[Snakeibf]

I confirmed the O2 sensor on the MSX is indeed a zirconia type Narrow band sensor and does put out a DC signal.

Range is 0.0v at full lean at room temperature and i got .96v at full rich at 600F. It did not register a frequency on my meter, but I don't have an Oscilloscope so I'm limited to around 1000KHz.

This means a regular resistor probably isn't going to trick the ECU.

Here you can see the voltage output when the lighter flame is simulating a rich condition.

So now its time to work on an ECU emulator circuit, but with a constant .5V output.


Sent from my iPad using HondaGrom.net

 

[Snakeibf]

Making progress, i just need to make a voltage divider now.

Sent from my iPad using HondaGrom.net

 

[Grom'n-BR]

lol i'm so happy your the determined mad scientist with this stuff!
bc some of this stuff is way over my head but I get the jist of what your doing

 

[Grom1202]

Have you had a chance to test the bazzaz ground wire o2 eliminator/make one .. i want to see what stock ecu does with a/f when o2 signal wire is shorted to ground..

Thank you.

 

[Snakeibf]

With the resistors i had on hand I was able to get pretty close to the .5v i was aiming for. Ill just add a variable resistor so i can adjust the voltage offset. This will trick the ecu into thinking the o2 sensor is there and running close to stochiometric.

In the picture I used a 5v regulator to begin which will give me a constant 5v even when the bikes voltage fluctuates. The capacitors prevent high frequency oscillations and smooth the power to the ECU. The voltage divider is just two resistors in series which drops the voltage down to the .45v this will connect to the O2 sensor wire. It only draws 50 milliamps with the resistors i used. Once i solder it permanently to a PCB ill connect it to switched power and add an LED so I can see when its on.

Sent from my iPad using HondaGrom.net

 

[jvalason]

On the road to figuring this ECU out! Great job man :cool2:

 

[Snakeibf]

I believe Bazzaz is just sending a wire to connect to ground to close the circuit. This will give the ECU a 0 voltage reading and may still work,but the ECU will assume the bike is running full lean and probably try to add fuel which is ok you will just have to adjust the map in the Z-FI to account for it. With this setup ill be able to control a wider range of variables.

Time to sleep ill continue tomorrow...

Sent from my iPad using HondaGrom.net

 

[Grom1202]

Awesome Info.. Thank you for sharing...

 

[THE TECH]

Good stuff as usual Snake.

 

[CiscoSanJose]

With the resistors i had on hand I was able to get pretty close to the .5v i was aiming for. Ill just add a variable resistor so i can adjust the voltage offset. This will trick the ecu into thinking the o2 sensor is there and running close to stochiometric.

In the picture I used a 5v regulator to begin which will give me a constant 5v even when the bikes voltage fluctuates. The capacitors prevent high frequency oscillations and smooth the power to the ECU. The voltage divider is just two resistors in series which drops the voltage down to the .45v this will connect to the O2 sensor wire. It only draws 50 milliamps with the resistors i used. Once i solder it permanently to a PCB ill connect it to switched power and add an LED so I can see when its on.
View attachment 7690


Sent from my iPad using HondaGrom.net

are you using a 9 ohm resistor?

 

[Snakeibf]

No, R1 was at 98 (100) ohm and R2 9.9 (10) ohm to be precise the lower readings are due to the 5% tolerance of the resistors. A 9 watt resistor on R1 and a 1ohm at R2 would have also gotten me to the target voltage, but then would have drawn over 2 watts so not very efficient for what I'm doing here and would have probably required a heat sink on the regulator. Since there should be no additional load when i hook up the o2 wire this should be sufficient in theory, but ill test it out this evening.


Sent from my iPad using HondaGrom.net

 

[Snakeibf]

I will be building another one using a 555 chip so it outputs an oscillating signal using a schematic I found, but using the same values as the voltage divider and adding a variable resistor. This will not only work on the grom, but on pretty much any vehicle with a narrow band zirconia O2 sensor. It looks as if the ECU averages the input of the voltage fluctuations when the sensor is in the exhaust.

Sent from my XT890 using HondaGrom.net mobile app

 

[Snakeibf]

I just finished the new improved circuit for the O2 sensor emulator. Im sure some of you are wondering what the hell I'm doing and why this is better than a simple resistor. Here is a picture of it working on the test board. Ill be transferring everything to a prototype board and it will be much more tidy once its done. The blue LED is just showing its on and the red LED shows the output signal is on. This circuit only draws 30 milliamps and creates an ideal oscillating pulse just like the stock O2 sensor.

This is not just some crazy thing i dreamed up, these are commercially available. I already had all the components in my kit so i though why not just build one.

Sent from my iPad using HondaGrom.net

 

[Snakeibf]

The unit is on the prototype board. I just need a variable resistor for R4 so i can adjust the output, but it is working good and its ready to try on the bike after i cut the board smaller, but unfortunately not today. Ill explain more later, but i have go get up early so its time to sleep.

Sent from my iPad using HondaGrom.net

 

[Snakeibf]

On the attachment the Narrow band output graph illustrates the output of the O2 sensor relative to the A/F mixture. When the oxygen sensor reaches operating temperature it generates a voltage, the voltage on a narrow band sensor puts out between .1 and .9 volts, but at an oscillating pulse most likely because of the pulsating exhaust gasses in the manifold. You can prove this by holding a propane torch or lighter to the sensor end and heating it. With a multimeter attached put the flame to the sensing element. The flame consumes the oxygen around the sensing element and heats it. After a few seconds the voltage will slowly rise until it peaks to around .96 volts. When the sensor is in the exhaust the exhaust pulses cause a fluctuation in the signal the ECU then averages these to come up with the current A/F ratio and changes the Pulse Width of the injector to either add or reduce the fuel being injected. With the simulator one can choose what A/F ratio you want to the ECU to think it's running at by changing the output voltages. This will hold the signal steady at one voltage so the ECU doesn't lean or Richen the mixture. Basically it will think the fuel being burned is staying the same and won't try to adjust the PWM. This is good because then you can tune it with a piggyback tuner and when it goes closed loop it won't lean your mixture back out.


This video is a bit weird, but does a good job of explaining how the simulator works:

If you have some basic soldering skills, can read schematics and an understanding of electronics it's easy to build.

If you want to buy the stuff to make your own check out:

Jameco Electronics - Electronic Components Distributor
or
www.digikey.com

If this prospect scares you or makes you go cross eyed just buy one here. These units look pretty nice because they use all surface mounted components.

http://www.o2simulator.com/o2-simulator-c-76/universal-o2-simulator-singledual-output-p-184.html

It's really easy to connect, one wire goes to the positive terminal on the battery post, one on the negative and the output goes to the O2 sensor plug.

Note: This thread is only for information purposes. I have no affiliation with any of these companies. What you choose to do with the information is up to you.

 

[Snakeibf]

Tomorrow comes the first running test to see what the A/F ratios do with the unit hooked up, next week I'll have the trim pots I ordered so I can adjust the output voltage and later, once I get all my other engine parts, I'll be able to test this and compare it to the wire that Bazzaz is sending. I'm very interested to see how this "resistor wire" that nobody seems to have seen yet will affect the ECU or if it will throw a code 21 eventually. I'm assuming they did plenty of testing and I would imagine it will work coming from a reputable company like Bazzaz. At 0 volts at the O2 sensor input i'm guessing the computer will assume a full lean condition, because the sensor at it's cold, open air state puts out no voltage. From what I've read the ECU only goes into closed loop when under low load at cruising speeds. I guess only more testing will tell us in the end, but either way it should be really interesting.

 

[HardRacing]

.

So your confident that simply grounding it out, is not a resolve ?


BTW, very impressive knowledge of how all this works.

.

 

[Snakeibf]

.

So your confident that simply grounding it out, is not a resolve ?


BTW, very impressive knowledge of how all this works.

.

I'm not saying that necessarily, but if it doesn't work this may give us another option. It will also just allow you to adjust an additional variable. I believe the voltage offset circuit on the EJK is similar to this, but it's just integrated into their unit.

If it were as easy as adding a resistor i think Dynojet would have done that on the power commander.

 

[FXCLM5]

wow did bazzaz go down the wrong road again?

only time will tell if this guys setup will actually make the ecu accept the programmed fuel maps with the o2 disabled..... and not try to fix it.

hopefully your solution will work with any aftermarket fuel box controller.

 

[HardRacing]

I'm not saying that necessarily, but if it doesn't work this may give us another option. It will also just allow you to adjust an additional variable. I believe the voltage offset circuit on the EJK is similar to this, but it's just integrated into their unit.

.

Cool.
Regardless, keep up the great work.

This is Very interesting stuff.

.