Autosport Labs

THIS THREAD MOVED FROM GENERAL DISCUSSION

I have cobbled together an EDIS4 stand alone (at least for now) system for a 4 cyl aircraft engine. Dual ignition - two sets of spark plugs. The VRS is aligned with TDC to give a 25* limp mode. The engine is started on the remaining magneto, which uses an impulse coupling to achieve a temporary 0* advance for starting. Engine starts and runs reasonably good at low RPMs. The engine actually idles (600 RPM) more smoothly when on the EDIS only. BUTTTTT . . . when one throttles up to 1700 RPM and performs a "mag check", the remaining mag, when run alone, shows a 50 RPM drop - same as before. When the EDIS is run alone at that engine speed, there is about a 250-300 RPM drop - and, to state the obvious, that is totally unacceptable. There is some problem with the system. Finally, at the moment, I do not have access to an oscilloscope.

I have a list of about ten things to check to track down the problem. Here are the system components:

1. aluminum trigger wheel (9.875" diameter) with thirty-five 1/4" flat head steel rivets squeezed into it. The 36th hole has a brass rivet. The trigger wheel is 1/8" thick and the shaft of the rivet is 1/8" diameter and 3/16" long - not much mass. The distance between rivets is about .6".
2. VRS was sourced from a Ford Fiesta (SMP PC914), chosen because of physical form factor. The magnet at the tip appears to be about 1/4" so I think there is a good match between the tooth and the sensor magnet. Resistance tested on this sensor and also on a backup sensor at about 390 ohm. The consistency between the two sensors gives me some confidence in their overall health.
3. Gap between sensor and steel rivets is about .045" based on measuring with feeler gauge. Definitely less that 1/16" (.0625")
4. Two conductor shielded wire for sensor to EDIS
5. EDIS4 module from a Ford Escort and is mounted in the engine compartment close to the coil and fairly close to the spark plugs.
6. MSD 8.5 mm super conductor spark plug wires
7. Gen2 wasted spark coil from a Ford Focus - based on a slightly different firing order between the Ford and the aircraft engine, the No. 2 and No. 4 plug wires were switch between coil sides.
8. Denso M24S spark plugs regapped to .040" - 18mm threads which is a direct fit. This is not a resistor plug.

QUESTIONS:
a. Can a non-resistor plug cause problems with the EDIS operation? How likely is this outcome?
b. Is it possible that the 1/4" steel rivet does not have enough mass to create a good or acceptable sensor performance?
c. VRSs can test anywhere from 2000 ohm to apparently 390 ohms. Would one with more resistance (more wire) put out a better/stronger signal than one with less resistance (less wire)? Maybe the Ford Fiesta ECU is designed to take a weaker or different signal than what the Escort EDIS required?

Thanks to all that have read this far. I could go through my list of things to check but that would be too much for now. Any input is greatly appreciated, especially on a, b and c above.
Larry

Diagnostic procedures on Thur, Jan 27:

1. Reversed polarity. Engine would not run at all on the E.I so polarity was correct. I immediately put back to the way it was.
2. Installed another new set of plugs with the original Denso gap (.027). No change. I put the other set (.040) back in. The thinking was that if the signaling and/or firing was marginal, plugs that were easier to fire might show improved performance.
3. Checked the continuity of the spark plug wires. The longest with the most resistance: 3.5' with 185 ohms.
4. Tried another EDIS4 module. No change.
5. Had timing checked by local airplane mechanic. He thought he saw a "final" advance of about 33*. I was in cockpit so could not confirm what he observed. For the timing check the engine was started on the remaining magneto and then the E.I. was energized with a toggle switch. When I say final, he would see something around 10* less for 1 or 2 seconds and then it would jump up 10* more advanced. Well, that's good news in that that is just how the EDIS module is supposed to work.

This item 5 is a good area to investigate and I'll pursue the timing until it's figured out. First thing will be to manually locate TDC (find the furthest throw of No. 1) and then confirm that the TDC marking on the flywheel (generally called ring gear) is at the correct location. Without relying on the TDC marking of the ring gear, I'll also check the timing on the remaining magneto. Both systems need to be at 25* advance (static until an ECU is added).

Will report back.
Larry

I pursued the timing issue and have found that I'm getting about 6* to 7* additional advance in all instances.

Trigger wheel geometry: the trigger wheel is 9.875" in diameter. That's about 31" in circumference. Coming inside that diameter 3/8" for the 1/4" flat head rivets, the circumference is slightly less but still almost 30". The distance between the 1/4" rivets is about 6/10", more that twice the size of the rivet.

Question: does or can the imbalance between the size of the ferrous object (the tooth) and the distance between these ferrous objects cause the EDIS module to do strange things or to act differently? How important is it to size the tooth and the distance between the teeth the same?

Thanks,
Larry

Answers to your first post:

a) yes. Always use resistor plugs.
b) yes.
c) unknown. Use the recommended sensor.

With your creative design of the trigger wheel you are coloring outside of the lines. The better recommendation is to use a conventional trigger wheel design (all steel) - following the original design as much as possible, so you can verify a properly working baseline configuration.

Once you've established that, you can start getting creative with your design and do A:B comparisons with the reference.

Hope this helps,

Sorry for the long quiet period!

Good news - I've tweaked the system to the point that it is working fairly well and, for an EDIS standalone, I'm fairly satisfied. I'd like to keep this readable so I'll be as brief as possible. Ask questions if something is obviously left out.

1. Every wiring diagram that I could dredge up showed the capacitor downstream of the coil. Many posts say it isn't needed. What's the downside? - none. So I made a trip to Pic-N-Pull and harvested a few capacitors. Installed one on the aircraft. I have no idea if it helps or not but the cost and effort are nothing. Follow the wiring diagrams.

2. The other area of attention was the gap on the sensor versus the ferrous tooth. 1 mm seems to be the sweet spot.

Beyond the above, I retraced all quality control points and improved or resolved where I found sloppiness.

I flew the airplane today after recent encouraging tests. Had never run the engine above 1700 RPM so did not know what to expect. Did a high speed taxi test and got good results. Then immediately taxied to the launch end of the runway and took off. All was good.

Today, I flew twice, once with resistor plugs and once with non-resistor plugs. My sense is that the engine ran better on the non-resistor plugs. Again, there was no Megajolt to be impacted by the noise put out by the non-resistor plugs.

I welcome your posts and questions. There was a lot of experimentation that took place since my previous post. Way too much to detail out here. But I'm willing to explore any of it.
Cheer,
Larry

Thanks for the update.

One note, we always recommend resistor plugs for RF noise suppression for the electronics.