P-Mag bummer!

My inspection is on Tuesday. Started engine yesterday and my Pmag is dead!....swapped wiring to Emag on the other side and it ran great...Yup Pmag is dead as a dodo.

Strabge thing is it timed up OK...the LED came on when it should etc. But when it runs it won't even give a tach signal...

So even if I pass inspection there is no way to make Van's homecoming next weekend.
I guess the moral is don't wait till the last minute to statrt the engine!


Darn it!

Frank

Corvallis, Oregon

Is it possible that it's still in timing mode? IE the red/green LED is on? I know that in that mode it won't make sparks...

Pmag dead

Good thought Chuck but I pulled the plug (complete with wiring) off the Pmag and plugged it directly into the opposite Emag and it ran great on the Emag.

I think to stay in timing mode the P lead has to stay grounded....As the wiring worked on the other Emag then by default the P lead could not have been grounded.

Looks like i'm dead in the water for a week or so...Not a big deal in the grand scheme of things but boy its hard to have a running airplane you just can't fly....

Frank
Inspection Tuesday

Call Brad

Call Brad first thing this morning. I'm sure he can Fedex you another one.

Mine (Emag/Pmag) set up has worked flawlessly for 135 hours.

I understood that, the check I'd make is to see when you have it in 'run' mode that the timing LED is off, that will 'prove' that PMAG is not stuck in the timing mode. That is a 'different' test than what you did. Your test *might* be susceptible to an intermittent connection. If you did this test and LED stayed on you would chase wiring, and if it went off you'd chase Brad...

In thinking about this it the EMag should have the LED indicate the state it is in with the LED, currently there are 4 states Run/Timing/Generator and Off. Right now only the timing mode uses the LED. It would be nice if it gave you an indication for each of these states...

Good luck

Chuck

Good idea

Yes I will check to see if the LED goes out when the mag lead is ungrounded...

At least the FEDEX building is right on the airport...

Frank

I think i solved it ..DUH!

My dead P mag worked great back at the factory.

Then I had a brain wave.

I had tried to combine the off-power fail test-Run modes onto one DPDT switch.

Unfortunatly there is a momentary change in state when you go from OFF to one of the other states.

What happens is the Pmag becomes ungrounded before it gets power.

This does not work...You have to have a seperate switch for the PMag power fail...Preferably a normally closed momentary switch.

Frank

Frank,

This is not good news to me. Are you saying that when I take ship's power away and momentarily lose ground too, that it doesn't come right back on when things are restored? It'd be great if you could elaborate just a little on the specifics of what went wrong.

Correct

Its not when you lose ground...The P lead is UNgrounded in flight.

Its no big deal in terms of flying. I.e when your cruising along the PMag will be ungrounded....If it then loses power it will operate as normally.

I.e not an issue.

The problem is if when starting you Unground the Plead before you apply power...The Pmag then won't work...At least mine didn't.

As this will never happen in flight its a non issue.

The only reason it was an issue for me was that I combined the OFF-TEST-Run modes onto a single switch. You cannot do this because no matter how you wire the switch you will always unground the Plead before you apply power.

Long story short you need a seperate switch to test the power fail mode...or pull the CB or whatever.

Right now I just have the power connected permanently...I.e no way to test it (unless I pull the fuse) .

I want to find a normally closed momentary switch that I can't leave in the OFF postion...Haven't found one yet.

Frank

I test my P-mag at shut down rather than at start up.

The P-mag is powered from my main + buss. At shutdown, I turn off the master and the left (Slick) mag. If the engine continues to run, the P-mag is functioning.

Ben Beaird
-6A

frankh said:

The only reason it was an issue for me was that I combined the OFF-TEST-Run modes onto a single switch. You cannot do this because no matter how you wire the switch you will always unground the Plead before you apply power.
Frank

Click to expand...

You need to look for a "make before break" switch----you are using a "break before make" switch.

For the momentary "normaly closed" switch, try a DPST, and just leave one side---the NO, unused. Probably easier to fine than the SPST you are looking for.

Good luck,
Mike

So, if I understand correctly, taking away ship's power while the engine is spinning and then returning it causes no problem, but taking away power before engine is spinning confuses the unit, requiring grounding, restoring power and then ugrounding to get it going. Is this it? Dang, I think this means that my wiring (a la Nuckoll's) is going to need to be redone.

Just when I think I'm out, they pull me back in!!!

Not before its spinning

It's UNgrounding before applying power...If you do the power and ungrounding on one (break before make) switch, it becomes ungrounded 1/10th of a second before power is applied.

That for me didn't work.

Frank

According to the info I have, that shouldn't be a problem. In my earlier version of the manual (corresponding to an early firmware version), selecting the run mode starting delay is done by ungrounding the P-lead before or after applying power. Ungrounding the P-lead before applying power should just bring up the P-mag in run mode. I'm confirming with Brad at E-mag Ignitions.

And I don't see any issue with the B&C 2-10 switch that I'm using. In the down position, the P-lead is grounded, power off. Middle position, P-lead ungrounded, power off. Switching to the top position, the P-lead remains ungrounded the whole time, and power turns on.

I don't really want two switches for each ignition.

HUH

A 2-10 is what I have and wired as you describe....

Since the Pmag came back I simply permanently connected the power...Maybe I'll have to re-try this...Do tell us what Brad says.

Thanks

Frank

P Mag

I have a 3 Amp Circuit Breaker switch on mine, and it works fine.
Al Grajek

Here's my dialog with Brad at E-MAG.

Me: Somebody on the VansAirforce forum had some trouble with using a switch that ungrounded the P-mag before applying power to it. That's the same way I have my switches set up. That should work, right?

Brad: In order to enter setup mode, the ignition needs to be powered up WHILE the p-lead is grounded (and stays grounded as long as you want to continue in setup mode). If you unground the p-lead at any point, you will leave Setup mode and go into RUN mode.

Me: So after setup is done, and I'm just starting up the engine normally on a given day, does it matter if the p-lead is ungrounded before applying power to the P-mag? It'll just come up in run mode, right?

Brad: correct


Sounds like the 2-10 switch should work. Let us know Frank what you find out if you go back to using the switch.

Dave

Thanks Dave for reporting back. It is certainly easy enough to keep it grounded for setup mode with a temporary jumper wire.

Will do

Sounds like the 2-10 switch should work. Let us know Frank what you find out if you go back to using the switch.

Dave[/QUOTE]

When I get round to it..

Frank

Yeah, I can see how flying might be more fun than tinkering at this point.

Indeed

MUCH better than tinkering

Did you follow the single switch p-mag wiring suggested in Bob Nuckoll's Aero-electric Connection? This suggested wiring has been out for awhile, and I would be surprised if it does not work, as Bob has many followers and it seems that this would have been previously reported. I have forwarded your report on to the Aero-electric list for further consideration by others, and will report back with any findings. THis concerns me, as I have also used a single switch for my p-mags, but am not yet to the test start stage.

regards

erich weaver

Ya I saw it

On the Aeroelectric list.

Actaully it was not THAT big of a deal for me. I got the E/Pmags back from the firmware update the night before the inspection. As the single switch did not work before and Brad told me it worked just fine on the test stand...AND...as I had very little time I just connected the power feed permanently and it has run great since. Of course I currently do not have a way to check the power fail mode but with a 50 mile day VFR test radius it was pretty low on my priority list.

This week I am grounded (on call for work) so I am doing the chores I been putting off while building the RV (timing belt on the car was Monday, replacing rotten siding on my rental trailer this weekend!..YUK)...So next week I'll be back boring holes in the sky.

I wire up the switch again and report what happens.

Frank

P.s I'll copy this to the aerolectric list

Failed P-mag

Today I joined the club of those that have experienced a P-mag failure. Total time on the ignition is 13hrs. When only this mag was selected during run-up, the
engine died. The engine was warm from a recent 1hr flight, before
which the ignition ran fine before takeoff. Luckily I was at my home
field and returned to my hangar for diagnostics. All wiring and
switching was checked and confirmed to be fully functional. Grounds
are good, power was good on the correct pin, no LED indications in run
mode (LED off), in setup mode (grounded with power applied) timing
looked correct with the proper green LED at the correct prop location.
After these diagnostics, the engine was started two separate times
within the following 2.5 hrs. with the same failure.

Just a data point. The unit is going to e-mag in the morning.

Sorry to hear that Scott. My 2 P-mags have been running strong for 50+ hours.

Update

I heard back from Brad today. They ran may failed P-mag for over an hour and it ran fine. It had not exceeded their temperature specification (which he said is the same as a mag, whatever that is). The only weird thing that they noticed is that the run time counter only turned on at operating speeds and showed only a few minutes of run time. They replaced the electronics board and are sending it back to me. So far as I can tell, the only difference between when I removed it (failed) from the aircraft and shipped it back and them testing it today (not failed) was temperature. When I took it off, it was still warm, but not so hot that I couldn?t handle it. He assured me that they would do whatever they could to keep me happy, including offering to send another e-mag to have on the shelf while they were at Oshkosh just in case that would make me feel better. I declined the standby unit. I will install the returned unit and continue flying with an abundance of caution.

Update: Second failure

Emag replaced the electronics board in the failed unit and sent it back. I re-installed it and did a one hour flight yesterday. The new unit has software v.26
Now for the kicker... The other p-mag ignition failed today before the first flight of the day. Here is a clip from the email I sent Brad today:

I re-installed the repaired #634 unit and flew with it for one hour
yesterday. Both ignitions performed fine throughout the flight. I
checked that both units were firing after the flight before I put the
airplane away. Max temp on the indicator sticker that you put on the
repaired unit currently reads 190deg. F. Today I prepared for another
flight. This would be the first flight of the day. I started up and
was at the run-up pad within about 3min. The other mag (serial# 633)
was not firing at all. After identifying the ignition failure, I
switched off power to that ignition, then grounded the P-lead (full
ignition unit shutdown). Then turned it back on, un-grounding the
P-lead and applying power. It was still not producing ignition.
Back at the hangar, I pulled the cowl and checked all wiring and
switching. It was all fine. I checked the timing and it was correct
with normal LED indications. I pulled the aircraft back out and
started it again (no cowl). The failed ignition (633) ran fine. I
shut down and put the cowl back on and prepared to taxi back out to
the runway. Before I taxied out, I did a quick mag check and it
(#633) was failed again. Obviously, P-mag serial# 633 and I will not
be flying together.

Obviously the likelihood of both of these units failing at the same time was very great. If that had happened, the world would look very different for me right about now.

??

Do you know what software is in your 633 mag?

cawmd82 said:

Do you know what software is in your 633 mag?

Click to expand...

Yep. v.25

In my message to Brad, I also invited them to come inspect my installation of they were interested.

Third pmag failure

And we thought this conversation was done... NOT! I had my third pmag failure this weekend. Each failure has been on a different unit and with all wiring re-done once. Believe it or not (I don't care at this point), but this is not a failure of my installation. I have never seen the reported timing failures. All of mine have been that the unit simply stops firing at all.
My denial is over and I feel like I won the best prize of survival to get this far with these ignitions. I'm done messing with them and have one Slick mag and one Lightspeed on their way. No more P-mags for me. My full details are on my build log if anyone is interested in them.

I'm glad you got out before they hurt you. One of the happiest days of my life was when I saw those P-Mags going south of the ups truck.

Pmags

The same thing happend to me.
Had 2 Pmag failures during the 40 hours I flew with them .
Both failures on run up. Slick's are back on.
A friend with a RV-8 had 3 Pmag failures in flight after 10 hours.He's back to Slick's.
Emag guys have made a couple of software changes since my failures.
Hope they got the bugs out.They are nice guys and will help you anyway
they can. But when the engine quits your on your own.

Rich Rudolph
RV-6A
800 RV hours

Scott,

What are the serial numbers on your P-mags?

I'm wondering if they were an early version or not.

Sure hate to hear about your problems but thanks for keeping us informed.

I now have 77 hours on my two P-mags since the last software upgrade and have had no issues of any kind. (I'll post my SN's later, when I get home.)

Last week I checked the timing while the cowling was off for inspection and both P-mags were spot on.

N941WR said:

Scott,

What are the serial numbers on your P-mags?

I'm wondering if they were an early version or not.

Click to expand...

My oldest one is Serial# 634, the newest one that recently failed is Serial# 1079. Both Model 113 with current software.

Using the search feature on my build log for "P-mag" pretty well organizes my experiences with P-mags.

Scott:

Do you know whether or not your recent failure included failure of the p-mag's internal power generating ability? That is, was your RPM high enough so that the unit should have been able to produce a spark on its own without being supplied power from the battery?

thanks

erich

erich weaver said:

Scott:

Do you know whether or not your recent failure included failure of the p-mag's internal power generating ability? That is, was your RPM high enough so that the unit should have been able to produce a spark on its own without being supplied power from the battery?

thanks

erich

Click to expand...

If you're interested, you should really read the full account on my build log. I didn't rake full through the exact details here on VAF like I did on my own space. Just trying to be respectful of Doug's space here and not get into much vendor bashing, yet make sure that relevant experiences are shared.

That is a really odd coincidence that 3 different boards went south. Thanks for the report.

On the advice of another builder, I sent my 2 P-mags back a few weeks ago for the latest firmware. When I re-installed them, I used the factory timing mark. Actually, I used a blow-in-the-tube timing mark on one of them that was close to the factory timing mark, but that caused some real issues. I went back to the factory timing mark on both. No problems.

Davepar said:

That is a really odd coincidence that 3 different boards went south. Thanks for the report.

On the advice of another builder, I sent my 2 P-mags back a few weeks ago for the latest firmware. When I re-installed them, I used the factory timing mark. Actually, I used a blow-in-the-tube timing mark on one of them that was close to the factory timing mark, but that caused some real issues. I went back to the factory timing mark on both. No problems.

Click to expand...

I'm not sure "coincidence" is the right word. We're dealing with an equipment engineering problem here. I'm an engineer as are many people here, whether it be software, mechanical, or aeronautical. We all know the drill. I'll even present something more than that generality. I feel pretty confident that people are seeing a mix of logic and HEAT related issues. With a little more work, I feel like emag could prove the heat side of the hypothesis. I think I was closing in on it pretty quickly, I'm just running out of lives to test it to the end. I have no more interest in being the guinea pig or developing proofs for the manufacturer. If you've seen those fins on the model 114 units... If these units are so sensitive, they have no place in an airplane. Those fins should be much bigger from what I've seen.
I wonder if those that are experiencing failures, not of the logic type with timing "slip", are in high temperature conditions?

Here is an opinion piece that I wrote in 2005: It's full of both sceptism and hope for Emagair's success. I'm sorry to say that I will not be upgrading at this time. Perhaps when my mags need rebuilding I will reconsider.
========================================================

The story of P-Mags and E-Mags

When I heard about EMagair, the maker of P-MAG and E-MAGs, I thought 'what a great idea!'. I designed out my conventional mags and designed in the EMagair products with their assistance.

I placed my order and waited, but the products were delayed several times and I finally had to get my engine built so I cancelled my E/P-Mag order and switched backed to conventional mags. I'll leave all of the wiring in place for a future upgrade, so if EMagair can ever deliver and proved their reliability, I can upgrade. It will cost me more money, but it's a safe way to go.

I have been involved in dozens of new product introductions in my career, and many of them had technical problems and delays at the beginning. The company I have worked for the last 17 years (PMC-Sierra) was very proud of it's record on getting new products out on time with minimal technical problems, but sometimes we had some problem cases.

My opinion is that EMagair is having technical problems, and they want to take the time to get their products right. That's what they should do. In my discussions with them, however, I became concerned about how they are going about things. It's entirely possible that they will have a great product, but a broken business plan. I hope they are getting good business advice.

My background is engineering, marketing and business analysis (mergers and acquisitions). I have seen a lot of companies fail, even though on paper they have a killer product. It's all about the right plan, the right people, execution and documentation.

Documentation?? Yup, 90% of the businesses I analyzed failed because of poor documentation. When you have 50 people designing a 20 million gate integrated circuit, you better get the engineering documents, manufacturing and customer documents right or you will never, ever find all of the bugs, never ship the product for revenue and never have a successful business.

EMagair is a different animal, I agree... but being successful with a product like this (eventually to be certified) will swamp them in paperwork. For example, what is their qualification test plan?

Are they going to let their customers debug their products for them? I hope not. If I were them, I'd get 3 or 4 aircraft test beds in the air as soon as possible with backup magnetos, and run them through a battery of tests, collect data and analyze it before I shipped any products to customers.

Things to think about...

- will they work at temperature extremes?

- will they work with a lot of temperature cycling?

- what about high altitudes?

- what about vibration (normal and abnormal)?

- what about contamination from engine fluids or dirt?

- what about reliability in corrosive atmospheres?

- what about susceptibility to electrical interference (comm radios, transponders, etc.)?

- what happens when the internal micro goes insane?

.... and so on.

There are many questions. They have been good at answering some, but not others. Food for thought. Good luck to them. If they are successful I will upgrade.

========

I feel your pain; at 219 hrs on my dual Pmags I've had 2.5 failures... the 1/2 was not a complete failure, but it sure wasn't running right.

One was a lemon right from the start; sent it back to the shop several times during initial ground runs and test flights. Finally, at 13 hrs it quit again and I was done with it. Brad replaced with a new unit; I put another 170 hrs on that one until it too quit in flight. Once on the ground I shut everything down to reboot, and it worked fine. I sent it in anyway; they couldn't duplicate the problem, but replaced the board and it's running fine since.

Here's a strange thing about the other unit though; it was running well for 180 hrs or so, but when I sent the failed one in, I figured I might as well send them both in for the latest updates and firmware. After that, the original that had been running fine from the start began giving trouble; runup on the ground was fine, but in flight there was some roughness... when I tried to run on just that one unit, it started running rough, sputtering and popping and losing power. I was also using that unit for my tach feed; after the update the tach readings started jumping around on my EIS from time to time. I began to wonder if something was wrong with my installation, so I swapped sides with the good unit; the problem moved with the unit. Timing lights were green. So, I sent it back in. Again, they couldn't duplicate anything on the bench, but Brad said the temp had spiked pretty high... dunno why, I've had blast tubes on 'em from the start. They replaced the board on that one, and now have 8-10 hrs on it so far, no problems.

I haven't had any timing shift problems at all; I always use the manual positioning method, and each time I've had trouble, I check the timing and see two pretty green lights. So, according to the lights it's not a timing thing, it seems that they quit running or just run rough. Brad and Tom have been very helpful to try and resolve these issues, but I have to say it's been a bit frustrating lately. I'm thinking that if they give me any more trouble I might ditch 'em and buy a pair of Slicks...

Testing, Testing, Testing....

Vern's write-up hit the nail on the head, and this holds true of so many products and projects. It was a discipline pounded into me by the old men of Apollo - test, test, and test again! It is expensive and time consuming, but the system or component doesn't go into service until you have tested it throughout it's design envelope, and well beyond - to figure out where that envelope really ends.

In the space business, flights are too costly and dangerous to leave to chance - for ever component flown to the moon, many, many more were destroyed in testing, and the folks in control had the failure analysis data in hand when it came to getting the most out of the overall system in flight. Today, we have lost that mantra of "test, test, test" due to the cost of doing things that way....and we pay the price by having less data to look at to make good decisions. The same thing happens in our little corner of experimental aviation with very tiny start-up companies - many of them can't afford to test by themselves, so they get customers to do it. If it is done in a formal Beta-Test program, and the testers are really, truly understanding of this, it can be fine. But I think that sometimes we mistake a pretty box and laser-graphic manuals for a "seal of approval" that this thing or widget it a commercial product like the iPod you bought at Best Buy......but it's not.

I have chosen to fly behind a lot of "archaic" equipment - a Lyclone with Mags and a carb - because I personally wasn't interested in experimenting with that part of my airplane, and there are millions of flight hours and decades of testing that we can rely on to know how they are going to operate. I enjoy testing and being a Beta tester on other parts of my airplane, but I go into it with my eyes wide open.

Experimenting and Testing is important - it's why many of us are IN aviation - but go into it with your eyes wide open and be prepared for the consequences of it. It's really fun to be on the leading edge of a new product. But If the company I'm dealing with has great ideas yet is still operating out of a small garage or hangar, I have to realize that I might be a Beta tester, not a "fly it away and forget it" customer.

Paul

Quality testing

Good point on the testing thing.

If any of you want to know how the E-mag boys are testing their equipment, call Brad. He will answer any questions you may have.

The good news is, they are going after FAA certification for the P-mags so we will all benefit from their testing.

BTW, years ago I read Michael Collin's book Carrying the Fire: An Astronaut's Journey, which I highly recommend. In it he talks about the quality control measures NASA took at the time. He stated something like, "If we had a 99.99% success rate on all of our systems we would still have 10,000 failures." (It has been years since I read this book so my numbers might be off but you get the idea.) Complex systems are a challenge to debug.

Little testing in the X world

I'd submit there is very little actual flight testing that goes into the majority of products in the Experimental world. As someone pointed out, many of the companies are small and do not have the resources to do "certification," type testing.

As one example, remember all the problems Dynon had with the early 10 series EFIS? All the bench running didn't do anything to reveal problems until the units got into the field and used in real world flying.

I know that Emagair has a test engine used in their business and use it extensively. Still isn't like flying though.

Expecting 1000's of hours of testing for Experimental is completely unrealistic. If so, we'd still all be using vacuum gauges, round engine instruments and VOR's Only the big buck companies, Garmin comes to mind, can afford to do the testing. There is a definite "rush to market" mentality. One major Subaru company immediately comes to mind.

Buyer beware. We assume these risks, hopefully after careful review and investigation.

I've been following the P-mag story with interest and have refrained from comment until now. I cringed a couple weeks back when someone posted that a new software upgrade would solve all problems. In the EI field, people should never post stuff like this. You just don't know until that version has accumulated a lot of bench and more importantly actual use time to validate. Rushing products to market is often a short, quick path to bankruptcy.

I would hope that anyone offering electronic products even for experimental aviation at least does high and low temperature testing, low voltage testing, moisture/ humidity testing and EMI/ RFI testing before release. They should also get a few hundred hours minimum actual running time in the most severe conditions they can. Real world testing often uncovers faults that will never be discovered in bench tests or in a programmers wildest dreams.

While this costs some money, it will cost far less than a damaged reputation, replacing hardware or software or worse...

Companies producing digital EIs should also consider automatic reboot features.

I hope P-mag can lick their problems and certify their products but I see an uphill battle for them now with the historical facts in hand, increased FAA scrutiny and the costs involved.

scard said:

I'm not sure "coincidence" is the right word...

Click to expand...

True. Although I would say it's statistically unlikely to have 3 failures at random. I'd guess there is something about your installation that exacerbates an issue in the P-mags. Could be heat, vibration, electrical, etc, or a combination thereof. You obviously did more than due diligence to make it work.