How to drive high-side N-channel FET to saturation??

does a mosfet actually go to saturation? i thought that was for a bipolar transistor.

you gave no other info to explain your situation. just the word 'controls'.

maybe you can ask in the context of the device you are examining.
 
ekline309 said:
Based on observation, it seems many controls use dual N-channel instead of P-N FET pairs. How can the top NMOS ever be saturated?? I don't get this....


Saturation is the wrong term, but I guess you mean "how is the high side FET turned on?"

The answer is by using a bootstrap circuit that increases the gate drive voltage to around 12V above the voltage of the battery + supply line. This is done via a diode and capacitor. The diode allows the bootstrap capacitor to be charged to 12V (from the FET drive supply) when the low side FET is turned on and stops the bootstrap capacitor from discharging when the low side FET turns off. The negative end of the bootstrap capacitor is connected to the junction between the low side drain and the high side source. When the low side FET turns off, because the negative end of the bootstrap capacitor is connected directly to the high side FET source, it can provide the gate voltage needed to keep the high side FET turned on, even when the FET is fully conducting and the high side source is close to the +supply rail voltage.
 
highside n channel??????? why,why would you wanna do that.
 
nevermind- i figured it out myself,i only 100 microwatts
 
ekline309 said:
Based on observation, it seems many controls use dual N-channel instead of P-N FET pairs. How can the top NMOS ever be saturated?? I don't get this....
Jeremy answered this question correctly, but the other answer is: Although N-channel FETs are "better" FETs in some respects, a lot of people use P Channel FETs in simple controls because theya re soooo much easier to drive and they are "good enough" FETs. A 2n2222 connected to a high side P channel FET will drive it just fine, for about a nickel, vs a complex high side driver circuit with all kinds of parts.
 
The snag with P channel FETs, though, is that you just can't get them with specs that match those of N channel FETs for a reasonable price. Take the ubiquitous IRFB4110, for example, with it's 100V Vgs max and 4.5 mohm Rdson. There just isn't a TO220 package P channel FET that comes close to that spec, so the only choice is to use a pair of N channel FETs and put up with the need for bootstrap drive on the high side.
 
And remember, standard bootstrap circuits/high side drivers are useless if you need the high side FET to be on continuously...
 
i thought that the H-bridge ICs had a special charge pumping circuit that allowed the high side to be on continuously. gotta google when i get a chance and see what wikipedia says.
 
mikal said:
highside n channel??????? why,why would you wanna do that.
my thoughts exactly.

llile said:
A 2n2222 connected to a high side P channel FET will drive it just fine
this is what I've been using as well.

Jeremy Harris said:
The snag with P channel FETs, though, is that you just can't get them with specs that match those of N channel FETs for a reasonable price

so really the only reason is cost, but if you add the extra gate drive componentry its a wash...
 
dnmun said:
i thought that the H-bridge ICs had a special charge pumping circuit that allowed the high side to be on continuously. gotta google when i get a chance and see what wikipedia says.

Almost all the data sheets mysteriously don't discuss the matter. Ask their FAEs and you get the song and dance routine. You do find some app notes that have circuits with external devices that address the matter. I don't know of any chips that for sure can keep the high side FET on at DC.
 
ekline309 said:
Jeremy Harris said:
The snag with P channel FETs, though, is that you just can't get them with specs that match those of N channel FETs for a reasonable price

so really the only reason is cost, but if you add the extra gate drive componentry its a wash...

If you can find an available P channel FET, in a TO220 package, that matches the spec of the ubiquitous IRFB4110, for example, then sure, you can go down this route. What you'll find is that low Rdson, high Vds, P channel FETs are virtually unobtainable in practice. For example, the closest 100V P channel TO220 FET to the IRFB4110 that IR make is the IRF5210, but this has an Rdson of 60 mohms, versus the 4.5 mohms for the IRFB4110 N channel FET.

There simply aren't low Rdson P channel FETs available that will do what we want in practice.
 
Back
Top