2 issues: Could they be a grounding issue?

[blueflyer]

My #4 CHT has started intermittently telling me "shock cooling" during all phases of flight, to include taxi after first startup. In cruise, my #4 CHT is within a few degrees of the other CHT's and it does not fluctuate more than 5 degrees or so. I still get intermittent "shock cooling" indicator. I have the shock cooling warning system set at 50*/minute.

Also, my #1 EGT has started reading much lower than all the others. Previously, in cruise, all EGT's were within 20-30 degrees of one another. Now, #1 is always reading about 100* lower.

When leaning, I can sometimes get #1 to get to peak EGT at the same time as the others. But, it will still end up reading about 100* cooler than the others.

I noticed on the last flight, when I turned my battery on prior to starting, that the #1 EGT was jumping around from ambient to 1100*.

I have not touched anything under the cowl that would cause these issues. Could it be a poor ground causing both these issues?

 

[Dynon]

Not necessarily ground, but the first thing I'd probably check are your CHT and EGT crimp connections. EGTs, because of the environment that they're subjected to, do get consumed over time in some installations. So that one, more so than the CHT, is a bit more likely to be the sensor itself.

 

[vasper]

A possibly related question: I am just installing my EMS (Lycoming IO-360) using my old EGT and CHT sensors. I noticed that, when shrinking the heat shrink tubing over the connection between the probe cable and the EMS harness, the displayed temperature goes up by 80F or so. I read up on thermocouples and found that the junction between the probe wire and the harness wire is a "reference junction" of sorts and that the temperature of this junction is important.

So, is this a factor on our installations? Will I get different measurements if this junction is in the engine compartment, where it's hot, or in the cockpit, where it's cool? Is there any way to "calibrate" this effect? Or is the normal engine compartment not hot enough for this to be a factor (the heat gun is, admittedly, really hot!)?

Thanks!

 

[Dynon]

As long as there's no temperature change across that junction (which you probably created while heating it) you're OK. In other words, junctions across firewalls are bad; junctions in a place where both sides of the wires AND the connection are all the same temperature and you're OK.

 

[vasper]

Sorry, but I didn't quite follow that. Are you saying that the junction should be on the same side of the firewall as the probe or on the same side as the EMS? Right now, I have some of both; the left side cylinder junctions are on the cockpit side and the right side cylinder junctions are in the engine compartment. This is a Defiant so I have another engine to wire up; where should I put the junctions?
Thanks,

 

[dynonsupport]

In a thermocouple junction, you need to make sure all the parts of the junction are at the same temperature. This means a junction that is just sitting in air in the engine compartment or fuselage is fine. It doesn't matter if one junction is at one temp and another junction is at another temp. All that matters is each individual junction has the same temp all over.

When you heated the junction with a heat gun, you created a difference across it since you were likely heating one end, and the heat was flowing out the wire, creating a temperature gradient across the junction.

Thus, it's fine to have some junctions in the plane and some in the engine compartment.

The only thing that is really not OK is firewall connectors- they put the junction through something that clearly has a big temperature differential across it, which is an issue. If you do this, you need to use specifically designed connectors.

 

[vasper]

Wow, what an excellent reply; I totally get it. In the past, I did have some junctions go through a bulkhead connector in the firewall but now all of them go through a protected hole so all junctions are either wholly inside or outside.

Thanks so much for this reply.