Project: Build an On Board Diagnostics System for a carb'd car

[Buzo]

Made one 5 feet long harness, Labeled each wire with the name of the signal it will carry.



Decided to modify my ECU instead of the car's harness, so if I screw up something, I just buy another ECU.



Each cable is soldered to the ECU, An advise if you ever solder something, put some solder to the wire and to the PCB, then join them and apply heat with the iron. It will allow the solder to reflow better, also, this way there is no need of a third arm to hold the solder while you are holding the iron and the wire.







How the PCB looks after the wires were soldered





Putting together the ECU again.



There is plenty of room for the wires below the PCB

[Buzo]

Cont'd...

Had to cut a hole in the side of the ECU.






The modifyed ECU already connected to the OBDZ.



Installed back in the car.

[Buzo]

A quick note here: If I ever finish debugging this OBDZ, I will buy a yunk ECU and its matting connectors, so the OBDZ can be installed quickly in a car, get the code (s) and then put everything back to normal.
I soldered directly to the ECU in mine just because I din't want to go to the JY today. Football time!

[lostforawhile]

i'll sell you mine for a few bucks plus shipping with the connector, i have no use for it with the SU's

[Buzo]

This is what the book says about the cranking leak solenoid valve.



And this is what actually happens in the car during start up.



So my code will be:

Ign ON, Cranking Sol ON
Start ON, Cranking Sol ON
Start OFF, Cranking Sol OFF (after a 1 sec delay)

Anything else will cause the LED to blink and point you to that.

[Buzo]

This is what the book says about these O2, Ta and TW sensors







And this is what I captured.
The temp when I measured these signals this morning was 10 C.



Easy to put pass/fail limits!

[Buzo]

i'll sell you mine for a few bucks plus shipping with the connector, i have no use for it with the SU's

Sure! I'll let you know when I'm ready, and thanks for the offer.

[Buzo]

Here is the EFE relay information.

First, the book information and below the ECU Output measured in the car.

[Buzo]

This picture is amazing. It shows what happens in the first 6 to 7 seconds during start up.




I think I finished capturing the Digital signals, I need to start capturing the frequency solenoids and control valves. So I need to do some more code for the microprocessor.

[Buzo]

The theory belonging to below post.



It makes sense the Vacuum B gets activated first, since it's calibated at a higher vaccum compared to the Vacuum Switch A.

[Buzo]

Here is a picture of the user interface, written in LabView.
At the end, each LED will say the name of the sensor/solenoid and will have a second LED with the expected status, so visually one can compare if a Solenoid should be off/on, blinking or something else in a given time.

This is if a laptop with a serial port is available, but the OBDZ will have one single LED indicating the problem, so no computer is needed.

By the way, Would you guys mind if I use your 3Geez logo as background in my software? It will look better than the default gray.



I have logged lots of data during my trips to my work and back home. So I plan to keep explaining every single actuator in the car, so anybody with a multimeter can do the diagnostic of a failure.

As I said, I am currently working to determine frequency and PWM values for the control solenoids...

[Buzo]

Continuing with some theory, here is the Primary Slow Mixture Cut-Off Solenoid Valve.

The book explains better than anybody the function of this solenoid. Also, if it is working fine, it means all the needed inputs are working OK as well.

[Buzo]

Just to confirm what the service manual and I just said...

[Buzo]

Just one more... (for today)

This is the Oxigen Sensor. I increased the acquisition speed, so now it reflects better whats happening.
As long as it is at .5 average, the car and the systems are well carbureted, and should pass the smog test!

I have always wondered if a guy with a non-stock carburetor can use the O2 signal to make the carburation adjustments. Why not?

[integratuner77]

Hey man I just wanted to say great work. Coming from an engineering view this is awesome! Absolutely an ingenious idea and very well executed. Thank you for the constant updates.

[Buzo]

Couldn't resist and made another run with the Solenoid and Control valves connected to the OBDZ.

And just because I found this graph very accurate I think it is worth to post it.

[carotman]

I don't see a problem if you use the 3geez logo as long as you continue your project till the end

I will buy one of your diagnosis boards for sure.

You should make the thing "configurable" so one can disable sensors depending on where they live.

Like I said, Canadian Accords didn't get any O2 sensor for 86-87. This means the frequency solenoid A and B that alter mixture aren't there.

[Buzo]

Hey man I just wanted to say great work. Coming from an engineering view this is awesome! Absolutely an ingenious idea and very well executed. Thank you for the constant updates.

Thank you for your comment. I think I got here a little too late, but still all this information might be valuable to understand any other fuel system.
You may not be controlling the air suction in a FI system, but the injector's open/close time. At the end it is the same thing if one has the bases.

[Buzo]

I don't see a problem if you use the 3geez logo as long as you continue your project till the end

I will buy one of your diagnosis boards for sure.

You should make the thing "configurable" so one can disable sensors depending on where they live.

Like I said, Canadian Accords didn't get any O2 sensor for 86-87. This means the frequency solenoid A and B that alter mixture aren't there.

OK great! Thank you. Hopefully somebody has the 86-87 Service manual around here, it will help to understand the differences, but if not, as simple as you said, put a way to dissable inputs and outputs, like configurable switches in the PCB to select options like automatic/manual, etc, O2 yes/no and so on.

I will finish posting the control valves graphs I captured today, and the next step will be to cause controlled failures in the systems and see what happen, then find a way to catch it.

[lostforawhile]

Just one more... (for today)

This is the Oxigen Sensor. I increased the acquisition speed, so now it reflects better whats happening.
As long as it is at .5 average, the car and the systems are well carbureted, and should pass the smog test!

I have always wondered if a guy with a non-stock carburetor can use the O2 signal to make the carburation adjustments. Why not?

it's best to use a good air fuel meter and a wideband sensor for non stock carb setup, this is a very simple narrow band sensor, it's also not heated, so it's going to be inaccurate for a few minutes until it reaches operating temp. those sensors are designed to run almost red hot, this is the reason for modern heated sensors, the electric element heats the sensor immediately

[Buzo]

In case you guys wonder how the pass/fail decisions are going to be made, I'm including here the concept of the flow chart I have (still) in my mind.

The goal is to use more than one signal to define the current state of the car, so if one is missing the others will tell the OBDZ the truth.

So multiple codes could appear, each one separated with a LED off for 2 seconds. When only the the ignition is present the LED wil be blinking really quick so you'll know it is thinking, instead of a steady (software cycled?) LED.

[Buzo]

I have found a couple of real failures in my car so far. I should get my LED blinking if the project were complete at this time

1) Idle Boost Solenoid ON and OFF, so the idle speed rises to 1100 and back to 900 once per second. I have a problem with the Park shift indicator, it sometimes won't turn on after a trip. So it may be affecting it.

Its hard to notice but the parking signal has double line when changing from 5 to 0V. That's exactly when the false contact ocurrs.



2) I can't read the frequency of the ignition coil, even though the RPMS in my instrument panel is reading OK, the signal is changing too fast so can't measure the frequency. I know I have problems with the coil, so its time to buy a new one.



Those two issues are not a job stopper, so i'll keep going with the project while I fix them.

[2ndGenGuy]

I'm still following. This is totally badass.

[Buzo]

Enough testing!! I got enough information to complete this project. I know when a valve/sensor should be ON and when should not.

I felt some inspiration today and completed the front page of the software.
Sorry didn't resized the screenshots, but will lost some resolution.

[integratuner77]

Nice work on the VI. Its all coming together very well. I'm really interested in how this will turn out. I love all the data your collecting and find it very interesting. Keep it up.