Problem with Linear Coolant Guage Mod
 

So living in the Mojave Desert, and having my dash apart, I just could not turn down actually knowing what my water temp was, and dove into the linear coolant guage mod.

So I soldered the jumper in place, did not damage the guage, Put the dash back in, then wired as shown in my attachment.

So I fire the car up with my OBD-II software on my laptop, and the guage does not budge till I have a coolant temp of ~195. At 210, when my fan comes on, the guage is barely touching the bottom of the C.

I have the sensor wired in parallel with a 100Ohm resistor, which is grounded on the firewall.

I have both of those wires coming together in one side of a connector, with the other side going to the guage, with 2 10 Ohm resistors in series on the other side.

Anyone see where I went wrong?

-Jason

Where's your potentiometer? The actual adjustability of the gauge needle's position and reaction to coolant temps is part of the mod.

The 20-ohm in series may be the problem. Not all cars are the same.

A small change in series resistance could give you what you want - that's why the potentiometer is needed.

I don't remember what OHM pot I used, but the nice thing is...you can adjust when the needle moves.

You might want to be at 16 or 17 Ohms.

My guess is the 20 ohm series resistance is probably way too high...

Instead of altering the gauge, I simply put 100 ohms in parallel with the sensor., just to try something to start. This would change a 20 ohm sensor resistance to 16.667 ohms, which from memory moves it up the response curve about 20 deg F. The net result is that the gauge starts to move just before the 221 fan kicks on... The minus is that the gauge starts to move off cold right at stat up also.

The temp sensor is higly non-linear with temp. It is 200+ ohms at room temp and goes down to about 18 ohms when operating. The temps of interest occur in a very small range of resistance from the sensor. That is why you need to tune the circuit to your situation. Too many variables otherwise stack up (bad connections, sensor variability etc.) and the reading makes little sense. The key is to know what the gauge is telling you and make it so there is some early warning, as opposed to the factory late warning.

When you read the mod documentation, take a close look at the response curve that is published and use it to figure out your operating point on the curve. The high series resistance will push you back down the curve to a cooler reading.

I am going to try a 10 and a 5 Ohm in series today, it seems that the 20 Ohms is way too much between the guage and the sensor...

I was going to try and do it without the potentiometer, as it looks like 20ohms is what I need, but that appears to be wrong.

-Jason

I ended up with 75 ohms is parallel with the sensor and 4 ohms in series (calibration). With my measured and calibrated sensor the above values produced 191F at the gauge center point, but as others have said the values change some depending on the specifics of your vehicle.

Radio Shack has a nice wire-wound potentiometer that I think is 0~20 ohms or so. You'll definitely need the wire-wound version for the heat involved; electrical current not engine.

I got it...

pulled one of the 10 Ohms out, so I am running 100 in parallel, 10 in series... I may step that down to 7.5 if I could find one for the needle in the middle for other people driving my car.

Top of the top C mark is 190, 200 is at 8 o'clock, 210 is at 9 o'clock, and 220 is at about 10 o'clock. The LS1 likes to run at about 194 while moving, 205 stopped.

Long as I am below horizontal, I am gravy...

Im stoked to actually know my coolant temp at all times.

Its accurate enough at a stoplight you can see the guage move...

Welcome to the world of the properly-functioning coolant gauge linearization mod. :)