What's new
Van's Air Force

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

Lean of Peak Problems with Dual P-Mags

Lt Dan

Well Known Member
This is the second of two issues I'm having with my P-Mags on my RV-8 with a Lycoming IO-360-M1B. I've been running dual P-Mags ever since I bought the airplane about 150 hours ago. The other issues is erratic RPM readings, which I posted about separately, so as not to cross streams. I believe this to be completely unrelated.

It has recently become harder to run lean of peak. As I approach peak EGT, the engine seems to develop a slight stumble. It's not what I would describe as roughness, but rather an almost rhythmic stumble. Today I attempted Mike Busch's LOP mag check in flight. I ran lean of peak and alternately switched off each mag. The engine ran very rough and exhibited lots of popping on each of the left and right mags when running individually.

I thought that the problem might have been old spark plugs getting weak, as I only recently learned to replace my automotive plugs after about 100 hours. The old ones had approximately 150 hours on them and I just replaced them with new NGKs: BR8ES Part Number 3961, but alas, the engine acts the same with the new plugs. I gapped them all to between 0.032 and 0.035. Any ideas on what might be causing this behavior?

Thanks in advance for the help!
Dan
 
The only times I have noticed a bit of stumble with my P-Mags I traced it down to a deteriorating plug wire - check the resistance (per the manual) on all the wires and see if that might be the issue.

Paul
 
Hey Lt Dan, I've been a dual PMAG user for about 1500 hours on 3 different planes. Here are some ideas:

1. Time your mags. It's easy (blowing in the tube) and takes about 5 minutes. Follow the instructions in the user manual.

2. Check your fuel flow. Your PMAGs could be perfectly good, but you're simply not leaning to where you "think" you are because of invalid fuel flow indications. It wouldn't hurt to review your fuel flow indicator manufacturer's instructions on how to check and re-calibrate the K factor if it's wrong.

3. Are you using a Red Cube fuel flow transducer with your IO-360-M1B? They have been known to slowly fail (it's happened to me twice). They don't just die, but they become slowly inaccurate for awhile and then one day they'll just start sending a totally erratic signal.

4. I recommend following the PMAG user manual's spark plug gap recommendation. I also ran the NGK BR8ES plugs in my IO-360 engines, and I always gapped to .032. I never had a problem with this. For reference, I could always run those engines LOP at ~7.5 GPH, WOT and they were as smooth as silk.

5. A suggestion for your erratic RPM reading (if you're using a digital tachometer): See pages 4-6 of the PMAG Installation and User guide. Wire 6 in the control lead plug is your digital tach signal. If it's loose in the control lead plug, then you could get an erratic tach. Also check the 2 control plug anchor screws. If those are loose, the plug can work loose and give you tachometer (and other) problems.
 
I run two p-mags also and can lean to 100 LOP perfectly smoothly. You mentioned Mike Bush. If you pay for a subscription and have an EMS, download the leaning data and let his team diagnose the issue. They really know what they're doing. Your ignition should easily pass the leaning health test, especially with new plugs.

Maybe the previous owner timed the P-mags incorrectly. If a mechanic tried to time them like normal mags it would cause all sorts of issues. That's a good place to start.
 
Meter the Pmag spark wires. They should all be in the same range (300-500 ohms as I recall). If you get some above 1000 ohms (or fluctuating when twisting and pulling on the ends) replace or take them apart and carefully replace the connectors. I’ve had some crimps at the plug end go bad causing high resistance. Engine was still running fine when I found the poor crimps.
 
I had a similar issue. The tiny spring steel half circle clip on the end of the spark plug wire that holds the electrode was broken.

IMG_5086-M.jpg


It is the piece on the right.

I found a new one at West Marine, and got it installed on the existing plug wire. I had to destroy the boot to access it. I got a replacement boot from E-mag.

I did the same inflight mag check and it pointed to Cylinder #1 and the top sparkplug. I knew something was wrong when the engine would just burble a bit here and there. Temps on the #1 cylinder were just a bit out of line from previous experiences. Only real indication was that the clip didn't feel tight on the end of the sparkplug, but with the rubber boot, it was difficult to really see what was going on until I cut off the boot.
 
timing and plug wires

All good advice you have gotten here. I also had a bad plug wire. It showed up with sound in the headset and similar symptoms that you have.
 
Non-builder with apparent Pmag issues - have u:
1- obtained the build docs from VANs ?
2- wiring schematic from the builder ?
3- downloaded the Pmag install & maintenance doc from the Emagair website ?
U are going to need all of these. U need to learn how they work & how yours were wired. U may have to get a multimeter and trace wires. Are they both connected to manifold pressure with good connections ? U definitely need to re-time them (EZ!) so u know how to do it and how much advance they will be set for.
I suggest u do this first.
Recently did the GAMI lean test and changed one Nozzle restrictor for less than 0.2 gph differential. Works great.
 
All great advice here. I appreciate it.

I just got done setting the timing. It was good, (maybe 3° past TDC) but now it’s perfect. Of note, I used the “crimp and pinch the tube” method (as mentioned on the E-Mag Air website) instead of blowing and it was MUCH easier!

I also just got done putting an ohm meter to every ignition wire and the results I got were rather shocking. So much so that I don’t know if this is actually possible. Two wires showed good resistance. ALL of the others showed open loops or nearly so. Could the engine have even run if this data is valid? Maybe my ohm meter couldn’t complete the circuit, but the thousands of volts the P-Mags are sending could get the job done (barely)? Is that possible?

The plugs that were in the engine previously were all NGK BR8ES’s with removable terminals and every one of them was coming loose. Is it possible that (as I have heard) these plugs caused arcing issues which fried my ignition harness?

By the way, upon inspection, I found in each of my cylinder 2 and 3 wires, one half of those of those clips as in Bruce’s post above. They were broken, very rusty and corroded, and falling out. Seems like this harness is toast.

I’m still not sure my resistance data could even be valid. I did make sure I had solid connections, correct voltmeter settings, etc.

Cyl 1
Top
Length: 36 in
Nominal: 540 Ω
Actual: 517 Ω

Bottom
Length: 40 in
Nominal: 600
Actual: 1 (Open Loop)

Cyl 2
Top
Length: 40 in
Nominal: 600
Actual: 1 (Open Loop)

Bottom
Length: 31 in
Nominal: 465
Actual: 4050

Cyl 3
Top
Length: 29 in
Nominal: 435
Actual: 1 (Open Loop)

Bottom
Length: 34 in
Nominal: 510
Actual: 1 (Open Loop)

Cyl 4
Top
Length: 30 in
Nominal: 450
Actual: 1,046,000

Bottom
Length: 26 in
Nominal: 390
Actual: 378
 
Last edited:
U may want to buy a new set (2 sets) of wires from Emagair. When I got mine they were already assembled and have held up very well. I ohmed mine about 3 months ago and all 8 were within tolerance.
The plugs u should replace yours with are NGKs 3961. They are the same as yours but with solid tips.
Sounds like u are learning your plane. Good work.
 
Here's an additional update - after seeing those broken clips, I started pulling terminal boots off. Here are some photos of the three that I pulled. Looks like every one of my terminals is bearing a striking resemblance to Bruce's. After that I decided to just purchase new harnesses...


DmrlyUz.jpg


Bjr77jJ.jpg


vhyrbK1.jpg
 
I also just got done putting an ohm meter to every ignition wire and the results I got were rather shocking. So much so that I don’t know if this is actually possible. Two wires showed good resistance. ALL of the others showed open loops or nearly so. Could the engine have even run if this data is valid? Maybe my ohm meter couldn’t complete the circuit, but the thousands of volts the P-Mags are sending could get the job done (barely)? Is that possible?

Yes. Your spark jumps an .035" gap on every event. However, this stack up of resistance will greatly decrease spark energy, so no suprise that there wasn't enough energy to consistently lit a lean mixture (They are harder to light than richer mixtures).
 
A good lesson for us all. Modern electronic ignitions will fire plugs, even with bad plugs and wires, up to the point where things go south. Check the wires and plugs at every annual.

With magnetos, I could watch the waveforms on my portable oscilloscope and determine a problem pretty quickly. I developed a gizmo that turn on a light whenever a plug fouled or a wire degraded, but a rough engine did the same thing. Electronic ignitions don’t tend to have P-leads, so this is harder to do.

I bought a roll of spark plug wire, terminals and crimp tools from Summit Racing for DIY cables to fix some exhaust-burned plug wires on my Rocket. Cores were exposed, discovered at annual. Engine ran fine, but I’m glad I caught it.

V
 
Update - Not the Harness

UPDATE:

Unfortunately, the problem persists! The ignition harness was not the problem. I believe it was contributing to the roughness, however the roughness remains with the harness fixed.

I finally got to fly the plane with new terminal clips on the end of each wire. I bought 8 of them from Napa. For future reference the P-Mag wires are the 7 mm type (hard to find that data anywhere). I replaced each clip on each wire and all were then within resistance specs with solid connections all around.

Conducted another LOP mag check and the roughness persists. It was definitely better (smoother), but not fixed. On each mag, as EGTs rose, the popping and roughness quickly started again.

Back to the drawing board. I'm interested to hear your thoughts...
 
UPDATE:

Unfortunately, the problem persists! The ignition harness was not the problem. I believe it was contributing to the roughness, however the roughness remains with the harness fixed.

I finally got to fly the plane with new terminal clips on the end of each wire. I bought 8 of them from Napa. For future reference the P-Mag wires are the 7 mm type (hard to find that data anywhere). I replaced each clip on each wire and all were then within resistance specs with solid connections all around.

Conducted another LOP mag check and the roughness persists. It was definitely better (smoother), but not fixed. On each mag, as EGTs rose, the popping and roughness quickly started again.

Back to the drawing board. I'm interested to hear your thoughts...

Did you re-time the mags? Sorry if you wrote this elsewhere and I missed it.
 
OK. I like that u are trying the mentioned suggestions. I’ve never had the popping problem u mention. But I did experiment with different settings back in phase 1, 600 hours ago. I settled on timing my Pmags 1 tooth AFTER TDC with the jumpers IN. This lessens the advance by 2.4 degrees (I think ?). Seems to provide the smoothest run and best CHTs for my setup. U may want to try timing yours similarly. Start with 1/2 tooth less advance. As always, be careful and YMMV.
FYI - my LOP today due to Nozzle/Restrictor selection assistance from Don at AFP was 0.0 gph.
 
... U may want to try timing yours similarly. Start with 1/2 tooth less advance. As always, be careful and YMMV.
FYI - my LOP today due to Nozzle/Restrictor selection assistance from Don at AFP was 0.0 gph.

Thanks for the advice. Before I re-timed them about two weeks ago, I actually found that they were set to about where you described. 1/2 to 1 tooth less advance. Same results.

Impressive GAMI spread. How far are you able to run LOP (in terms of degrees)? I’m only able to get 10-30 deg LOP before the engine runs rough at the present time. (That’s with mags on BOTH, of course.)
 
Not sure if you mentioned this, but have you had a chance to upload your engine monitor data to savvy? They might be able to provide some hints based on those graphs. What kind of FI do you have? Air filters clean? Perhaps turn off your squelch on your headset to see if you hear any popping - that could indicate a spark plug wire that has a small break in it which only appears from time to time. Any consideration of getting an enginebridge or a eicommander to see if your timing on the pmags looks good during this lean test?
 
Not sure if you mentioned this, but have you had a chance to upload your engine monitor data to savvy? They might be able to provide some hints based on those graphs. What kind of FI do you have? Air filters clean? Perhaps turn off your squelch on your headset to see if you hear any popping - that could indicate a spark plug wire that has a small break in it which only appears from time to time. Any consideration of getting an enginebridge or a eicommander to see if your timing on the pmags looks good during this lean test?

Good question regarding the engine data. This airplane has Electronics International engine gauges. I’m not a fan at all. The EGT/CHT is provided by the UBG-16. I found a serial port under the panel and attempted a data download, but it seems that this panel was not built with a data recorder. It’s located opposite the engine gauges closer to the autopilot. Any ideas what that random serial port could be if not for a data downloader?

I believe it’s stock fuel injection that came with the IO-360-M1B new about 600 hrs ago. That was going to be my next route. From the outside, they appear somewhat dirty. But I’m not so savvy on fuel injector nozzles. I need to do some more homework before I pull them and investigate.

As for the enginebridge or eicommander, similar story. I don’t know much about those. I’ll have to look into it. Could anyone provide a snapshot of insight as to what is involved installing one of those?

Thanks!!
 
I think you’re going through this the right way - tune up every piece of the engine and accessories until you find the problem. Sounds to me like you’ve nailed down the ignition - now clean the fuel injectors and see what that does.
 
Thanks Paul. You guys have all been great.

I'm thinking in terms of what could possibly cause a periodic and almost rhythmic hiccup. Are there moving parts inside the flow divider? I did some preliminary reading and sounds like it has a diaphragm. Do these things ever need to be rebuilt? Not jumping to conclusions, but how does one test the condition of their flow divider?
 
I don’t know what brand injection system you have, but there’s a final filter near the inlet of the fuel control that should be checked annually. I had an issue with my AFP divider when it got some gunk in one of the metering slots. AFP(Airflow Performance) has a field instruction on how to clean it if that’s the type you have. YMMV. Bob
 
More Data

I was able to run two test profiles today in hopes of gathering more data. I have to gather it all manually, as I have an EI UBG-16 but no recording capability. Long story short, I believe it points more towards the fuel system, though I could use some help. Questions in bold.

Mike Busch's Induction Leak Test:

My EGT drop between high power (25" MAP) and low power (15" MAP) were: 89, 109, 72, and 95 deg F (cylinders 1-4 respectively). I know they are supposed to fall uniformly and if not you can suspect an induction leak. What Mike doesn't say is what is "normal." I figured my EGTs would fall more uniformly (for example, all by 100 +/- 2 deg). Instead, the fall was about +/- 20 degrees. Is this a normal spread?


GAMI Lean Test:
12,500 ft, 65% power.

Now here's where I'm suspecting the fuel system. To start the test, you go to full rich. I recorded the fuel flow at 12.4 GPH. When I looked back at the gauge after recording the value, it had jumped to 14.0 GPH and then settled back down and remained at 12.4 GPH.

I thought this was strange. I did not detect any change in power, nor note any change in MAP or EGT. (It may have very briefly occurred but I don't have recording capability.) Therefore I suspected the fuel flow sensor may be going bad. However, I doubt that's the case, as after any cross country, I can reliably fill the tanks to within 0.1 gal of what's reported by the totalizer.

As for the test, I slowly reduced fuel flow with the mixture control, moving it towards lean in the smallest increments I could and counting to 15 in between movements. The first time I did the mixture sweep, all cylinders simultaneously peaked at 7.1 GPH. A GAMI spread of 0.0.

Upon repeating the test, the first cylinder peaked at 7.4 GPH and while I was recording data, the fuel flow rapidly fell off to 7.1 and the EGT gauge was showing that all cylinders had peaked. This was without any further movement of the mixture control.

I threw out that sample and repeated. On the third attempt, the first cylinder peaked at 7.5 GPH and before I could remove my eyes from the gauge, I watched the fuel flow rapidly drop off to 7.1 and all cylinders again indicated they had peaked.

Now I suspect that my fuel servo may be going bad. The mixture has always be extremely, extremely stiff on the ground prior to start and I suspected that it was the cable, but now I think that the fuel servo may be suspect.

Next time I pull the cowl, I will check the previously mentioned things (injectors and flow divider), as well as the mixture cable, and the fuel servo control linkages. I'll gather more data on what setup I have and report back, but I'm pretty sure it's whatever comes stock with an IO-360-M1B new from Lycoming in 2004 (Precision Airmotive?).

Are there known failure modes of the fuel servo that can explain the behavior described above?
 
it seems that this panel was not built with a data recorder. It’s located opposite the engine gauges closer to the autopilot. Any ideas what that random serial port could be if not for a data downloader?

Thanks!!

If you have a serial port, there is probably a "black box" stuffed in a corner somewhere. I keep a very old laptop around and a serial to usb adaptor just to handle this. The system predates the ubiquitous thumb drives most systems use now. http://befcstl.org/wp-content/uploads/2017/10/OI-II-MUX.pdf

Good luck!
 
If you have a serial port, there is probably a "black box" stuffed in a corner somewhere. I keep a very old laptop around and a serial to usb adaptor just to handle this. The system predates the ubiquitous thumb drives most systems use now. http://befcstl.org/wp-content/uploads/2017/10/OI-II-MUX.pdf

Good luck!

That is what I thought, too. However, there's no "Record" Light on the panel, nor is there a Data Recorder Control Switch ("Burst/On/Download") located anywhere on or behind the panel. I even went so far as to hook up my old computer to the serial port via adapter with EI's software, but it requires you to switch the switch to "Download" in order to pull data. Unless I'm missing something, there's no way to get data without that switch...
 
Mike Busch's Induction Leak Test:

My EGT drop between high power (25" MAP) and low power (15" MAP) were: 89, 109, 72, and 95 deg F (cylinders 1-4 respectively). I know they are supposed to fall uniformly and if not you can suspect an induction leak. What Mike doesn't say is what is "normal." I figured my EGTs would fall more uniformly (for example, all by 100 +/- 2 deg). Instead, the fall was about +/- 20 degrees. Is this a normal spread?

Still wondering if anyone else has conducted an induction leak test with whom I can compare data.
 
That is what I thought, too. However, there's no "Record" Light on the panel, nor is there a Data Recorder Control Switch ("Burst/On/Download") located anywhere on or behind the panel. I even went so far as to hook up my old computer to the serial port via adapter with EI's software, but it requires you to switch the switch to "Download" in order to pull data. Unless I'm missing something, there's no way to get data without that switch...
Strange, this diagram seems to indicate that it's blasting out data on the rs-232 all the time.

electronics-international-ubg-16-page23.png

UBG-16 RS232.png

Savvy seems to support the UBG-16: https://resources.savvyaviation.com/monitor-specific-faqs/
 
Last edited:
Strange, this diagram seems to indicate that it's blasting out data on the rs-232 all the time.

You're right but the "black box" Stan is referring to is the MUX-8. I don't think I have one of those. Without one of those, the data from the UBG is unobtainable/lost/unusable... (?)
 
You're right but the "black box" Stan is referring to is the MUX-8. I don't think I have one of those. Without one of those, the data from the UBG is unobtainable/lost/unusable... (?)
Looks like this serial data connection feeds the black box, and it makes it a bit easier to read. But I think you can get the data from that connection - my guess is it streams so you would need something to read that data in real time - like a laptop or an arduino or a raspberry pi or some other small device that can listen on RS232 at 9600 baud.

Another option is to buy the mux-8, but I personally would use that 300 towards installing a better engine monitor system like a GRT EIS 4000 with one of their EFIS systems.
 
From above :
My EGT drop between high power (25" MAP) and low power (15" MAP) were: 89, 109, 72, and 95 deg F (cylinders 1-4 respectively). I know they are supposed to fall uniformly and if not you can suspect an induction leak. What Mike doesn't say is what is "normal." I figured my EGTs would fall more uniformly (for example, all by 100 +/- 2 deg). Instead, the fall was about +/- 20 degrees. Is this a normal spread?

Went back to my data back in December that I downloaded to Savvy. My MAP was 22 then 12. EGT dropped 131, 184, 133 & 184. 1 thru 4. Odds are same and evens same. Normal ??? No idea. But CHTs and TAS were all good and they looked normal. Data points for the future I guess.
 
A little thought on induction leaks. Any leak in the induction manifold is going to be greatest when the differential between ambient (ambient is the air outside of the engine) and manifold pressure is greatest.

The largest differential between ambient and manifold pressure is at idle! The least is during WOT operation (WOT= wide open throttle). Therefore if you have a induction leak it is almost non existent at full throttle and most pronounced during idle.
 
You're right but the "black box" Stan is referring to is the MUX-8. I don't think I have one of those. Without one of those, the data from the UBG is unobtainable/lost/unusable... (?)
Correct. You have to have the MUX box for memory storage, the UBG-16 doesn't store its own.
However, what I get from this whole thread is an expectation that EGT values be more consistent and predictable than they are, while not noticing that the fuel flow values are bouncing around without control inputs. While even a vernier mixture control is pretty good, you aren't going to nail a setting on an exact value. I am not sure with what you are seeing if it is normal variance or not. The EI "red cube" flow sensor has some small, wires that can break if not well supported, and the spade connectors that used to come with that sensor are not very good at the very low current involved. For more than 5 years EI has been supplying either an OLC-1 or OLC-2 barrel connector, which is reddish round material with either 1 or 2 hex set screws. When I changed my system to those connectors the readings were much more stable, and had to actually reduce the K factor somewhat.
They are available from EI in packages of 5. (you need at least 3). For the unusually low price of $1 per connector. You can also get them from Spruce for a bit more money.
Main point is be sure your readings are as consistent as possible. I would increase your time at each fuel flow setting to at least 30 seconds before changing anything. Don't assume that what can be a sensor problem is an engine problem unless all readings and your seat of the pants agree that the engine isn't running right.
 
Developments!

We have some developments...

Yesterday I went to go fly and the engine ran extremely rough on the ground. So much so that I turned around and returned to parking. The only time I’ve ever done that with this plane. Felt that same rhythmic stumble but rather amplified. Experimenting in front of the hangar, I tested each mag individually. On the left, the engine eventually coughed and died!

I was fully about to finally remove both mags in desperation and send them in to E Mag Air for bench testing as they suggested I do next on the phone.

Uncowled it, discovered what you see in the picture. Left mag ground had become disconnected. The screw backed out.

I took the opportunity to troubleshoot my tach issue as well. E Mag Air suggested I run it withe either tach wire disconnected, one then the other. No progress. The tach was still intermittent on both wires individually. Weird.

However, when I put everything back together it shocked me how loose each terminal screw had become on the mag connectors. I decided to say screw it and cranked them down WAY more than the installation manual spec recommends. Put everything together and... Voila! Started on the first blade, no stumbling on the ground, tach issues were completely solved, ran smoother than I’ve ever felt it before, no problem running deep LOP. In-flight LOP mag check passed with flying colors...

I’m skeptical that all these problems could have been caused by (barely) “loose” wires, considering they were torqued to factory specs but I’m hopeful the issue is solved. Strange lesson learned on this one...

Definitely not counting my chickens yet. I’ll update if anything changes but 5 hours later, so far so good.
 

Attachments

  • 350C5B66-55F4-4B9B-80F1-CFFF174FF297.jpg
    350C5B66-55F4-4B9B-80F1-CFFF174FF297.jpg
    180.2 KB · Views: 176
Mike Busch's Induction Leak Test:

My EGT drop between high power (25" MAP) and low power (15" MAP) were: 89, 109, 72, and 95 deg F (cylinders 1-4 respectively). I know they are supposed to fall uniformly and if not you can suspect an induction leak. What Mike doesn't say is what is "normal." I figured my EGTs would fall more uniformly (for example, all by 100 +/- 2 deg). Instead, the fall was about +/- 20 degrees. Is this a normal spread?

I'll call it normal. High probability the minor spread (20F) is a normal function of the flow divider/nozzle relationship. You did not mention fuel flow at 15" MP, but I'll venture it wasn't much, and here's the story.

At high fuel flows, flow balance is function of restrictor diameter, the little insert in each injector nozzle. As flow is reduced, balance becomes more and more a function of the flow divider. It's normal to see a zero GAMI spread at 9 GPH (for example) become less than perfect at 6 or 7 GPH, because even the best flow divider isn't perfect, and unlike nozzles, it can't easily be tweaked to change distribution.

GAMI Lean Test:
12,500 ft, 65% power.

Now here's where I'm suspecting the fuel system. To start the test, you go to full rich. I recorded the fuel flow at 12.4 GPH. When I looked back at the gauge after recording the value, it had jumped to 14.0 GPH and then settled back down and remained at 12.4 GPH.

I thought this was strange. I did not detect any change in power, nor note any change in MAP or EGT. (It may have very briefly occurred but I don't have recording capability.) Therefore I suspected the fuel flow sensor may be going bad. However, I doubt that's the case, as after any cross country, I can reliably fill the tanks to within 0.1 gal of what's reported by the totalizer.

The red cube works by counting pulses when a rotating shutter breaks a light beam. The rotor/shutter has very little mass, and can oscillate while rotating, adding pulses. The common manifestation is seen when the cube is mounted between the boost pump and the engine-driven pump, and turning on the boost makes the indicated fuel flow rise with no actual flow change.

In this case, I suspect you are sucking air somewhere prior to the engine driven pump. The bubbles would make the fuel flow indication do strange things, while not significantly changing the actual fuel flow. The bubbles would also upset fuel distribution to individual cylinders, in particular when running lean...the roughness. The rhythmic aspect? Air collects in the top of a high point, until some spills into the flow.

It's a theory fitting the symptoms. That said, long distance diagnosis has severe limitations.

Now I suspect that my fuel servo may be going bad.

I doubt it, but than again, I don't know how old it might be, or what sort of fuel it has seen, etc. It's shotgun repair, but sending the divider and servo to Airflow Performance will eliminate the concern.

The mixture has always be extremely, extremely stiff on the ground prior to start and I suspected that it was the cable, but now I think that the fuel servo may be suspect.

The mixture cable sticking is a wild card to eliminate post haste, just because it has the potential to confuse with erratic control.

Some system background here: https://www.danhorton.net/Articles/Bendix-and-Beyond (1).pdf
 

Attachments

  • Light or Receiver.jpg
    Light or Receiver.jpg
    101.4 KB · Views: 103
  • Trigger.jpg
    Trigger.jpg
    138.5 KB · Views: 115
Chronic Problem

Many years ago I ran across the same problem with P-Mag connectors and posted the following solution.

P-Mag employs a type of pluggable terminal strip that uses captive screws to retain the connecting wires.

ybVZJzS.jpg


WkzSJqz.jpg


While performing an inspection of the engine I found what appeared to be fraying and breakage of the stranded wires at the P-Mag connector.

2DE9pwA.jpg


5VfxV3A.jpg


I started to think about how I could make a more durable connection between the stranded wire and the P-Mag connector block. As luck would have it I explained the situation to a friend in the hanger next door. He is an electrical engineer and had worked many years before as a telephone lineman. He knew exactly what I needed to do and I am passing his advice along.

He told me to purchase a ferrule crimping tool and the right size ferrules for the stranded wire. I went to the local electronics store, explained what I wanted, and was astounded when they handed me just what I needed. What they sold me for less than $15.00 was an Eclipse brand Economy Wire Ferrule Tool with 150 assorted sizes of ferrules in the package. The model number is Eclipse 300-016 and it seems to be available at many sources on-line.

H1H9qzF.jpg


GiBKgJu.jpg


I cut off the frayed wire ends, stripped the wire and crimped a metal ferrule onto each of the wires. The wire ends fit perfectly into the P-Mag connector block openings and will not crush no matter how much you tighten the screw in the connector.

TlY7sLl.jpg


The end result is a nice looking terminal end that appears to be far more robust than the bare stranded wire connection. This whole issue may not be consequential under your particular circumstances, but adding the ferrules to the stranded wires inserted into this type of connector just seems to be the right thing to do.

6fNLR8f.jpg


I hope that this information is useful to others that have P-Mag ignitions.
 
Last edited:
Outstanding!

I have never been overly impressed with the connections to the pMags. The ferrule idea never occurred to me. Doing this mod has been added to the list for this month's condition inspection. Thank you!

Beautiful airplane! Not surprising with a last name of Ellis :D
 
D-sub pins

I have never been overly impressed with the connections to the pMags. The ferrule idea never occurred to me. Doing this mod has been added to the list for this month's condition inspection. Thank you!

Beautiful airplane! Not surprising with a last name of Ellis :D

I haven't reached that point yet so this is just an idea. How about crimping a D-sub pin on each?
 
I haven't reached that point yet so this is just an idea. How about crimping a D-sub pin on each?

That might work, but the ferrules seem to be crimped flat - versus the round D-sub pin, so my instinct says they will retain better under the screw. Your thoughts?
 
That might work, but the ferrules seem to be crimped flat - versus the round D-sub pin, so my instinct says they will retain better under the screw. Your thoughts?

A female D-Sub pin could be smashed flat, if it works better. Either that or solder the wire ends prior to insertion.
 
Great idea on the ferrule crimper. I, too, have had issues with the pmag wire connections. I ordered the tool! Thanks, Jim Ellis.:cool:
 
Either that or solder the wire ends prior to insertion.
DO NOT SOLDER THE WIRE ENDS!!!

The solder will wick up the stranded wire that is still covered by insulation. That creates a hard connection failure point inside the insulation. With typical airplane engine vibrations it will eventually break inside the insulation and cause a major troubleshooting issue. That is why you see mostly crimped wires and very few soldered wires in certified airplane installations. This is also why most communications problems in airplanes wind up being a broken wire at the headset jack. One of the few places you have to solder a wire.

:cool:
 
Last edited:
DO NOT SOLDER THE WIRE ENDS!!!
:cool:

Bob Nuckolls and the Aero-Electric wiring guide disagree.

"CONNECTORS and CONNECTIONS: As the various
wires wend their ways about your airplane, they have to start
and stop somewhere and somehow. There are basically two
ways to attach wires to things, crimp the buggers with some
form of solderless connection or warm up the soldering iron
and stick them together."

There are many references to soldering in the manual, including the recommended solder.
 
Nothing wrong with soldered connections, but does a P-mag offer a way to fixate the wires against vibration? If not, use the crimp caps, no solder.

Returning to the original, it's fascinating how the reported symptoms don't seem to match the presumed cause. And the roughness when going LOP is due to a bad power ground? It's a P-mag. In theory it doesn't need external power, but a loss of ground makes it run badly?
 
Last edited:
...
Returning to the original, it's fascinating how the reported symptoms don't seem to match the presumed cause. And the roughness when going LOP is due to a bad power ground? It's a P-mag. In theory it doesn't need external power, but a loss of ground makes it run badly?
The pmag manual seems to say that lack of ground could cause "bad things" to happen.

pmag ground - manual - 20210413.png
 
Nothing wrong with soldered connections, but does a P-mag offer a way to fixate the wires against vibration? If not, use the crimp caps, no solder.
AFAIK there is no way to fixate the wires against vibration in a typical P-mag installation.

:cool:
 
Back
Top