hi
is it better to use 2 pnp bjt in high side and 2 npn bjt in low side of an inductive load drive h-bridge ?
or using 2 pmos in high side and 2 nmos in low side of h-bridge?
what are the advantages and disadvantages of using each one ?
if our system is intended to drive a particular inductive load continuously for 10 years with switching frequency below 50Hz, which one seems to be more reliable and durable?
What's the power level?
Reliability depends a lot on the design, not only on the device. BJTs might seem more rugged as they don't have such a delicate part as the gate, but there are mosfet devices running for many years without problems, protect and drive propperly and you shouldn't have problems. Mosfets runs at loser temp, as a low rds on one will have way lower losses than a saturated. Bjt.
JS
JS
this is my schematic :
i want to excite a (0.5H 200OHM) load with +- 20ma; with switching frequency below 50Hz
for protection i used two back to back 4.7v zener diodes connected to the load terminals as illustrated in my schematic .
is this configuration of zener diodes a good way for protection? or i should use following configuration?
FETs are extremely reliable, as long as you know how to drive then properly.
can you please explain more ?
i would be appreciated for any schematic .
If you put clamping diodes across each BJT, then BJT can be as reliable as FETs.
FETs are extremely reliable, as long as you know how to drive then properly.
BJTs suffer from various issues, mostly related to SOA, both FBSOA and RBSOA.
Adding clamping kinda gets rid of the RBSOA issue, and since it is used for power switching, FBSOA is naturally not an issue.
For the sake of simplicity, I will go with BJT with diode.
in most bjt datasheets i just can find maximum rating of Collector −Emitter Voltage as a positive value and it seems that it's just corresponds to the positive voltages that one can apply to the Collector −Emitter.
but what about negative Vce voltage which a bjt can tolerate?
but what about negative Vce voltage which a bjt can tolerate?
Very little, ~7V, and this can alter hFE and frequency response irreversibly of the BJT even if it doesn't blow up immediately.
That's why you always need a diode in parallel.
so why bjt datasheets don't mention anything about this rating?
so why bjt datasheets don't mention anything about this rating?
It's kinda well known. When you apply voltage across E to C for an NPN, the EB junction breaks down at ~6V (common magic number), and the diode junction from B to C has ~0.7V (again, common magic number).
so as log as negative voltage that applies to the Vce is less than 6 volt i am in the soa and i don't need clamping diode?
so as log as negative voltage that applies to the Vce is less than 6 volt i am in the soa and i don't need clamping diode?
You need one regardless. Without a inductor current freewheeling path, voltage can get much higher than input voltage. What you are building is essentially an unintentional boost converter.
i provide inductor current freewheeling path by using two back to back zener diodes.
pleas look at my schematic :
if l want to use inductor current freewheeling path by using 4 diodes like in following schematic , then the problem which i confront is that during switching time a negative current will go through +20ma current supply .
then the problem which i confront is that during switching time a negative current will go through +20ma current supply .
Add a 100uF on the output of your regulator. Problem solved.
i think you mean add a 100uF on the output of the current source not the 5v rail regulator. am i right?
i think you mean add a 100uF on the output of the current source not the 5v rail regulator. am i right?
No. Add a diode in parallel with the Isrc and make it to conduct reverse current, then add caps to 5V rail.
thanks i got it.
my current source composed of a single nmos and a shunt resistors and op-amp like this:
can i rely on nmos body diode and don't add any other extra diode?
but what about negative Vce voltage which a bjt can tolerate?
Very little, ~7V, and this can alter hFE and frequency response irreversibly of the BJT even if it doesn't blow up immediately.
That's why you always need a diode in parallel.
so why bjt datasheets don't mention anything about this rating?
They do; it is Vbe.
For power transistors, base-emitter breakdown is not as important as it mostly affects hfe at low currents.
Using bipolar transistors or more likely Darlingtons even with external diodes may be less expensive than power MOSFETs simply because for a given current and voltage they use less silicon area making them less expensive. The advantage in speed that MOSFETs have does not matter at low frequencies.
Bipolar h-bridges are also easier to drive because the same signal and driver may be used to turn on pairs of transistors.