The amazing cryogenic JPI EDM 930

[Katamarino]

I've got a weird one for the group. A few weeks ago my A&P and I completed installation of an EDM930 bought new from JPI. Initial power-up was normal - but after a few power cycles and leaving it on for longer, all temperatures start dropping from ambient rapidly and in unison. EGTs, CHTs, OAT, Oil T, Carb T. See the video.

JPI said that the temperature control circuit must have malfunctioned and shipped a new unit, which we installed. It worked completely normally for about 4 flying hours. Then, it started behaving like the old one did.

In flight, it is as if the unit is applying an ~360 degree negative offset to all temps. CHT and EGT read about 360 lower than normal, but with the same relative offsets between cylinders. OAT/Oil-T/Carb-T all go so low that the unit Xes them out, and EGT/CHT X out on the ground when the engine is off. Occasionally in flight the temps will all revert quickly to normal for a minute or less, and then drop again. See video 1, of this behaviour:

Once home, my A&P and I replaced the ground connection to a different part of the engine case. We confirmed good continuity of ground all the way from the pin on the connector at the unit, to the terminal on the engine, and then through to the engine itself (e.g. cylinder barrel). There was less than 0.5 ohm resistance across any of these points.

We then unplugged the EGT/CHT connector from the unit; the same behaviour continued on the other temps when powered back on.

We ran the unit with every other circuit breaker on the aircraft pulled; the behaviour was unchanged. See video:

We disconnected the OAT, Carb and Oil-T probes one at a time and powered on the unit. Temps were still displaying "off the scale" low. Visual inspection of all wiring in the harness that we could see revealed nothing abnormal.

We checked voltage differences between instrument case and panel ground, airframe ground, and engine ground when powered on. All were 0V.



Some further *really* odd behaviour; if the unit has been turned off for a while, then at power-on the temps read normal. After some time, they start to drop in synchrony. The longer that the unit has been turned off, roughly the longer it takes before the temps start to drop after power-on.

It almost feels like some kind of capacitance is building up a voltage offset that affects the temp unit, slowly dissipating when you turn the unit off. I have no idea if that's even possible or how it would occur, though.


JPI don't know what's going on, but are working on it. The first unit arrived back to them today and I'll be interested to see the bench test results. The avionics installer who sold me the unit is very experienced and has never seen anything like it. Does anyone have any idea what could be causing this? It seems unique.

 

[edessa]

How close were the serial numbers of the units JPI sent you?
I wonder if they are have an issue with a certain batch of 930's.
It would be nice to find a known good 930 and install it for troubleshooting in your plane and see what happens.

 

[Ed Haywood]

WAG: engine ground to frame and battery is being disrupted by vibration, force, or heat. Engine builds up a static charge, which skews ground to EDM. Every so often the static discharges, which allows temp sensors to work normally until static builds up again.

Just a guess.

 

[forsonsinc]

WAG: engine ground to frame and battery is being disrupted by vibration, force, or heat. Engine builds up a static charge, which skews ground to EDM. Every so often the static discharges, which allows temp sensors to work normally until static builds up again.

Just a guess.

Watch the video?

The suggestion does fit your choice of acronym

 

[Half Fast]

I’m not familiar with the unit. Are there separate connections for the power ground and the signal ground? If so, those grounds could be drifting apart, creating a potential between the engine case and the aircraft chassis. You might check the engine’s ground strap.

 

[Katamarino]

WAG: engine ground to frame and battery is being disrupted by vibration, force, or heat. Engine builds up a static charge, which skews ground to EDM. Every so often the static discharges, which allows temp sensors to work normally until static builds up again.

Just a guess.

I have been thinking about ground issues and charge build-up, too.

However, the first unit exhibited the behaviour without the engine ever running. The second unit, after the initial 4 hours, also shows the problem without the engine being started. With battery on, we did measure voltage difference between the instrument case and various ground points (engine, panel, airframe) and all were 0V - even while the unit was showing the odd behaviour. So if there's a ground issue it must be internal to the unit.

JPI are adopting the principle of "third time lucky" and sending another one, but doing a couple of days of serious testing of it first. They'll also turn down the returned unit and see what they can find. I asked them if there was anything at all we could have done in the install to cause this, and they maintain that there isn't. It's very odd.

 

[Katamarino]

How close were the serial numbers of the units JPI sent you?
I wonder if they are have an issue with a certain batch of 930's.
It would be nice to find a known good 930 and install it for troubleshooting in your plane and see what happens.

The serial numbers were 250 apart! We had the same thought.

My concern is that some peculiarity of our install is killing EDM temperature circuitry, although JPI is adamant that can't happen.

 

[Katamarino]

I’m not familiar with the unit. Are there separate connections for the power ground and the signal ground? If so, those grounds could be drifting apart, creating a potential between the engine case and the aircraft chassis. You might check the engine’s ground strap.

Only a single ground for the whole unit, which runs to the engine block.

MP, RPM, Oil-P, fuel flow, fuel levels all indicate normally and rock solid so the issue seems isolated to the temperature circuitry.

 

[Bell206]

so the issue seems isolated to the temperature circuitry.

Is the correct type of wire installed for the temp side?

 

[Katamarino]

Is the correct type of wire installed for the temp side?

Should be; the wires used came from JPI already connected to the plugs, as most of a harness. All we did was cut them to length at the sensor end and connect them in line with the instructions from JPI.

 

[Salty]

This is a fascinating and frustrating problem.

 

[Katamarino]

This is a fascinating and frustrating problem.

You said it..!

 

[Bell206]

Should be; the wires used came from JPI already connected to the plugs, as most of a harness. All we did was cut them to length at the sensor end and connect them in line with the instructions from JPI.

And per your video the temp decrease also happens on the ground with just the JPI powered up?
If so, would there be an easy way to remove the EGT, CHT, OAT, OilT, CarbT sending probes and isolate from the engine/airframe for a separate power cycle test?

 

[Katamarino]

And per your video the temp decrease also happens on the ground with just the JPI powered up?
If so, would there be an easy way to remove the EGT, CHT, OAT, OilT, CarbT sending probes and isolate from the engine/airframe for a separate power cycle test?

Correct.

We have tried the unit with the EGT & CHT harness disconnected and get the same issue; although haven't done that from a "cold start" so that might be worth a try.

The Oil T, Carb T and OAT are on the same plug as the power and ground so for them, we disconnected the probes one at a time to see if it made a difference (it did not). Again, it might be worth connecting just one at a time from a cold start and seeing what happens, I'll suggest it to my A&P.

 

[GeorgeC]

Is the correct type of wire installed for the temp side?

Low oil temperature during flight

Hello Everyone, For this thread is important to mention I live in Panama City (Central America) and the weather here is 80ºF to 95ºF all year round. I fly a 1980 C182Q with recently installed all-six brand new Superior Millenium cylinders in February this year. After the cylinder change, as...

www.pilotsofamerica.com

 

[Katamarino]

Low oil temperature during flight

Hello Everyone, For this thread is important to mention I live in Panama City (Central America) and the weather here is 80ºF to 95ºF all year round. I fly a 1980 C182Q with recently installed all-six brand new Superior Millenium cylinders in February this year. After the cylinder change, as...

www.pilotsofamerica.com

It's an interesting thread and I appreciate you pointing me to it. However, JPI provide the wiring harnesses with the unit, almost fully complete - we have not added any wiring ourselves. So I have trouble seeing how this could cause our issue, especially as unit two worked fine for 4 hours!

 

[Bell206]

EGT & CHT harness disconnected

disconnected the probes one at a time to see if it made a difference (it did not).

By disconnect you mean at the harness connector and not by removing the probes from the engine?

 

[Katamarino]

By disconnect you mean at the harness connector and not by removing the probes from the engine?

For the EGT and CHT, yes. We just removed the big plug connector from the back of the instrument.

For the single other probes, we disconnected the wires where the probes connect to the harness; they use ring terminals and little nut/bolt combos for ease of swapping probes.

 

[Bell206]

For the EGT and CHT, yes. We just removed the big plug connector from the back of the instrument.
For the single other probes, we disconnected the wires where the probes connect to the harness; they use ring terminals and little nut/bolt combos for ease of swapping probes.

So when disconnected, that specific input reading becomes all dashes but the other temp readings continue to decrease?
And when the individual readings go beyond limits you get a red X?
Do you find the problem with the 2nd unit is identical to the 1st unit problem? Or only similar?

 

[Katamarino]

Bell, that's correct. The disconnected probe goes to dashes, the others all still show the odd behaviour. Once the temp goes well beyond the normal range, the readings go to a red X.

The problem seems to be identical - if not a bit more severe on the second unit. The first one, the temps only dropped 50 - 100 degrees. This time around it's 300 - 400.

 

[Bell206]

the others all still show the odd behaviour.

So from a "cold start" all temp readings are normal, but after 4 hours of power applied to the JPI, all temp readings start to count down until they X out?
And then it repeats provided you let the system "cool" down enough?
Have you verfied all thermocouple wires follow the same and correct polarity?
Are any of the thermo wires bundled with existing wires or run separately?
Where is the ground wire for the 930 unit connected at?

 

[Katamarino]

Not quite.

The first unit exhibited this behaviour from the beginning, before we'd even run the engine. After a few minutes being powered on from cold, the temps all started to drop.

The replacement unit was fine for about 4 flying hours, and then started to do the same. Today we powered it on, with every other circuit breaker on the aircraft pulled. The JPI was the only thing powered. It read normally for 4 minutes, and then the temps all dove off the scale.

The thermocouple wires are all correct polarity, and read normally on power on. They are mostly run separately (although bundled with each other). On the engine they run alongside the ignition wires in a few places, but the first unit showed the behaviour before the engine was ever run.

The JPI is grounded to the engine case. Measurements with it powered on show no voltage difference between instrument case and engine.

 

[Bell206]

The JPI is grounded to the engine case. Measurements with it powered on show no voltage difference between instrument case and engine.

The only thing common to all the temp probes is the 930 and the aircraft electrical ground system. You changed the 930 once with the same problem. That leaves the ground system. I've seen thermocouple voltage issues where there was a "differential" in the electrical ground values that a VOM will not detect due to the millivolt levels thermocouples operate at.
Perhaps establish a new external ground path and see if that changes the results after 4 minutes?
And just to throw in there, it wouldn't be the first time a system was completely removed from an aircraft and reconnected on the bench to figure out a problem. However, I wouldn't recommend it at this point.

 

[Katamarino]

We could try running a ground directly to the battery negative terminal maybe (I'm not an electrical engineer so I don't know if this is a sensible idea)?

JPI's installation instructions do stress very clearly the need to ground to the engine block and not the panel or airframe, but I'm open to ideas.

 

[Ed Haywood]

From the JPI FAQ:

"Engine ground only – the same ground that the EGT probes are using – which is engine ground. Have them attach the ‘black’ ground wire off the P1 connector directly to the engine case. NOT panel ground, battery ground or airframe ground. These can vary from engine ground by 1/4v. This small voltage can cause issues such as all EGT or CHT readings seem High or Low, or Unsteady."

Gotta be your ground.

 

[Mikey52]

Or it could be in the wiring harness, or was that changed out too? Bad harness with two units showing similar results.

 

[Katamarino]

Or it could be in the wiring harness, or was that changed out too? Bad harness with two units showing similar results.

Indeed. JPI are sending a third unit, which they say they will stress test "extensively" first in the same configuration as my aircraft. They've also received the first unit back so I am eager to see the results of their inspection. I will discuss the harness option with them.

 

[Skyrys62]

Good to see how well they are supporting their product.
Not the case with some other companies.

 

[Katamarino]

Yeah - so far, I have been very happy with their response.

 

[Bell206]

JPI are sending a third unit, which they say they will stress test "extensively" first in the same configuration as my aircraft.

Interesting. And they have not required you check anything else with the installation prior to installing the 3rd unit?

 

[Half Fast]

From the JPI FAQ:

"Engine ground only – the same ground that the EGT probes are using – which is engine ground. Have them attach the ‘black’ ground wire off the P1 connector directly to the engine case. NOT panel ground, battery ground or airframe ground. These can vary from engine ground by 1/4v. This small voltage can cause issues such as all EGT or CHT readings seem High or Low, or Unsteady."

Gotta be your ground.

It does sound very much like a grounding issue.

@Katamarino , is there a way to connect your ground wire very close to one of the temperature probes and see whether that one particular probe then behaves normally?

 

[Katamarino]

Interesting. And they have not required you check anything else with the installation prior to installing the 3rd unit?

Correct. I specifically asked them if there was anything at all which we could have done on the install to cause this; they said they'd be asking the same thing, but that they could not think of anything.

 

[Katamarino]

It does sound very much like a grounding issue.

@Katamarino , is there a way to connect your ground wire very close to one of the temperature probes and see whether that one particular probe then behaves normally?

Agreed. It's just odd that every probe is affected the same, from the OAT in the wing to the various probes on the engine. It's as if the temp circuit in the EDM itself is building up a slight charge.

I will suggest that to my A&P. I'm sure we could attach the ground directly to a probe and try that.

 

[Half Fast]

Agreed. It's just odd that every probe is affected the same, from the OAT in the wing to the various probes on the engine.

I will suggest that to my A&P. I'm sure we could attach the ground directly to a probe and try that.

Probably find that there's a ground discontinuity caused by a crack in the engine block....

 

[Katamarino]

Probably find that there's a ground discontinuity caused by a crack in the engine block....

 

[Ed Haywood]

It's just odd that every probe is affected the same, from the OAT in the wing to the various probes on the engine.

Actually that's the most telling indicator that it is a grounding issue. Each probe generates a voltage. The gauge compares the voltage to ground to determine the temperature. If ground is off, all probes will be off. At least that's my theory.

 

[Half Fast]

Actually that's the most telling indicator that it is a grounding issue. Each probe generates a voltage. The gauge compares the voltage to ground to determine the temperature. If ground is off, all probes will be off. At least that's my theory.

It seems likely.

The whole method of grounding in our aircraft is pretty poor, pretty much the same as in an automobile. In all the avionics equipment I designed for the military (mostly night vision equipment), we never used chassis ground as a signal reference. In fact, most contracts had a prohibition against using the chassis as a path for current returns. Every device in a system had its own dedicated signal and signal return lines, as well as power and power return lines. Signal returns and power supply returns were tied together and connected to chassis at exactly one spot, called a single point ground.

It's no wonder our planes have so many problems with electrical interference.

 

[Albany Tom]

All the above makes sense. The observation I have is that the commonality between all those different temperature systems may be that they're maybe supposed to be isolated from the aircraft. Guessing on the setup, but the thermocouples operate at low voltages, so they may be run "floated" a bit off ground by the monitoring system, so that the differential voltage is easier to read and not affected by local ground issues in the aircraft. If that's the case, they *may* be all tied together on one side or the other, to make reading all the different temperatures simpler. They'd just have to read a bunch of single ended voltages. So all that said...if that's the setup, and something happens to connect just one of those sensor wires to anything, either hot or ground or something in between, it might be enough to pull the common reference voltage up or down and invalidate all the readings for every temperature.

That's a LOT of iff's and assumptions. If they're supposed to be isolated, then simplest way to check might be to check for resistance between the sensor wires and aircraft ground with a DMM and with them not connected to the panel. Or just connect one sensor at a time to the system and see if that lets you track down one cable/sensor that's affecting all of them.

 

[Katamarino]

Tom, interesting points.

We did measure continuity between the probe wires (the little tails attached to the probes) and ground on the OAT and the Carb T. Both had continuity to ground which we found a bit odd - but we're not experts on thermocouples.

The fact that they work normal some of the time suggested to us that this connection to ground must be normal, but I'll explore it with JPI next time I talk to them.

 

[mondtster]

Try methodically disconnecting a few thermocouples at a time and see if it makes a difference.

Did you or your mechanic have to do any harness building/pinning that could have gotten reversed? The JPI manual does a good job of not referencing positive or negative for the thermocouples but the red wire on traditional thermocouple wiring is the negative one and in some measurement systems when the positive and negative are reversed the indicated temperatures will go down instead of up.