Electronics > Projects, Designs, and Technical Stuff
Toggle switch for small aircraft charging system in case of failure
<< < (3/14) > >>
ITman496:

--- Quote from: BradC on April 16, 2020, 02:57:43 pm ---
--- Quote from: ITman496 on April 16, 2020, 08:14:55 am ---My worry for A is that with nothing to sink the load into, the voltage coming out of the regulator will spike to whatever the p2p voltage of the stator winding is. Which may be above the rating of certain components in the regulator.
--- End quote ---

You are asking for a way to disconnect the charging circuit in the case of rectifier/regulator failure, and you are worried about damaging the regulator?

I'd be more concerned about an unrestrained stator generating enough voltage to cause insulation breakdown in the windings.

I'm with Mike. Switch it over to a dummy load small enough to be viable and big enough to prevent huge inductive spikes. A couple of well calculated high power resistors somewhere in the prop airflow should do the job and weigh bugger all.


--- End quote ---

Agreed.  Also, the reason for the disconnect without damaging the regulator was to allow me to disconnect it on demand, on the preflight run-up of the engine on the ground before takeoff to test to make sure its working.  But in retrospect, that's kind of dumb because I can see that its working by the voltage rise on the system.  If the battery is 13V or higher, it HAS to be working.

I'll instead use a SPDT switch and route the positive output of the regulator to a power resistor pack under the aircraft to take the edge off the pulses.
richard.cs:
Personally I would go for a shorting switch before the rectifier. Output voltage remains low so prevents diode or insulation damage and the load on the engine will be much lower than with a resistor connected. This kind of stator is usually designed to have high output voltage compared to the load and high series inductance so they behave pretty much as a current source (it's a nice trick - open circuit voltage and impedance both scale linearly with RPM so it's constant current independent of revs). They're specifically designed to be regulated by shorting so it does them no harm. Yes there's the I2R loss in the winding, but that current won't be a great deal higher than the normal charging condition.

Using a load resistor draws more power from the engine (a shorted stator looks largely reactive) and gives you the problem of needing to safely dissipate that power somewhere without a fire risk.

There is one advantage to an opening series switch though -  it protects you from shorted regulator or rectifier discharging your battery which is a failure mode that hasn't been discussed. An optimal solution is probably a DPDT switch with one pole wired to open the battery connection and the other wired to short the stator. You need it to be break before make or (perhaps better) have a shunt zener on the rectifier output to soak up the voltage transient during switching.
David Hess:

--- Quote from: ITman496 on April 16, 2020, 12:48:36 pm ---That's interesting David!  It makes sense though, those windings can sure put out a lot if you let them!  I had no idea that was ever a thing.  I guess I know why it continues to not be one..
--- End quote ---

Back then there were fewer options.  The flux-switched inverters available back then only put out a square wave and not a modified sine waves and they were pretty heavy compared to just modifying an alternator.  Today we have cheap and powerful high frequency inverters.

What I might do is use a switch to open the output but also include a big power zener diode to limit the voltage to 24 volts or something.  Big stud mount power zener diodes used to be available but no longer.  That leaves a more complex power shunt regulator although it is possible to configure a common voltage regulator like an LM317 as a shunt regulator by shorting its output; use an LM338 instead for higher power capability although I am not sure it will be high enough.
ITman496:

--- Quote from: richard.cs on April 16, 2020, 04:05:52 pm ---Personally I would go for a shorting switch before the rectifier. Output voltage remains low so prevents diode or insulation damage and the load on the engine will be much lower than with a resistor connected. This kind of stator is usually designed to have high output voltage compared to the load and high series inductance so they behave pretty much as a current source (it's a nice trick - open circuit voltage and impedance both scale linearly with RPM so it's constant current independent of revs). They're specifically designed to be regulated by shorting so it does them no harm. Yes there's the I2R loss in the winding, but that current won't be a great deal higher than the normal charging condition.

Using a load resistor draws more power from the engine (a shorted stator looks largely reactive) and gives you the problem of needing to safely dissipate that power somewhere without a fire risk.

There is one advantage to an opening series switch though -  it protects you from shorted regulator or rectifier discharging your battery which is a failure mode that hasn't been discussed. An optimal solution is probably a DPDT switch with one pole wired to open the battery connection and the other wired to short the stator. You need it to be break before make or (perhaps better) have a shunt zener on the rectifier output to soak up the voltage transient during switching.

--- End quote ---

Interesting. I'll have to raid the parts pile and see if I can find anything like that, I know I have a few stud mount components. 

Right now the plan is to use a dual throw switch to redirect the regulator output to a ~50W 10 Ohm power resistor mounted under the plane on the bottom skin in the air stream for cooling.  Just enough to take the edge off the high spikes, and assuming this switch is only thrown in regulator failure, it should.. be okay?  The idea is that it only has to last until I get back home and not damage the stator in the process.

I do like your idea to have it short the stator instead though.  The only annoying thing with that is that I now have to route a very noisy high frequency AC line through to the middle of my plane, past all the guts and electronics, to the switch.  I suppose I could use a relay up front, but now we're getting very complicated again and I'm trying to avoid that.


--- Quote from: David Hess on April 16, 2020, 04:37:02 pm ---
--- Quote from: ITman496 on April 16, 2020, 12:48:36 pm ---That's interesting David!  It makes sense though, those windings can sure put out a lot if you let them!  I had no idea that was ever a thing.  I guess I know why it continues to not be one..
--- End quote ---

Back then there were fewer options.  The flux-switched inverters available back then only put out a square wave and not a modified sine waves and they were pretty heavy compared to just modifying an alternator.  Today we have cheap and powerful high frequency inverters.

What I might do is use a switch to open the output but also include a big power zener diode to limit the voltage to 24 volts or something.  Big stud mount power zener diodes used to be available but no longer.  That leaves a more complex power shunt regulator although it is possible to configure a common voltage regulator like an LM317 as a shunt regulator by shorting its output; use an LM338 instead for higher power capability although I am not sure it will be high enough.


--- End quote ---

I went on ebay and they seem to be.. not awfully expensive? 
richard.cs:
If you route the stator in on twisted pair then the noise shouldn't really be to much of a problem. The AC is low frequency, and whilst it will have switching harmonics from the regulator the energy is quite limited - it's impulses when the SCRs turn on and the repetition rate and therefore the overall energy content will be low. I'd expect it to mess with poorly shielded audio but not to have enough high frequency content to cause radio problems - SCRs are just not very fast devices.

You could put the short circuit after the rectifier. It'll obviously stress the rectifier diodes a bit more but the current won't be much different to charging a dead battery. That would allow you to have a simple changeover switch with alternator output on the moving contact, battery one side and short circuit the other. You would need it to be an ordinary break-before-make switch and a permanently connected Zener, or better a TVS diode (basically a pulse rated Zener) on the alternator output to clamp the voltage for a few tens of milliseconds as the switch moves. This is exactly the same as your proposed load dump, just with a zero Ohm dump resistor. The power resistor would get you slightly lower (perhaps 20%, dependent on engine revs and chosen resistor value) rectifier currents when dumping at the expense of increased engine load. With or without a power resistor, a Zener or TVS to cover the switchover is required - 10 ms of switching time is forever as far as insulation breakdown and the semiconductors are concerned. Thinking about it a bit more this is probably the approach I would take, with a zero-Ohm dump. It's simpler than switching both output and stator, and the rectifier should cope just fine.

PS: you can replicate the behaviour of a high power Zener with a low power Zener and a big transistor, though at >100 W it might not be that sensible. If you really want a load dump then it should probably be a resistor.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod