possible voltage.   

[zoomkat]

I am trying to have a mathematical basis for choosing values,

I hope you are good with calculus and differential equations. Any of the components whose electrical characteristics are expressed as "curves" will involve complex math to get hard numbers.

[polymorph]

Rules of thumb and merely looking at the curves can get you pretty close.

If your transistor switching circuit is so touchy regarding parts values, you've done something wrong.

[michinyon]

It seems to me,  your calculation in reply #5 needs to include the apparent capacitance of the FET gate.

[nickgammon]

How would you factor that in?

[dlloyd]

This P-Channel MOSFET is fast and powerful (and well, uhm, costs more):

Parameter             FQP47P06  IXTH96P085T   Units
Turn-On Delay Time         110           23      ns  
Turn-On Rise Time          910           34      ns
Turn-Off Delay Time        210           45      ns
Turn-Off Fall Time         400           22      ns
RDS(on)                     26           13      mΩ
ID                         -23.5        -96       A
VDSS                       -60          -85       V

Digi-Key     IXTH96P085T Datasheet

[polymorph]

Ouch! $6. Not much point in getting a stinking fast MOSFET if the circuit driving it isn't going to switch it faster than 5uS anyway.

[nickgammon]

I've added more to my thread describing using BJTs for switching, if anyone wants to check the calculations.

[dlloyd]

Ouch! $6. Not much point in getting a stinking fast MOSFET if the circuit driving it isn't going to switch it faster than 5uS anyway.

Hmm ... I thought it was the slow MOSFETs that begin to "stink"

• high switching speed = high power conversion efficiency
• 50% RDS(on) = 50% less steady state power loss
• Larger safe operating area
• Cost savings on heatsink requirements
• Much lower power loss from DC to 5µs (or faster)

[polymorph]

My point was about diminishing returns.

[polymorph]

BTW, that rule of thumb of base current being 1/10th collector current is in the datasheet for the 2N2222 and 2N3904. It isn't that the gain is 10. It is higher than that, albeit much less than the linear hfe. It is that we need the transistor heavily overdriven into saturation, taking into account that you may have a transistor on the low end of the specs regarding gain, and gain drift with time and temperature.

https://www.fairchildsemi.com/datasheets/PN/PN2222A.pdf

https://www.fairchildsemi.com/datasheets/2N/2N3904.pdf

[nickgammon]

BTW, that rule of thumb of base current being 1/10th collector current is in the datasheet for the 2N2222 and 2N3904. It isn't that the gain is 10.

Where in the datasheet?

[polymorph]

Ctrl-F type in saturation.

[nickgammon]

Out of the 3 datasheets I have I could only see one reference to 10 and that was on the graph for "Base-Emitter Saturation Voltage vs. Collector Current" where it mentioned β = 10.

Is that what you are referring to?