Inconsistent AFR behavior issue

[BlackBird_SR71]

Here is a revised Tune based with Idle VE Cells locked out so Auto Tune can't touch them. Don't know if you had Auto-Tune enabled when this happens, but this will ensure it doesn't mess with the idle Cells ever.

VE cells don't change in the Data Log though... so this might be a Red Herring. Worth trying though.

2018-02-03 SR71 Idle VE Cells locked #3.msq

(118.77 KiB) Downloaded 32 times

[turbo conversion]

I also think it is something to do with the o2 sensor not reading correctly.

Afr stuck at 19.6 for longer than it takes the sensor to warm up is not correct obviously.

As mentioned before map pw rpm etc. are to stable for 19.6 afr to be correct for that long.

David

[dbyrd]

Agree that the O2 sensor readings seem off. Don't consistently see this, but then that seems to go along with the whack-a-mole list of items I have run into since starting the EFI conversion adventure. Not a complaint. It has been an interesting learning experience.

In my current msq I have the 3x3 area of cells around idle already locked. As BlackBird_SR71 had suggested in a previous post.

Here is the O2 Controller I am using. I have it mounted in a small enclosure under the dash.
https://www.14point7.com/products/spartan-2-oem

I had sent a note off to Alan with 14point7 about the behavior. He suggested trying temporarily connecting the controller to a different 12v source as a test to see if that resolves the issue.

The O2 controller is connected to the fuel pump relay which is standard fare on the MS wiring diagrams. However, given that the MS sends out a brief priming signal to the fuel pump relay. Then shuts off. Then starts again when it sees an RPM signal this could potentially cause some odd power signals to the O2 controller. Call it a hunch but suspect that the MS could also very briefly shut off the Fuel Pump relay during some save operations. The hiccup referenced earlier in an earlier post.

The 14point7 controller sends out two calibration signals. One at 13.3 and one at 16.6 during startup. These values are used to enter custom WB calibration settings in TunerStudio. The signals can be seen in the log just after the engine start.

After the second calibration signal the controller then typically starts sending the normal AFR ranges and not a full lean indication. There is a heater status LED hooked up on the Spartan 2 and shows if the O2 sensor is in the correct temp range. When the O2 sensor is in operating temp range it is solid otherwise it flashes. During my testing periods it pretty quickly goes through warm up to a solid indication.

[BlackBird_SR71]

I'm thinking the " Hiccup " just resets the EGO Delay function in Tuner Studio and the TS gauge jut goes to a default setting of full Lean. Which is 19.6 on your Spartan gauge configuration. Is that correct?

That would make sense as the EGO Warm-up delay is set at 30 seconds. From the time of starting to Data log reads correct is about 30 seconds. So I'm thinking that you are seeing a false DATA reading. Change the Warm up delay to 5 seconds and see what happens . This doesn't physically alter the O2 sensor delay. Juts alters when Tuner Studio will turn on sensing of WB controller outputs.

Keep in mind that the Tuner Studio delay, is only to ensure that the O2 sensor is fully warmed up and is putting out a stable signal. Bear in mind that any AFR readings while the O2 sensor is warming up could very well be inaccurate. 30 seconds in TS is just an arbitrary default.

When remote tuning I see the TS AFR go RED and Full Lean all the time. I've always taken this as a Warmup Reset of TS and juts ignore any readings or logs until it starts reading normally. I think this may be the same situation. You can't find a fault... because no fault exists. It's just a delay for the O2 signal to settle down.

I've logged my Audi with VAG-COM and you see the same thing. O2 correction is disabled for about 30 seconds until O2 sensor goes thru it's warm up period. Very common.

[dbyrd]

Well just finished going through testing of the O2 controller including temporarily disabling the priming pulse and connecting it to an alternate 12v source.

At this point pretty sure it is acting up. Have checked the voltage on MS O2 wire (pink) from the controller and it matches what I see in TS dash EGO voltage gauge.

One of the things I also tried was pulling the O2 sensor fuse for a few seconds power cycling the controller while the engine was running. Was able to get it to go from reading full lean to the expected AFR once. During this the PW, MAP and RPM readings were all consistent.

Checking about getting a replacement controller.

Appreciate all the help as I have been working through this. Will report back once I get a replacement controller how things are going.

David

[rickb794]

Where have you grounded the O2 controller?

[dbyrd]

Where have you grounded the O2 controller?

The 14point7 Spartan 2 OEM controller provides for an electronics ground and a separate heater ground.

The electronics ground is connected to the engine block along with the MicroSquirt and the IAC stepper controller. The engine block ground location is also the same attachment point as the main ground wire going back to the battery.

The O2 heater ground is connected to the chassis.

David

[rickb794]

The electronics ground for the Spartan controller should go to the sensor return on the MS. The heater ground should go to the motor.

[dbyrd]

The electronics ground for the Spartan controller should go to the sensor return on the MS. The heater ground should go to the motor.

I hooked up the controller based upon the 14point7 materials on how to get maximum accuracy.
https://www.14point7.com/blogs/news/165 ... megasquirt

Install and wire up spartan 2 according to the user manual here, On my MicroSquirt unit wire #34 (pink wire) connects to Spartan 2's linear output (green wire). Grounding is very important, Spartan 2's black wire (electronics ground) should be grounded where the MegaSquirt ECU is grounded, Spartan 2's white wire (Wideband Sensor Heater Ground) should be grounded to the chassis far away from where the black wire is grounded. The white wire carries upto 3 Amps and if grounds are not properly chosen, it can cause very large ground offset errors. The first line of defense against offset errors is proper installation and proper grounds, there may still be small offset errors present which we can easily compensate for in the next steps.

I can move the electronics ground to be connected with the other sensor grounds. Not following how moving the O2 heater ground would impact much. Are you thinking the issue I have described is because of a grounding issue?

Just trying to reconcile info from multiple sources.

[BlackBird_SR71]

Connecting the 14Point7 AFR gauge ground to the same Star point ground on the engine as the MS ECU is just fine. You have it wired correctly. AFR gauges are not the same as a sensor ground such as a TPS sensor, CLT sensor, IAT sensor etc.

An AFR gauge or WB is an entirely separate, self contained electronic device. same as you do not you do not wire any Solenoid ( Idle solenoid, Stepper Motor, Boost solenoid ) through the MS ECU ground circuit. The sensor circuit ground is for ECU powered +5v or +12v sensors only.

Wire it as 14 Point seven recommend with a Star ground on the engine for the Gauge and a separate Chassis ( or engine ground ) for the O2 heater circuit. Juts don't ground the O2 heater circuit to the same Star ground as the Gauge and MS point. There should be minimal voltage offset that way.

[BlackBird_SR71]

The electronics ground for the Spartan controller should go to the sensor return on the MS. The heater ground should go to the motor.

I disagree... WB Controller is not an ECU powered sensor.

[dbyrd]

Thanks BlackBird_SR71.

In the MicroSquirt hardware manual it does show a diagram connecting the controller signal ground to the sensor ground, but also says follow the instructions that come with the controller. I did the later. It made sense to me for the reasons you outlined.
http://www.msextra.com/doc/pdf/html/Mic ... .4-24.html

Hence my reconciling info comment earlier. Can make things confusing for those of us just starting out.

[BlackBird_SR71]

Did you ever try it with EGO Delay set to 2 seconds. If the time period that the AFR gauge in TS stays Red and stuck on 19.6 AFR suddenly shortens, then you know that it is just the default behavior of TS.

The other possibility is a bad connection of the Auxiliary channel to the MS WB input.

And of course there is the possibility of a bad WB controller. Although Allen's products are pretty darned good.

[dbyrd]

Did test out the EGO delay dropping it significantly from the default of 30 seconds. Didn't change anything in terms of the TS gauge display. Also included a EGO voltage gauge to the TS dash. Both show correctly the two calibration signals coming out of the O2 controller. Then depending on how the magic 8 ball lands either the AFR shows correctly or incorrectly.

In parallel I verified the voltage signal coming out of the controller with a meter. It also showed the save value I see on the EGO voltage gauge on the TS dash.

Based upon these two things would say that TS is behaving correctly.

I had checked the connections and nothing obvious jumps out. All indications are that the controller is intermittently sending out odd values after the calibration signals.

Agree on the quality of the 14point7 products. After extensive research, I landed on purchasing a controller from them. Alan has been very good to work with. The construction quality of the Spartan 2 OEM unit is very good. Very obvious he takes pride in his work.

Getting a replacement unit from Alan later this week.

[dbyrd]

Just wanted to post a follow up. Haven't had much of a chance to tinker over the past few weeks.

Have installed the new Spartan 2 controller and still experienced the intermittent issue. Additionally, replaced the Bosch 4.9 WB and no change either. Per a previous comment on the electronics ground this has also been moved.

The electronics ground for the Spartan controller should go to the sensor return on the MS. The heater ground should go to the motor.

Put a new set of NGK plugs in the motor as well. Hadn't changed these out since rebuilding the motor and them living through the Weber's fuel issues.

Net is still experiencing the behavior. Next thing to check will be a possible exhaust gasket leak between the header and head. The pace setter headers on the samurai have the O2 sensor in #1 cyl tube only. http://pacesetterexhaust.com/86-93-suzu ... aders.aspx The header was already installed on the car when I bought it. Granted it isn't black any longer and has a nice rust patina.

For grins I installed 2.920 firmware on the Microsquirt. It actually seemed to do better through several rounds of testing, but unfortunately that release won't support the GM stepper IAC controller. Motor started, went through ASE and WUE, but didn't idle down right because of the IAC being opened. Will put the previous msextra release back on it.

The whack-a-mole exercise continues.

[dbyrd]

Well after taking the old header off found a nice crack on the #1 cylinder tube. It was upstream of the O2 sensor and couldn't be seen or felt while on the car.

Installed a new header from Low Range Offroad http://www.lowrangeoffroad.com/suzuki-s ... esign.html. This is much better constructed, but also moves the O2 sensor to the collector.

Installed msextra 3.4.2 back on the microsquirt and reload the previous tune. Engine fired up fine and idled consistently.

After a few hours of engine restarts and driving around with autotune. Have been unable to reproduce the previous issues with the AFR readings.

Have to assume this ended up being a combination of a few things. Unequal fuel distribution through the intake from the TBI in particular to cyl #1 and #4. Coupled with the exhaust leak in cyl #1 that introduced additional air into the stream causing odd readings. Then resetting the microsquirt back to 2.920 and reflashing to 3.4.2 would remove any potential firmware corruption.

Appreciate all the help. Now time to put some miles on the Samurai. With spring time finally coming on can enjoy it more.

[jamies]

the issue of an engine sumble when burning settings, or changing a setting in a dialog box then opening another box or closing the existing one in tunerstudio is normal on ms2 / micro and nothing to worry about...