Hello,

This is my first post on this Forum. I am an ex Mechanical Design Engineer (Engineering Recruiter last 15 years). I reside in Sydney Australia.

My knowledge of Automotive electronics is in all honesty, quite rudimentary so please bear with me. . .


I recently completed an Auto > Manual conversion on a 2002 Lexus IS300 Sportcross. The IS300 platform was only ever offered from Factory with a manual transmission in the USA. As part of the conversion I replaced the original (Automatic) ECM with a unit from a USDM (California) vehicle.

The ECU is throwing DTCs (0450 & 0451) because it is looking for inputs from an Evap (Fuel Tank) Pressure Sensor that non-USA cars don't have.

I have been advised that re-configuring the ECM is not feasible so my only other option is to give the ECM the inputs it is looking for via an emulation / simulation circuit. From my research to date I have read that the sensors 0-5V output signal is best approximated by a potentiometer.

The ECM supplies the sensor with a regulated 5V reference voltage and the sensor returns a signal voltage between 0.5V and 4.5V according to the pressure / fuel level in the fuel tank. When the pressure in the fuel tank is low, the output voltage of the vapour pressure sensor is low; when it is high the output voltage is high.

Referencing the Lexus IS300 Diagnostic & Wiring Manual (Page DI-251) and the link below;

http://www.tsienna.net/evaporative_emission_control_system_pressure-2236.html

1. For DTC P0450 the ECM checks for abnormal voltage fluctuation from the sensor circuit; which it's getting due to a non-existing sensor / open circuit.

2. For DTC P0451 the ECM checks for both a "noisy" sensor and a "stuck" sensor

The ECM checks for a "noisy" sensor by monitoring fuel tank pressures (time monitored = 10 sec) when car is stationary as there's little variation in the tank vapour pressure. If the indicated pressure varies beyond stated limits (0.5-4.5V) the ECM will illuminate the MIL (2 trip detection logic) and a DTC set

The ECM checks for a "stuck" sensor by monitoring the fuel tank pressure over an extended period (time monitored =10++ seconds). If the indicated pressure does not change over this period the ECM will conclude that the fuel tank pressure sensor is malfunctioning. The ECM will illuminate the MIL and a DTC set

So; even if I knew the resistance across the Terminals of the Sensor & inserted a potentiometer into the circuit; I would still get a fault code because the output voltage would be static & so would trigger a P0451 as the ECM would deduce the sensor was stuck.

As far as I can determine what I need to do is insert a Signal Generator module into the ECM's sensor circuit that has the following characteristics;

i) 5V input / 3.3V (approx) output
ii) Sinusiodal signal output; because voltage change in a square wave signal might be too abrupt & trigger DTC (can anyone confirm/ refute this?)
ii) Adjustable frequency ( Min 0.01 Hz?) and amplitude (1.5V Max)

Some Modules also have an adjustable Duty Cycle. As I stated at the beginning I'm an electronics novice so have no idea if adjusting the Duty Cycle would lower the effective minimum frequency of say; a 1Hz -30KHz module to say 0.01 Hz / (2+ minutes per pulse) to stop the ECM triggering a noisy/stuck sensor fault.

I have spent quite an amount of time online & come across various modules (i.e. link below) but as yet; have not been able to find a PnP module with the above specs & would like to know if any members can help me resolve this problem.

https://www.amazon.co.uk/SainSmart-Frequency-Function-Generator-Triangle/dp/B011NVCKBG

Regards
George

 

The ECM requires a signal indicating a vapor pressure from the fuel tank. This signal would be used as one of several in formulating the fuel flow to injectors, or possibly other functions. So, is there a reason that a compatible vapor pressure sensor could not be installed, independently, yielding the appropriate signal to the ECM, with a vapor connection to the fuel tank?

 

If only it was that easy. . .

Referring to the attached diagrams; aside from both a Pressure Sensor (and VSV Solenoid); the USA model has more emission plumbing + an extra hard-line. All this takes up more space & necessitated a protective cover over the Canister assembly.

Yes I could buy the Sensor & install it but unless I 100% replicate the USA cars' EVAP layout (which I don't think I have the space for..); I'm just going to feed the ECM erroneous data..

 

Apparently, that sensor operates a valve when fuel tank pressure (FTP) is too high or too low. Of course, years ago, fuel caps were merely vented. While it may be anathema to some, I see no reason not to mimic an open tank or one with a slightly positive (i.e., normal) pressure.

I do not have a Lexus wiring manual. My references here are the Ford F-150 and Fusion manuals. Check to see whether the sensor is three-wire. If so, one wire is a reference voltage (say 5V), one is ground, and the third is signal. Many automotive signals can be viewed as a variable resistor, e.g., the wiper on a potentiometer provides the signal. Certainly the pressure sensor in my Fusion A/C is that way. Here is a rater long video that says the same ting about the FTP: https://www.scannerdanner.com/chapt...an-evap-pressure-sensor-problem-p0452-gm.html

I did not listen to the whole thing, and only scanned it. But is seems confirmatory. Maybe the DIYLexus crowd has something similar, and you may be able to find the normal wiper voltage and total resistance of the sensor. A very friendly Lexus dealer might have that info from the factory service manual. As an alternative, find a junked Toyota that has an FTP sensor and measure it. As a last resort, I would use the values of the FTP sensor from any car of that approximate vintage.

EDIT: My apologies. I am a very slow reader and didn't check your link. Apparently, you have the appropriate voltages. If you want a sine wave output, quite easy to do with a microprocessor or even an NE555/LM555: http://www.learningaboutelectronics.com/Articles/Sine-wave-generator-circuit-with-a-555-timer.php

 

Thanks for this information. . .

https://www.amazon.co.uk/SainSmart-Frequency-Function-Generator-Triangle/dp/B011NVCKBG > Vref = 9V but I just included it as an example

https://www.ebay.com.au/itm/253143569154

As stated; I've limited knowledge in this area so perhaps I'm missing something; but AFAIK signal generators are quite commonly available "off-the-shelf" items (i.e. above link I shared previously + another); so I rather than re-inventing the wheel, what I was hoping to do (with a bit of searching..); is find an adjustable signal generation module with the parameters I'm after and plug it right into the ECM...

Having said that I would not be able to install the DC 5V TLC5615 10bit DAC Module (2nd link above). Aside from VCC. GND & REF as I don't u/s the functionality of the other terminals let alone how to wire them up:
DI = Digital Input | VCC = Signal
SCK = Serial Clock | GND = Ground
CS = Chip Select | AO = Analog Output
DO = Digital Output | REF = 5V

I think getting the correct signal frequency, amplitude & attenuation such that the ECM didn't generate Fault Codes would be an iterative process so I would prefer a module with an LCD screen that showed the values in real time. Of course once I had refined/ confirmed values that worked; I then could, if I wished; construct a small OEM-looking module and insert in into the engine bay within or adjacent to ECM casing.

PS: I have attached IS300 Electrical schematics for both EU-UK versions (w/out an EVAP VSV & Pressure Sensor) & USA versions. . .

 

MONITOR STRATEGY | P0451
Frequency of Operation: Once per driving cycle
Duration: Signal fluctuation (noise) monitoring: once every 10 sec
No signal change (stuck sensor)monitoring: once every 20 minutes
MIL operation: 2 driving cycles (i.e. Malfunction Indicator Light will trip if above conditions are detected in 2 subsequent driving cycles)


Referencing your 555 timer link, and the above Monitoring Strategy from the Lexus Manual; I would* be looking at a frequency ( 10 sec << F < < 20 min.) of say; I pulse per minute; I.e. F = 1/60 = 0.0167 Hz

Calculation of Resistor Size for required frequency:
For a sine signal of frequency = 0.0167 Hz I calculated the following (assuming capacitor size remained unchanged at 10nF)

F = 0.016 Hz = 1 / ( 2 x 3.14 x 10e8 Ohms x 10e-8 Farads)

I.e Required resistor size is 100 MegOhm

Calculation of LC Network values such that Resonant frequency = 0.0167 Hz. . .

i) Regarding the sizing of the Capacitor I used the Authors examples (I also came across the following (quoted) information I found online but ignored it as the corresponding Inductor size would have been 300 K Henrys; which is obviously not right ???

"The rule of thumb is to put in 1 Farad of capacitance for every 1,000 watts RMS of total system power. But there is no electronic penalty for using larger value caps, and in fact, many see benefits with 2 or 3 Farads per 1,000 watts RMS. The larger the cap, the faster it gets ready for the amp's next big hit."

I do not know of any equation that determines the size of the inductor relative to the capacitor in a sine signal generation circuit so I just used the same ratio as the author of the article, namely: 1microFarad V 470 microHenrys < I.e. L = 500 C >

So for F = 0.016 Hz. . .

F = 1 / ( 2 x 3.14 x SqRoot (LxC)
LC = 3.09 = 500 C (squared)
C = 0.08 Farad > 80 milliFarads
L = 40 Henrys < which still seems quite large. . ..>