I always thought MOVs were sloppy, but they are actually pretty cool.
We blew up a channel in our alternator simulator, probably some crazy flyback voltage from the simulated load, so I'm going to add an MOV across each of the output phases. This one can absorb 22,000 amps and
300 joules.390 volts max at 100 amps.
The low part of the curve looks very exponential. Someone once sold a MOV type part as an analog computing element.
Just keep in mind that MOVs are "consumables". They get eaten up a little bit with every jolt until one fine day ... KABLOUIE.
-- Regards, Joerg http://www.analogconsultants.com/
Also the leakage at the low-voltage end probably deteriorates quite a bit after some punishment.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Better a later kablouie than a sooner one.
I don't expect much integrated clamping energy over time.
Here's my ECO:
Each phase of my sinewave generator is transformer isolated, and during a test we blew up some 1206 divider resistors, a big RC snubber, and cracked an ADUM7703 isolator chip rated for 5700 VRMS. All that was clearly impossible and can't be reproduced. There was way too much energy for it to be ESD. So the MOVs and caps might help.
So each phase now has an MOV hi-lo and another lo-gnd. The caps across the transformer secondaries may help too, and also reduce feedthru fuzz from the big class-D amps.
Very nice rework, almost looks like it was always planned that way.
It can be ok as long as this transient load isn't repetitive. Otherwise I'd rather use TVS.
-- Regards, Joerg http://www.analogconsultants.com/
This MOV can gobble 300 joules and 20K amps!
PM alternators are inductive AC voltage sources. As speed goes up, open-circuit voltage and frequency both go up linearly, so they approach constant-current sources. They are typically regulated by shorting a 3-phase full-wave bridge (3 diodes, 3 mosfets). Un-shorting had better be done carefully or you might see a lot of volts.
Buy a 3-terminal MOV with built-in overheating protection. That way you will be able to issue a warning that the MOV is dead and the channel is no longer protected.
Best regards, Piotr
It can, but this energy fuses more and more oxide grains and the current goes up. This heating will cause a regular MOV to explode eventually or disconnect a thermally protected one from the circuit.
You are wearing out your MOVs that way.
Best regards, Piotr
The MOVs are not expected to conduct in normal operation. If they do, they save a lot of expensive parts. If they eventually short, it's a lot easier to fix than the other things that would have exploded.
We were simulating a FADEC voltage regulator and can't imagine what went wrong. And we hope that MOVs will keep it from happening again.
We have software snooping our output current, and it's supposed to shut down our amps if we see too much current... before transformers or MOVs melt down.
Power electronics can be fun, until it blows up. That tends to wreck a
*lot* of stuff that's hard to find and fix. Luckily, this event only blew up a few parts, and they were obvious.
The 3phase diode bridges in automotive Alternators must handle the load dump from the Joules in e-clutches, ACU's and starters. Yet the diodes are most often the 1st parts to fail in an alternator. Is it possible to choose MOSFETs gated by overvoltage to shunt the BEMF?
When it blows up, if you're lucky, you get the technician to fix it.
Better yet, get the tech to do the test. We call them expendable technicians.
We use a LOT of MOVs. 35mm ones for lightning protection (SPDs)
Here is what happens, during a UL test, when they are tested for overload before the correct MOV fuse is designed in with about 500V DC and a current limit of 10,000 amps....
This is not what happens during a typical lightning strike, assuming it isn't a direct strike to the system grid wires or PV array, etc.
30,000 amps no problem for a typical short-ish strike.
Tee-MOVs. That's what we use in our SPDs and also LEDs to indicate. If the LED turns OFF, then there is now power or the fuse is open.
Unfortunately, for the UL 1449 certs, that internal melting solder protection didn't work for these some times asinine requirements so we had to make our own low temperature solder fuse that would open under the specified conditions for the test.
When a thing blows up, it's my responsibility to find out why, and see if a design mistake caused the failure. I can't delgate that, although I can often get help running experiments.
We don't have engineering technicians.
The aircraft regulators have lightning protection, even thought the FADEC is just a couple of feet from the alternator, both inside the engine housing.
I guess a strike can make a big ground loop.
-- John Larkin Highland Technology, Inc Science teaches us to doubt. Claude Bernard
Car alternators are usually controlled by adjusting the field current, so all they need is a plain diode bridge. PM alternators have no control, so they are regulated by PWM shorting. The usual config is an open wye connected to a bridge made of three diodes and three mosfets. When the fets are off, their substrate diodes complete the bridge. If you turn all three on, it shorts the alternator and there's no output from the bridge. The fet sources can be grounded, so the single gate driver is easy. That's fairly clever.
-- John Larkin Highland Technology, Inc Science teaches us to doubt. Claude Bernard
On a sunny day (Fri, 31 Jul 2020 12:13:37 -0700) it happened John Larkin wrote in :
I have a little mains protector thing that has both VDRs and spark-gaps,
The spark gaps are from both mains poles to mains ground, the 3 VDRs are between the mains, neutral and ground. The neon is just to show there is power. Was 5 $ or so 20 years ago on the local market...
Nice little spark!
The traditional lightning arrestors in open country are (1) a non-resistor spark plug (expendable) and... (2) tie a knot in any important line cord. Every little bt of indictance helps.
Correction, looks like 2 VDRs are in series with the spark gaps... and the third one between neutral and ground.
I use neon bulbs for the same purpose, but wired differently. If the bulb lights on, you need to replace your MOV. The color is right for a warning and there is smaller current consumption when everything is OK.
Best regards, Piotr
Here's my FADEC regulator simulator.
I inherited the proto board, so it's not up to my usual gold standard of Dremel'd things.
I took out the previous fast gate drivers and added series gate resistors and caps to slow down switching and bypass things better. And bent the metal to bolt everything to.
The phase A current waveform, blue trace, kinda threw me, but it's right. Yellow is the phase A voltage monitor. Both are picked off by isolated delta-sigma ADCs.
I work with big power at times and if that is truely an alternator (Hardware rotor etc) Then you must be removing loads during critical cycle times where as the regulator does not have a chance to redcuce the field.. Alternators are current devices so removing the load abruptly is going to produce some interesting results if the field was energized at some higher levl.
Our run of the mill generator had such issues with the way we were using it. It would at times cause a little damage to Mov's, fuses for small devices when a large load was removed.
The fix was to figure out the L of the generator output, I used a meter for that and then put some AC caps of good amp service across the output. Using Welding caps are good for this, they should be low ESR types.
But the results are the caps will absorb the sudden fly back effects which is really an effect of unloading the alternator abruptly.
It isn't the best but but its better than nothing.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.