mosfet driver question

[mribble]

I want to drive a large mosfet with microsecond pulses.  I want to use 12V to drive the mosfet.

I used this mosfet driver and everything worked as expected: https://www.mouser.com/datasheet/2/268/21419c-23818.pdf

However, when I used this mosfet driver the output was only 5V when I expected it to be 12V.  https://www.mouser.com/datasheet/2/115/ZXGD3009DY-611094.pdf

I'm using a 5V micro to drive the mosfet driver, but in both cases I'm powering the mosfet driver with 12V.  I seem to be missing something on the second data sheet that tells me about this limitation.  Could somewhat tell me what in the datasheet tells me I can't drive the output at 12V when using a 5V input?

[mribble]

Thanks for the reply.  I understood the implementation was very different, but could you explain why the second which you described as BJT connected as emitter follower doesn't work with 12V when the input is 5V?

If I look at the internal schematic in the data sheet I'm just switching 2 transistors which would seem to work.  But I'm sure I'm overlooking something so I'm asking for a bit more of an explanation on why my assumption is wrong.

[james_s]

Because that's what an emitter follower does. It is effectively a unity gain buffer, the output voltage will follow (be the same as) the input voltage. If you put 5V on the input, you will get 5V on the output, which is precisely what you're doing.

[mribble]

Thanks James!  I was just thinking about it more and came the the same realization by thinking through what happens at the transistor level.

[mribble]

Could you experts confirm that this chip would function for my needs?  https://www.mouser.com/datasheet/2/115/ZXGD3002E6-74529.pdf

It looks like it would allow me to drive the big mosfet with 12V even though the signal from my micro is at 5V, right?

It is nice because it is pin compatible.  The footprint is slightly larger, but should solder onto my existing pcbs easy enough.

[Audioguru]

The test circuit printed in blue shows that it is the same very simple pair of emitter-followers. The voltage output is a little less than the voltage input even if it has a high supply voltage.

[mribble]

Thanks for checking.  I didn't really understand that test circuit before, but now I think I see they are actually showing the internals of the chip without actually saying it.

I have another one to check.  This one seems seems to allow different input voltages and while they don't explicitly state the input on the graph they do have high and low values for input in the datasheet which I think means it's what I'm looking for.  https://www.mouser.com/ProductDetail/Microchip-Technology/MCP1416T-E-OT?qs=sGAEpiMZZMvQcoNRkxSQku%252bA3lkO5S2CnAPOHg65zB0%3d

I'm really sorry for my mistakes, but I'm having trouble understanding enough on these datasheets to understand them, but I am learning a lot.  Thanks!

[james_s]

Yes that looks like it will do what you want, note the line in the datasheet that says it can be directly controlled by either TTL or CMOS logic, and also the internal block diagram shows logic gates. If the internal diagram shows only transistors then it is just an emitter follower.

[T3sl4co1l]

Easy to confirm, multiple ways:
1. Check the V_IN spec on the output voltage parameter.
2. Note that it's the same basic thing, ZXGD (Zetex Gate Driver?), 3000 family.  They only differ by amperage.
3. Incidentally, if you're following Mouser's catalog information, be very skeptical.  It's mostly correct, but outliers and simply-wrong data are always present.

Keep in mind, ultimately, what you're buying is exactly two things: a. the manufacturer and part number in the listing, and b. whatever the manufacturer says that is.  By "says", I mean preferably obtaining the datasheet directly from the source.  (Although I don't think I've seen an inaccurate datasheet from a supplier.  I have seen some datasheets that don't exist on the manufacturer's website!  There was a time when Panasonic's website was basically unavailable for over a year.  Wurth Elektronik still seems to be figuring this out in some places...)  And even then, sometimes not even that, as stocking errors are infrequent (usually in the parts-per-thousand rate).  Any added data in the catalog, may be helpful, but may be incorrect!

Anyway, it's always been strange to me that they advertise it as if it were a fully featured gate driver, and conveniently leave off the part about how you'd happen to go about generating a 10V (or more) "logic" signal with a cool nanosecond risetime.

They used to make complementary duals that were specified as you'd expect (ZXTC2045, etc.).  Maybe they still do, I haven't checked in a long time.  One would assume the reason is as any: cheapness; why bother testing and specifying them as BJTs, when you can save a few seconds of test time as "gate drivers"?

Well, beware such steps.  Sometimes they're okay.  (Here, as long as you're aware that it's just stupid BJTs, it's unlikely to spread from that.)  Sometimes it's a plan to loosen specs that you've been depending on.

But such is life, buyer beware.  Yes, you must be like a lawyer, reading datasheets carefully.  In time you will come to understand them (or at least, better able to puzzle over the poorly written ones..), though you won't really spend less time reading them (if you're not spending 10+ minutes reading a datasheet before making a final decision, it's probably not a good decision).

Tim

[mribble]

Thanks James and Tim.  Both of your answers were very helpful!