The times it starts off at 3 or 4 volts I didnt catch it on camera.

Thanks for the confirmation. I will check the sensors soon and adjust the afm if the sensors are good.

 

Your system seems electrically noisy and more erratic than it should be. Seems like it's working well enough to pass smog, but not up to Toyota's original design expectations. Could be still some bad wiring, bad contacts on the AFM or TPS, or the O2 sensor being flakey, or some other sensor delivering erratic data. Just speculating at this point.

I've been working on a computer program which reads the serial data stream from the ECU via the diag port. It's working now, but not quite ready to publish yet. In about a week I hope to put up a link to a zip file which will allow you, if you have a serial port (or USB to serial adaptor) on your laptop, read the real time ECU data. It provides updates every second on injector timing, ignition timing, VAFM position, TPS position, engine coolant temp, rpm, mph, and the status of a few flags and switches. It also calculates instantaneous and average mpg (not sure about accuracy yet). It will log everything to a time stamped file in Excel format for later analysis.

I'll let everyone know when it's ready. Might be useful to smoke out some erratic sensors.

 

RJR, that computer program would be quite interesting to use. I am looking forward to it!!

I purchased some aluminum heat shield wrap, which by nature doubles as shielding electrical noise shielding. I am going to do the following in the listed order (this list is so I don't get ADD while i'm working again!). If these don't produce fruit, I will wait for your program and see what that shows.

1) Swap O2 sensor with my old one and test while driving as before and record results. if results are very similar to the previous posted tests then the O2 sensor is most likely not the issue. If results are significantly different, then I will buy a new O2 sensor and test that.

2) Adjust VAFM to lean out mixture while meanwhile working on the following (this will force the mixture to be proper until I figure out if there is another problem).

3) If O2 sensor is good, I will check all of the other sensors for the 5th time and make sure they are up to spec. If any are out of spec, replace, test O2 system response again for erratic operation.

4) If above 2 steps don't show results, Wrap the whole harness in aluminum tape and see if that provides shielding from noise outside of the Harness (assuming no other shorts in the harness are present!) Also assuming that the individual wires that need shielding from eachother within the harness are properly shielded as they should be.

5) if non of the above steps work go on vacation. Forget about the system not working 100% perfectly and allow mind to relax about tricking the system via AFM adjustment

 

RJR, would that program be useful for estimating air/fuel ratio while adjusting the VAFM? I would like to mess with mine, but have been worried about going too lean and blowing up the motor.

 

Sruemcrow, I am figuring that I can lean out the VAFM and use the same type of tests with the O2 readings to see if the engine is running rich or lean. SO, if I lean it out too much and the engine can't compensate for the additional air then I will have the opposite readings I am having now thus I would adjust the VAFM back a click here and there until I find the O2 readings on 2.5 volts steady.

That's my plan in the short term. If RJR's program has a more straight forward way, obviously I would be drivign around with a laptop plugged into my car

 

Bit banging? My measurements show it isn't standard RS232, so it won't read through the regular UART. But it's so slow (around 110 baud) that a clever programmer should be able to pull it off. (Alas, if I were only a clever programmer ...)

 

Programming is definitely a weak point... Hate code language, can't remember any of it! I survived in college by writing the algorithms and my buddy writing the code (same buddy who did my harness... he was a better programmer!) lol.

 

It's actually 122.1 baud (8.192 mSec/bit). You are correct, a standard UART won't read it because of the odd baud rate and also because the first word of the string is only 5 bits long, and the rest are 11 bits (counting start and stop bits), so the UART gets all confused. Current PC's are so fast, though, that it's easy to parse it all out in code and pull out the information. I'll post the code when I get the package together. It's written in VB.net 2010 Express.

As far as using for tuning the mixture, I'm not sure the right data is available in the bit stream. You really want to be able to read the O2 sensor as well. I'm working on adding that, but that's a little more complex.

I am interested in seeing how cumulative mpg is affected by different things. Combining injector pulse width with rpm and mph gives a measure of fuel consumption per mile, and small changes should be fairly easy to spot if you drive the same route every day.

Stay tuned ...

 

I checked the sensors, rsistance wise they check out as before.

I pulled the #1 plug and inspected it. It looked to perfect. No indication of rich condition. So I need to understand what is the range of a good looking spark plug??

I set the afm 3 more clicks lean, total of 7 clicks. I went out for a ride and sure enough without the TE1 connected I was showing very near 2 volts on steady 2500 or so rpm accelleration. At this point I am assuming the engine overall with the orange flamethrower injectors and its already rich running stock setup is most of tje reason here. However I dont want to damage my engine. I am going to keep it at 7 clicks lean and run it for a couple of tanks of gas and measure mpg. Meantime this weekend after a few miles I will pull the plugs to confirm consistency among cylinders. I have a nagging feeling that one of the plugs isnt doing its job right!

But my main issue is I dont want to hurt the engine by running to lean incase the voltage readings from the O2 sensor are off for whatever reason. Meaning, now im reading 2 volts which says tevhnically I am still slightly rich, but I am afraid if that 2 volts is actually 3 or 4. Although the comparative fuel mileage seems to suggest rich condition as well.

What are other signs to look for short term to warn me of a lean running engine? Engine temp and spark llugs are all Iv got right now.

Note: my emgine has been running less than halfway on the temp gauge. Assuming it is accurate, is it the case that I am running slightly cooler than I should? Rich condition does result in Cooler engines... so it kind of makes sense...

I have officially put most of my current thoughts in this thread now

 

I would think you're just fine, and probably in a better place now than before. The chemistry of the O2 sensor is such that it's much more likely to lie to the ECU and make the ECU enrich the mixture too much, rather than the converse. The O2 sensor is basically a battery, activated by a mismatch in oxygen in the tailpipe vs. free air. It puts out 1 volt if the mixture is too rich (no leftover oxygen in the tailpipe), and 0V if it is too lean. Since batteries almost always fail by putting out less than their specified voltage (seldom more), a failing O2 sensor would tell the ECU the mixture is too lean, causing the ECU to enrich it. So, it would seem unlikely even a bad O2 sensor could cause the engine to run too lean - much more likely the other way.

It's a great confirmation of theory that adjusting the AFM spring tighter causes the VF1 voltage to move higher. I'm going to pull the top off my AFM soon and try the same thing. It's bouncing between 0V and 1 volt (which I think is partly due to operating at 5000 feet elevation), so some adjustment seems warranted to try to center things up.

I doubt you'll see much difference in coolant temperature lean vs. rich. Coolant temp is largely controlled by the thermostat. It is true that, up to the point of misfire, combustion temps are higher with leaner mixtures, but you really need an exhaust gas temperature probe to see that. The higher exhaust temps can be an issue with airplane engines, for example, where they run at 75-80% power continuously. Not so much with automotive engines, where normally you're running the engine at less than 25% of full power, around town closer to 10-15%.

 

Here is the plug.


And video of the O2 sensor reading. uhumm.. motion sickness warning!

 

Thanks for those thought Ron,

the more I learn about this O2 sensor the more I am impressed by it.

I understand now that a failing O2 sensor would most likely enrich the mixture. I will still be keeping a close eye on the plugs to make sure things are well.

The AFM adjustment is really easy, and it is nice to see the effect so clearly

 

Update. Been driving a little more with voltmeter on. I notice that it goes to 2 amd sometimes 2.5 volts under any acceleration. However once I hit steady speed but not idleing (still rpm around 2500) the voltage drops to 1 about half of the time.

Also upon cold start it wasnt as immediate as before, so I plugged my CSI back in thinking it probably was running rich therefore CsI was superfluous but now its needed....

I am going to do the wiring shielding this weekend without any other changes to see how different it behaves.

 

another update:
Engine didnt feel right, I unplugged CSI since it was only changed variable today. Engine went back normal. (of course this is engine at operating temp). Drove some more and O2 sensor readings were back to a 'more' normal case. I was getting some erratic info but mostly it was consistent. This tells me CSI is open during normal operating temp.

Am I correct in saying that the CSI switch is most likely at fault here (although I've tested it per fsm like 6 times)? Does the engine temp sender effect the CSI circuit through the ECU at all? Does the CSI switch influence the ECU's fuel mapping at all? I would assume not since they have specific sensors for each circuit.. but....

Also, filled up this morning to start fresh tank on new AFM setting, and I was at 14.2mpg. Which makes sense because this tank i drove a little less aggressively.

RON, will your program show the CSI circuit? This program will be an awesome tool!

 

The program shows the state of the CSI as the ECU sees it. On mine it turns off 5-10 seconds after starting. Also what will be useful is the ECT reading that the ECU is seeing. I don't think ECT affects the CSI, but it will affect mixture control.

 

That's awesome. Big thank you for making it available to us!

Gevo -- if you're trying to tune by your 02 sensor, you're very much ahead to use a wideband over the narrow band stock readings for a variety or reasons, foremost among them precision. With narrow, you get readings from 0 to 1 volts. With wide, you get readings from 0 to 5 volts.
Read more here: http://forum.ih8mud.com/80-series-te...r-summary.html
And here: http://www.autometer.com/tech_faq_an...x?sid=1&qid=48

 

RSR, that was an interesting read.. furthering my knowledge on O2 sensors Thanks. So, by your first sentence and the fact that the Denso O2 sensor I have is giving voltages in the 0 to 4 range, I have a wide band air fuel ratio sensor..

But The problem that is nagging me with having to force the mixture by afm adjustment is that the 02 sensor is sending the correvt signal, so wjy isnt the ecu adjusting to stoich???

 

Brake and light test results- dash lights not coming on..??? Rear wheel cylinders bad, marker and license plate lights need to be fixed. Pretty minor stuff, fixed license plate lights yesterday, can't figure out why markers don't have power yet but I will work on it today. I suspect the dimmer switch in the dash is bad but that will be tomorrow's project.

 

Gevo,
Have you tested your temp sensors... both coolant and iat?


my 22re last month blew a head gasket, and the first few days before it turned into a 'easily diagnosed issue'.. I had a wildly fluctuating engine. I imagined the bubbles going past the sensor was causing problems, not to mention the 'steam' going down the tailpipe was instant death to the O2 sensor.
.....not saying you are witnessing a blown hg, just the issues that a faulty temp sensor can give.. your rich conditions sound like it's reading 'cold'???
btw, you 'hard wired' your flame throwers? (no adapter?) cut the original plug off, and SOLDERED the correct plug?

 

Ramcherman, yes the new flamethrower plugs are soldered on to the harness. The existing plugs werent all in good condition.

I may just change the temp sensors again just to rule them out for good. Ill wait to see if the various other tweeks dont get mee the results I am looking for.

As far as news... I connected the analog voltmeter to the Cold start injector connector and was looking at it while in the car starting. Contrary to my previous thouths... it only spiked once for a second on a cold start. I willl test it at other cold starts intermittently to get better spread of data.

I shielded some of harness this weekend and funny thing.. its helped the erratic behavior of the O2 sensor reading. I also tested the stock shielding of the sensor wire and found it to have 2 ohms of resistance to the battery's ground. Other grounds are .3 to .5 ohms. So I added a temporary ground to see if Notice anything different. Ill remove it later to see how voltage from o2 sensor behaves.