Why You Need Flyback Protection Diodes
Вставка
- Опубліковано 15 жов 2024
- I demonstrate the need for protection diodes when switching highly inductive loads at high current. Small coils in small relays are usually a non-issue, but larger high current devices can slowly destroy the switching transistor.
A few things:
- dI/dt is not steeply flanked because there's no base-resistor to earth. Some base-charge keeps left in the transisor itself which makes it shut-off slowly. Using a complementary Cmos driver or low-R to earth would already make a difference to fastly uncharge the transistor-base. Know this from transistor-ignition in automotive.
- But more of a factor in this case: According to you voltage current ratio there a substantial ohms resistance in the big coil, like 200R or so. So, your flyback pulse seen as a powersource has a high internal resistance which reduces ability to generate high current-pulse.
- A semiconductor on top of its rating behaves like zener. Its actually NOT the voltage which causes the damage, its local diffusion caused by temperature which makes the damage. Could eventually be very local because of 'tunneling'-effect.
- Previous two factors together with the fact that the huge reactance of the coil makes that any currentpulse from flyback is a smoothly buffered one.
- All could be compared with blowing a normal (thermal) fuse. There's a current specified, but the specified current is NOT what's doing the damage. On short term it's energie (joule) for a specific fuse expressed in I * t^2, on longer term its power which relates to I^2 * R, but specified in Amp though.
- Surprisingly, over-current in semiconductors can highly over-rate its specs as long as the short pulse doesn't exceed the semiconductor's I * t^2 rating. So, for instance: a 1A pulse in a 100mA transistor is not a problem as long as it's short enough, if the momentary quantity of dissipated energie which matters.
Just read I mistakenly wrote I * t^2, I meant I^2 * t. srry.
OK, this makes sense. As I was messing around with different coils and voltages I did notice behavior that indicated to me that there was probably a lot more going on here. At the time, I suspected that the coil resistance might have something to do with it. I didn't think that the switch off time of the transistor would be so slow in this configuration. I s'pose the impedance/resistance to ground from the base would be on the order of Kohms or Mohms.
I wonder what the max peak voltage flyback you get on your multimeter, or maybe use a ptc?
The only reason I didn't try to use a multimeter is because I thought the pulse would be too short to read out properly. Some oscilloscopes can do one-shot readings to capture a pulse like this but my old analog scope can't do that very well. I'm not sure a PTC would work in this case since they are usually used for overcurrent protection for feeding something. It might protect better against an external transient or protect other circuits on the same power rail rather than protect the transistor driving the relay. Truth be told, I haven't used PTCs much, so I don't know what their limits really are. A quick and dirty way to see if there is high voltage is to put a neon tube across the relay coil and see if it blinks.
That sounds right. Ive been trying to figure out the best way to prevent some fuses from bust in my cap leakage tester, they pop if I change the voltage too quickly on the rheostat.
That's sounds really weird. Does it blow only if a cap is hooked up? Have you tried a slow-blow fuse? Maybe the inrush current is somehow blowing fuses.
Yeah it has a 350mA slow.blo or it did. The manual warns not to make quick voltage changes otherwise the fuse could blow
I wonder if a NTC (negative temperature coefficient) thermistor would work in series with the fuse. I think they make some just for inrush current limiting.