Using a real magneto start switch with FSX

[Kerbo]

I snagged a real Cessna switch on eBay for $5 so have abandoned my scratch built design. You can see the full details on my blog (http://workbench.freetcp.com/archives/666).

Once I get the wiring figured out I will update the thread with how I connected it to FSX.

I can hardly wait to crank up my virtual Cessna. :)

[Kerbo]

The continuity tests were thoroughly confusing and I was starting to think the switch was re-assembled wrong somehow or broken. A bit of googling led me to this site: http://www.ercoupe.net/ads/76_07_12.html

The formatting is all jumbled but near the bottom is a continuity test for the switch that I bought. It all checks out and works fine. Unfortunately the terminals don't connect in a straightforward manner when it comes to simulator input. I think with some diodes I can make some sense of it.

This is for a switch with Cessna part number C292501-0101, Bendix part number 10-357200-12. I noted the document at the URL above states that my part number switch has a PRIME function, but it does not.

Here are the continuity settings.

Position Between Terminals
OFF
R and GRD
L and GRD
L and R
S and PR
R
L and GRD
R and unmarked
L
R and GRD
R and unmarked
GRD and unmarked
BOTH
R and unmarked
START (twist and hold, do not push)
GRD and unmarked
S and BAT
L and BO
L and LR
BO and LR
PRIME (twist, push and hold)
Same as above, plus
BAT and PR

[Ronson2k9]

Probably don't need (priming) as that is a starter function. FSX doesn't simulate fuel difficulties save that of running out and pump loss.

The control only starts the engine. Unlike a car there is no battery or other function that is controlled through the switch. You have to turn those on with the master switch on the MIP. While I'm thinking in the real plane you can have ignition from one magneto in the sim that isn't simulated.

So you need to have Off and Both and Start.

Have you considered taking the switch a part cut through some of the confusion? On second thought that could add to that..

I would think though a small test rig could help out. With lights for each of the output terminals.

In looking at the terminals on the back of the switch you have a PR.. Which would be for Prime? Atleast it's in the terminals if not in the function.

There is the one terminal that is unmarked in there.

Shouldn't the 'R' Right magneto also connect ground to unmarked as the left one does?

I'm taking a wild guess here but Unmarked could be the master switch (Alt) in the cockpit. Sure would help to have a schematic of the switch itself.

I'm thinking.

'S' = Starter (starter motor)
'L' = Left magneto
'R' = Right magneto
'PR' = Primer pump
'Bat' = Battery
'GRD' = Ground
'Unmarked' = Could be Alternator (as it's a power source that is unlisted)

This may help though.. Page 7-24 from the Cessna 172N POH .. There is quite a bit of data in there if you take a look.. I sent you the link in previous email.
http://i416.photobucket.com/albums/pp244/Ronson2k3/Cessna172NElectricalSystem.png

Edit: BO could = Bus Out as it shows the bus connection in the diagram to the switch.

[turbopropdriver]

I wired up real mag switches in my old Queenair sim but it's so long ago that I can't remember how I did it. I do remember though, that part of the prob is that a mag switch just connects the mag "p" cable to ground to cut the ignition.
That's why you do a mag drop test just before shutting down, to make sure one of the "p" cables hasn't broken, thus leaving a "live" prop.

Dean

[Kerbo]

I haven't hooked it up yet but it looks like I found the necessary pairs of contacts to connect to my input board for all the positions. When I verify it is working I will post what I did.

Ron, I'm pretty sure all positions of the switch are modeled in FSX, at least with the stock 172. I get a slight RPM drop when running off just one of the mags. Has your experience been different?

[Ronson2k9]

When flying - the manual says that there are the two mags one on the left and one on the right. They supply ignition coil power to the spark plugs. So with one off the engine is running with only half power.

Only supposed to fly that way in an emergency or to test when on the ground.

Your right about them being simulated. I had forgotten. Did you get my email I sent you BTW?

[turbopropdriver]

When flying - the manual says that there are the two mags one on the left and one on the right. They supply ignition coil power to the spark plugs. So with one off the engine is running with only half power.

Only supposed to fly that way in an emergency or to test when on the ground.

Your right about them being simulated. I had forgotten. Did you get my email I sent you BTW?

You're not running on half power, just half the spark plugs. It's a safety system.Magnetos aren't like the ignition system in your car etc. They make power and fire the plugs as long as they are driven by the physical connection to the engine, no battery power is required for them. That's why in order to shut off the ignition you have to ground a mag.
You would expect a small drop in engine RPM when you turn off one mag.

Dean

[Ronson2k9]

Learning every day.. So you are loosing some power not half power.. I think I get it now.. It's less efficient as some fuel is not being combusted. The engine is working harder to make nearly the same RPM.. I did a little searching and found that each cylinder has 2 spark plugs one powered by each mag so if you switch off one mag or one fails the other is still firing both cylinders. That was the missing piece of the puzzle.

2 plugs per cylinder one mag for each one.

I had tried that in the C206 by the way just to experiment. There was a slight loss of RPM but not that much. So your right all switch points in the mag switch are modeled in the sim.

You can still do the switch bypass and create a new switch. Running off the other one.

I found some more info on the mag switch
http://www.aeroelectric.com/articles/MagnetoSwitchOptions.pdf

It's not the exact same switch but perhaps will be helpful?

[Kerbo]

I found some more info on the mag switch
http://www.aeroelectric.com/articles/MagnetoSwitchOptions.pdf

It's not the exact same switch but perhaps will be helpful?

That helped some, thanks for the link.

[Ronson2k9]

I did this Mag Switch bypass design for you.

http://i416.photobucket.com/albums/pp244/Ronson2k3/Cessna%20T206H/MagSwitchBypass.jpg

The blade has a key way and the nylon bolt is filed flat to accept the key way of the blade. That way you can turn the bolt have the blade sweep. The spring is to keep the tension on the blade as it passes by each of the contacts. From the looks of the switch it's a 1/4 turn from OFF to BOTH. I used nails as contacts in the design but if you have something that is a better conductor that would be better.

You would need to figure out how to connect the bolt to the switch you currently have. You could use a small piece of sheet metal in a 'C' shape with one end bolted to your mag switch and the other bolted to the bypass. If you wanted to be really slick about it. You could use the same strip for the blade too. Just a thought. It would be good to have the blade free floating though so that all the contacts are positive (contact is always made).

I used nails because they have flat thin head. You may want to counter sink them a little. so the blade is prevented from moving.

- A small hole for the pointed end of the nail to pass through. Then you use a 'nail head' diameter bit to counter sink the head. flush as you could reverse the transit problem with a hole that is too deep.

You could use some CA to hold the nail in place and connect the nail to the switch leads. for FS not the mag switch you have hehe.

This should let you set up the switch fairly easily if you are having troubles getting the actual wiring working..

[Kerbo]

Ron,
You are a wizard with the drawings, thanks.

[AK Mongo]

I haven't hooked it up yet but it looks like I found the necessary pairs of contacts to connect to my input board for all the positions. When I verify it is working I will post what I did.



William,
Did you ever get yours working, and what did you do? I just received a cheap one from Ebay today, and would love to use your solution.

Reid

[Tom_G_2010]

I am working on the same issue for my 172 sim. I decided to remove the back cover off the ignition switch and found that their are three triangular shaped contacts inside and I am thinking about replacing the back cover with a piece of printed circuit board cut to fit and etched to create independent switch contacts for each function. If it works I'll post pic's. May be a few weeks though . . .

[iwik]

Hi Guys,
If you have the real switch with the spring Start option, the start switch is isolated(found with multimeter). The remaining contacts appear as a 2 bit code.
That is 4 combinations 00 01 10 11,i then used a ic to decode. Use each for Off Lmag R Mag and Both or You can either place resistors across these switches and as a result of what switches are closed the output will change and can be fed into an analogue input and decoded by your software. Its been a while since i did my one but above is the jist. You dont have to modify the switch.
Les

[AK Mongo]

Hi Guys,
You can either place resistors across these switches and as a result of what switches are closed the output will change and can be fed into an analogue input and decoded by your software. Its been a while since i did my one but above is the jist. You dont have to modify the switch.
Les

This is an interesting idea Les, but can you assign the mag functions to an analog input? Would the size of the resistors matter? I would probably use an axis on the BUO836x.
Anyone know if it could be used as is with FSUIPC, and in effect have multiple keypresses for each position, but only have fsuipc give the command based on the combination?

Reid

[Ronson2k9]

One thing you will need to do Reid is check the continuity first as you have a used switch 'pretty cheap' from ebay so before going through the pain of getting it to work I would make sure the switch is in good working order first. As the spring is what most important part that's why I thought of the bypass. This takes all the switch action out of the switch itself and puts it somewhere you can control it easier. If your mag switch goes bad then you will end up with the bypass anyway. I know they must be built to take a ton of use but you never know..

Just my thoughts.

[AK Mongo]

Absolutely right, Ron. I will be doing that tonight. Probably do some testing with it and fsuipc as well to see if the inputs matter. Fsuipc may just be able to see the grounding of several inputs at once and "interpret" them as a single command. Wont hurt to try at least.

Reid

[iwik]

Hi Reid,
If you look at page 48 of Fsuipc Users Manual it explains how to do it. Basicaly you can use up to 10 zones of an axis to trigger an event that a button
or switch operates. I havnt used it like this but a friend of mine who used resistors across the switch idea with SIOC.
Resistor choice should be such that gives a reasonable zones.
My hookup with the IC fed into i/ocards input. So if you are using 836x cards then Petes method should work.
Hope this helps.
Les

[AK Mongo]

I am working on the same issue for my 172 sim. I decided to remove the back cover off the ignition switch and found that their are three triangular shaped contacts inside and I am thinking about replacing the back cover with a piece of printed circuit board cut to fit and etched to create independent switch contacts for each function. If it works I'll post pic's. May be a few weeks though . . .
Tom,

Mine appears to have the three wedge shaped contacts as well. It looks like it will be possible to modify it easily by leaving out 2/3 of the triangular pieces, reducing the remaining piece to 2 contact "dimples, drilling a hole through the switch back between the existing BAT and R mag GRD, adding a contact in the hole, taking off the jumper plate between the GRD and R contact and adding a washer or other conductor for the center ground to the inside of the plate to create a circular Grnd contact.

Connections will then be:

Switch(as viewed from back clockwise)-----sim

S-----Start
Bat-----Both
New contact-----L
GRD-----R
R-----Off
Center GRD---to ground on i/o card

The remaining contacts will be unused.

This solution appears like it will work mechanically and electrically.

Sounds like a lot, but I think it will work. Will report results in 24 hours.

Reid

[Cessna172sim]

We are also using a real magneto switch. I can ask from my brother how connected it.

Some parts in the test bench:
http://www.lentosimulaattori.1g.fi/kuvat/assorted+construction+pictures/throttle+and+controls/181_8102.JPG

[AK Mongo]

I was able to successfully make it work without PCB's, IC's or any other acronyms except KISS*. It involved a screwdriver, a 6-32 x 3/4 SS machine Screw, a drill and bit, a washer, some super glue and a 1" piece of electrical tape. Took about 10 minutes to do. I wrote up a basic tutorial on my blog, please check it out there if you are trying to do the same thing:
http://juneaucessnasim.blogspot.com/2011/06/start-your-engines.html

Reid

*KISS= Keep It Simple, Stupid!

[Ronson2k9]

My philosophy exactly.. Nice work..

[Marsh]

I know this thread is kind of old now, but I found the same thing. I didn't have the key so I had to pull it apart to yank out the tumblers and thought I put it together wrong. After probing with a multimeter I figured out the matrix, I use 3 inputs on my OC master card without messing with the switch.
I connected the terminals labled "R", "L" and "GND" to the card and supplied one of the center terminals labled "GND" with +5VDC. Then some SIOC script to recognise the positions:
OFF = R & L on
R = only L on
L = GND & R on
BOTH = all off
START = only GND on

[Tom_G_2010]

I know this thread is kind of old now, but I found the same thing. I didn't have the key so I had to pull it apart to yank out the tumblers and thought I put it together wrong. After probing with a multimeter I figured out the matrix, I use 3 inputs on my OC master card without messing with the switch.
I connected the terminals labled "R", "L" and "GND" to the card and supplied one of the center terminals labled "GND" with +5VDC. Then some SIOC script to recognise the positions:
OFF = R & L on
R = only L on
L = GND & R on
BOTH = all off
START = only GND on

Marsh,

I'm glad you posted on this thread. I had attempted a couple experiments including a pc board and had a lot of trouble. I went with Reids solution and it worked great. Now that I've got the switch installed in the panel and wired though I'm experiencing problems. The contact positions all work IF I put some forward and upward pressure on the ignition key. Without that pressure I only get reliable contact on a couple of them.

No flaw at all in Reads design, it's a very simple and easy build. I think the problem might be in the thickness of the washer I used as a center contact. So, I'll be pulling the switch out and attempting to replace it or thin it down. The other possible solution would be to stretch the spring a bit to see if that gets me a little more contact pressure.

I think that will do the trick and I should be all set. But, should that fail I'll be looking for a plan B. As such I was trying to follow your switch logic pattern. I'm missing a detail and it's not making sense to my pre-morning-coffee brain...

Did you wire to 3 or 4 terminals? Is there an outer terminal labeled Gnd and another in the center?