Once Again Science and Engineering (finally) Overcome Fear and Superstition

Many thanks to all those who wrote on line and directly to me.  A lot of excellent ideas were proposed. 

The story got complicated.  Sit back with a cocktail and cigarette, relax, and all be revealed.  It was almost as elusive as the leaking O ring at the fuel pump inlet reported here a few weeks ago. 

Recall that I began to get an intermittent miss on the electronic ignition during run up.  I would increase power, lean mixture to burn the plugs, and that would usually make the missing  go away.  But it became more persistent but intermittent missing badly sometimes and not at others. 

Time for corrective action.  First I put in new plugs although the old plugs looked pretty good.  It was a little better, but still the missing would come and go.  Then I lined up some friends and we decided to go for it and fix it once and for all. 

What a battle.

There are rumours circulating around the airport that we sacrificed a large number of farm animals and studied their entrails for insight and guidance, burned incense, and held hands while standing in a circle around the engine while under a full moon and singing Kum-by-ah.

These are exaggerations. 

Only a limited number of animals were sacrificed including a cocky RV builder who made one snide comment too many. 

In the end “Persistence and rationality can always overcome inanimate objects.” [Brent Regan, 1843]

The maddening part was that we found the mis-firing would occur with the cowl on, but not with the cowl off.  Most of the time.  Sometimes it went away with the cowl on until a mag check at the end of the flight when it would reappear.  We figured the cowl effect must be air blast on the wires wiggling some intermittent connection.

Wrong.

In the end we switched plugs, switched plug wires, rerouted all the plug wires, rerouted all the primary wires going to the coils, replaced the a coil with modest improvement (sometimes we could see it was number one missing, other times the miss was too slight to detect with the EGT), replaced a lot of wire terminations that might,  just might be suspicious, tied down all wires so they could not wiggle, checked the magnetic  sensor board, magnet clearances and connections, and then started sacrificing the farm animals.

Sometimes our changes made things better.  Sometimes they did not.  The only consistent trends were that it was worse when wet and cold in the morning (first start of the day) and fairly consistently putting the cowl on caused the problem to reoccur.  Usually.   Not always.  Grrrrr. ..

My Lightspeed electronic ignition box lives on the front face of the firewall inside of its own protective fibreglass box which is cooled with blast of air carried in from the upper cowl via a one inch diameter SCAT tube.  It must be the air blast on the electronics, right?  The supplier said no way, it would not affect only one cylinder, but we tried anyway, with and without blast tube.  No difference.  The supplier was right.  Grrrrr…

When psychic invocations failed, we decided to disassemble the entire system and inspect every wire, connector, termination, routing, in short, everything we could think of.   This meant having a close look at the Lightspeed ignition box which lives on the front face of the firewall in its own fiberglass box.  Off with the box and then we removed all the connectors carefully inspecting as we went.  No problem found. 

We pulled the ignition box off the firewall.  We looked down into the 25 pin Dsub connector, and there visible under a 500 watt work light was something.  A magnifying glass showed that the something was a sliver of aluminum lying at the bottom of the connector next to the base of the tiny gold pins.  And the sliver was located in proximity to pins nine and ten which carry the magnetic pickup signal for channels A and B to the ignition box.

AHA! We got it!  Our theory was that the sliver of metal was moving around causing problems on the channel that fires cylinders one and two and occasionally three and four.  When we put the cowl on, it changed the vibration characteristics of the firewall, so that would explain the cowl effect.

No doubt about it, we got it this time for sure (third time we had said that).

Reassemble, push plane out of the hangar between rain showers, test, no problem.  Install plenum and cowl, push back out, and test again.  No problem.  Got it for sure! 

Taxi up to the apron, do a normal full run up, and number one cylinder pops and bangs and drives the EGT bar way up.   Grrrrr… 

How could we be so wrong? 

On a hunch, we pulled the cowl, pulled the inner plenum, pulled number one plug which was somewhat fouled.  Put in a new plug, reassembled everything, plenum and cowl back on, run up checked out OK, test flight checked out OK, run up after flight, tested OK. 

The following morning we did the acid test:  a cold, wet morning start.  Started OK, run up OK.  

At last.  

Maybe. 

Lessons learned:

1)   Persistence pays off.

2)   Keep your connectors covered.  If you lose the plastic covers, use tape.

3)   Electrical problems can propagate down the wire causing more failures elsewhere.  The sliver caused an intermittent signals which made intermittent pulses which ultimately ate a coil and then a spark plug while we continued the chase.  As a result, you seem to be going in circles.

4)   I have met the devil and he is me.  I did it to myself.

Fred Moreno