On the linked page is a suggested amended circuit involving a zener diode and a couple of resistors. Does this look sound? It seems to me that perhaps the current through the zener is too low, but I might be wrong.
Does anyone here have any recommendations about the best approach to doing this? Assume a higher supply voltage (like 24 V) and a lower VGS(max) like +/- 15 V.
Common to use a protection zener across source/gate if the supply voltage is close to max Vgs.
Maybe ~18volt in this case, since absolute max Vgs of this fet is 25volt.
The circuit in that link is bad.
They mention a 6v8 zener. That might not turn the mosfet fully "on" (heat).
The 10k collector resistor and 100k source/gate resistor slows down switching (heat from switching losses).
I would have used a 1k gate/source resistor, and a 100ohm collector resistor.
I wouldn't worry about your circuit if the 12volt is good and bypassed.
If I had to switch ~24volt, I would use your circuit and add a 18volt (max) gs zener.
And add a 100ohm current limiting resistor in series with the collector of the level shifter.
Leo..
I would need a base resistor for Q1 though, would I not? It seems without it the current would be too high, particularly if it was turned on by a 5V pin.
Yes, I think so.
12-4.5 = 7.5volt drive voltage. A logic fet should be happy with that. Look at the datasheet for other fets.
~16mA charge and ~4mA discharge, that still can be increased if you have to.
Watch the temp of the drive transistor and zener. Both ~200mw on a 24volt supply.
Leo..
One thing I don't get about this is R1. That limits current both through the base and the collector, right? So if VBE(sat) is 0.95 then we have:
(5 - 0.95) / 200 = 20.25 mA
But what proportion of that flows through the base and what proportion from the collector? If the transistor is fully switched on, would that be 1/10 through the base?
VBE is more like 0.7volt.
With 5volt at the base, 4.3volt at the emitter, and 200ohm, you have 21.5mA emitter current.
Base current in lineair mode is 1/Hfe.
If Hfe is 100 at that current, base current is 1/100 of emitter current. ~215uA in this case.
Collector current is ofcourse emitter current minus base current. ~21.3mA
Leo..
The common-base current gain is approximately the gain of current from emitter to collector in the forward-active region. This ratio usually has a value close to unity; between 0.98 and 0.998.
Thus if there is 21 mA going through the emitter resistor (R1) then 21 * 0.98 must be coming from the collector, assuming that there is a voltage at the collector, and the remaining (21 * 0.02) from the base.
However as someone else pointed out to me, if the supply voltage (12 V in my diagram) is not connected then the entire 21 mA would come from the base.
MarkT:
Depends on the voltage across the collector resistor, that limits the collector current, all the
excess must come from the base.
But what you are saying, are you not, is that since the collector resistor R2 also limits current (in this case to say 10/1000) then that would be 10 mA through R2 and therefore the rest through R1. But how much would flow through the zener diode?
There are always gremlins you don't think of.
Yes, if the collector can't provide the emitter current, then the base has to supply it.
The volt drop across the 1k resistor can't be any higher than 12volt - 4.3volt.
Practically 12volt - 4.5volt, when the transistor is fully saturated.
21.5mA emitter current, 7.7mA collector current, and 13.8mA base current.
Nothing through the 10volt zener, because there is only 7.7volt across it.
Leo..
I wonder why you went from a simple grounded emitter level translater (post#1) to a more complex Arduino voltage dependent constant current driver (post#5).
The first circuit can drive any mosfet fully "on", while in the second circuit you loose 4.5volt (with a 5volt Arduino).
I think the circuit from post#1 is more suitable for 12volt supplies, while the circuit from post#5 is better for 24volt supplies.
Leo..
I can't answer your question, but I have one of my own. Why do you add the zener diode from Vcc to Gate instead of from Gate to GND ? (with the collector resistor serving as the zener current limiting resistor) .
I'm sure the answer is obvious to anyone who has used similar circuits but I'm not seeing it.
Is it really a "6 of one, half a dozen of the other .." situation ?
