Carbon Cub EX-3 Engine Failure (N40DT)

[Neal]

The NTSB preliminary report is attached and another article linked below. For aircraft owners with Light Speed ignition systems (or any electrically driven ignition system) I think this is an important one to learn from and discuss.

Excerpt:

The airplane departed from the Thompsonville Airport (7Y2), Thompsonville, Michigan, about 0917. Shortly into the local area flight, the pilot noticed a flicker on the Garmin G3X Touch display, but a low voltage warning never appeared. The pilot then looked at the engine monitoring display which indicated the alternator was not charging and it showed 12.8 volts direct current (VDC). The pilot immediately started a climbing turn back to 7Y2. After reaching about 4,000 ft, he tried several methods to get the alternator working including cycling both the field and main alternator circuit breakers several times. He checked the voltage in the backup ignition battery which read 12.8 VDC.

Since he had about a 15-minute flight back to 7Y2, he decided to save the backup ignition battery and run off the main battery until he received a low voltage warning. The airplane was on final approach when a low voltage warning appeared on the Garmin G3X Touch display. The pilot switched over to the backup ignition battery, he watched as the voltage dropped from 12.8 VDC down to 0 VDC, and the engine sustained a total loss of engine power. The pilot noticed that the backup battery was faulty, and it only lasted about one or two minutes.

One question I have is "how old was the ignition backup battery" as this was a kit built Carbon Cub and did this battery come with the kit at time of order or was it purchased closer to date of completion as this is an item we're supposed to replace annually at condition inspections. I don't know if CubCrafters ships batteries with the kit.

I'll also add that I don't personally trust the ignition backup battery and there is a test that can be done and is supposed to be done to load test the battery which I bet 99% of us don't do. During runup this battery is tested (backup ignition switch) and it runs down fast. I hope I never have to rely on it.

My understanding is when the ignition backup battery fails it will fall through to the starter battery. I welcome any CubCrafters' owners to chime in to help all better understand the system. With that said, it seems to be your best starter battery is insurance and that's why I have the EarthX in mine. The "upgraded" (now standard) SBS-J16 has always been weak in my experience. Mike Patey did a video on how turbine failure in how the EarthX saved him and FlightChops also did a video along the same line.

So let's discuss. First off, the sad situation of a ring terminal failing. The cheapest item in the entire aircraft cost the aircraft. There are a lot of spades and terminals like this in the aircraft and from what I hear it's the most common cause of failures. I had one separate in my coil pack saga, a slight tug test can ensure they are secure. Is there such a thing as marine grade terminals and if so maybe they should be used instead instead of these cheap connectors we see as commonly used in the build. It may be wise not only to inspect the ring terminal that failed in this accident but consider replacing it with a marine grade terminal. I always use a flashlight during my preflight to look into the side vents to inspect coil pack connections. I'll now focus also on this terminal going forward.

Did the pilot do wrong by waiting for the starter battery to draw down first before going to ignition backup (ign backup battery)? When is the right time to engage the ignition backup battery? It seems the pilot's plan here may be a good one to not simultaneously draw down both batteries? Something I'd like to better understand.

In addition, I recently did a review and adjustment to the items that appear in my EIS sidebar when I realized I could add an additional one as recently posted here which added AMPS as it's a critical item to monitor in my opinion. I had my "field" breaker pop once and amps went negative and I caught it right away. I believe this was caused by plugging a USB accessory (GoPro) into my GMA 245R over the rear seat USB port.

There has been some discussion on voltage cutoff settings programmed into the G3X I believe, maybe @Cactus Charlie can help us better understand as I think this comes up when EarthX's are installed and how this is affected.

Unfortunate incident especially for a kit builder that spent years building their dream shattered by a $.25 part.

Accident Cub Crafters CCX-2000 Carbon Cub N40DT, Sunday 7 September 2025

On September 7, 2025, about 0932 eastern daylight time, a CubCrafters Carbon Cub EX airplane, N40DT, sustained substantial damage when it was involved in an accident near Thompsonville, M...

asn.flightsafety.org

 

[Cactus Charlie]

"My understanding is when the ignition backup battery fails it will fall through to the starter battery. "

There is no automatic switching of right ignition power source in an FX-3 or a standard EX-3. Right ignition is powered from MAIN bus with the emergency ignition switch "down" and by the 2 AH emergency ignition battery when the switch is "up".

I run a 10 second test of the emergency ignition battery at every run-up. I watch the rate of voltage decay and I periodically plot the data. Over 5 years of FX-3 ownership I have collected a lot of ignition battery data. I have concluded that the emergency ignition battery is unfit for purpose (crap in plain English).

Even if replaced every year the new one may not last until the next change. These batteries are typically rated for 5 years in float charge standby applications but the longest I have had one last is about 18 months.

It is essential for anyone operating an FX-3 to understand their electrical system and have a plan for load shedding and ignition management if the alternator fails. That plan will be different depending on whether the main battery is AGM or EarthX. (EarthX hard cutoff is very different from AGM voltage decay.)

I started a thread on this subject on the CubCrafters forum. It was very disturbing to see the lack of system understanding in some of the follow up discussion.

 

[Neal]

Thanks @Cactus Charlie and that's why I posted this to help better understand the system and develop a load shed plan. I would love to hear your load shed plan and how you plan to handle a scenario like this. I'm curious what the biggest loads are to be aware of.

I agree the ignition backup battery is more of a placeholder than a functional backup. The question lies is there a better option? It seems to just be a 12V power source. I wonder if there is a LiFePO4 alternative now in 2025?

 

[Cactus Charlie]

From the accident report -

"He checked the voltage in the backup ignition battery which read 12.8 VDC"

The backup ignition voltage scale reads MAIN BUS voltage when the emergency ignition is not active. The only way to read the battery voltage is to switch to the battery! That is why monitoring ignition battery voltage during the emergency ignition test is essential.

If the report says what main battery was in use I must have missed it. For the standard SBS-J16 AGM battery - shed non essential loads, force reversion to IBBS, run ignition on main battery until engine runs rough or main voltage drops to about 6.5 V.

What I don't yet understand is why the engine stopped. If ignition battery was bad and zero volts the engine should still run on the left ignition until main battery was exhausted. Was this an EarthX battery and engine stopped when the battery cut off?

I would very much like to see the SD card log for this flight.

Revised Edit - "The backup ignition voltage scale reads MAIN BUS voltage when the emergency ignition is not active. " MAIN BUS is connected through a diode to the ignition battery. The voltage monitor point is the junction of the diode and the battery positive terminal. With the ignition switch in the normal (down) position the indicated voltage will be MAIN bus voltage less the diode drop. This voltage does not provide any indication of the state of battery charge unless MAIN bus volts is lower than the battery terminal voltage. When the switch is moved to the emergency (up) position MAIN bus is disconnected and no longer influences the voltage reading. Indicated voltage is now battery voltage and does provide some indication of the battery charge state.

 

[Cactus Charlie]

EarthX now has a "backup" battery that is not rated for cranking. I had considered fitting this but the voltage drop in the blocking diode would likely prevent the battery reaching full charge.

I plan to fit a right ignition IBBS this winter. I have bench tested my design and am now working on the implementation details.

Fitting the EarthX backup battery would be much easier to do as it would require no wiring changes.

 

[Cactus Charlie]

Link to the new EarthX backup battery -

ETX104 Back Up Battery

The ETX104 back up battery provides power for critical loads with current limiting built in to protect from excessive discharge/charge current.

earthxbatteries.com

 

[Neal]

Link to the new EarthX backup battery -

ETX104 Back Up Battery

The ETX104 back up battery provides power for critical loads with current limiting built in to protect from excessive discharge/charge current.

earthxbatteries.com

I like it. Would be worth experimenting with.

 

[Neal]

Okay, I think we covered the insufficient and untrusted (by me) ignition backup battery. To the root cause of this accident. A broken ring terminal. All of the terminals/spades used to me appear to be "Radio Shack" grade (Amazon) items. Should "marine grade" terminals be used, assuming that's a thing? The point is something that can withstand vibration, corrosion, etc. as best as possible. I'm not about to go changing anything on my plane beyond visual inspection but I just wonder if the quality of these terminals could use a better choice by the factory, builder, etc.?

 

[Cactus Charlie]

I don't doubt there are different quality terminals but I have seen nothing to suggest that the ring terminals used by CubCrafters are substandard.

These terminals are usually made of copper and any copper will work harden and fracture if flexed enough times. I have no knowledge of how this terminal was installed, how it was protected from vibration, or why it fractured.

AC 43.13-1B says - "d. Copper Terminal Lugs. Solderless crimp style, copper wire, terminal lugs should be used and conform to MIL-T-7928. Spacers or washers should not be used between the tongues of terminal lugs." I found no reference to "Marine grade".

I found no reference to MIL-T-7928 in the Aircraft Spruce listings of ring terminal but I would expect them to be suitable for use on aircraft.

If I had been involved in the investigation of this accident I would have concentrated on the condition of the main and emergency ignition batteries and would have run a charge/discharge test on both of them.

 

[Tsquared]

Interesting discussion.

I appreciate you guys documenting this and developing potential solutions, but for those of us who lack an Electrical Engineering degree, is there a simple way I can check to see how long my ignition backup battery will last? Keep in mind my electrical knowledge starts and stop at plugging in things to a wall outlet.

For example, if I do an engine run at 1900rpm, on the right mag, with the right ignition battery set to emergency and run it for some set amount of time (5min, 10 min or whatever). Will that be a valid test of the emergency ignition battery and will it provide me with any useful data that could be useful in an emergency?

 

[Cactus Charlie]

For example, if I do an engine run at 1900rpm, on the right mag, with the right ignition battery set to emergency and run it for some set amount of time (5min, 10 min or whatever). Will that be a valid test of the emergency ignition battery and will it provide me with any useful data that could be useful in an emergency?

I have tried two methods to test my emergency ignition battery:

1. In flight test running at about 2100 rpm (my typical cruise) right ignition to emergency, left ignition normal. Monitor ignition voltage decay for 30 minutes. Return right ignition to normal. Take the SD card home and plot the voltage curve.

2. Observe ignition battery voltage during a 10 second emergency ignition test at every run-up. Periodically plot the voltage to see trend.

I only did method 1 once because I found the ignition battery charge current when the test was terminated exceeded the recommended maximum charge rate for the battery. I found that method 2 gave a good indication of battery health and did not abuse the battery.

The best way to test the battery is to remove it from the aircraft and run a loaded discharge test (at least 1 A discharge current) for at least 30 minutes. Record and plot the voltage. Then charge the battery and put it back in the aircraft. No one is going to do that as it's easier to replace the battery once a year. I have run this test on batteries I have taken out of service.

The problem with changing the battery once a year is that there is no guarantee that the new battery you fitted is any better than the one you removed. Fitting a new battery does not mean the run-up discharge test (test 2 above) is not needed.

You must shed non essential loads and carefully monitor MAIN bus voltage after alternator failure.

An EarthX battery will self isolate (cut off) at about 11 V and the engine will stop unless the emergency ignition battery is selected. An AGM battery such as the SBS-J16 should keep the engine running until MAIN volts drops below 7 V.

Edit - The fact that an AGM battery will keep the engine running down to 7 V does not imply that it will keep the engine running longer than an EarthX battery that cuts off at 11 V. The rate of change of voltage, and the shape of the voltage curve, is very different for these different battery chemistries.

 

[Neal]

An EarthX battery will self isolate (cut off) at about 11 V and the engine will stop unless the emergency ignition battery is selected. An AGM battery such as the SBS-J16 should keep the engine running until MAIN volts drops below 7 V.

I think LiFePO4's shutdown at 10.0 volts, not 11, but depend on the BMS programming. The voltages can't be compared but instead "state of charge" is the better number. LiFePO4 can use 90% of the available capacity whereas AGM's typically are damaged below 50% which can be around 11.9-12.1V depending on battery.

 

[Cactus Charlie]

I want to know at what MAIN bus voltage my engine will stop. That information helps me determine where I set the amber and red CAS alerts for the MAIN Volts scale.

I may estimate how long that will take based on rate of voltage decay and knowledge of the battery characteristics but I can't program that information into the alerting system.

The EarthX Lithium Battery User’s Manual does not specify the self isolation voltage but the discharge curves for the ETX680C all terminate at approx 11 Volts.

Earth X also says "EarthX ETX-Series batteries have over discharge protection which will disconnect the battery from the active load (drainage) when approximate 95-98% which is typically around *10.5-11V (double for a 24V battery) to protect the cells from damage. At this point, there are only seconds left of runtime left, which is equivalent to running out of fuel."

and "An over-discharged battery typically has an internal voltage less than 11.0V, but when the BMS disconnects, the voltage reading at the terminals of the battery will be zero volts."

I have chosen to use 11 V as the self isolation voltage.

 

[Neal]

I'm curious why CubCrafters, when they went to glass panels (G3X) they did not add a shunt for alternator current. Apparently this is common in certified aircraft and glass panels. I wonder if this is/was just an oversight? Some added one and I looked at this but decided it's beyond my desire to mess with. But the ability to get a CAS alert of an alternator failure seems like a critical alert we should have in glass panel aircraft.

Current shunt - CubCrafters Forums

CubCrafters Builder Assist program aircraft

forum.cubcrafters.com

 

[Cactus Charlie]

You don't need a shunt to add alternator current monitoring. Simply disconnect the alternator wire at the circuit breaker, slip a Hall effect sensor over the wire, re-connect the wire to the circuit breaker. That part is a 5 minute job.

What takes the time is fitting the sensor wires to the GEA 24 connector and configuring the custom GP input definition.

Some would argue that Battery Amps tells you if the alternator is working. Battery Amps will go negative immediately if the alternator fails. I prefer to have battery and alternator current displayed and recorded.

I used the Amploc Key100 sensor for alternator current.