Water temp sensor (Lambda Sensor)

[Spittybug]

I need a little technical help here please. As part of the EFI conversion, I use the stock lambda sensor switch (located in the Y pipe in the VOD, passenger side) to give Megasquirt the engine coolant temperature. Before doing so, I calibrated it by taking a resistance reading at boiling water, icy water and room temperature. I inputted these values into a table. Well, for some reason the MS unit always says I'm running about 20 degrees hotter than either my car gauge or my infrared thermometer say I am. The car's temp gauge gets its reading from the temperature switch located on the passenger side of the cylinder head, close to the firewall.

To pull out and recheck the switch means draining the coolant and ripping off the manifold again. Before I do that, I thought I would see if anyone knows the specs on this switch. My readings were: boiling (100c)=1060 ohms, room (21c)=8,590 ohms and near freezing (0c)=20,000 ohms.

Since the lower resistance means a hotter temperature, I can always make Megasquirt report the same value as my car gauge by raising the resistance in the table. I would feel more comfortable with some basis for doing this though. Does anyone have a switch they could do a quick check on or perhaps you already know these values?

An added benefit of this is that MS can be configured to trigger a relay based on any monitored sensor. This means a simple buzzer can be added to give me an audible warning if I'm overheating. I know lots of people have suggested this feature!
Thanks,
Owen.

[opethmike]

What did for my (failed) conversion was I picked up a generic GM temperature sensor (diyautotune has them for around $20, I believe), and then I drilled/tapped the adapter for the thermotime switch to accept the sensor.

I went that route since Megasquirt is by default configured for GM sensors. Not really the answer you are looking for, but I thought I would offer up the suggestion nonetheless.

And yes, the extra triggering in MS is really cool. When I try my conversion again, I am planning on doing the same thing for an overheating warning.

[robvanderveer]

The water sensor is not the lambda sensor. The lambda sensor is an O2 sensor that measures your exhaust fumes to drive the ECU.

[Bitsyncmaster]

The water sensor is not the lambda sensor. The lambda sensor is an O2 sensor that measures your exhaust fumes to drive the ECU.

Yes, I think the lambda temp sensor is just a thermal switch. The idle ECU thermal sensor is a resistance sensor. I did a table of voltage readings with that sensor on my ECU so don't have the resistance. But off the top of my head I think the cold resistance is much higher than you stated.

[Spittybug]

The water sensor is not the lambda sensor. The lambda sensor is an O2 sensor that measures your exhaust fumes to drive the ECU.

Yes, I am aware that the lambda sensor is indeed an oxygen sensor. You are equally aware I'm sure that it only works at certain temperatures (above 15c), hence the need for and the existence of, the lambda water temp sensor, to which I make reference in my thread title.

[Spittybug]

Yes, I think the lambda temp sensor is just a thermal switch. The idle ECU thermal sensor is a resistance sensor. I did a table of voltage readings with that sensor on my ECU so don't have the resistance. But off the top of my head I think the cold resistance is much higher than you stated.

If it were just a switch, I wouldn't have been able to get the readings that I did, would I? It would have just been on or off.
The other insert into the Y pipe is indeed a vacuum switch.

[content22207]

The Workshop Manual thinks it is just an on/off switch.

Variable sender is in the passenger head.

Bill Robertson
#5939

[AdmiralSenn]

Yeah, I also used the GM sensor in the TTS adapter. Works great, no issues whatsoever with adapting or tuning resistance values. That's probably the least-hassle option available to you, even if it does mean more work.

I really don't know whether it's a good idea to adjust those values as you've described - I never had to deal with them.

[Bitsyncmaster]

If it were just a switch, I wouldn't have been able to get the readings that I did, would I? It would have just been on or off.
The other insert into the Y pipe is indeed a vacuum switch.

You got the resistance readings but I think you are using the idle speed thermal sensor. Nothing wrong with using that for your purpose. Does your circuit use a pull up resistor and feed that voltage divider into your MS unit? Maybe the error is in that ratio divider.

[Bitsyncmaster]

Are you still running the idle ECU? If so than that would cause your calibration error with both the idle ECU and MS using pull-up resistors on that sensor. It might affect the idle ECU starting in the cold but just wait and see. But that would explain your error from the MS so just adjust your tables to compensate.

[Spittybug]

Are you still running the idle ECU? If so than that would cause your calibration error with both the idle ECU and MS using pull-up resistors on that sensor. It might affect the idle ECU starting in the cold but just wait and see. But that would explain your error from the MS so just adjust your tables to compensate.

Nope, idle ECU gone. I fiddled with the calibration value in my table yesterday. To get it to read within a couple of degrees of what the car and infrared thermometer do, I changed the 100c table entry to about 500 instead of 1060 ohms. It's pretty sensitive in this area; 400 is too low.

[Bitsyncmaster]

Nope, idle ECU gone. I fiddled with the calibration value in my table yesterday. To get it to read within a couple of degrees of what the car and infrared thermometer do, I changed the 100c table entry to about 500 instead of 1060 ohms. It's pretty sensitive in this area; 400 is too low.

Yes, I found the hot measurements are sensitive on the idle ECU. But for the idle ECU nothing much changes only with the engine starting and warming up. You would not want to use this sensor to indicate an overheating engine.

[Spittybug]

I'm thinking an exponential curve like that is probably not the best for engine management such as afterstart enrichment, warm up or any of the other temperature related calculations that MS makes. Best bet may be to find a more linear one. I have the data on several others (including the GM) and will look at them. Not sure which if any will be direct screw ins for the Y pipe however.

**edit** I just entered the data from both a GM and Ford sensor into my spreadsheet, and their curves are very similar to ours, if not steeper. Must be common to have non-linear resistance to temp change.

[Spittybug]

:ehh:

My mistake. I was using the picture on the DMCH site of the Y pipe for reference rather than using my own photo. http://www.delorean.com/store/images/CATEGORY/large/1-3-0.gif

Why does the shown Y pipe only have two sensors in it while mine has three? I reused the water temp one as you can see. I don't see it on DMCH's schematic.
5090

[dmc6960]

My mistake. I was using the picture on the DMCH site of the Y pipe for reference rather than using my own photo. http://www.delorean.com/store/images/CATEGORY/large/1-3-0.gif

Why does the shown Y pipe only have two sensors in it while mine has three? I reused the water temp one as you can see. I don't see it on DMCH's schematic.
5090

While that specific image has only the thermal vacuum switch and the LAMBDA temp switch, the idle thermistor (temp sensor in question here) does show up on the ECU/Motor diagram. Chalk it up to being a new parts diagram for a new car.