Subpage of MembersPage/WolfTm
Summary:
The 100C/V (0-11V) output of the [EGT amplifier] can be used with some care, if the ECU was not ordered with the sufficient number of EGT inputs or analog inputs without pullup. See "GOOD granularity" below.
When ECU is purchased with 1,2 or more (4 or 6 is sometimes requested, some logger was ordered and delivered with 8 EGT inputs),
- EGT amplifier outputs are connected to inputs with no pullup
Confirmed:
- the cause of the higher reading was the internal ecu pull up
- 51K (really ? Usually 2k7 pullup on analog inputs)
Unfortunately NO input with no pullup resistor in this ECU.
- is there a way to scale it mathematically based on the pull up?
- yes, since the function is monotone, it is possible to calculate. However...
Rough granularity/resolution
The EGT amplifier (internal voltage divider for convenience =) 233C/V (= also called "5V output") has an output protection equivalent to
- 1 / (1/75+1/100) = 42.86 k Ohm resistance
- 2.7 / ( 2.7+42.86) = 0.0593 ... this means 2k7 pullup causes a strong bias, and resolution is reduced
- say 20C => 1102 C (at 0.02V measured on the 0-11V output )
- say 1170C => 1170C (at 5V)
- The theoretical formula for that case is:
- EGTconv1=1170 - (1170-20)/(1170 - 1102) * (1170 - EGT)
- EGTconv1=1170 - (1170-20)/(1170 - 1102) * (1170 - 1147) = 781 C for example
- Note: this is rough estimate, not based on actual measurement or calibration.
- even if calibrated, the granularity would be rather rough.
Best to use input with no pullup.
- input with 2k7 pullup will likely make the EGT reading granularity annoyingly rough, above.
Achieving GOOD granularity, using the 100C/V (= also called 0-11V) output and 2 resistor voltage divider:
- reasonable formula (VT ini can be tweaked to output real calibrated result) becomes possible if
Connecting the analog input (with internal 2k7 pullup) via/to:
- 6k3 to the 100C/V (0-11V) output of the EGT amplifier
- 4k7 "pulldown" to GND
- this way it acts like output divider with effective of 2k7=1/(1/6.3 +1/4.7) which means that half of the range is available, which allows precise measurement after simple calibration.
- Something like 2*EGT-1170 conversion will be sufficient (refine it at least with the low temperature reading) and good granularity:
- 0C => 585C (2.5V on input) => 0C
- 1170C => 1170C (5V) => 1170C
Needless to say: take care when using the 100C/V output, do NOT short to GND or 12V (only connect to the 1 end of the 6k3 resistor).
- when using a DVM to measure voltage of the 0-11V output, make sure the DVM is in DC voltage mode (not current mode, which shorts the signal to GND, and might cause damage)
Orig report:
Engine off, ambient sensor temp (~20C)
- Tested Pin7 DVM output (1V/100°C) and getting 0.7 volts
- corresponding to 70C
- not 6 volts, roughly corresponding to Pin6 2.5 volts.
- I'm getting 2.5 volts out of Pin6 (0-5V output, 1V/233.3C)
- Thus getting 580°C instead of 20°C.
Hint: The output is almost certainly connected to an input with pullup (possibly strong pullup).
- Diagnostic: disconnect anything from output, and measure DC voltage with DVM (the DVM has 1 to 10MOhm input impedance).
- Solution: make sure to connect pin6 output to an input without pullup resistor
Sensor connection (seems good):
EGT sensor is K type yellow-red wires:
- yellow = green (plus wire) to board GND
- red = white (minus wire) to the board cathode on corner