Single, spurious CHT alarm cause?

[swatson999]

Here's a new one...I got a high CHT alarm for just a fraction of a second, once, on my most recent flight.  Thing is, the data shows rock-steady CHTs at that time (data at 4 Hz).  Alarm log is as follows (the time is actually 20:37:00 for all of these, but Excel is only displaying hh:mm, but the field actually is hh:mm:ss):

5/28/2018 20:37 34.78963 -120.39211 7822 170.3 146.8 -12.5 20:37:17 AUDIO PLAY EMS_CHT
5/28/2018 20:37 34.78963 -120.39211 7822 170.3 146.8 -12.5 20:37:17 AUDIO PLAY EMS_CHT
5/28/2018 20:37 34.78963 -120.39211 7822 170.3 146.8 -12.5 20:37:17 TEXT HIDE CHT 3 HIGH
5/28/2018 20:37 34.78963 -120.39211 7822 170.3 146.8 -12.5 20:37:17 AUDIO STOP EMS_CHT
5/28/2018 20:37 34.78963 -120.39211 7822 170.3 146.8 -12.5 20:37:17 AUDIO STOP EMS_CLNT_AS_CHT

Any idea what could have caused  this spurious alert?  Again, the data log shows absolutely no change in any CHTs around that time.

 

[Dynon]

I don't have an obvious answer. I suppose it could've happened somewhere in-between the interval, but 4hz is a pretty good rate to catch everything. Is there anything at all in the fuller data log at the same time that looks amiss?

 

[swatson999]

I don't have an obvious answer. I suppose it could've happened somewhere in-between the interval, but 4hz is a pretty good rate to catch everything. Is there anything at all in the fuller data log at the same time that looks amiss?

Not a thing...rock steady CHT values all around that time.

The only thing even remotely unusual was that I had the airport Info page open at the time of alarm, although I doubt that had anything to do with it.

 

[preid]

Funny, I thought the random CHT alerts were from my Vertical Power connection to the Skyview NMEA port, but seeing someone else has this random issue, make that 2 of us now. I cant duplicate it, it just randomly says "check CHT" when they are in the lower levels and no where near a warning zone.


Probably a timing issue, but started when I upgraded to touch screen.

For reference I'm just flying with nothing (extra) on the screen when the alarm(s) go off. As Steve mentions, very quick alarm. I happen to have a VP200 that keeps a red line for the CHT, otherwise I wouldnt know which one it is. Next time I'll look and see the # and if its the same one each time or different.

 

[Raymo]

I have gotten the same CHT warning due to a poor connection on one of my cylinders that always goes away after a few minutes (usually). I disabled the audio on that one to avoid hearing the noise until I find time to replace the blade connectors with something more reliable.

 

[swatson999]

I have gotten the same CHT warning due to a poor connection on one of my cylinders that always goes away after a few minutes (usually). I disabled the audio on that one to avoid hearing the noise until I find time to replace the blade connectors with something more reliable.

I think that's different...I've had EGTs fail and jump all over, and IIRC, I had a CHT fail as you describe (replaced several hundred hours ago), but this was just a single, instantaneous "blip" which tripped an alarm. As reidvaitor says, I suspect a data error internally (timing, a dropped bit, something like that) more than any actual problem with the sensor itself. Which is why I brought it to Dynon's attention, in case they have other reports of something similar (may indicate a latent hardware timing issue, like a cycle slip or something).

 

[Dynon]

So a few thoughts here. Other than the reports here, we don't hear of errant alerts of this variety. So I don't think we have a systemic software bug here. All of that said, after talking to our engineering team, I'd suggest upping the data recording rate to 16hz to see if you can catch anything in the data at the highest rate available. All of that said, if even that doesn't shed some light on it, it might make sense to get our support team a diagnostic file within an hour of the event so that we can help you uncover the source of the issue.

 

[swatson999]

So a few thoughts here. Other than the reports here, we don't hear of errant alerts of this variety. So I don't think we have a systemic software bug here. All of that said, after talking to our engineering team, I'd suggest upping the data recording rate to 16hz to see if you can catch anything in the data at the highest rate available. All of that said, if even that doesn't shed some light on it, it might make sense to get our support team a diagnostic file within an hour of the event so that we can help you uncover the source of the issue.

I will do all of that...may be a while, travel overseas coming up, but will get to it once I return!

Thanks for the help...and with luck, it'll never happen again (in which case we'll chalk it up to a GCR ).