Electronics > Projects, Designs, and Technical Stuff

Toggle switch for small aircraft charging system in case of failure

<< < (6/14) > >>

floobydust:
I think it's important to look at the root cause of the regulator SCR's failing. For what you are trying to do, the added wires, switch and complexity I think would lower reliability maybe worse than the voltage regulator's alone. I see a main output circuit breaker anyway on airplanes to fuse/disconnect the output winding which seems best.

Having a TVS diode or snubber or diode-clamp at the winding will prevent large voltage spikes from stressing the regulator. Some older designs had almost nothing for surge protection. Not sure of your part number. I looked at a Rotax/Ducati RV12 and was not impressed, they are terrible bad solder joints and connections, nothing for surge protection.  Also I've seen a huge capacitor 22,000uF 40V added to Rotax 912/914 engines, so it's possible ignition noise and spikes are high. Heat is also the other enemy, so located away from the engine in a place of cool airflow. Vibration is not good for them either.

Never short the output of a PM generator (dynamo). It will just load the engine and burn up the stator winding. Whoever said the short-circuit current is small is out to lunch. I have seen Kawasaki voltage regulators fail shorted and the engine was loaded and the stator (oil immersed in the crankcase) burned up as a black tar covered piece of carbon.

Example schematics of PM generator voltage regulator- Kubota, Yanmar Tractor RS1105
https://www.eevblog.com/forum/projects/ac-generatoralternator-charger-rs1105-replacement/msg1256470/#msg1256470

Just a note, a vehicle has a few grounds- the battery (-ve) connects to the engine block but also it's important to have a thick wire braid from engine block to chassis, where the regulator, nav, radio stuff is. If you instead ground the chassis at the battery, there is a surge (at the engine block) during cranking that can raise havoc - like your USB troubles it can damage some electronics connected to the block.

SkyMaster:

--- Quote from: ITman496 on April 17, 2020, 09:52:42 am ---
--- Quote from: richard.cs on April 17, 2020, 08:59:24 am ---It should do no harm to test it, from the stator's perspective it's the same as normal regulation, and the rectifiers will see only a modest current.

You're doing the project I've wanted to do for a long time, but the rules here are very different which makes it much more difficult, plus there's the question of space. Got any photos?

--- End quote ---

Sure do!  Keep in mind its a little rough still because I inherited the built from someone else and am going over it.  The fabric will be coming off next week and replaced and repainted with new stuff.  Also, these photos are in no particular chronological order, covering the span of 2 years. Hence the different looking engine, different instrument panel, windshield, no windshield.. etc.  The First picture is the most current.

--- End quote ---

When looking at your instrument panel, a Glass Cockpit no less, it is not possible to imagine that this is for a Minimax.

FBW CTRLS... Fly-by-wire  ;)

Electric trims, on all three axis; wow.

Your project looks like a lot of fun.


What is the EFIS you are using? It this your own creation?


 :)

richard.cs:
If his regulator is of the type he described then it is a current source, designed to be shorted. That's what the SCRs do in normal operation, and the stator has deliberately high leakage inductance so to make everything behave. This was the normal design on engines of this era (I read 80s on wiki). In fact this type predates regulation at all and battery overcharging at sustained high revs was normal.

There are designs meant for use with modern regulators that have low inductance and act like voltage sources, the regulator for one of those is essentially a buck converter. This type shouldn't be shorted but you're unlikely to find that type of stator on an 80s engine.

If there's any doubt the two types can be distinguished by measurement.

floobydust:
I've never seen or encountered a constant-current dynamo. I guess they would be on a moped or small engine, where do you see them used?
I see the regulator uses SCR's as a switchable rectifier diodes, doing cycle-by-cycle control, not shunt regulation.
KZ440A service manual gives 75VAC at 4,000RPM with no load on the dynamo - but this is a 4-stroke bike and OP said he has a 2-stroke engine. Anyhow, some idea of the open-circuit voltage.

ITman496:
@richard.cs and @floobydust

At this point I'm leaning towards it being the shortable coil style, but I could be wrong!  The points you brought up about crummy regulators is also an interesting point.   
The solder looks okay at least, but I can not check any further as it is epoxy potted and I can not get it out without likely destroying it.

Identifying marks are sparse, but I can see:  "2" "TYMPANIUM CORP"  "620403" "4990"

Googling reveals this similar product with some semblance of spec sheet.  What do you think?  https://www.classicbritishspares.com/products/tympanium-regulators

Also, what measurements can I do to test as you suggested, richard?

@skymaster

Hahahaha yeah, right?  I love it.

The fly by wire controls are a backup system I decided on to raise the travel limits of the trim motors and raise the speed so god forbid something happens to the teleflex control cables, I can fly the plane home via the trim motors.

It is tremendous fun!!

The EFIS is my own creation, running on a pi 3b+ (well a 3b, i got a b+ to swap in though for more performance) running on processing 3.  Which is why it looks like garbage.  The screen is a 10.1" 1200nit brightness non touch with a glass multitouch overlay stuck on top.


Still tidying the wiring on the switches on the right!  I had connectors on it to remove them (glad I did since I had to twice to shuffle things around)


Here is a custom board I made to run power for everything.  On it is an arduino to listen to the rotary encoder and backup mfd hardware buttons, and to run the power supplies.  On startup it listens to the power switch, turns 12v and 5v on (for the pi and 12v lcd) and the pi raises a gpio high while the software is running.  Then when I flip the power switch on the panel down, it lowers a gpio to the pi and the pi sees that and shuts itself down, eventually powering down and the gpio its holding high drops, which signals to the arduino its safe to cut power.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod