MS3 Set Up: Thermistor & AFR Tables Calibration - Electrical Issue?
 

I bought Rev's plug'n'play MS3 Enhanced and have slowly been trying to set it up through the summer.

Idle smoothly but got pretty concerned about my coolant temperatures - the TunerStudio read out using the base map/standard CLT sensitivity (20C=2500ohm, 40C=1170ohm, 80C=320ohm, 2490ohm bias resistor) has me idling at 215F even at 80F ambient temperature. However, the temperature gauge on the dash is at "11 o'clock" still. Driving around, the coolant temperature easily heads to 220F+ - but the dash gauge has never gone past 12 o'clock. No other signs of overheating that I noticed besides the readout in TunerStudio. I replaced the rad cap and CLT sensor and pulled the thermostat and confirmed it opens at 180F. I've got the Koyo 55mm rad and dual fans are running in parallel.

On a separate note, I noticed my TunerStudio AFR readout was noticeably different from my AEM UEGO gauge readout (set to P0 0-5V analogue voltage output). I had to adjust the AFR Calibration Table by 15% from what the AEM manual says the voltage would be so that TunerStudio matched the gauge readout.

I bought the Etekcity 774 12:1 Infrared Thermometer to check my engine bay temps. While idling at "215" on TunerStudio, the thermometer says the temp at the CLT housing is 190F, the heater hose send is 170F at firewall, heater hose return 155F at firewall, valve cover 170F at back, engine side of thermostat is 170F and rad side of thermostat is 140F. If I reduced the sensitivity of the CLT thermistor by 15% it would conveniently be pretty close to the thermometer read out...

WOuld it make sense that the CLT and AFR voltages both off from reality by ~12% because of a different ground? I guess I could try to check the air temperature gauge to see if that is off by a similar amount. Any suggestions to resolve the issue?

The dash gauge is very much not linear. It will usually hold a position somewhere near the middle for a very wide temperature range, and only goes above that if it is badly overheating. You might not even see the needle move before the radiator cap opens.
s for the coolant temp sensor, even if you have "the right settings" for the sensor type, there is a fair bit of variance between them. It doesn't matter much if the CLT reading is off by a few degrees, as long as it is in the ballpark, and is a consistent value. The exact value has some minor effect on tuning (things like warm-up enrichment and open-loop idle rpm, and at which point certain other functions are enabled). You need to get the fan to turn on at a reasonable temperature value (typically a few degrees above normal operating temperature) but other than that, the only really important thing with the actual temperature value is for you to know if the engine is for some reason running hotter (or colder) than normal.

As for electrical issues, it is absolutely critical for any kind of analog sensors to have a reliable ground connection. If any sensor or the Megasquirt isn't properly grounded, the reading can be wildly wrong, and fluctuate from a number of strange factors.

you miscalibrated.

select rx7 clt s4 and s5.

Thanks guys. Those cal values are basically the same as RX7 CLT S4&S5 (176F/80C=300ohm is the same as 80C=320ohm), and makes sense to me that there's fluctuation between sensors, but this is not a few degrees, it's >10%...

I understand that the AEM makes its own 5V and that's not the case with the CLT so I guess that's not it. Even though the thermostat opened for me when I had it in a pot maybe it's sticky/not consistent. Seems crazy that idling at 225F right now at the CLT could be possible!

OTOH... i just had my air temp sensor that came with the MS3 out of the engine bay - clearly in the ambient air - and it is reading 80F in TS when it should be reading 70F... again, conveniently 12.5% too high of a value. If I adjusted the CLT sensitivity to be 12.5% lower then it'd be reading 188-197F instead of the 215-225F I mentioned above.

the AEM always reads off because it's a horrible WB with a really shitty scale that's tuned to each controller and the default AEM Uego selection in TS isn't even close.

the AIT and CLT...dunno what to tell you, never had an issue with those reading off the normal values. sounds like there's a grounding offset between the MS and your sensors.

unsure how Rev builds his boards, but it's possible the bias resistor used as a pull-up is to blame here. Maybe the wrong ones installed, or the tolerances just being too far off the measured value. If you notice when you calibrate the sensors, there's a value: 2490. That's the resistor being used for a 5v pull-up. They are typically 1% tolerance resistors so they are always very close to that value. If he used the wrong ones, somehow (not saying he did), that would throw off the scale of your sensors.

Unless I'm missing something else, I think you're right Braineack. I pulled the coolant sensor and put it in a shot glass, while connected to the car, with a grounding wire between the sensor housing and the engine, and the temperature in TS got up to 231F while I poured boiling water into it! If I change the pullup resistor value to 3150 ohms then it gives the correct 212F. I also checked at lower temps as the water cooled down to ambient and it seemed like 3150 ohms is what it needs to be - case solved?

If so, then I need to figure out what sensitivities to put in for the air temp sensor so that it gives the correct readout. All I know is that it's Bosch. Any ideas?

use GM. and change to 3150 as well.

Rev says he's sure the pullup resistor is 2490 ohm... not sure what else I'm supposed to do when both the MAT and CLT have similar offsets from reality though! :(

Ugh, no replies from Reverant, great!

if calibration them both at 3150 seems to bring the curve scale back to a realistic value, that's probably what I'd go with. Sounds like there's a voltage offset between the sensor ground and ECU or something.

Sorry to bump an old thread, but did prog_rawk figured out the issue? I have the exact same problem. I bumped the resistor value up too but I hope the voltage offset doesn't affect any other electronics in the car.

Can Braineack chime in on this?

No, sorry, still assume I have a difference between sensor ground and MS3 ground but haven't found it yet. :/

I am having the exact same issue right now... MS3 from MSLabs and my ECT shows 210-220 at idle or cruise but my stock gauge is at the 10-11 position. I am fixing to put a thermometer to check the thermostat temps and radiator temps. I have a aftermarket radiator, parallel wired fans, and the whole coolant system flushed and filled 50/50. Driving at 70 or idling its the same temps and heater works just fine. My lower hose feels cool to the touch.

I have been going crazy

I am fixing to try the values to calibrate it at:

-4 16150
104 1160
176 330

Hi..i am a new user here. I would like to say the coolant temp sensor, even if you have the right settings for the sensor type, there is a fair bit of variance between them. It doesn't matter much if the CLT reading is off by a few degrees, as long as it is in the ballpark, and is a consistent value. The exact value has some minor effect on tuning. You need to get the fan to turn on at a reasonable temperature value but other than that, the only really important thing with the actual temperature value is for you to know if the engine is for some reason running hotter than normal.

https://www.7pcb.com/