What's new
Van's Air Force

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

PMag Backfire

Manchu16

Active Member
Will call EMAGAir on Monday but I wanted to check if I am missing something simple I can fix this weekend or if anyone has experienced this.

1800 RPM, test left which is the traditional mag. It checks ok, then go to both. Test right, the PMAG and it dies. I leave the key switch on right to let the prop stop, but a second or two later the PMAG comes to life with a big bag (backfire) and the engine regains power.

I switched to both and waited a few seconds before repeating the process. No issues this test at all. Even the PMAG generator test passed.

I know the switch positively engaged during the initial test, which was my flying copilot’s immediate question.
 
Will call EMAGAir on Monday but I wanted to check if I am missing something simple I can fix this weekend or if anyone has experienced this.

1800 RPM, test left which is the traditional mag. It checks ok, then go to both. Test right, the PMAG and it dies. I leave the key switch on right to let the prop stop, but a second or two later the PMAG comes to life with a big bag (backfire) and the engine regains power.

I switched to both and waited a few seconds before repeating the process. No issues this test at all. Even the PMAG generator test passed.

I know the switch positively engaged during the initial test, which was my flying copilot’s immediate question.
Have you double checked your timing for the PMAG? A reminder that the PMAG timing is at TDC whereas the timing for the traditional mag (slick) is at 25 or 20 depending on the engine model which you are not identifying here.
 
Are you sure the P-mag is getting aircraft power. Perhaps it not and then starts working on its own back up power. Un burnt fuel present from the few seconds with no fire then ignite when the backup power comes online.
No expert here, I’m sure someone that knows a bit about P-mag will chime in.
 
Back fire through induction or after fire (pop) through the exhaust?

An inspection of both potential paths is warranted. Best of luck.
 
Quick answers

Great questions and I appreciate the thoughts.

Yes timing is TDC. Timed 26 hours ago and confirmed after the flight.

Ship power to the PMAG is good at start up, confirmed after the flight. Needed ship power to check timing.

Based on sound and its location, most probable exhaust backfire. Assuming engine rotation was pulling air and fuel into cylinder for multiple cycles with no spark. Unburnt fuel expelled in the exhaust pipes. After the pmag started firing again the fuel in the cylinder and the exhaust stack ignited. I didn’t check the intake path, but would guess if the backfire was there I would have blown a coupling or the filter out and the engine would run like ****. Was a single backfire.
 
Likely not the problem by what you are describing but perhaps spark plug leads not correct?
 
Check the connections on the back of the key switch for the PMAG wire.
I had a screw fall out on the starter wire connection, but because the wires are bundle tight back there, the terminals made contact and I would only have an occasional starter issue. Frustrating to troubleshoot but easy fix, I even found the screw.
 
Did you observe an RPM difference between Left and BOTH. If so, the Pmag was likely firing when set to both. If not, it probably wasn't. There is nothing in an EI design that should cause this type of issue, so I speculate something is not right inside of the Pmag itself or the wiring controlling it. The Pmag is somewhat notorious for bearing failure, but unsure if that can cause intermittent spark/no spark conditions. I speculate that it wasn't running during all of the tests and due to coincidence it started a few seconds after switching to R OR it was intermittently working and not working and the 1-2 second thing was just a coincidence. I would suspect an internal fault or a wiring issue (less likely if working well for a long period of time). I would send it in, as I have good confidence that it will happen again. It is clear that the entire EI stopped working and then began working completely. Don't bother with plugs or plug wires

A after fire is typical if you stop sparking for a second and then begin sparking again (only happens when that on off cycle happens when running on one plug). The after fire was not the problem, but a symptom of a different problem.

Larry
 
Last edited:
My guess is you have a switch problem. I would remove both P-leads and then connect two tempoary leads that you can ground directly to the air frame one at a time.
 
My guess is you have a switch problem. I would remove both P-leads and then connect two tempoary leads that you can ground directly to the air frame one at a time.
+1 - and the rest of the wiring, particularly the control plug, making sure anchor screws and hold down screws are snug and no broken wire or something. Seems obvious but it's surprising how often I will jump to the most complicated thing without checking the simplest first.
 
Update

Just wanted to share an update. I have been working with EmagAir. I can replicate the issue with a new ACS switch and temp wiring (taped to the side of the airframe for testing). I can not replicate it using toggles in lieu of the ACS switch. It appears that the short period of time that it is grounded when I rotate the acs switch is causing the pmag to occasionally go off line just before I isolate it.

I have tried to go slow or go fast on the switch from both-left-right and it seems to be agnostic of the speed of switching. Once again, not consistent. I believe that I have ruled out a bad switch and bad wiring. Everything from the green terminal into the pmag can not be tested as I only have one. More to come, but I didn’t want this thread to just hang open.

I haven’t been that good on updating these. Overall I have a great aircraft, just working on knocking the cobwebs off from it being down for so many years.
 
Just wanted to share an update. I have been working with EmagAir. I can replicate the issue with a new ACS switch and temp wiring (taped to the side of the airframe for testing). I can not replicate it using toggles in lieu of the ACS switch. It appears that the short period of time that it is grounded when I rotate the acs switch is causing the pmag to occasionally go off line just before I isolate it.

I have tried to go slow or go fast on the switch from both-left-right and it seems to be agnostic of the speed of switching. Once again, not consistent. I believe that I have ruled out a bad switch and bad wiring. Everything from the green terminal into the pmag can not be tested as I only have one. More to come, but I didn’t want this thread to just hang open.

I haven’t been that good on updating these. Overall I have a great aircraft, just working on knocking the cobwebs off from it being down for so many years.

I’m a little confused. Are you saying that you swapped out your original ACS switch with a new one and were able to replicate the problem on a new ACS switch?

The points in an ACS switch do wear and there is a rebuild kit available. But assuming there is nothing wrong with the new ACS switch, would imply a different root cause with the switch. Especially if using toggles resolved the issue.

Did Brad share any conclusions? Also, which PMag do you have? I think it’s important to identify if it’s a four or six cylinder Pmag. We can longer reference just Pmag when discussing Pmag issues.
 
The process for defeating the 3-revolution firing delay in the PMAGs entails switching 'on' (open) the P-Leads, THEN turning on ignition power.

It is possible that your wiring or starting procedure is doing this, causing the backfire. The purpose of the 3-revolution delay is to prevent backfires.

The correct procedure is: Power 'on' the PMAGs, THEN switch them 'on' (open) the P-Leads.

So check your wiring or procedure.... best to leave the PMAGs powered on (through breakers) whenever the master bus is on. Once running, you can pull the breakers or flip the switches to test the internal generators.

V
 
The process for defeating the 3-revolution firing delay in the PMAGs entails switching 'on' (open) the P-Leads, THEN turning on ignition power.

It is possible that your wiring or starting procedure is doing this, causing the backfire. The purpose of the 3-revolution delay is to prevent backfires.

The correct procedure is: Power 'on' the PMAGs, THEN switch them 'on' (open) the P-Leads.

So check your wiring or procedure.... best to leave the PMAGs powered on (through breakers) whenever the master bus is on. Once running, you can pull the breakers or flip the switches to test the internal generators.

V

So what does that mean for those with an ACS switch? Leave the key in "both" for a few seconds before engaging the starter?
 
The process for defeating the 3-revolution firing delay in the PMAGs entails switching 'on' (open) the P-Leads, THEN turning on ignition power.

It is possible that your wiring or starting procedure is doing this, causing the backfire. The purpose of the 3-revolution delay is to prevent backfires.

The correct procedure is: Power 'on' the PMAGs, THEN switch them 'on' (open) the P-Leads.

So check your wiring or procedure.... best to leave the PMAGs powered on (through breakers) whenever the master bus is on. Once running, you can pull the breakers or flip the switches to test the internal generators.

V
Hi Vern, do you know if this is documented anywhere? I have what I believe is the latest PMAG book and it does not say this. In fact, if I read the book right, then applying power before ungrounding the pmags will put it in setup mode.

My version of the documentation says that RMSD must be activated by the EICAD software, which you can adjust with a PC or EI Commander or Enginebridge.


Emagair-LC114v28 said:
Run Mode Starting Delay – for engines using composite sumps:

Note: The factory default has Run Mode Start Delay DISABLED. See EICAD below for instructions on how to enable Run Mode Start Delay.

Series 114 ignitions can be programmed with a Run Mode Starting Delay (”RMSD”). RMSD is a ONE-TIME firing delay that will suspend plug firing until the ignition sees “X” number passes of the TDC index. At start up, RMSD can help ventilate the intake and exhaust chambers and minimize the accumulation of fuel vapor left over from a prior run (a hot start), or other causes. At the time of this writing, this only known to be an issue for those using composite oil sumps.

After the RMSD (one time) delay routine has been executed, the plugs will fire every revolution. If the engine fails to start on the first attempt and if the operator wants to re-enable RMSD the unit must be powered OFF and then ON in order to reset the RMSD delay.
 
So what does that mean for those with an ACS switch? Leave the key in "both" for a few seconds before engaging the starter?

All the ACS switch does is to connect or disconnect ground to the plead depending on the switch position.

What Vince recommended was to ensure power is applied before doing anything with the ACS switch.

I have an ACS switch with 6cyl Pmags and have not experienced your symptoms, but the process Vince describes is what I follow.
 
...I have an ACS switch with 6cyl Pmags and have not experienced your symptoms, but the process Vince describes is what I follow.

They're not my symptoms. They're the OP's. But thanks for the reply.

It is possible that your wiring or starting procedure is doing this, causing the backfire. The purpose of the 3-revolution delay is to prevent backfires.

The correct procedure is: Power 'on' the PMAGs, THEN switch them 'on' (open) the P-Leads.

I"m just trying to figure out how it would be likely that one wouldn't have "Power 'on' the PMAGs, THEN switch them 'on' (open) the P-Leads". I realize there are several switch possibilities but it seems one would really have to go out of their way to not have "Power 'on' the PMAGs" before they "switch...'on' (open) the P-Leads".
 
Last edited:
Hi Vern, do you know if this is documented anywhere? I have what I believe is the latest PMAG book and it does not say this. In fact, if I read the book right, then applying power before ungrounding the pmags will put it in setup mode.

My version of the documentation says that RMSD must be activated by the EICAD software, which you can adjust with a PC or EI Commander or Enginebridge.

It's documented in Brad Dement's brain! I spent a long time with him working through the behaviour of the PMAGs, in particular this 'undocumented' mode. The purpose of introducing the 3-revolution delay was to prevent the destruction of composite sumps by backfiring.

The reason for being able to defeat the 3-revolution delay was to allow for hand-propping. I hope I never have to hand-prop, but I was walking through all of the failure scenarios and this came up. He did comment that all of the 'fancy' switchology that folks have tried may not allow certain functions (such as hand propping) and to follow Emagair's recommendations.

NOTE: I am working on the development of a new ignition controller program for my huVVer-AVI devices... but it will take some time.
 
Last edited:
Clarifications

All great info shared. Just to clarify my process.

On start up, I energize the main bus, then the PMAG (4 Cyl) power. After fuel pressure builds I start the engine with the ACS Switch.

To do a mag check. I run the engine up to 1800 RPM, rotate the ACS ignition switch from Both through left to right in order to ground the traditional mag (left) and validate the operation of the PMAG. If all is well I remove power from the PMAG to test the internal generator and then reapply power. Afterwards I rotate the ACS switch from right through left to both.

Here is what is happening occasionally. I know the PMAG is online and firing the cylinders. I can prove that because I see a drop when I ground the PMAG using the ACS Switch when I move from both to left. If the PMAG was offline, I would not see a drop going just to the traditional mag. Returning to both, I see the rise in RPM as I would expect when both mags are online.

Here is where the occasional drop happens, about 1x in every 30 attempts.

When I rotate the ACS switch from Both to Left, the PMAG is temporarily grounded. Continuing the rotation from Left to Right the PMAG is ungrounded and the traditional mag is now grounded. It is at this point that all spark is lost to the engine. The engine coasts for 2 seconds and then roars back to life on the PMAG only. I have not touched any switches after the ACS Switch is rotated to Right.

In order to rule out my ACS switch which demonstrated the issue before and after the rebuild kit as well as the ships wiring, I ran a temporary set of wires to a new ACS switch and the PMAG to test. I can replicate the issue on both configurations.

It happens in both directions rotating the ACS Switch. I notice it when I rotate from Right to Left to Both. I dont see any increase when I rotate from Left to Both and then a second or so later the RPM rise shows up you would expect from running on both mags. The dark period for the PMAG is about 2 seconds.

I was going to do some tests recommended by Brad this weekend, but my hangar mate was at work. I need to have someone with eyes on the PMAG led while I run through these tests or I need to set up a gopro or something that can capture the LED status.
 
Back
Top