Strange indications? Read here first for some tips.

[Anticept]

I know this post is a little long. The first part will help you understand the thought process of troubleshooting, the second part is actually the tips as a summary of bullet items.

 

Preamble:

 

So a lot of you see me posting around here about some of the electrical systems in our CTs. I'm not an expert like a couple guys here on electronics (Chanik), but I do know my way around quite a bit and just need more time than those experts to figure things out. I'm not a fan of how Flight Design and Rotax works with their electrical and electronic systems. A few things are engineering issues, while others are quality control issues. For the most part, everything works though.

 

Except when it doesn't.

 

And, because of some of these engineering and QC issues, and because Rotax and FD like to hold everything real close to the chest, it can be a real pain in the ass to diagnose sometimes. Dynon Avionics isn't off the hook either due to their heavy use of resistor jumping (not as bad in skyview) on the older panels, but at least they provide a lot of information and diagrams in their installation manuals, which are VERY useful.

 

Basically, I've been having an odd CHT split on one airplane, but it was never great enough to cause me to tear everything apart. It's just annoying. I now know why. Read on.

 

The Story:

 

Over the past couple days, I had popped in an older style oil pressure sensor that I had sitting around on an engine that had the honeywell sensor which went bad. Whenever you change from one P/N to another, you must carefully read your indicator's manual, especially with glass panels! In this case, because the honeywell and newer sensors work differently from the old VDO style, there has to be a resistor jumper installed from ground to the oil pressure input pin when using the new sensors in accordance with the installation manual for the D-120 EMS. Well, guess what? There was no resistor, it was just jumpered with a wire! I had been told that the oil pressure has always read close to the upper arc of the green zone, and this explains why. The customer never let me get into it though before because $$$ :-).

 

In addition, I was having a very high right CHT reading after switching the oil pressure sensor. Took the voltmeter, and tested the voltage drop across the sensor, and the resistance of the sensor (I like taking two related measurements, it helps to verify that my meter is working correctly). Voltage matched the other sensor, resistance did too, and so did the amperage. Even pulled out another sensor to test, and it all matches. Well crap, might be a bad EMS, but wait! Again, READ THE MANUAL. The D-120 EMS Install Manual stated that Rotax CHT sensors must have a 1.21kΩ resistor jumpered from the +5V excitation pin. Aha! So they are shunting! This means if the resistance is off by more than a few percent, it will affect CHT sensor readings! So I pulled the DB37 connector, put a couple paperclips (very useful tools!) into the pins, and clipped my voltmeter on in resistance mode. Now, mind you, I had a newbie mistake and set the scale wrong (was still set for testing the CHTs, which are much higher resistance), which caused me to dig deep into the electrical system thinking there was a wiring problem and spent a days time on it. Don't make the same mistake, always check your settings!!! Anyways, once I cleared that up, Pins 4 (Left CHT, GP 1) and 18 (+5V) showed 1.22kΩ, but Pins 18 and 22 (right CHT, GP 2) showed erratic resistances. Turns out while I was clipping the oil pressure jumper wire, I was messing with the connector a little too much and broke a resistor connection. Those little bastards were a bit hard to find because EVERYTHING is shrink wrapped and I thought they were further downstream in the electrical system, but nope, they were right there, hidden in the bundle, well camouflaged in shrink wrap.

 

Replaced the broken resistor, voila!

 

 

The lessons and the tips you should abide by:

• Faulty indications are not always probe problems (but usually are). A great way to quickly eliminate possibilities is to compare against another known variable, such as a second sensor. NOTE: The second sensor doesn't even have to be working right, as long as you KNOW (and can therefore compensate), hence why "known" is italicized! Anyways, things like switching connections (alligator clips are very useful as quick extensions!) will help you isolate where the problem is very quickly. For example: if the readings swap with the connections, it's the sensors, but if it stays consistent on the indicator, it's further upstream, such as a wire short or a problem with the indicator.
• Take your time. Carefully think it through. Evaluate possibilities, and double check your work and your readings.
• Seriously, double check your work and readings. Troubleshooting is a very logical process, and if you have faulty logic, you cannot troubleshoot effectively or solve the issue at hand. This also means you need to consider if your test equipment is faulty, or your methods (such as incorrect settings). This is especially true if the readings are not making sense!
• There's a lot of manuals out there, and the magic missing piece of information is usually buried deep. A good part of troubleshooting is understanding what you need to be looking for, and especially, WHERE. Sometimes you have to ask the manufacturer to provide additional drawings and diagrams.

In closing:

 

I hope this helps if you have weird indications in the cockpit. In my case with the aircraft that has a CHT split, I now know to check the jumper resistors, I'm willing to bet one has gone bad and started drifting.

 

If you want help with troubleshooting, just ask. I love a challenge. I'm working with FastEddieB on an intermittent trim issue he's having and set him on a track to hunt down a couple possibilities, and helped him consider a few points that he hadn't thought of :-). Just remember, I might not be able to help much in the first phone call or message since we've got to set up the "troubleshooting environment" and exchange a bunch of information and gather information material such as manuals.

 

Finally, please message first and wait for a response. I don't like cold calls. This image should help explain why, and yes I do get that deep into things (I do dabble in programming too), and it's easy to make it all go "poof".

[Johnny]

Hey Corey, your post is way over my head, but it does mention weird indications.  I am definitely getting one of those.  Upon cold engine startup, my right CHT immediately alarms and shows temp over 900 degrees!  Then the temp reading steadily declines ... through the normal range and then below normal range.  This pattern now happens every flight, exactly the same each time.  

 

My mechanic has checked the probe, and says it is ok.  I know Roger says to check all grounds and connections, which I will encourage my mechanic to do, unless you have some other diagnosis.   It just seems like a specific problem with a specific solution, so that I thought I would ask if you have seen this sort of thing before.  

 

Thanks in advance for any help! 

 

~~ Johnny

[Anticept]

Don't start up with the alternator on. See if that solves it.

[Runtoeat]

Corey, thanks for the post.  I know I appreciate your offer for electrical advice and imagine others do too.  My deceased father-in-law was a "wire wizzard".  He could make or fix anything electonic and helped his flying friends fix their avionics.  When he worked on my stuff and was ready for the "big test", he always would make the statement, "OK, now we'll go for maximum smoke".  Almost always there was no smoke but once in a while he would have to go to "plan B".

[Anticept]

I'm not at wire wizard yet, more like wire apprentice, but I do know my way around things well enough that I can figure it out given enough time.