DavidHarris
Well Known Member
I had an engine puzzler that kept me occupied for two months. I'll post the story and clues (and red herrings) today, and then follow up with my solution and reflections in a few days.
In March I received the airworthiness certificate for N47HM, an RV7-A with a Superior XP-IO360 engine, dual PMAGs, and a Precision Airflow fuel servo.
About 8 hours into Phase 1 testing, my engine quit during stall testing (at the top of a stall at 48 KIAS, with partial flaps). Much to my surprise, it would not restart in the air despite playing with the boost pump, mixture settings, throttle, and starter, and pitching the plane down to increase the windmilling. I shut everything off and made a forced landing on the sandy bed of the dry Mojave River. Fortunately I landed into a good headwind, the sand was firmer than usual because of recent rains, and I held the stick full back, so the nosewheel didn't touch until I had come to an almost complete stop. The aircraft was undamaged. On the ground, I observed the right PMAG switch was off. I likely bumped it while entering a transponder setting during turbulence, although it is not impossible that I failed to turn it back on after turning it off during my runup test.
I had been on flight following and had reported the engine out to ATC. On the ground, I called the Victorville tower, and they sent the fire department out to meet me. I jogged about 3/4 mile to the nearest named road, met the fire department and sheriff, and then talked with the FAA. All was well, and they referred me to Jack Vanderman at Southwest Aircraft Recovery.
My family came up to meet me. The area was dubious, with burnt-out houses and ranches resembling junkyards. I found a sandy service road on the riverbank that reached within 150' of the airplane, and spent the night there in my truck to guard against looters. In the morning, I met Mr. Depue at Depue Airpark, a private dirt strip near Barstow, and got his permission to tow the plane over there. Jack Vanderman arrived with his flatbed trailer and aircraft crane, and we spent about 6 hours winching the airplane out of the riverbed, up the embankment, and onto the flatbed, then winching the trailer along the sandy road until it was firm enough to drive. He carefully drove 10 miles on Old Highway 66 to Depue, and the police car that passed us in the other direction kindly didn't bother our odd caravan. Jack's elderly father and my 14-year-old son both helped with the recovery and it was quite an adventure for four generations of aviators.
In hindsight, I would have done stall tests at 7000 AGL over Dagget airport rather than 3000 AGL over the desert. I would have looked for a dirt road rather than a riverbed. I would have aggressively tracked down engine issues on the ground before flying. I was thankful to have taken glider training last summer and to have memorized the emergency checklists and thought about them deeply. I was grateful for the help from the FAA, the Victorville Fire Department, Mr. Depue, and Jack Vanderman.
Two days later I went back to Depue, removed the cowling, did a lengthy inspection, found nothing unusual, put it back together, did a runup, and then departed the short dirt strip and flew back to my home base at Cable Airport. Flight was normal.
I had observed some issues during Phase 1 testing before the incident, and didn't troubleshoot them on the ground as much as I would have wished because I feared long ground runs while breaking in the new engine.
* The engine would sometimes quit during the rollout after landing when I was at low RPM. I attributed this (not quite correctly) to the idle being too low. Adjusting the fuel servo idle to 900 RPM prevented the problem, but caused a long landing, so I had set it back lower for further testing.
* The engine was also sometimes very difficult to start, even though I felt I was following the Superior and Precision instructions. Sometimes I would have to prime and crank up to 6 times, following the cool down instructions for the Skytek starter.
After recovering the airplane, I contacted Superior. Bill Ross, VP of Product Support, took decisive steps to help, at no cost to me, and gave instructions for testing before further flight. I wiggled the wires while monitoring the LEDs on the PMAGs to verify connectivity. The grounds both go separately and directly to the aircraft central ground stud on the firewall. The power lines go through test switches on the panel to a short shared wire going to the main power bus. The kill lines go through the key switch to ground. The PMAG LEDs turn on when the power line is at 12 volts and the kill and ground lines are properly grounded, showing connectivity. Aggressively wiggling either end of the wires on either PMAG wouldn't turn the LEDs off. I then removed the PMAGs and sent them back for bench testing. I removed the fuel servo, spider, and injectors and sent them back to Precision Airflow for bench testing. Everything came back looking normal. I reviewed the excellent Dynon engine logs and observed 0 RPM from both PMAGs after the engine failure, even though the prop was windmilling, which is an indication that the PMAGs were not powered up during the descent, despite have internal alternators that should provide backup power. I spoke with Brad at EMAG and Alan at Precision and both were very knowledgable and patient and helpful, but none of us could find a theory that explained all the facts. I was impressed by how all the vendors stood behind their products and took the time to walk me through everything even though I'm not a mechanic.
I reassembled the engine, changed the oil, replaced all the spark plugs, and started the engine again. Starting remained difficult at times. Superior confirmed from the engine logs that the engine break-in was satisfactory and that it was ok to do longer ground tests. I tuned the idle throttle and mixture settings using a screw and thumbwheel on the mixture servo, in consultation with Alan at Precision. I had a difficult time getting the idle right; if I turned it much below 800 RPM, the engine would quit at idle, while the manual called for 700-750 RPM idle. After a bunch of testing one hot afternoon, I finally found a setting at about 840 RPM where the engine would never quit on the ground at idle. Out of an abundance of caution, I waited until the cool morning to fly the aircraft. My son stood by as an observer with a fire extinguisher, and I started the engine again, but found the engine quitting when the throttle was pulled back to idle. I spoke with Alan again and he was puzzled and said there was no known temperature sensitivity for the idle RPM. My son adjusted the idle higher. Then I found the engine wouldn't start at all despited repeated priming and cranking.
Then my son told me he saw something. His statement was the clue that resolved the whole mystery. He and I made a small fix, and the aircraft has been starting, idling, and flying well for two weeks since then.
Happy puzzling,
David
In March I received the airworthiness certificate for N47HM, an RV7-A with a Superior XP-IO360 engine, dual PMAGs, and a Precision Airflow fuel servo.
About 8 hours into Phase 1 testing, my engine quit during stall testing (at the top of a stall at 48 KIAS, with partial flaps). Much to my surprise, it would not restart in the air despite playing with the boost pump, mixture settings, throttle, and starter, and pitching the plane down to increase the windmilling. I shut everything off and made a forced landing on the sandy bed of the dry Mojave River. Fortunately I landed into a good headwind, the sand was firmer than usual because of recent rains, and I held the stick full back, so the nosewheel didn't touch until I had come to an almost complete stop. The aircraft was undamaged. On the ground, I observed the right PMAG switch was off. I likely bumped it while entering a transponder setting during turbulence, although it is not impossible that I failed to turn it back on after turning it off during my runup test.
I had been on flight following and had reported the engine out to ATC. On the ground, I called the Victorville tower, and they sent the fire department out to meet me. I jogged about 3/4 mile to the nearest named road, met the fire department and sheriff, and then talked with the FAA. All was well, and they referred me to Jack Vanderman at Southwest Aircraft Recovery.
My family came up to meet me. The area was dubious, with burnt-out houses and ranches resembling junkyards. I found a sandy service road on the riverbank that reached within 150' of the airplane, and spent the night there in my truck to guard against looters. In the morning, I met Mr. Depue at Depue Airpark, a private dirt strip near Barstow, and got his permission to tow the plane over there. Jack Vanderman arrived with his flatbed trailer and aircraft crane, and we spent about 6 hours winching the airplane out of the riverbed, up the embankment, and onto the flatbed, then winching the trailer along the sandy road until it was firm enough to drive. He carefully drove 10 miles on Old Highway 66 to Depue, and the police car that passed us in the other direction kindly didn't bother our odd caravan. Jack's elderly father and my 14-year-old son both helped with the recovery and it was quite an adventure for four generations of aviators.
In hindsight, I would have done stall tests at 7000 AGL over Dagget airport rather than 3000 AGL over the desert. I would have looked for a dirt road rather than a riverbed. I would have aggressively tracked down engine issues on the ground before flying. I was thankful to have taken glider training last summer and to have memorized the emergency checklists and thought about them deeply. I was grateful for the help from the FAA, the Victorville Fire Department, Mr. Depue, and Jack Vanderman.
Two days later I went back to Depue, removed the cowling, did a lengthy inspection, found nothing unusual, put it back together, did a runup, and then departed the short dirt strip and flew back to my home base at Cable Airport. Flight was normal.
I had observed some issues during Phase 1 testing before the incident, and didn't troubleshoot them on the ground as much as I would have wished because I feared long ground runs while breaking in the new engine.
* The engine would sometimes quit during the rollout after landing when I was at low RPM. I attributed this (not quite correctly) to the idle being too low. Adjusting the fuel servo idle to 900 RPM prevented the problem, but caused a long landing, so I had set it back lower for further testing.
* The engine was also sometimes very difficult to start, even though I felt I was following the Superior and Precision instructions. Sometimes I would have to prime and crank up to 6 times, following the cool down instructions for the Skytek starter.
After recovering the airplane, I contacted Superior. Bill Ross, VP of Product Support, took decisive steps to help, at no cost to me, and gave instructions for testing before further flight. I wiggled the wires while monitoring the LEDs on the PMAGs to verify connectivity. The grounds both go separately and directly to the aircraft central ground stud on the firewall. The power lines go through test switches on the panel to a short shared wire going to the main power bus. The kill lines go through the key switch to ground. The PMAG LEDs turn on when the power line is at 12 volts and the kill and ground lines are properly grounded, showing connectivity. Aggressively wiggling either end of the wires on either PMAG wouldn't turn the LEDs off. I then removed the PMAGs and sent them back for bench testing. I removed the fuel servo, spider, and injectors and sent them back to Precision Airflow for bench testing. Everything came back looking normal. I reviewed the excellent Dynon engine logs and observed 0 RPM from both PMAGs after the engine failure, even though the prop was windmilling, which is an indication that the PMAGs were not powered up during the descent, despite have internal alternators that should provide backup power. I spoke with Brad at EMAG and Alan at Precision and both were very knowledgable and patient and helpful, but none of us could find a theory that explained all the facts. I was impressed by how all the vendors stood behind their products and took the time to walk me through everything even though I'm not a mechanic.
I reassembled the engine, changed the oil, replaced all the spark plugs, and started the engine again. Starting remained difficult at times. Superior confirmed from the engine logs that the engine break-in was satisfactory and that it was ok to do longer ground tests. I tuned the idle throttle and mixture settings using a screw and thumbwheel on the mixture servo, in consultation with Alan at Precision. I had a difficult time getting the idle right; if I turned it much below 800 RPM, the engine would quit at idle, while the manual called for 700-750 RPM idle. After a bunch of testing one hot afternoon, I finally found a setting at about 840 RPM where the engine would never quit on the ground at idle. Out of an abundance of caution, I waited until the cool morning to fly the aircraft. My son stood by as an observer with a fire extinguisher, and I started the engine again, but found the engine quitting when the throttle was pulled back to idle. I spoke with Alan again and he was puzzled and said there was no known temperature sensitivity for the idle RPM. My son adjusted the idle higher. Then I found the engine wouldn't start at all despited repeated priming and cranking.
Then my son told me he saw something. His statement was the clue that resolved the whole mystery. He and I made a small fix, and the aircraft has been starting, idling, and flying well for two weeks since then.
Happy puzzling,
David