What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Ford edis with megajolt

It's a Thing

Sure. This would be the 4-cyl orientation, slip-over-hub style... [snipped]

Trigger%20Wheel%20Small.jpg

I took the design DanH's post above CAD'd it up to create a Thingiverse for it. I included my source CAD file too. I went with the under-sized version as that will fit larger as well.

NOTE 1: The 7 small holes on the top are counter-balances for the missing tooth. Their volume exactly matches the missing tooth.
NOTE 2: The alignment of the ring to the flywheel center is very critical. Inside three teeth are small 2mm alignment holes to be filled with pins or screws. These should match the inner lip

https://www.thingiverse.com/thing:3816677

-Bruce
 
Last edited:
Average, just a Fluke multimeter in series.

Me too, but I have not put a scope on it.

I "third" this. You'll need to average the current to get "effective current rate" the battery would see. You could put a $5 5A DC/DC power supply module in front of it (just for measurement) and read what that tells you. It will be smooth.

-Bruce
 
Messed around, current sensing

I compared the shunt amazon Chinese ammeter/ voltmeter to a Fluke 117 ‘true RMS’ meter. Voltage as right on, current was about a third; the shunt meter showed 4.5A at 2400 rpm, the Fluke 1.29A. This reading includes the coil and the EDIS, but not the Megajolt.

I played with the shunts some more, using a wave form generator I have, and while I can only generate 250ma out of that unit, it’s enough to prove the shunt is displaying 90%-95% peak current of most wave forms, not RMS, and that there is heavy filtering in the frequencies the ignition generates. The Fluke was actually very accurate with its RMS calculation.

My conclusion is that the shunts are only useful for Consistent DC loads. my second conclusion is a 10A circuit is sufficient for the ignition system. From there, it’s about controlling the electrical noise.
 
Last edited:
First off, thanks to everyone for their input to this thread. This is a project I plan to take on in the near future. I am going to start with a fixed timing setup to replace the right mag and then go from there, one step at a time.

Stepping a back a number of pages to the topic of run-up testing of the electronic ignition system. Rather than shutting off power to the electronic system and relying on a smooth return to timing. What about grounding the primary coil (just like a mag), but through a resistor to eliminate a current spike. That way the timing circuit is still active and you simply dissipate the spark energy. It should be a smooth transition then.
 
Stepping a back a number of pages to the topic of run-up testing of the electronic ignition system. Rather than shutting off power to the electronic system and relying on a smooth return to timing. What about grounding the primary coil (just like a mag), but through a resistor to eliminate a current spike. That way the timing circuit is still active and you simply dissipate the spark energy. It should be a smooth transition then.

Simply interrupting power to the system as a whole (EDIS module and Megajolt/E, coil, everything) works fine if the other ignition system is operating. That means separate toggles. There can be a hiccup if the two ignitions are routed through a left-right-off switch.

The coil has one power lead (EDIS module pin 8 to coil pin 2) and two coil grounds, pins 1 and 3. The EDIS box switches the grounds to control coil current.

You can insert a switch in the pin 8 to pin 2 coil power wire to interrupt coil power without shutting down the EDIS and Megajolt. That approach has been flying without issue. I suppose you could switch to route that coil power to a resistor and then to ground, although I don't see the point; there would still be a spike of some value while the switch contacts are in transit.

However, in addition to system power switching at the panel (which you need anyway), any coil switching adds a wire from EDIS pin 8 back to the panel and then from panel to the engine compartment coil location. It adds component risk for no practical gain.
 
Dan, that makes a lot of sense. I had read about people having blips as the system reacquires timing, but it sounds like that is generally a non-issue. It will require the addition of the toggle in place of the L/R switch, but it is just a minor modification.

I have been researching the possible incorporation of a permanent magnet generator into the mag drive as part of the modification (similar to PMag). Overall, it seems pretty doable so far, but adds mechanical risk. It will be important to incorporate a shear device into the system in case of a mechanical failure. Fun projects all of it....
 
MJ/EDIS play code / test bench

I think my best decision was to build a bench setup. It?s a pretty steep learning curve with the minimally documented megajolt, especially the two map switch programming. I created test maps just to prove rpm readings, map function, rpm switching, some wiring failure modes, taking data, etc. i really feel pretty good about the risk level. It probably took me twelve to fifteen hours start to finish and $300 of amazon purchased questionable quality stuff to build the thing, not including the power supply, the old PC680 and the MJ.

You might consider starting with a map or two somebody who is already flying has or can make for you and learn from that. I?m sure there are a few of us who would do that for you.

As I plan my install, I?m still not sure I?m going to go with a seasoned EDIS module or a new AM knockoff. Even a 20yo NOS EDIS has risk....
 
I do plan to bench test everything before installation, I hace a lathe that can carry even the flywheel trigger wheel for testing, so the system will have at least some checkout in advance. However, I cant replicate the vibration and heat easily, so it will only be functionally tested.

We do have full vibe, environmental, and emf test chambers at work, but I don't think I could convince them to let me use them.
 
So, I have slowly continued on with this project for my plane. I decided to go with a mag-hole trigger. I have sourced most of the parts and designed the assembly. Obviously this is a copy of Dan's setup, though I don't know all his details, so I'm sure there are plenty of small differences.

I have one dimension I failed to gather while I had my mag off, and that is the exact position of the Mag gear face relative to the shaft shoulder, so I haven't finished that portion of the design.

Mag-Hole-Trigger-Mount-v02.jpg

Mag-Hole-Trigger-Mount-v02-3-D.jpg
 
Last edited:
I had some fun this week after work. Machined the body for the mag replacement. I am still waiting on a couple of parts, but it is coming together nicely.

20191205-235821.jpg


20191205-235823.jpg
 
Like the o-ring accessory case seal. Nice touch, should have done that to mine.

The snap ring will be hard to install with the toothed wheel pinned to the shaft.

The 25x15x12 spacer must also be pinned to the shaft. If not, the shaft can shift axially in the bearings.
 
Dan, all good feedback. The spacer is a shaft collar, so it does retain the shaft. Yes, the gear will have to be removed in order to remove the snap ring, but that shouldn't be a problem. The shaft has two existing holes from where the magnet was attached, I'm hopeful those will line up close enough to be used for the gear retention, but I missed one critical dimension when I pulled my mag to measure, so I don't know for sure yet.
 
Dan, all good feedback. The spacer is a shaft collar, so it does retain the shaft.

Set screws? Better with a roll pin.

Yes, the gear will have to be removed in order to remove the snap ring, but that shouldn't be a problem. The shaft has two existing holes from where the magnet was attached, I'm hopeful those will line up close enough to be used for the gear retention, but I missed one critical dimension when I pulled my mag to measure, so I don't know for sure yet.

I briefly considered attaching the wheel as Electroair does, i.e. a center screw and a tiny index pin. In the end I elected to cross pin it like the bearing spacer, as it allows the use of a press fit at the wheel-to-shaft interface. The cross pin is insurance.
Shaft%20Assembly.jpg


Using a surface retainer (the black bar below) allows dropping a complete assembly into the body cavity.

Finished%20Triggers%201000w.jpg
 
That is why I am hoping the existing holes in the shaft will align with the gear. Right now it looks like it will be very close. My plan depends a bit on that alignment. Either a slot in the botton of the gear that will fit over a roll pin in the shaft and a bolt on top, or a threaded pin in place of the set scew if all the planets align.

For the shaft collar I am replacing the set screw with a threaded shear pin that will fit into a hole in the shaft.

The lower bearing in my assembly is and interference fit with the housing, while the upper is a slip-fit. So the lower bearing is installed first, then everything else drops in on top of that.
 
Dan,

Do you happen to know either:
The distance from the shaft shoulder to the face of the mag drive gear, or the distance from the magneto-to-acc housing face to the shaft shoulder? I failed to get that measurement when I pulled my mag.
 
Dan,

Do you happen to know either:
The distance from the shaft shoulder to the face of the mag drive gear, or the distance from the magneto-to-acc housing face to the shaft shoulder? I failed to get that measurement when I pulled my mag.

Not sure of what your mean by "shaft shoulder". Got a drawing for dummies?
 
Not sure of what your mean by "shaft shoulder". Got a drawing for dummies?

The sholder where the shaft taper transitions to the threaded portion for the mag drive gear. Basically I need to figure out the position of the mag drive gear on the shaft. Hoping to do so without pulling the mag again, but that may be what I have to do.
 
The shoulder where the shaft taper transitions to the threaded portion for the mag drive gear. Basically I need to figure out the position of the mag drive gear on the shaft.

I did not accurately reference that location. I referenced the mounted and torqued drive gear to the accessory case sealing surface. My drawings say 29mm from seal surface to center of the gear tooth width.
 
Thanks Dan, I took that same measurement. I just don't have a spare gear, so I'll have to wait until I pull mag again.
 
The sholder where the shaft taper transitions to the threaded portion for the mag drive gear. Basically I need to figure out the position of the mag drive gear on the shaft. Hoping to do so without pulling the mag again, but that may be what I have to do.

A caution here. The gear is secured via a morse taper interference fit. The gear goes on as far as the morse taper interface allows it. The gear is not secured or stopped by butting up to a step. When I made one, I cut the taper on the shaft first, put a gear on it and then took measurements from the gear as I worked up the shaft. I suspect you could take the ID at the large side of the gear bore and use that as a reference when machining the taper to place the gear, if all of the other machining is already done. I would not try machining the taper without a test gear in hand. Once you get the angles and cross slide set for the taper (cut longer than what you need), you can just move along the x axis a specific amount after test fitting the gear. A more experienced machinist I am sure would be more confident in just putting the end of the taper in a specific spot, if the gears taper precision could be trusted.


Good luck.

Larry
 
Last edited:
Colin isn't cutting a taper. The idea here is to re-purpose shafts taken from old mags...with an existing taper, keyway and threads.
 
Good advice Larry, but as Dan noted, this is a repurposed magneto drive, saves a bit of work over making one. I don't have a quick change gearbox on my lathe, so setup for thread citting takes forever.

I'll just have to pull the Mag and take the measurements. With 2 little ones and a decent drive to the airport, I'm lucky to get up there once a week.
 
I received the gear I ordered and modified it to use for timing. Also installed the pickup. Took the assembly for a spin on the lathe.

20191213-161624.jpg


20191214-001925.jpg
 
Last edited:
Dan,

What wire gauge did you use for the coil pack? The megajolt manual recommends 14awg, but that seems like way overkill given the reported current draw. Also, 14awg shielded is proving hard to find. 16awg should be plenty sufficient. Thoughts?

On a side note, pickup assembly is all together and ready to roll. Need to get the rest of the system wired

20200101-200547.jpg
 
Last edited:
I'm too lazy to go thru the thread but I would like to make a couple of these on my CNC turret lathe. Can someone point me to a non-metric drawing or a Solidworks file?
 
Last edited:
Thanks guys. Great information.

I'm thinking at this point I will probably just get an 8 cylinder plug new from your side of the pond (half the price of UK) and cut the keys of it as Kirk suggested.

Be nice to have one that has NOT been pre wired like this one or a source for replacement female pins perhaps? Any Ideas?


I bought the new EDIS 8 connector off ebay for $19. I was able to remove the pins from the cut-off EDIS4 module (bought used). Now I can use the pins that came with the new connector with the edis4 body, so no cutting the keys.

20200102-160050.jpg
 
I'm too lazy to go thru the thread but I would like to make a couple of these on my CNC turret lathe. Can someone point me to a non-metric drawing or a Solidworks file?

I'm happy to repost my drawing in inches. However, my setup is yet untested. Also, I made a few minor tweaks that aren't on the drawing. I made the o-ring grooves slightly deeper and made the engine-side bearing press fit and the aft bearing slip-fit.
 
Removing the rotor?

How do you remove the magnet/rotor assembly from the shaft? Is it a press fit?

Thanks
Richard
 
How do you remove the magnet/rotor assembly from the shaft? Is it a press fit?

Thanks
Richard

The one i used was from a slick mag. I tried to machine it off with poor results. Then i realized it is cast on (maybe aluminum) with the magnets embedded in the casting. I ended up just melting it off with a propane torch. Then there were two brass rivets that I drilled out.
 
Last edited:
Where are you guys mounting your coils? In the upper plenum area would make for short plug wires (for one set), plus it is pretty cool during flight. On the other hand, on the firewall gets them away from the vibration.
 
Where are you guys mounting your coils? In the upper plenum area would make for short plug wires (for one set), plus it is pretty cool during flight. On the other hand, on the firewall gets them away from the vibration.

On the 6, I put the coil on the firewall. Not enough room to put on top of the engine with the coil type I had (top facing posts). On the 10, I put one on top of the engine and the other on top of the engine mount. The 10 cowl area has a lot more room to work with though.

Larry
 
On the 6, I put the coil on the firewall. Not enough room to put on top of the engine with the coil type I had (top facing posts). On the 10, I put one on top of the engine and the other on top of the engine mount. The 10 cowl area has a lot more room to work with though.

Larry

Thanks for the input. I think i am going to mount them to the engine mount on angle atrached via cushion clamps. I could always move thrm to the firewall later, but this looks to be the most modular and least intrusive.
 
The one i used was from a slick mag. I tried to machine it off with poor results. Then i realized it is cast on (maybe aluminum) with the magnets embedded in the casting. I ended up just melting it off with a propane torch. Then there were two brass rivets that I drilled out.

Thanks Colin, I had issues with the first one too. On the second one I found an angle grinder removed most of it. I then carefully removed the remainder with the lathe.
 
Thanks Colin, I had issues with the first one too. On the second one I found an angle grinder removed most of it. I then carefully removed the remainder with the lathe.

The torch ended up working great. Took just a couple minutes and everything came off clean.
 
I'm happy to repost my drawing in inches. However, my setup is yet untested. Also, I made a few minor tweaks that aren't on the drawing. I made the o-ring grooves slightly deeper and made the engine-side bearing press fit and the aft bearing slip-fit.

Yes that would be great.

Would like a drawing of the shaft as well. I think it would be ok to make it out of 4140 or 8620 steel.
 
Yes that would be great.

Would like a drawing of the shaft as well. I think it would be ok to make it out of 4140 or 8620 steel.


Bob,

Sorry for the delay. Here is my housing drawing in inches. I don't have a detail drawing of the shaft since I repurposed one from a slick mag.

Mag-Hole-Trigger-Housing-v02.jpg
 
Bob,

Sorry for the delay. Here is my housing drawing in inches.

Interesting, so you have an O-ring seal at the top or is that grove for bearing retention?

How are you attaching the gear to the shaft? Loctite 638 might be an option with interference fit, or is there still room for a grub screw? There must be a secure solution involving a minimum of machining. It looks like Dan may have used grub screws.
 
How are you attaching the gear to the shaft? Loctite 638 might be an option with interference fit, or is there still room for a grub screw? There must be a secure solution involving a minimum of machining. It looks like Dan may have used grub screws.

No grub screws for me. Both the trigger wheel and the bearing collar are a light press fit locked with ASME B18.8.2 split pins in double shear, 0.125"D.

The idea was to create a cartridge, a complete shaft assembly for insertion into the body. It is retained with a surface plate, rather than the internal snap ring chosen by Colin. An internal ring requires installing the trigger wheel after the fact.

Shaft%20Parts.jpg


Shaft%20Assembly%20Drawing.jpg


Trigger%20Body.jpg


All dimensions are nominal. Detail all bearing fits. Trigger wheel is a KSS1-36 from from qtcgears.com.

Usual caution; experimental, use at your own risk.
 
Last edited:
Interesting, so you have an O-ring seal at the top or is that grove for bearing retention?

How are you attaching the gear to the shaft? Loctite 638 might be an option with interference fit, or is there still room for a grub screw? There must be a secure solution involving a minimum of machining. It looks like Dan may have used grub screws.

The groove at the top of mine is for a snap ring to retain the bearing.

The spacer between the bearings is retained to the shaft by a dog-type grub screw. The tip of the screw has an unthreaded portion that fits into a matching hole on the shaft. The tip extends about 3/8' into the shaft and is 3/16 diameter of grade 12.9.

The top gear is held on in a similar fashion. However, since my gear position lined up perfectly with the existing rivet holes in the shaft (from where the magneto was cast on) that pin passes all the way through the shaft. There is also a regular grub-screw on the gear as well. I just left that since there was already a second threaded hole from the gear manufacturer, so why not. According to the spec for Grade 12.9, that little 3/16 dog should have a shear strength close to 19k pound.

In my case, the bottom bearing is pressed in. and everything else is assembled on top of that.
 
Last edited:
Had a private note with a very good suggestion. The trigger body could have an oil weep hole to vent the bearing space. It would offer an early indication of oil seal failure. Although oil is unlikely to bother the magnetic pickup, oil washing is not good thing for sealed bearings.
 
I have one, wasn't planned for that purpose. It is and access point for the dog in the center bearing spacer, just in case. At least I was smart enough to align it with the timing, so I can stick an allen wrench in there snd it will hold the gear properly aligned with the pickup.
 
Detailing Bearings

No grub screws for me. Both the trigger wheel and the bearing collar are a light press fit locked with ASME B18.8.2 split pins in double shear, 0.125"D.

The idea was to create a cartridge, a complete shaft assembly for insertion into the body. It is retained with a surface plate, rather than the internal snap ring chosen by Colin. An internal ring requires installing the trigger wheel after the fact.

Thanks for the updated diagram Dan. I'm wondering what your thoughts on detailing the bearings are? Figuring to reverse engineers something with some service history, I measured 0.002" interference fit on the outer race of a slick core (6002 40mm OD Bearing). Given the relative expansion rate of the aluminium housing, vibration and estimated operating temperature of ~100C that seems like a reasonable number to shoot for on the outer races.

One of the other vendors appears to use Loctite retaining compound to hold the bearings in their housings however I think that is pushing the friendship with an aluminium housing exposed to the heat of the engine bay.

The shaft measures 0.0003" over the inner race which would put it towards the tighter end of a js5 fit etc.

Regards
Richard
 
Had a private note with a very good suggestion. The trigger body could have an oil weep hole to vent the bearing space. It would offer an early indication of oil seal failure. Although oil is unlikely to bother the magnetic pickup, oil washing is not good thing for sealed bearings.

This is a good idea. It is also possible that the unit could be exposed to water from an enthusiastic degreasing. It is always best to allow the fluids out. I have seen electronic circuit boards corroded by sitting in water in similar circumstances.
 
Thanks for the updated diagram Dan. I'm wondering what your thoughts on detailing the bearings are? Figuring to reverse engineers something with some service history, I measured 0.002" interference fit on the outer race of a slick core (6002 40mm OD Bearing). Given the relative expansion rate of the aluminium housing, vibration and estimated operating temperature of ~100C that seems like a reasonable number to shoot for on the outer races.

One of the other vendors appears to use Loctite retaining compound to hold the bearings in their housings however I think that is pushing the friendship with an aluminium housing exposed to the heat of the engine bay.

The shaft measures 0.0003" over the inner race which would put it towards the tighter end of a js5 fit etc.

Regards
Richard

Do your research here. Bearings in these applications typically have one interference fit and the other with slip fit. I built a similar device and used interference on the shaft and .001 slip on the outer shell (recommendation from bearing company driven source gets interference). I did use wave washers on the bearing retainer to add a bit of tension on the bearing to prevent any spinning. Also a recommendation from the bearing engineers. Every alternator or similar device that I took apart was done the same; one interference and one slip. which one varies with application.

I am told that two interference fits will take too much clearance from the ball/race interface and is a problem for high RPM applications.

Larry
 
Last edited:
Do your research here. Bearings in these applications typically have one interference fit and the other with slip fit. I built a similar device and used interference on the shaft and .001 slip on the outer shell (recommendation from bearing company driven source gets interference). I did use wave washers on the bearing retainer to add a bit of tension on the bearing to prevent any spinning. Also a recommendation from the bearing engineers. Every alternator or similar device that I took apart was done the same.

I am told that two interference fits will take too much clearance from the ball/race interface and is a problem for high RPM applications.

Larry

The data I have here indicates that at 100C you can expect the deviation of clearance/interference of .0018" on a 40mm bearing bore. To me that indicates why the Slick is bored 0.002 under as once up to temperature the fits are correct. I am making an assumption here that the Slick bearings generally run to TBO without failure.

The following document makes mention of previous problems with Champion magnetos caused by loose fitting bearings, severe side loadings and/or high temperature (page 6).

https://www.casa.gov.au/sites/default/files/_assets/main/airworth/awb/74/005.pdf?acsf_files_redirect.
 
Last edited:
The data I have here indicates that at 100C you can expect the deviation of clearance/interference of .0018" on a 40mm bearing bore. To me that indicates why the Slick is bored 0.002 under as once up to temperature the fits are correct. I am making an assumption here that the Slick bearings generally run to TBO without failure.

The following document makes mention of previous problems with Champion magnetos caused by loose fitting bearings, severe side loadings and/or high temperature (page 6).

https://www.casa.gov.au/sites/default/files/_assets/main/airworth/awb/74/005.pdf?acsf_files_redirect.

Does slick use and interference fit on the shaft AND the bearing housing bore? My point was not the amount of interference, but that the bearing ID should be interference and the OD slip.

My intent was not to question your effort here, only to point out what I learned from engineers that do bearing application engineering for a living.
Larry
 
Back
Top