lakeffect2

I'lkl be adding a Bosch wideband sensor kit from Glowshift Gauges soon.

While I have the 7730 ECM out and accessible, I'd like to repin whatever is needed to get the signal into the 7730 on an &8D mask. I'm also using the Super_AUJP_ V4 file as well. I found lots of scattered piecemeal info on a search, but didn't see a coordinated thread.

I've set up a wire pinned to F4 on the ECM.

Links appreciated.

No doubt there is some code to add to the files which I've never done before. I only made changes to data like setup, spark and fuel which already existed.

Knowing many here have done this before, can anyone walk me through this? (be gentle, I have a broken foot)

DENN_SHAH

iTrader: (1)

with S_AUJP_V4 i believe you'll want to use F14 or A3.
down near the bottom in flags/switchs look for "WB OPT 2, bit 7, Input "A3", Else "F14"

as for which WB O2 to select in TunerPro, im not sure which one you'll need. you may have to make up lookup table for it.

there is some additional info in the document files that come in the S_AUJP_V4 package

lakeffect2

As I dig into the s_aujp_v4 documentation, it appears that either pin B3 or F14 might be already set up in the ADS file.

Can anyone else confirm?

JP86SS

iTrader: (1)

Its all setup for you already.
Use F14, some have had issues with the A3 line input.
select/unselect the bit for that selection.

Easiest thing is to use the "Raw" type and set the scale in the ADX to match the WB since its output is linear.

Setup the WB output for 0-5 volt range of 7.35 to 22.39:1 AFR
The total span of the AFR is 15.04 AFR (22.39-7.35)
Setup the ADX "Value" equation for 15.04/255 counts = X * 0.05898 +7.35
7.35 to 22.39:1 AFR will be displayed.

lakeffect2

Thanks! JP to the rescue again!


Once the unit arrives and it's put, in I'll post up with success or issues.

lakeffect2

I found the flag/switch marked "WB opt, bit 7, input A3, else F14"
The bit selection box is blank, not checked off, My assumption being it is defaulting to A3 as an input source

Since I want to use F14 as my input, am I correct in thinking I must now put a check mark in the box to change the default over to F14?

I purchased a Glowshift WB with a bosch 4.9 sensor. It is hooked to to it;s controller and gauge so far, and is displaying well. The last step is making the ECM connections. My headers had an extra Bung available, so I currently have both WB and NB sensors in place.

I was wondering if it is a problem to simply leave the NB in place and still connected. Since I am Using a Prominator with 8 bins available, I thought of keeping one bin selection available using the old narrow band sensor as a "plan B" in case the WB craps out. Any thoughts?

Dave

JP86SS

iTrader: (1)

You would leave the box "un-checked" for F14.

That is the best way.
Let the ECM use the NB as usual and have the WB for reference as to what is really going on. That's how mine has always been and I like it that way.

lakeffect2

Thanks for the info, although it takes me by surprise that your ECM is not using the real time wideband data on a direct basis. I thought that was the whole point of having a wideband system in place.

Granted it's nice to have for review of datalogs, but why wouldn't you want that data to be used directly by the ECM in real time for control purposes? Did I miss something?

DENN_SHAH

iTrader: (1)

the 7730 with 8D can't use the 0-5 volts from the wideband for fuel control.
if the code was majority changed, i believe it can be done. but i think the code would have to be pretty much completely rewritten from scratch which would be a huge job.
it seems like i read that code 59 can use a WB for fuel control, but im not sure.
i only have 1 O2 bung in my exhaust, so i have the NB output from my wideband controller hooked to the ECM O2 sensor input wire.
when i got the tune working pretty good, i put a new stock O2 sensor back in and the ECM added a ton of fuel. i thought i may have gotten a bad o2 sensor, so i got another one and had the same results.
i don't know what the deal is, im thinking a grounding problem somewhere when running the stock sensor. if i had a second bung and had been running the motor off the stock sensor and tunning with the WB, i would have known about the problem.
i want to get my headers put on some day and add another O2 bung on the passenger side. with headers i know i'll need a heated O2 sensor so i haven't really looked into what the problem might be.

JP86SS

iTrader: (1)

I actually have them both in the drivers side collector.
I put a bung in the passenger side just in case I wanted to move it around.
The NB works good for part throttle fueling so I just want to see what the effects of shifting the o2 tables and such do to the operation.
I still have a gain problem in that I have a rolling AFR now that I need to get smoothed out. (before I swap in the 406 I just built )

ULTM8Z

Sorry to ressurect an old thread...

I have my AEM WB 0-5V output wire hooked up to F14 on the ECM, but can't for the life of me get it to show up on the TP dash.

I'm using the S_AUJP V5 ADX file, and have tried setting the AEM option bit b6, but that didn't work.

My b7 bit is not set, so from reading above, it means ECM should be reading F14.

I started to try JP86SS's suggestion above, but can't figure out where to enter this equation...

84Elky

You have the correct settings for the bits (b6 set=AEM, b7 NOT set=Pin F14).
A couple of questions:
1. Is WB data in your logs and just not showing in a dashboard?
2. Are you using the ...D_2 WBo2 Innovate LC-1.adx file found in Post #1 here: https://www.thirdgen.org/forums/diy-...available.html
If so, in the values area, just change the name of the "WB AFR (LC-1 0-5V)" value replacing 'LC-1' with 'AEM' and then add that value to a dashboard.
That should cause everything to work. You shouild see 4 different WB values in a log (raw with JP's formula, raw +10, raw, and raw * 0.39..... JP's gauge. If you are not seeing those in a log, hopefully JP can chime in to assist.

ULTM8Z

Hmmm.... I must be seriously dense or something

Here's what I'm getting...

1. I'm not seeing it in the logs or the dash board

2. Yes, I'm using the ...D_2 WBo2 Innovate LC-1.adx file from that thread...

Here's the process I'm using...

Modify bin to set the flags (bits) per what I did above.

Burn chip with bin

Turn on ignition and connect up tuner pro. I'm using a AA (1.5V) battery hooked up to the F14 input (+ to F14 and - to ground). This is basically to conveniently simulate an output from the WB. According to AEM instruction sheet, 1.5V corresponds to ~13:1 AFR.

The dash gauges are basically stuck at 7.35V and 10.00:1 AFR. When I export the log file, I get the same thing.

JP86SS

iTrader: (1)

Use the "Raw setting" and see if your input goes up and down.
The engine must be running possibly to get the reading to indicate when you are simulating the input. Used to be anyway.
Could try the alternate inputs with S_AUJP V5 to see if the channel on your ECM is not working right.

Need to see the input move with the simulated signal, If not then you do not have it wired into the input correctly, Please recheck that the grounding is connected good too.

ULTM8Z

Thanks JP. I'll check all that again. The car was not running when I was simulating the input. I haven't had a chance to do anything since I last posted on this. I'll see what I can do after Christams.

ULTM8Z

Ok, I'm still at a loss here... had some time to play around with this.

Engine running... re-verified I have the white wire (WB 5V output) connected to pin F14.

3 different ECM's, no results.

Tried every WB bit option available in TP, no results.

Attached is my grounding scheme for the WB. A DMM on the white 5V output wire and anywhere on chassis ground shows the WB is indeed producing the voltage output with respect to chassis ground.

Just the ECM is not processing it through to TP.

ULTM8Z

Just tried pin A3, but still nothing...

ULTM8Z

I started looking at this again now that I have the PE mode straightened out.

Still having the same issue...

Either Tunerpro isn't reading the ECM output, or the ECM is not recognizing the WB input.

I'm using the adx file S_AUJP v5 2030_D_2-WBo2 Innovate LC-1.

Any thoughts?

Vanilla Ice

Can't read through all this but there's steps to getting it to work.

1. Hook up wire to ecu pin.
2. Change code to read pin and store it at ram location.
3. Change aldl output to include ram location of AFR.
4. Change ADX to read that ram location and convert it accordingly.

I don't know this hack version as I created my own but in my skimming I didn't see step 3 mentioned yet.

ULTM8Z

I'm using the SAUJP adx file with the WB mod. Supposedly from what I've read, all you need to do is turn on the option for your type of wide band and (if you're using pin F14, you don't set the pin number option).

RBob

iTrader: (1)

Need to use the S_AUJP code (ECM firmware) as it has the code changes required to read the pin. The stock AUJP GM BIN won't work for this.

RBob.

Vanilla Ice

If these guys would let me see their backend work I'd be able to help. The problem is they won't share so I'd have to reverse engineer their work. It should be a simple problem being overlooked.

ULTM8Z

Ahhhhh.... now we're getting somewhere...

Ok, I'll try that today. If it works then I guess i need to build a new bin based on the SAUJP.

Thanks!

Vanilla Ice

Step 2 & 3 in my list above!

ULTM8Z

Except I'm too code-illiterate to grasp what you said.

Vanilla Ice

That's totally fine, if you didn't understand the step just ask! It's easy to explain this stuff.

ULTM8Z

Success! The SAUJP bin works with either pin F14 or A3.

THANKS GUYS!!!

84Elky

Given you are using the D_2 ADX file, I assume you are also using the D_3 XDF file available in Post #1 here:
https://www.thirdgen.org/forums/diy-...available.html

It's important to use the D_3 xdf file because as stated in Post #1 above, the D_2 version omitted in error the entry for setting the Pin on which the WBo2 sensor is to be read. The BIN defaults to 0x99B b7=0 (=Pin F14), but there was no way via the D_2 XDF to set Pin A3 (b7=1) if desired.

ULTM8Z

Yes, I have the D_3 XDF.

ULTM8Z

Ok, so the good news is I'm logging data with the WB output being read by TunerproRT! Woo hoo!

The bad news is it's not reporting what I"m seeing on the WB gauge itself.

When I get into PE, WB gauge is reading about 12.3:1. When I go back and look at the Tunerpro data, it's showing around 15.5:1.

In fact there appears to be on offset of about 1.5 to 2.5 AFR points higher than the commanded AFR across the board (the offset seems to vary, but I'm not sure with respect to what.

Is there some calculation error that's occuring when converting volts to AFR?

ULTM8Z

Looks like the conversion equation is X * 0.058980 + 7.

The WB raw data is coming in at around 140... so given this equation, it makes sense that I'm seeing in the 15.5 AFR range..

Now the question is, why is the native input way up around 140? It should be around 100...

Is there some sort of ground offset in the voltage? I have the ground hooked up to ECM ground... The O2 sensor ground potential is the same as every other ground, so it doesn't seem like it'd make any difference to hook it up to that...

Hmmm maybe I just edit the equation such that the Tunerpro output matches what I'm seeing on the WB... figure if I get within 0.1 AFR it should be good enough for my purposes... The equation is linear, so it should just scale up and down I suppose...

Vanilla Ice

That voltage offset is wrong.
It's +7.35.
Don't know the scalar offhand so can't comment on that part.

Wideband should match narrowband at idle. If it does then your gauge is wrong. (Excluding other issues with your narrowband)

ULTM8Z

It was a copy paste error... I didn't grab the whole equation... it is in fact 7.35 in tunerpro.

I'll check NB vs WB.

84Elky

Believe the Innovate 0-5v ranges are:
0v = 7.35 AFR
5v = 22.39 AFR
So A/D AFR conversion to voltage steps for 8 bits would be:
(22.39 - 7.35) / 255 = 15.04 / 255 = 0.0589803 AFR per voltage step
or some prefer using 2^8=256 steps = 15.04/256 = 0.05875

So the conversion would be: X * 0.05898 (or 0.5875) + 7.35

The conversion is correct if the above AFRs at voltage are correct. But that does not explain the difference between gauge and converted voltage. Given the conversion formula is correct, I'd tend to believe the raw converted value than the gauge unless there is in fact a ground difference.

ULTM8Z

There's no way the car is going to pull as hard is it does with a 15.5-16.0:1 AFR in PE mode. In addition, the WB gauge seems to track the commanded AFR within about 1/2 point.... so I definitely believe the gauge when it tells me in the 12.3:1 range.

I'm running a AEM gauge... 0-5V as well.

When I look at the Tunerpro data, the NB AFR is pegged at 14.0 and doesn't move. I'm looking at the Channel 3 NB AFR (assuming htat's the one that Vanilla Ice was referring to).

Edit: Looks like the AEM gauge reads from 10:1 to 18.5:1, over 0 to 5V.

Doing the math from Elky's post, I get a new conversion of X * .033 + 10.0.

When I enter that in for the conversion formula, it drops the reported AFR by about 1 pt. But it still doesn't track what I'm seeing on the gauge.

The wierd thing is the delta between commanded and reported gets bigger as the commanded gets lower. I.e., at commanded of 14.5:1, I get a delta of about 0.5 point... but when the commanded goes to 12.75:1, the delta jumps to about 1.5-1.75 points.

ULTM8Z

Ok, so it's definitely not the gauge.

I hooked up a volt meter to the 5V output on the gauge with the ground hooked up to the ECM ground.

With ~13.5:1 reading on the gauge, the volt meter reads ~1.8V, which is smack dab right where it should be according to the curve provided in the AEM instructions.

So then for some reason, Tunerpro is processing the signal incorrectly or something?

To recap on my setup, I have the 5V output on the gauge plumbed into F14 on the ECM, with the ground going to chassis ground (same as ECM ground).

I have Opt2 b7 unchecked, and Opt1 b6 checked for AEM linear.


Maybe someone can take a look at my adx file? I had to change the file exension to .bin so I could upload it here... when you download, simply change it back to .adx.

Vanilla Ice

If the bin is set correctly for that wideband then it's probably your ADX. Post it.

ULTM8Z

I posted the adx in my previous thread.

Basically it's the (I believe) latest SAUJP adx file set for the AEM WB...

84Elky

ULTM8Z -- The file in Post #36 has a .bin extension but it's an ADX file. If you'll post the actual bin, adx and xdf being used and causing the problem, I'll take a look at them.

ULTM8Z

Yeah it wouldn't let me upload an adx, so I changed the extensiom to bin in order to upload it. You'd have to change the extension back to adx upon download.

​​​​​​I'll put everything into a zip file and the post the zip file.

ULTM8Z

Ok, here it is...

adx, xdf, and bin I'm using.

RBob

iTrader: (1)

If you are setting option bits in the BIN, then the ECM firmware is doing part of the conversion. Since you are getting the value of about 140 in the data stream while the gauge is at 13.5, my guess is that the ADX needs to do a divide by 10, giving 14.0:1 AFR.

But I'm also confused about which WB setup you are using. Supposed to be an LC1 with an AEM gauge? Too much mix & match here. Also, the ECM doesn't go to chassis ground, it goes to the engine block.

Is the LC1 output set up to match the AEM gauge input? Or is the AEM gauge input set up to match the default output of the LC1?

RBob.

ULTM8Z

I'm using an AEM gauge. The SAUJP adx file has several option bits available for different WB gauges. I selected the AEM which runs on 0-5V range over a range of 10:1 to 18.5:1 AFR. The only thing is (as noted a few posts up) the formula employed in the as-downloaded adx appears to be incorrect for the AEM AFR range. So I came up with a corrected formula based on the math from Elky's post. But that didn't really seem to do the trick.

For the grounds, when I back probe the O2 sensor ground to the ECM ground (which I agree is hooked up to my engine block), I get a ~.5 ohm connection. So electrically right now, there is no difference between those two electrical nodes.

ULTM8Z

Ok, problem solved by 84Elky! 84, please add more detail if you think it's needed... I figured what I pulled below from your e-mail was the gist of the solution...

Here's what I did per an e-mail he sent to me...

The WB gauge and the reported AFR are within about 1/4-1/2 pt of eachother now. But I know I can fine tune that by fine tuning the formula within TP. The main thing is that TP is actually reporting the WB AFR now (and accurately, at that!).

Hats off to 84Elky who spent a lot of time with e-mails back and forth and pouring through data logs, adx and bin files, etc!

Thanks 84!!

https://www.thirdgen.org/forums/newt...newthread&f=16

84Elky

Glad all is working and you now have similar AFRs (gauge and TPro). The key here is to specify use of raw WB sensor voltage counts and NOT specify the type of WB being used which causes S_AUJP code to compute the AFR. Then all that's needed is to tweak the A/D conversion formula so everything is in synch.

For all S_AUJP v5 users using the supplied WB ADX file, it is clearly labeled as being applicable only to the Innovate LC-1/2 product. If another WB is used, always best to use raw counts from the sensor and convert them in the ADX with the manufacturer's formula rather than use the S_AUJP code to do the conversion. A lot of time was devoted to writing AFR calculations, but it is unknown how much in-car-testing was actually done on other than Innovate. ULTM8Z's experience clearly shows that the code's AEM and PLX WB AFR calculations are incorrect.

ULTM8Z

Ok figured I'd post this up in case anyone else is running an AEM gauge and TP RT with the WB option on the SAUJP adx file.

I fine tuned the formula for the AEM WB and now I'm within about 0.1 points on AFR reported by TP vs the WB gauge across the entire WB range.

What I did was hook up three AAA batteries in series to get a 4.6V source (as close to 5V as I could get). 0 to 5V is the voltage output range of the WB gauge.

I than grabbed a throttle position sensor and wired up the 5V input to the 4.6V source (since the TPS's are meant to run on a 5V input anyway).

I then hooked up the TPS output to the to the WB input on the ECM at F14 and hooked up the ground wire to chassis.

Basically I spoofed the WB gauge with a potentiometer so I could plot the voltage vs AFR in terms of how the ECM is telling TP what it's seeing.

Then fired up the car and TP, and simply started mapping the voltage output from the TPS (using a DMM) as I moved the lever, and then reading off the counts as reported by TP WB AFR Raw Data (in counts).

So yeah, what that did was tell me how TP is reading and reporting the ECM output for every voltage input from the potentiometer. Basically allowing me to map it without having to drive the car.

Looks like the formula works out to be (X*.038)+10 for the AEM gauge.

Should be able to get my AFR's dialed in with very high fidelity now.