Help! EGT is wacky

[Sregorb]

Everything I know so far: Everything has been working great to this point.

Got in today to drive to the movies and EGT was off, way off. Read around 200' and fluctuated a fraction of what it normally would have under normal operation. It's as if the scale was greatly reduced and 250' now equaled to around 1000' normally. Got to the movies and popped the hood. Wiggled the wires from the post-turbo sensor and got back in. Started up and was back to normal. Great loose connection maybe. Got out of the movie and drove home with normal indications. Went to get one of the kids and had wild fluctuations under 500' and when stomped on the go pedel it smoothed out from 500 up. Dropped down and same flaky jumping. Got home and traced wires from pyro along firewall, into cab, up dash and into pod. Everything looks clean and tight. There is a little plastic/rubber sleeve that goes into a (tension relief-looking spring from the sensor along the first couple inches of the wire run that was worked back away from the sensor. Pushed that back to stop and haven't had symptoms again for two quick trips. Are these symptoms something that someone experienced would know exactly what's going on with or has run across before? I just hate not having a reproducable problem that can be hard-fixed and closure brought around... . Could it have been a little sleeve at the sensor causing all of that or am I chasing a ghost in the wrong area?

I originally started on the path to a lost/loose power supply to the gauge thinking that the sensor was still sensing but the "gain" wasn't available to push the needle to it's proper ratio. But everything there looks nice and tight.....



Brian



edit: Oh! and they are Autometer Ultra-Lite series gauges if that helps diagnose... .

 

[illflem]

You pretty much diagnosed it yourself. You must have a short where you are wiggling the wire to produce a change. Whether or not it's in an area that you can isolate, cut and slice I couldn't say without seeing it. You may have to replace the sensor.

 

[tntway]

just had this problem......

started doing this after a year. took it back to ted jannetty and they fixed it. said it was a bad connection at the gauge.

hope this helps.

 

[Sregorb]

More info....

Ok, it hasn't worked since yesterday's quick test drive. Wiggles don't do any good and there are no splices that I've found. There are basically only 3 parts to the whole system:

1. The sensor-all one piece run with no splices and two leads coming from sensor to power box. Sensor to wire junction is prefab and not openable as it's a sealed unit.

2. The power box-has two leads inbound from sensor, three leads outbound to meter and two leads for power. Power checks good, ground good, and all leads are tight. Outbound to meter all show three diff levels of voltage (it lights a probe on each lead). No real way to check voltage variation from sensor and don't know tolerance of resistance/voltage levels to look for anyway.

3. The meter itself-Getting power fine to light bulb. Slight deviation of meter corresponding to acceleration and expected "heat up" says it's responding. When under heavy load and accelerate hard it will sometimes jump to expected level and then drop off again so I don't think it's the meter.



So, now I'm left with a bad sensor (in my mind not very likely as it's just two pieces of metal hooked to wires right?) or a bad meter powerbox that seems to be working. How can I be sure of either? How can I test either to eliminate one or more of the three? What's a 'safe' voltage to send thru leads to meter to test meter reaction (peg it)?



Brian

 

[Sregorb]

No gauge experts on the board?



Brian

 

[drees1]

Try this

Ok, like you said it has 3 parts to the system, but that does not make it any easier to figure out sometimes. Beg, Borrow, or buy an electrical meter with thermocouple input. Many have them these days. I paid $29. 00 for mine. Most everyone is using K thermocouple wire today, but it could be J too. Your meter will most likely be set up for K. Pull the gage wires and put 'em to the meter and go for a drive. If it still acts up, replace the probe and lead wire. You can also feed the meter a specified voltage that corresponsds to the temperature to check it too. Search for Omega Scientific and get their huge temperature book for free. It lists the voltage vs. temp for all the different wire types. But then you would need an adjustable voltage source to check it. I know we had these in the lab, but I do not know where to get one. Omega Scientific maybe. Any one you know in a manufacturing plant that handles a lot of temp controlled processes might have access to a voltage supply that could test the gage. Good luck!!



Doug Rees

 

[Sregorb]

Doug,



Great thanks! I called Autometer and got the expected millivoltage readings from the sensor at various temps, hooked up the meter and went for a drive. Sure enough the sensor read exponentially lower than what should have been there. Sounds like a toasted sensor/lead. Given the fact that the first time it went bad it was 'fixed' by wiggling the wires by the sensor I'd venture to say I have a broken wire in the lead somewhere. Looks like I'll be pulling a sensor, putting a plug in, and waiting for a replacement to show up... ... . Bummer, no temp readings for two weeks. I hadn't realized just how much I unconsiously use that gauge til now.



Brian