ECT Home    AP1 Details    AP2 Details    Buy Here    Install Info

General Info

The ECT Display Driver is a small module with a removable wire harness that mounts behind the driver's foot well cover. All wiring connections are near the ECM and only 4 wires are needed for basic operation:

If you want to use the Alarm Output feature 3 more wires are needed:

The module comes pre-programmed with your choice of temperature set points but may be re-programmed by using a Windows PC equipped with a serial port. The programming process assigns each gauge segment to an ECT voltage that corresponds to a specific coolant temperature. Voltages in the range of .010v to .999v are allowed, which covers temperatures between approximately 155°F and 275°F (for the stock sensor and ECM). For those doing their own programming a chart is provided to convert temperature to voltage, allowing complete customization of the gauge display and alarm output.

In addition to the 6 (or 19 for the AP2 version) set points for the gauge segments there is a separate "Alarm Output" set point designed to activate a sounder or indicator light when the set point is reached. This alarm set point is completely independent of the gauge set points, meaning it can be higher, lower, or in-between gauge segments. Typical uses would be an "OK to VTEC" light or an "Engine Overheat" sounder. The Alarm Output provides a GROUND output when in alarm and is capable of switching up to 1/2 amp of current.  If more current is required or you need to switch +12v you will need to use a relay (not provided).

During normal operation an LED on the module blinks to indicate the unit is operating. The frequency of the blinks gives a rough indication of the coolant temperature, ranging from 1 blink every 2 seconds (cold) to 5 blinks a second (hot as hell).  Faster blinks = hotter coolant, so this can be a good troubleshooting tool.

Diagnostic Mode

For more sophisticated testing or troubleshooting there is a "Diagnostic Mode" available, activated by pressing a small button on the module. To enter Diagnostic Mode press and hold the button until the LED makes one long blink.  To turn off Diagnostic Mode, remove power from the module (turn the ignition OFF).

During Diagnostic Mode the LED and Alarm Output will activate each time the coolant sensor voltage crosses one of the set points. The LED/Alarm Output blinks out the number of gauge segments that should be lit.  In addition, it blinks fast for a rising temperature and slow for a falling temperature.  This is easier to understand with an example:

Assume you have activated Diagnostic Mode and the car is warming up and reaches the set point for the "third bar" on the gauge.  As soon as that threshold is reached the LED will flash 3 short blinks, indicating the temperature is rising and the new gauge reading will be "3 bars"   If the temperature cools down below the "3 bar" threshold you will see 2 long blinks, indicating a falling temperature and after a few seconds you will see a new gauge reading of "2 bars".  If you have a sounder connected to the Alarm Output it will activate in the same pattern as the LED.

Diagnostic Mode will be especially useful for those doing custom programming or using the module with a non-stock sensor or ECM.  Verification of your set points can be done by connecting a piezo sounder to the Alarm Output and a voltmeter to the coolant sensor input.  This allows you to easily verify each transition point by watching the voltage on the meter as you listen for the beeps.  Without the beeper you would have to watch the dash temperature gauge (with it's built-in delay) and make a guess as to when the transition really occurred.

Note: Diagnostic Mode is NOT to be used for normal operation because it delays updates to the gauge display.  As such, the module is programmed to reset back to "normal" mode when ignition power is turned off.

Programming

Programming is performed with a PC using a standard 9-pin serial port and Windows HyperTerminal software.  A programming cable is provided and programming can only be done with the module powered up (installed in the car).  You will need to configure HyperTerminal for a direct serial port session at 9600, 8,N,1 (9600 baud, 8 data bits, no parity, 1 stop bit) and send a simple text file using the "transmit \ text file" menu option and selecting the proper file. Sample text files and a HyperTerminal session file are included in the Installation Manual downloads for AP1 and AP2 applications.

While downloading the text file to the module the LED will light continuously.  When the file is completely transmitted the LED will go out while the built-in EEPROM is programmed with the new set points.  A successful programming sequence is indicated by 2 slow blinks on the LED after the EEPROM is programmed.  Failure to get the 2 slow blinks indicates an incorrectly formatted text file.  If you see 5 quick blinks that means at least one set point in your text file is out of range (more than the allowed number of characters).

Programming can only be performed a limited number of times, typically 100,000.  So try to avoid re-programming the unit more than 50 times a day or you'll wear it out in less than 5 years.

For those not yet bored - a functional description

The module monitors the ECT input voltage to the ECM (either factory or AEM) and uses an A-to-D converter (analog to digital) to change the voltage to a 10-bit digital representation (a number between 0 and 1023).  It then compares this number to the programmed values to decide how many gauge bars to light up.  It then looks up the pulse width and frequency information in a table and sends out the proper signal to light the bars on the gauge.

The gauge transition points have no hysteresis and will have a tendency to oscillate if the ECT voltage is right at a set point.  This is not a problem for the module or the gauge, since the gauge does a very good job of averaging the pulses.  But it will be noticeable if you use Diagnostic Mode to test your programming.  Also, remember that the gauge can be pretty slow to update, usually taking 10-15 seconds to light the next bar after receiving the new pulse rate.

The Alarm Output DOES have some hysteresis built in so that you don't get "on-off-on-off" cycling of the output.  That could be annoying if you connect a sounder to the module and the ECT voltage is right at the alarm transition point (ask me how I know it's annoying).  With hysteresis, the alarm output will still activate right at your programmed set point, but won't turn off unless the voltage rises at least  5mv (.005volts) above the set point.

If you do any testing in Diagnostic Mode and monitor the ECT voltage with a meter you may notice the gauge transitions differ from your set points by 2 or 3 millivolts.  This is usually caused by the ECM voltage (the 5v supply for the module) being slightly different than exactly 5 volts.  These errors in supply voltage are somewhat self-corrected by the module since the voltage reference used for the A-D converter is tied to the 5v supply, but it's not perfect.   If you find the self-correction is not accurate enough, just tweak your programmed set points slightly till the transitions are exactly where you want.

How did I determine the correct ECT voltages?

Good question.  My initial tests relied strictly on the OBDII data port on the car for temperature readings, but because of the incremental jumps in the indicated temperature (like from 246° to 261° with nothing in-between) I decided to conduct additional testing to provide better precision and to validate the OBD data.

I purchased a new ECT sensor and immersed it a pot of synthetic oil (M1 10W-30) along with a fast-response thermocouple probe.  To keep the oil constantly circulating (to avoid temperature variations) I immersed a propeller-type stirrer into the oil bath, driven by a small electric motor. 

The ECT sensor was connected to the ECM in my S2000 in place of the sensor in the car.  The test was run with the ignition ON but the engine not running.  The pot was slowly heated using a propane torch and ECT voltage readings were taken every 2°, along with OBD readings.  When the oil was at maximum temperature I allowed it to cool slowly and repeated the temperature/voltage measurements on the way down, confirming they matched the "going up" readings.

From previous tests I knew the key to accurate measurements was to make sure the ECT sensor stabilized at the oil bath temperature.  If I changed the oil temp too quickly the ECT didn't change right away - it took a minute or two to catch up.  If I got impatient I knew I'd get inaccurate readings so I set up a lounge chair in the garage and took my time.  Excluding setup and tear-down this test took almost 6 hours.

The thermometer I used is a decent lab-grade digital unit with a basic accuracy of .2°F.  To measure the ECT voltage I used my trusty Fluke 189 that has DC accuracy of .025%, so I'm reasonably confident the numbers I came up with are as accurate as any of us are likely to require.  Overall, the OBD in my car reads between 2 and 5° high across the board.  I have no idea why, other than perhaps Honda didn't put circuitry in the ECM to compensate for an out of tolerance 5v sensor supply voltage.  My ECM 5v supply was slightly above 5v, and that would cause the ECT voltages to be higher than what the ECM was programmed for and may account for the slight error.

Bear in mind the ECT module DOES have compensation for the 5v supply, so you don't have to worry about the ECT being inaccurate if your sensor voltage is a little higher or lower than 5v.

top
 

Copyright 2005 by Modifry Products, LLC all rights reserved.
Modifry and the "squiggly arrows" design are trademarks of Modifry Products, LLC. 
AEM and AEM EMS are trademarks of AEM Performance Electronics. 
Neither Modifry Products, LLC nor Modifry.com are sponsored, endorsed, affiliated or authorized by AEM.