I am not sure if anyone else is doing this...but i just got off the phone with someone from Innovate Motorsports and after talking to Craig Moates this is what i have come up with.

Instead of Using a separate datalooger with a WideBand, like the LM-1, it is possible to use the LC-1 WideBand (use the following link)

http://www.innovatemotorsports.com/x...cat=253&page=1

Then using a hacked bin that is able to read from 0-5 volts like one that can be found on this website:

www.eecis.udel.edu/~davis/z28/WB_hacs/

It is possible to just replace the stock O2 sensor for the LC-1 WideBand O2 sensor, and hook it up to the ecm so that TunerPro RT will be able to datalog everything.

What do you guys think, anything wrong with this set-up, what about the stock o2 sensor wire? Plus the LC-1 is only 200 bucks.

 

Looks cool. Definatly a nice tool for people who just want the basics without all the bells and wistles.

Id still just stick with the NB rather then try to use a simulator or anything like that. The ecm is calibrated for its non-linear output as the AFR moves away from stoich. Not only that but I wouldnt want to subject an expensive WB-O2 to constant wear and tear.

 

I recently ordered a new replacement sensor (Bosch LSU 4.2) for my LM-1 after running it for over 40k all-weather continuous miles in simulator mode just to see how it would work and how long it would last. I must say that even in simulator mode it performed better than many other NB sensors I have tried. I also purchased an AC/Delco AFS 74 today to install in place of the WB experiment and paid $5 more than the LSU 4.2. Not too bad in my book. JMHO

 

Theyre that cheap? Well, I guess it makes sense since its made by bosch. Ive never had any luck with any of their products so far. Hopefully my WB wont be the same way.

 

That looks like progress. I'm impressed with the price of the botch, I mean Bosch sensors . Not a fan of them but you can't beat the price!

 

That’s a great price and the handheld display is not needed when you can use a scan tool with the WB patch.

The only problem I see is that I think the WB patch found at www.eecis.udel.edu/~davis/z28/WB_hacs/ is made for the DIY-WB which is not linear and I think the LC-1 is linear so the reading would be off.

 

I have never been a fan of Bosch sensors either and until now I have only had negative experiences with them, but that is what my LM-1 came with hence my putting it to the harsh test and comparing it to other units on the market. While there are better WB O2 sensors out there and the LM-1 will accomodate the LSU 4.2 and the NTK L1H1, the LSU 4.2 performed above and beyond my expectations despite the bad reputation on this board. In my mind, as mentioned, I could not beat it for the price, so I ordered another one. HTH

 

Does anyone have a WB patch for a $6E mask that would work with the LC-1?

 

That's exactly what I'm doing with my new setup. Have an LC-1 along with the XD-1 gauge pod, with the simulated NB signal going into the ECM (730) instead of the regular O2, and the 0-5v going into the ECM on an unused port.

As far as the 0-5v linear, that's an easy thing to do. I've done it on the 730 and 165 with great luck in both cases. Gotta do a little code hack of your own. Do a little search on here and both techniques should turn up. If you can't find them, let me know.

 

Do you have the 165 $6E bin hac that will work correctly with the LC-1 WideBand, preferably the ARAP. If not, do you know of a link where I could get it. If the Davis wb hac is only for the DIY WB, which is non-linear, it won’t work with the LC-1. Also, do you use the XD-1 gauge pod just for visual verification?

 

Mike Davis' hac mentioned in the first post will not work with the LM-1, LC-1, PLX, etc. It will only work with the DIY-WB. It was programmed specifically for the curve of the DIY-WB. Innovate's stuff has a different 0-5V curve.

Tim

 

Craig, I searched for the hac methods that u mentioned but could not find them could you please help!!!!

Thank you!!!!

 

Does anyone have the bin that will work with the linear LC-1????

 

TTT

 

i would like to see this "patch" or bin that would work with the $6E or ARAP bins myself. This is of course using the 165 ECM. There was a fellow named "dig" who has a site with WB patches, and for various tdf files, but he said his linear WB code mod did not work with the $6E. He was sending me some test patches, but they didnt quite work out, and now I havent heard from him.

 

OK, I had to dig this up. Found it :

https://www.thirdgen.org/techbb2/sho...highlight=MAP2

Looks like MAT2 and TPS2 are good search terms to churn it up.

Is someone interested to recreate this and establish it as a documented piece?

 

I'm working on making a bin for 8D that has a bit select for the sensor used like 1981TTA posted.
Might take a few weeks. Still need to get all of the release criteria down since it's not my code.
Sorry I'm to busy to work on 6E too. 8D & 58 take enough time already.

I like DIG's method of nested JSR's to do the ADC l/u.
Don't have to skip any code that way.
Should be a simple matter to take his patch and change the addresses and org's to match a specific bin.
I had started on this method for $8D when the $58 WB bin came out. Also, I've seen two posts talking about possible MAP2 introducing an offset into the WB voltage.
I haven't looked to see if this was corrected in the data table or not. But as Rbob said, TPS2 will fix that. I asked how the ADC l/u works a few months ago for anyone interested.

 

If you're just looking for a 0-5v datalogging solution for the $8D, I've got one in place on my own bin. It takes the WB 0-5v and brings it in to the TPS2 or MAP2 input (gotta check my notes) and then puts it in the Fuel Pump Voltage slot of the datastream. Then it makes sure the battery voltage is referenced instead of FP voltage for all internal calcs (injector etc). Works like a charm...

 

By the way, you can do lookup tables in TunerPro RT within the ADS datastream definition. That way, just a simple A/D 0-5v->0-255 you can use. No complexities to the assembly code.

 

After seeing how much soot was caked onto my new O2 when I pulled it out last, maybe converting over to use the wideband at all times would be better. After all, it would pretty much be 6 1/2 dozen to the other as the cost is similar to the heated NB. Not to mention that the WB is many times more accurate, and constantly glows a nice hot orange color, so it wont accumulate a thick pile of **** like the present one seems to be doing.

Craig, with the O2 volts you could also just do a simple table lookup using the built in lookup routine in the ecm to do the AFR conversion.

For anyone interested, you could then just stash teh small ammount of code in a litte cranny somehwere in one of the loops and put the table in an empty part of the bin.

It would pretty much be:

LDAA/B LO2VOLTS ;Load accumulator used by subroutine with O2 volts

LDX #LADDRESSOFTABLE ;table address in hex into X or Y or whatever

JSR L2DLOOKUP ;Jump to do 2D lookup

STAA LAFRADDRESS ;Save the AFR from the table lookup to memory for output

 

If you're motivated, you could do a compare on the Davis wb bins and extract his code. Then use an op code map to get all the commands down and probably just change the table values.

 

When doing the volts to AFR with an LC-1 and using the default
1-2v scaling for I think it's 10-20 afr.
I've seen a few ways of scaling it.
1. volts/.02 = cnts
(Cnt -(cnts/32))*2 =afr

2. Table l/u- which seems a waste on a linear unit.

3. volts/.02 = cnts
Limit cnts to 50-100
cnt * 2=AFR


Will # 3 work?

 

Not being sure of how the scale is setup within the unit, why not make the range 0-5 v and use the full counts of the input for 10 to 20:1 AFR.
This would give a simple multiplier of ((0.04)+10)
10 units between 10:1 and 20:1 on 255 counts = 0.04.
Add the starting point of 10:1
Seems easier unless I missed something in the setup like starting from 1 volt at which point the scale would shift.

If starting at 1 volt and going to 5 volts on the input would give 204 usable counts on the A/D input. 51 to 255.
10 AFR range/204 = 0.0490 AFR per count.
Sub off the initial 51 counts,
Multiply counts by 0.049, Add the base 10 AFR and go.
Yes?

Edit: ((X-51)*0.049)+10) = AFR Will work in TP.
I'm working to see if I can come up with something to output a 10 to 20 value directly from the ecm. That would be nice for any field it is output to in logging programs.

 

Expanding on that concept. rather than using a whole table, could a lookup table with only two points use the interpolation routine to output the correct value?

 

There are a million ways to do it, but if its linear a simple equation works. If its non-linear then a table lookup is the way to go. Table covers just about every possible setup, so thats what Im going to run with. The other option is to use the 2D lookup subroutine entry point (if its available) that lets you specify the number of lines in a table. Itll allow you to change the table size to whatever at will.

 

Guess I could have added that to my original.
I'm working on adapting 1981TTA's multi sensor code to 8D.
He used the divide by 32 approach for the IM linear WB.
Which on paper has a non linear output.
And the default for an IM setup is 1-2v. The code has a raw setting so you could use the logger software to do the math for all available scaling.
And full scale is in the area of 9-21.2 afr.

Let me know what you come up with JP.
I was planning on using the FPvolts location in aldl.

 

If you are looking into the whole solution you are much farther ahead than me.
I was looking at the easy way if it needed to be done quick for a linear output. I think I'll be getting one (LC-1) shortly and was just thinking how a "quicky" could get done.
1981TTA is really on the right track but I can't get enogh steam behind me to get something that ambitious done.
I'm still a little lost on what the std output config is and the adjustable values can be.
I "thought they would be 0-5V for WB and 0-1V for the NB.
What are some criteria that can be used? Start at 1 V to ensure its working etc.
Common thoughts so I don't go off into left field and then have to redo it. I do like Dimented's idea on the table for flexibility but again, if you can get the whole enchilda to pan out it would be the greater good.

 

How would one go about modifying Mike davis's hac to work with the innovative WB??

I have the hac installed and working, but the output that I am seeing in the scan is not correct due to the non linearity of the DIY WB.

Is there a way to make his hac work with the linear output??

p.s. If this has already been explained in this thread
I apologize Im having a little trouble following it.

 

Hey, you know what would be REALLY cool? How about an XDF definition that allows you to go in and edit the lookup table that is part of the Davis hack?

 

I'm not sure if this is Dig's or Davis's patch.
https://www.thirdgen.org/techbb2/sho...hlight=patcher
down at the bottom.

It's Dig's for the $58 (I can read )

 

Got to looking at the right one thanks to Marks graphical table finder
There looks to be a linear voltage table at $2040 to $2050.
Then there is a curved 17 point table at $2051 to $2061.
didn't break down the rest of it but those look to be them.

Now the addy are correct

 

I think the code changes to support various WB sensors I made for the $58 will work for $6E, too. I've disassembled ARAP to see what I can find. I notice that there doesn't seem to be much "extra space" laying around....? What's worse, I'm not very familiar with $6E/165.... So, I'm not sure what can be thrown out. My first inclination is to scrap the factory test logic to free up some space. I haven't done the math yet to see if I can fit support for *all* the WB sensors into this space even if it is deleted. It may be that there will be separate .bins for each sensor.

Has anyone tried to add code to the $6E mask? Is there some blank area floating around I'm not seeing? If there is room and I make the changes, are there people willing to spend time testing changes? Any thoughts on the pitfalls of deleting the factory test routines? Is ARAP a good version to use? It would be pretty cool if we could get this logic in $58, $8D and $6E masks!!

 

Those are 8D addresses btw. And are what I found.
The entire Davis Hack starts at $A000 with code and tables in one continuous string. Only the JSR change at CC90 in aujp is not.

 

for sure!

Z69, When addressing that location from TP be sure to use $2000 as the start. The $A000 values will cause a crash when accessing.
I originally posted the addresses as $A040 etc, but when entered it died.

 

I've had luck using Mike Davis' WB hac and the LM-1 when using tunerpro and putting the following equation in Tunerpro for the "Target AFR".

([X*-0.037994]+34.78)

Settings for the LM-1

1.246 Volts at Lamda 0.685
2.747 Volts at Lamda 1.404

Looking at the hand held unit and the computer shows a good match with an error of only .1 or .2 occationally.

 

Did you also splice the wire from your old O2 sensor to the new WB sensor to simulate the stock sensor? If so, did you have to alter the tables for the AFR?

 

I didn't replace my stock O2 with the WB. Sorry...

 

WOW! I wish i knew about this trick last fall!! I did have the Mike Davis WB hac working as well, but considerable testing gave me incorrect AFR numbers in the target AFR.

Is putting the math in TP a difficult thing to do?


How did you come up with the formula and the linear graph endpoints? (not really that important)

As soon as I get my laptop talking again to my ECM, with TP RT, I will be trying this method out immediatetly! Plus I will be checking out other code mods, but this looks the best so far!!

 

No, putting math into Tunerpro is very simple. Pick the item to edit and it'll show a section to edit the values. Select the right form of the equation...It looks just like the formula but the numbers are replaced by words.

I plotted the linear and curved responses and then used a program to create an equation that was a function of the curved line and resulted in the linear line. I came up with this last year sometime so I don't exactly remember how it did it..I can look it up if you're curious though.

 

Yea I just found the editor for the 'Target AFR' in TP. So it looks like you just use -0.037994 for the factor and 34.78 for the offset? Do you make a new ALDL datastream , or just save it over the 1227165_6E.ads? Probably dont really matter!

Dont need to explain how you got it, what you said is enough! Last fall i tried the Davis hac by putting in a DC voltage and reading the AFR. I took several points from 1.00 to 3.00vdc, and made a curve from the data, which was sort of like 2 linear curves hooked togethor. I then looked in the bin to try and tweak the 2 tables to make the curve come out linear, as well I was adjusting the LM1 analog output.

The only setback I see with the above method, is if I use my "other" scan tool, which is by Ease Diagnostics, I wont get the correct AFR in the datastream. What I get with Ease is called the "total air fuel value". Which is just a number that you divide into 6553.6 to get AFR. I just have a lookup table to do the conversion....but wait all I have to do is just change the formula in my spreadsheet to the new formula above....and should get the correct AFR.
I recall now, EASE gives you the TFAV, and TP does the math!

Cool stuff!

 

I made a new datastream. You may have to tweek the equation a bit. To do this I took the WB and set it to always display a constant voltage no matter what. Then look at the output on the WBO2 handheld thing and compare it to what the computer says. Take the difference of the two and change the 34.78 value. I think I had to change mine by .2.

 

I think i am going to be using the LC-1, instead of the LM-1, so will i need to adjust the ads differently than if i were using the LM-1? Also, how would i check to see the the AFR ratio is close, could i just check the voltage with a voltmeter and compare it to what tunerpro is saying?

 

Yes, I think so.

 

I need to decide where to put the AFR in the data stream for the 8D bins I post.
I only use TP so it doesn't matter to me.
I don't know what is available in the other logging software.
I planned on using the FP volts word 35 if that will work with the common software. Just have to move the decimal to get the actual AFR. If you pick the Raw cnts in the options
It will show the actual voltage in counts if needed.
Let me know it this will cause difficulties.

I have an AXYC bin out for testing currently. Waiting on results. I think he went on vacation....
And AUJP is done also. Still need to get the addresses
for one of the vette bins and assemble it.

 

Has anyone sucessfully intergrated the LM-1 Wideband O2 Sensor into the 165 ECU for datalogging purposes?

I have used Mike Daviss' patch for the 165 ARAP and entered in the formula ([X*-0.037994]+34.78) for the Target AFR, but the end results are off. It shows 17:1 at idle in TunerPro and it shows 14.8:1 on my WB display.

Do I need to tweak the formula or has someone else found a more accurate way to datalog the output of the LM-1?

I actually have a Zeitronix unit, but it uses the same Bosch sensor

 

Not sure if the 165 has this.
swap the bit if possible to use batt volts for injector offset instead of FP volts. Hook the WB to the FP volts pin.
Stick FP volts into your ads.
Adjust ads scaling to match voltage range of LM-1 output.
In 8D, I think the davis patch uses the FP word in the aldl, so might need to use a different non wb bin.

Alternatively email me as a zip your wb bin and the stock version of it.
I'll tweak it to stick the raw value into ram where it used to stick the calc'd afr. Maybe by this weekend.

 

If I send you both bins that you can just dump the raw Voltage read into the Target AFR field? That would be excellent, I can then simply create a lookup table in excel to convert my voltages into AFR. I have the table from Zeitronix:

Volts AFR
0.15 9.7
0.31 9.9
0.46 10.1
0.62 10.3
0.78 10.5
0.93 10.7
1.09 11
1.24 11.4
1.4 11.7
1.56 12.1
1.71 12.4
1.87 12.8
2.02 13.2
2.18 13.7
2.34 14.2
2.5 14.7
2.65 15.6
2.8 16.9
2.96 18.5
3 18.8
3.12 19.9
3.27 21.2

BTW the Zeitronix unit also comes with a NB O2 output to replace your stock NB Sensor with it. I personally only use the WB for tuning, so I haven't tested this but its there.

I'll send you my bins tonight.

 

Has anybody done any $32B hacks (Or have the knowledge needed to do one since I don't? ) that are compatible with the LC-1? I'd gladly provide a stock ABTB BIN to play with if someone wants to take a look at it.

I would love to be able to send a NB signal to the ECM and then use an unused channel in the ALDL stream to datalog the WB readings from within my datalogging software.