Products > Test Equipment
Blow oscilloscope measuring coil Back EMF
joeqsmith:
--- Quote from: Grandchuck on August 25, 2023, 01:46:24 pm ---Joe, why a MOV in one and a TVS in the other? Just curious.
--- End quote ---
Good question. I started my electronics design career in automotive. Back then we were using MOVs. This was before they became a fire concern and we ditched them. I wanted to learn more about microcontrollers and this power supply was the very first thing I ever designed using the MC68701. I wrote all the code, laid out the board and photo etched it at home. The keypad was from an old girlfriends phone that had died. There is an RS232 port to connect to my PC to control it and read back the current, voltage as well as a second voltage input.
The transformer I dare say was something I salvaged from an old Sperry Rand UNIVAC terminal. Case is powder coated steel. The front plastic came from a friend who had worked for a plastics company.
The HP mods were many years later.
HighVoltage:
The old FLUKE Combiscopes like the PM3394B seem to be immune to high voltage at the input. I had it several times that a 5kV spike jumped to the input and nothing happened to the scope.
I would not try this on a modern scope.
joeqsmith:
My love of the vacuum fluorescent display also came from automotive. The ADC is a whole 8-bits and supply still appears to pass cal. :-DD
ejeffrey:
--- Quote from: Fungus on August 25, 2023, 06:37:46 am ---
--- Quote from: ptluis on August 24, 2023, 10:34:05 pm ---What I'm trying to do is to design a portable module to protect the output of power supplies when I connect inductive loads like relays, contactors, electric egr solenoid, etc for testing purposes.
--- End quote ---
Power supplies will have a big capacitor across their output which will absorb EMF spikes, no problem.
--- End quote ---
Not necessarily. Lab power supplies or programmable power supplies usually have a relatively small capacitor to avoid dumping a huge current pulse into a load before current limiting kicks in. It's plenty to absorb ESD or small spikes from smaller inductors, but not anywhere near enough to absorb EMF from a power relay coil (which remember is more than just an inductor, part of the spring energy is converted back to electrical).
The output caps should give you some breathing room to let the protection circuit kick in. But a cheap power supply with bad or no protection circuit isn't going to be saved by a 10 microfarad capacitor.
T3sl4co1l:
I still see no mechanism for malfunction here. Say an inductor is fully charged, let's say fractional H at several A, and one lifts the wire to the power supply. In that instant, terminal current drops to zero in fractional ns; this launches a positive flyback up the PSU wiring, probably maximum amplitude of 20-100V as the initial (~nm) gap breaks down easily. The full route should be some 100nH to the input capacitor, if that. Very little energy. Could still be a fair amount of RFI around given the current and speed, but of note, this path is identical whether the load is inductive or resistive. If your supply can't handle a loosely-wired resistor, something else is wrong.
On the coil side: terminal current is zero, so inductor current charges the stray capacitance of the coil. A few dozen nanoseconds later, it's ramped to the breakdown voltage of the air gap -- mechanical things move slowly, this might be a few 100 nm at this point, and breaks down at maybe 100V. The spark discharges stray capacitance into the PSU, and we have an EFT (electrical fast transients) scenario. Each pulse is roughly characteristic impedance (the coil doesn't really have capacitance, it's a ball of wire, and that wire has characteristic impedance to itself), so, in the ballpark of 100V / 50 ohms = 2A. This energy dumps into the PSU wiring, recharging the output wires to coil current, until the spark cools too much, the connection breaks, and the cycle repeats.
As long as the coil remains charged, sparking continues as the gap slowly widens; the EFT response is a rapid-fire chain of increasing amplitude spikes; the repeat rate slows as it goes, as the coil runs out of current to charge the capacitance up to an ever-higher breakdown voltage.
In the process, no quasi-steady-state current greater than the initial coil current ever flows, and no peak current flows for longer than the characteristic length of the system (probably not the whole coil winding wire length, but a fraction of it; EFT is normally tested with 50ns (FWHM) pulses).
The impedance is high, so a small capacitor will suffice to bypass the terminals. There is likely common mode coupling as well, for which caps to a metal enclosure would be desirable. Even 0.1uF will have significant impact (the equivalent pulse capacitance of EFT is around 1nF).
It's far more likely to me that the EFT coupled upstream and blasted some control IC or transistor, especially on poorly laid protoboards with no ground plane. (Not pejorative, mind. I've made plenty of point-to-point and perfboard circuits. It's convenient. Susceptibility is simply a consequence of such open construction.)
This also suggests that a MOV does more than a TVS -- strictly on account of having more capacitance; breakdown of either is probably never activated.
Tim
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version