Temp Gauge Calibration

[3p141592654]

If you ever wondered what the stock temp gauge was trying to tell you, here it is.

It is designed so that it basically doesn't move over an operating range of about 90 to 110 degC, then it shoots up. Basically its useless except as an overheat indicator.


PS. the y axis is plotted as a percent of full scale deflection since there are no numbers on the gauge

 

[Piratetip]

Good info to know. thx

 

[Fletch124]

Interesting. Good to know but i bought an autometer 2 days ago.

 

[jetjock]

Nice job 3p. The problem is with the sender. The temp gage itself is of the cross coil variety, one of the most accurate and fastest responding types out there. In fact it's so fast silicone oil is used to dampen it. The fuel gage is also cross coil. It's slow because it's a retaining type gage and has even heavier oil in it.

 

[Ckanderson]

^ i heart our SME's

 

[steven89]

hmm, so is there a replacement sender that will actually work right? like one that we can drop in and actually see the temp fluctuations?

 

[3p141592654]

JJ, the nonlinear behavior is built in to the gauge.

To see this, I have attached a plot of gauge current versus FSD for the gauge by itself. I will have to pull it apart to see what is going on.

I agree the gauge is fast, and it seems well built to.

 

[jetjock]

Grrr. It's been years since I played with the thing so that's what I get for going on memory. That and thermistors aren't exactly known for their linearity.

Yes, you're correct. I just put my spare on the bench and checked it. Those last few ohms sure make a difference. I also pulled the book out for a quick refresher on how these things work. PM me an email addy and I'll scan and send you the info if you'd like. I noticed the resistors on the cross coils are different values. That may be a path to linearity but I'm not motivated enough to screw with it.

 

[CRE]

Doesn't really even have to be linear... A clean logarithmic scale would be fine. I'd like to see the materials you have on it. I'll PM my email addy, if you don't mind.

 

[Poodles]

interesting thread, would be neat if you could modify it to be accurate and not like an idiot light...

 

[CRE]

I have been wondering if it wouldn't be possible to do this with both temp and oil pressure gauges.

 

[Poodles]

definately be interested

hell, if you could figure out how to calibrate it, I'm sure speedhutt could make custom gauge faces that would make em as accurate as aftermarket gauges.

 

[jetjock]

I have been wondering if it wouldn't be possible to do this with both temp and oil pressure gauges.

You must have missed an earlier post where I calibrated my OP sender to solve a minor idle problem. I probably have the world's most accurate stock gage at this point. The gage itself has adjustments but messing with them is rarely needed. That said the OP gage and sender work on an entirely different principle than coolant temp. It's a bi-metal/bi-metal servo type deal that outputs a pulse. You'll have to machine open the sender to get at the adjustments and afterwards seal it back up. While the gage itself doesn't have oil in it like the coolant and fuel gages it's thermal design also makes it a bit slow. About 2-3 seconds to react.

 

[CRE]

Yeah, I missed that. I hadn't assumed they operated similarly. I was just expressing the interest in modifying the two.

 

[Poodles]

suck their's that much work involved... and that the oil pressure gauge is that slow...

so it's going to gloss over momentary pressure loss and then you get rod knock, BRILLIANT...

 

[jetjock]

Well, I've had the car 20 years today and so far haven't had momentary pressure loss. Perhaps it's happened and I didn't see it but since I've never had rod knock I'm assuming it's never lost pressure. Besides, even if I had an faster gage I don't spend enough time looking at it to see anything.