EDM 830 OAT error

[squeaky.stow]

My OAT readout on my EDM 830 has been unreliable since I bought my 252 last year, but I am now getting enough long flights to get a handle on what it is doing.

OAT always matches the analog gauge and the Aspen OAT values at startup, but once level in cruise it starts climbing. I typically see temperatures indicating 30C to 40C higher than the real OAT by the time it peaks about an hour into the flight.

Now here is the strange part. As the OAT on the EDM climbs, my oil temp and cylinder head temps on the EDM climb with it. Meanwhile my analog OAT, cylinder head temp and oil temp gauges stay rock steady. On my last flight, over four hours at FL180, the real OAT was steady at -5C as predicted by FDs and verified by the Aspen OAT. During this time the analog CHT was at the bottom of the green and the oil temp was in the lower half of the gauge. An hour into cruise, my OAT on the 830 started climbing and the CHTs and Oil Temp on the 830 started climbing in lockstep. OAT rose to about 43C and my engine temps rose by about the same amount,( or would have, if I hadn't opened the cowl flaps. I am doubting whether I am getting accurate cylinder head and oil temperatures from my EDM but I don't want to ignore them so I find myself gradually opening more cowl flap as the flight goes on, trying to keep my cylinders under 380. All through this, the analog gauges never budged.

Sometimes the OAT will drop back down by 20-30C for 10-15 minutes and as it does, my CHTs and OT drop as well. None of this is reflected in the analog gauges.

I downloaded some good examples of this behaviour from the 830 and sent it to EDM tech support to ask what was going on. When you graph it, you can clearly see that when the OAT starts going wonky, the CHTs and OT go right along with it at exactly the same time. I got a rather dismissive answer that my OAT probe was too close to the engine exhaust (which would not explain the CHT and OT changes) so I provided a picture of my probe location. It is in the NACA vent inlet for the cabin air, just ahead of the door. When I open the vent, that inlet air is definitely not even remotely warm at FL180, so I find it hard to believe that is the problem. They promised to ask their engineering folks to review my data but I have heard nothing in two months.

So here are my questions. For those of you with EDM 830s, have you ever seen this? Any ideas? Where do you have your OAT probe mounted? Is there any simple way to validate the accuracy of the CHT and OT probes? I know the OAT is out of whack, but why does it always start out exactly the same as the other OAT sources only to go crazy in cruise? Is it normal for CHTs to fluctuate in level cruise and steady power setting when the OAT is constant?

Puzzled and frustrated.

Mark

Edited July 12, 2018 by squeaky.stow
Typo

[Marauder]

My OAT readout on my EDM 830 has been unreliable since I bought my 252 last year, but I am now getting enough long flights to get a handle on what it is doing.
OAT always matches the analog gauge and the Aspen OAT values at startup, but once level in cruise it starts climbing. I typically see temperatures indicating 30C to 40C higher than the real OAT by the time it peaks about an hour into the flight.
Now here is the strange part. As the OAT on the EDM climbs, my oil temp and cylinder head temps on the EDM climb with it. Meanwhile my analog OAT, cylinder head temp and oil temp gauges stay rock steady. On my last flight, over four hours at FL180, the real OAT was steady at -5C as predicted by FDs and verified by the Aspen OAT. During this time the analog CHT was at the bottom of the green and the oil temp was in the lower half of the gauge. An hour into cruise, my OAT on the 830 started climbing and the CHTs and Oil Temp on the 830 started climbing in lockstep. OAT rose to about 43C and my engine temps rose by about the same amount,( or would have, if I hadn't opened the cowl flaps. I am doubting whether I am getting accurate cylinder head and oil temperatures from my EDM but I don't want to ignore them so I find myself gradually opening more cowl flap as the flight goes on, trying to keep my cylinders under 380. All through this, the analog gauges never budged.
Sometimes the OAT will drop back down by 20-30C for 10-15 minutes and as it does, my CHTs and OT drop as well. None of this is reflected in the analog gauges.
I downloaded some good examples of this behaviour from the 830 and sent it to EDM tech support to ask what was going on. When you graph it, you can clearly see that when the OAT starts going wonky, the CHTs and OT go right along with it at exactly the same time. I got a rather dismissive answer that my OAT probe was too close to the engine exhaust (which would not explain the CHT and OT changes) so I provided a picture of my probe location. It is in the NACA vent inlet for the cabin air, just ahead of the door. When I open the vent, that inlet air is definitely not even remotely warm at FL180, so I find it hard to believe that is the problem. They promised to ask their engineering folks to review my data but I have heard nothing in two months.
So here are my questions. For those of you with EDM 830s, have you ever seen this? Any ideas? Where do you have your OAT probe mounted? Is there any simple way to validate the accuracy of the CHT and OT probes? I know the OAT is out of whack, but why does it always start out exactly the same as the other OAT sources only to go crazy in cruise? Is it normal for CHTs to fluctuate in level cruise and steady power setting when the OAT is constant?
Puzzled and frustrated.
Mark

Mark - my JPI 900 OAT probe is mounted on the lower side of the left wing. I have not see any temp fluctuations.

Based on what you are seeing, it would be a weird combination of thermocouple failures. I would just keep the pressure on JPI technical support. It may be an issue with a temp logic board in the unit.

I was having an interference issue with the JPI opening the squelch on my brand new GTN. Eventually they sent me a ferrite filter to fix the issue.


Sent from my iPad using Tapatalk Pro

[Cruiser]

If possible disconnect the OAT probe on the JPI and see what the CHT and OT do. 

[Stephen]

EDM 900 feedback YMMV

 

1. Check to make sure the grounds are good on the unit; I think the 900 needs to be grounded to the engine block (if memory serves might be wrong) and make sure the bonding strap from engine to firewall is well connected at both ends.

2. If stilL doesnt work, Then  try switching the oat sensor Wires around.  The initial installer had the wires crossed to the wrong pins on my EDM 900 oat thermocouple and it did almostexactly what you describe. They tried to cross the wires at the thermocouple to undue this but I think the wires may have respectively different resistance properties or something. The OAT wires need to be correct at both ends. 

 

[carusoam]

I’m with Marauder on this one...

I didn’t see the OP mention where the OAT sensor is mounted...

So many are eagerly placed out of the windstream, near the instrument panel for short wire runs....

Expect warm air from the engine is being sensed...

If the sensor is a third the way out under the wing, out of the sun, out of the warm air...

Then things just got harder to identify.

Best regards,

-a-

[Stephen]

On 7/11/2018 at 10:00 PM, Marauder said:

It may be an issue with a temp logic board in the unit.

Agree with suggestion to engage their support. Due to multiple temps being effected, seems like a system wide issue. That is why I was thinking ground vs specific probe location.  I think the logic board is another good candidate because that could effect multiple readings. I suppose the 830 could use OAT for some calculation reference WRT the other items but that seems kind of odd vs directly sensing each one. What happened with mine (jr) installer got the wrong probe wires pinned to the wron terminations on the pins that click into the plug connector on the back of the unit. Once you click the barbed pins in they are hard to remove and he didn't have more pins.   I had no idea about this but know I'm having OAT sensing issues  .... so when I am looking under the cowl I see the colors on the sensor end are switched around and I say ...”. OH here is why it is not working”  Asked installer he told me the pinout story related above and added ..... so our guy got the pins crissed up..... blah blah...didn't have more pins  yada yada ........so we corrected it at the other end....yada.  He  is like ..... I'm like   [Chris @Marauder seems to have some more colorful, "off the official list" emoticons that involve sign language, but I'll go with this for now ]. So,  I called the JPI support   and they say my installers are  and that the wires have different properties (I assume resistance) and must be straightened out on both ends .  We did and it stabilized the readings. My sensor is in my fresh air inlet on the cowl so my OAT seeing is +7 °C vs windshield gauge in flight but matches it on the ground pre-start (need to move it to wing engine heat doesn't effect). Before the wire pin-out was corrected, my symptoms were OAT near match on the ground, as soon as I start the unit it starts climbing  until it was indicating crazy OAT's like 60 °C during cruise at altitude. *My other temps were not affected*, just OAT. That is a difference, but perhaps the above info  helps. 

[squeaky.stow]

Thanks so much all for some great feedback. 

I will try them in order of easiest to hardest.

1) Check the ground

2) Disconnect OAT and see what happens. This one may cause issues with the %HP display, but I already know that is wrong due to the high OAT readout. Worth a try. Don't know why that one never occurred to me!

3) Check the wiring pins are not reversed. That one is high on the suspect list for me after hearing that Stephen had similar symptoms.

4) Move the probe to under the wing. 3&4 will have to wait until the annual.

5) Talk to JPI about logic board.

On a more general "engine management" topic, what kind of variations in CHT should I normally expect to see if I am stabilized in level cruise with no changes in power settings or OAT? I am seeing CHT fluctuate between 340 and 380 over a period of 10-15 minutes, which seems like a lot to me if everything else is staying the same. 

Cheers,

Mark

[Aerodon]

14 hours ago, squeaky.stow said:

Now here is the strange part. As the OAT on the EDM climbs, my oil temp and cylinder head temps on the EDM climb with it. Meanwhile my analog OAT, cylinder head temp and oil temp gauges stay rock steady.

Mark,

If you have a 'bank' of temperatures that drift in unison, I suspect an instrument problem.  Can you send me a data file please?  We should try verify the problem, then just send it in for repair.

Aerodon supercub180@gmail.com

 

 

 

 

[N231BN]

The probe location in the fresh air vent is not a problem. If you are getting engine heat there, you have bigger issues than the OAT probe. Under the wing is better but not necessary.

[carusoam]

One problem...

The thread title points towards the OAT being incorrect...

The real issue seems to be more the whole JPI isn’t collecting data properly?

The analog gauges on the panel are reading steady as primary gauges are expected to do...

The JPI is giving inconsistent and drifting data that doesn’t match the ship’s gauges...

Squeaky,

1) Are you able to review the installation to see how everything is wired?  Sounds like something basic could be off

2) can you send the data file that Aerodon requested?

3) just know JPI customer service is a bit quirky...  If you haven’t heard from them in a while it is time to call them back again...

4) Do you have access to the shop that installed the JPI?

5) Do you have access to the prior owner? He may have some knowledge about the system... the shop... or something...

6) temperature signals are tiny. Measured in microvolts... a bad ground wire can cause enough noise to make challenging readings...

7) Check the ship’s grounding straps from the engine to airframe and instrument panel...

8) Are any of the sensors being shared? (Is this even possible?)

9)  A picture is worth a... If able... post some graphs of your JPI’s data... many people do this... it would be really cool to see the the voltage graph... if it is constant or varying with the other instrument readings....   if you look at other people's graphs of CHT and EGT they vary sinusoidally around a point...  the digital data from the JPI makes it obvious...  the ship’s analog gauges don’t have enough sensitivity to have it even show...

People have given feedback on

• crummy instalation challenges, which are common...
• JPI hardware challenges, JPI has a few in the record books...
• Aerodon sells JPI equipment and can be quite helpful...

PP summary of what I read around here... I’m Not an instrument tech...

Best regards,

-a-

[squeaky.stow]

Thanks Aerodon, I will download some flights and email the data, but it might take a few days.  I just did another 4.3 at 17000 today. This time the OAT stayed around 24C (still way too high, but not going crazy) for the first 2 hours and the CHTs stayed under 380 with cowl flaps closed. Then in the last 2 hours, OAT started climbing into the 30s and away go the CHTs and OT. The problem is that the graphs will not reflect the real CHT rise because I just can't stand to sit there and ignore it, so I start opening the cowl flaps to bring it back down. Until I know for sure that it is a false temperature reading I just don't want to risk my expensive engine.

I guess what I need to do is fly an entire flight with cowl flaps full open so that the graph will show the actual change.

More to follow.

Mark

[squeaky.stow]

Curiouser and curiouser! I sat in the hangar tonight and ran the EDM for an hour with the engine off. Starting OAT in my hangar was 27C/80F, as were the CHTs, EGTs, and OT  

By the end of the hour, OAT, CHTs, EGTs and OT had all risen to 40C/100F. 

I have emailed the attached photos to JPI tech support along with data from my last few flights. 

There is clearly something wrong with this monitor. The good news is that my engine is obviously running a lot cooler than I thought, but I need a monitor I can trust!

Aerodon, what is the best way to send you my downloaded data?

Mark

 

[squeaky.stow]

Forgive my clumsy uploading. Sideways and in reverse order!

[carusoam]

Well you nicely described a system that isn’t working properly...

Meaning any data it is getting, is suspect of being altered...

Now the question is what is altering all the data with time...

Is it a change of voltage...

Is it a change of the ground... something dirty, loose, causing a change in resistance....

Is it a change of something grounding the instrument panel with the engine.... (big probability)

Is the device itself got some unusual sensitivity it shouldn’t have...

Expect JPI to have seen every possible problem they have ever had...

Should take them five minutes to explain what could possibly cause this odd behavior with the data you have supplied.

PP thoughts only, not an instrument tech...

Best regards,

-a-

[carusoam]

What else was on when you were running the study?

The battery drained from 24v down to 23v during the hour...?

You might list everything that was on when discussing this challenge... in the event there is something else that is interfering with the proper operation...

nothing else but the temps are getting altered... the MP, RPM, and FF are staying constant...

CHT, EGT, and OAT all drift consistently higher... really consistently!

Do you know what type of sensor you have for OilT? The ship uses a thermistor that changes resistance in relation to temp. If you have a second OilT sensor, it may be a thermocouple that is drifting the same way as all the other thermocouples....

Nice pic of the thermometer... but to be useful we would want to see one before and after the test.  

Looks like something is warming up as the hour progresses... whatever it is... is probably related to the change...?

Feel like running a temporary ground wire as an experiment...? By grounding the engine to the Instrument panel you might see things settle, quickly...  (just a PP guess...)

 

Don’t forget to plug the battery back in for a charge...

Best regards,

-a-