Solution to LM-2 vs HP Tuners match

[remigi0bg]

Matching LM-2 with HP Tuners.

First you need to have wired the LM-2 output 1 with de HP Tuner Input, the correct way.

The Output of the LM-2 is linear from 0-5 Volts regarding how it is program will display and AFR value to a corresponding voltage.

Knowing this the equation is de following Y = M * X + B , to have correct relation we need to know 2 points to get the correct values.

In hp tuner log the raw voltage from de input or define one by yourself setting the direct input formula like this y = Volts / 1 + 0 .

Start your vehicle and wait until the afr is stable at 14.7 probably you need rev up a little bit the engine.

When the LM2 AFR is 14.7, record the voltage at HP Tuners. For Example, lets say we read 2.45 volts,

then we have our first point: AFR = 14.7 Volts = 2.45

Now the easy way to get our second point is to make a copy of your VCM program, and modify it to work in open loop, and in the Open Loop EQ Ratio table set it to 1.30 every cell above 40° ECT, with this, you are forcing to have a rich AFR, write the calibration, start the engine and start logging, wait until the afr is stable, (may need a little rev ups), and record the afr value, let say it is: AFR 11.3 and the volts of hp tuners log is 0.80 volts, Stop your engine, and write your original calibration file to return to normal.

Our second points is: AFR 11.3 Volts 0.80

With the 2 known points, we can make the proper calculation for our equation.

M = (Y2 – Y1) / (X2 –X1) so M = (14.7 – 11.3) / (2.45 – 0.80) is 2.0606

B = Y – M * X so B = 14.7 – (2.0606 * 2.45) is 9.6515

For the HP tuners the M has to be a divisor not a multiplier so

1/ M = 1 / 2.0606 is 0.4852

Now in the hp tuners make a user defined called WB-LM-2 and put this formula

Y = Volts / 0.485 + 9.65

And you are ready.

[S2H]

not really a reliable solution for getting AFR...


you do the math of setting your AFR to be what you want...
say 8AFR to 16 AFR
5 volts /8(AFR Range) = 0.625

so the equation becomes volts / 0.625 +8 for AFR

next is to solve in lambda
you will need to know your pcm stoich value
14.62857 in a 2001 LS1 Fbody as an example

our example above... 8 AFR would = 0.5468750534057669 Lambda
and 16 AFR would = 1.093750106811534 Lambda

then range = 0.546875053405767
so 5v/0.546875053405767= 9.142856249999999

so volts/9.143 + 0.5469 = Lambda



now you need to figure out voltage for AFR or Lambda
so (stoich-minimum range)*Divisor

(14.62857-8)*0.625 = 4.14285625 volts


using your setup of AFR or Lambda
you can set the voltage of the LM2 to be flat across the board so that at all AFR's it reads a set voltage...as above... set it to 4.14
you can then look at your HPT scanner, and see where it sits in relation to your original equation.. if its off, shift the Adder value over till it lines up
example.. if your AFR reading shows 14.60 with the voltage set to 4.14... then shift the adder over .3 so that it reads Volts/0.625+8.03

you can then set the voltages back to be 0 at minimum AFR and 5 at maximum AFR and proceed with your new corrected formula

same goes if you tune in lambda(which you should), set your folmula just the same way


the set range of your wideband determines the divisor(just because your wideband doesnt do the minimums and maximum ranges of the output very accurately doesnt change the slope of the formula... it just means the extremes of 0 volts and 5 volts may not read exactly correct)... the misalignment because of a weak or poor ground determines the voltage offset/AFR adder

[remigi0bg]

Theoretically you are right, i tried to do what you said, setting the slope fixed and adjust the cero just like you did, and yes I got equal values at sotich, but when I go WOT, there for rich i got error for example I read 10.4 in hp tuners when the lm-2 read 11.3, I´m agree that ground affect, that’s why I went for that solution, getting the theory to real world with real data.

In mi case the LM-2 is programmed to send 1 Volt at 9.7 AFR and 4 volts at 19.7 AFR
and the HP tuners user config (after been applied the way I said) finished like this

AFR = Volts / 0.444 + 9.3

When theoretically its supposed to be :

AFR = Volts / 0.3 + 6.36

Ground or whatever that its affecting, changes the theory values, so you need to tune using the two reference points and get the correct two new parameters, that are real.

The solution that I am mentioning, Should work with any linear input, not only
for LM-2.

It ´s simple math, applied to real data.

[Montecarlodrag]

I bought mine several months ago and had a big difference in AFR between scanner and LM2 display.

-I first did a sensor calibration in free air. It went 20.9 AFR, I measured volts at the analog output.
-Opened logworks and set up the device properly based on my measurements
-Next I logged raw volts on scanner and compared to the true volts at EIO input. I saw a 0.1 volts difference between them.
-I created an Excel spreadsheet to generate the formula for Hptuners.
-Then adjusted the Wideband aux input in the scanner, taking into account all my readings and measurements.

Now, I see the same value on scanner and LM-2 display, with 0.1 AFR max error. The closer to stoich the smaller the error ( think it's because I wasn't able to measure very rich AFR values, only free air)

[5_Liter_Eater]

I've also found that the ground differential makes the theoretical equations not accurate for real world application. I did something similar but used a lot more than two data points...

http://www.hptuners.com/forum/showthread.php?t=24181

[5_Liter_Eater]

I will also say that I have seen my offset need to be changed from day to day, even the same day sometimes, but the slope is consistent, but not the same as the calculated. My thinking is that it is the working voltage changing anywhere from 13V to 14V that is changing this. I need to do another round of calibrating as I now have the WB hooked up to a relay off alternator power.

My calculated should put me a V/.5+10 but my current equation is V/.474+8.6.

[DSteck]

I'm amazed at the depth of effort that goes into this stuff.

For the LC-1 I used, I just wired every ground together with good wire, and always connected straight to the battery. Then, I forced the LC-1 to output certain values, recorded them in HPT, and they ended up being what they should have... ie: when the LC-1 was told to output 2.93 volts, HP Tuners recorded 2.93 volts. When I pushed the values around, it always read correctly. With my NGK AFX, when I have the sensor in warmup mode, using my transfer equation based on the NGK's lambda, HP Tuners ALWAYS reads 1 lambda when the sensor is forcing it out. Maybe I just got lucky.

I also linked the ground wire for power on the NGK's harness to the 5V ground.

[omega_5]

If it's wired up to fix the ground offset, there is no funky math involved.
Plug it in, select the pre-made PID, and go.

http://www.hptuners.com/forum/showth...762#post213762


Recall, this whole issue is because of a hardware 'problem' with the LM-2.
Trying to fix it through software is just a tail-chasing band aid.
Fix it at the source, and all of your problems go away.

[Montecarlodrag]

If it's wired up to fix the ground offset, there is no funky math involved.
Plug it in, select the pre-made PID, and go.

http://www.hptuners.com/forum/showth...762#post213762


Recall, this whole issue is because of a hardware 'problem' with the LM-2.
Trying to fix it through software is just a tail-chasing band aid.
Fix it at the source, and all of your problems go away.

I checked that pin 5 ground and I found no difference. By default the Pin #5 is grounded through the DB-15 connector. There is continuity between Pin #5 and vehicle ground

I logged raw voltage with and without the chassis ground and found no difference at all. I use my Lm-2 with the extra ground anyway, but I find it interesting

[omega_5]

By default the Pin #5 is grounded through the DB-15 connector.

DB-15 connector?
Are you running a serial HPT box?

The point of this whole things isn't to ground pin 5 on the box... it's to ground the negative output of the LM-2 to the same point as that the unit is grounded to, thus eliminating the float between the negative and positive output.

I simply mentioned grounding it at pin 5 for simplicity, as it's a common connection point anyways.

[Montecarlodrag]

No, I have a USB MPVI

The connector where the OBD-2 cable attaches is a DB-15.
Shield of the DB15 connector goes to OBD-2 pin #5.
DB15 Pin #13 goes to OBD2 pin #4 (both 4 and 5 pins are grounds).

So, When the OBD-2 cable is hooked to the vehicle's port, there is ground on MPVI Pin #5. If you unplug the cable there is not.

Because there is 0.0 ohm continuity between MPVI pin #5 and vehicle ground with no aditional wiring, I think there is no difference if you plug an extra ground cable to pin #5

The (-) cable from LM-2 also goes to MPVI pin #5, so it's the same.

I'm not trying to be a smart ass, I'm just telling the things I found when I had problems with my LM-2. I measured volts at MPVI Input 1 with extra ground and without it, no difference.

Cheers

[omega_5]

Because there is 0.0 ohm continuity between MPVI pin #5 and vehicle ground with no aditional wiring, I think there is no difference if you plug an extra ground cable to pin #5

The (-) cable from LM-2 also goes to MPVI pin #5, so it's the same.

I know what you're getting at... but there is a difference between grounding through the OBD connection via another device and providing a direct ground to chassis.
Grounding at two different points, or through another electrical device can cause a ground offset, which is what we are trying to get rid of in the first place.

If you don't know what I'm talking about; just Google 'ground offset'.

In the end... if it's not working for you, then one of two things is wrong; your LM-2 is internally effed, or you still have a bad ground off the output.

[Alvin]

There are big time issues with the EIO ground offset but also slope error.

I don't know if its hptuners fault, the widebands, the car, etc. But I will tell you that recently I was doing a engine on a engine dyno at a big time race shop (think several million dollars a year). They exported a clean 0-5v signal and a clean ground from their daq so I could log a global lambda inside Hptuners. With their hardware they could define any output they wanted with a corresponding lambda they could also simulate it so we could check up on our end. Long story short we got it dialed in one day, did nothing but come back the next day and it was off by .04 or so. We double checked their output with a multimeter and it was right on the money. From that day on I calibrate anything I want to use with hptuners EIO right before I want to gather data from it. That means if I leave the car for lunch and come back later I recalibrate it.

I tell you slope is wrong just as often as offset. Be careful to calculate both. Attached is a sheet I use for any external input fuel pressure, wideband, etc. Nothing too hard just y=mx+b. As far as taking more than 2 samples, well unless your not confident in your samples there's nothing to gain unless you plan on building a look up table for the voltage and output. I suggest taking two samples near to where you plan to operate. Say you have a NA car.. A sample at 12.8 and a sample at 14.7 would be ideal. Pray its linear between those two!

[5_Liter_Eater]

Thanks for that spreadsheet Alvin. Very simple and effective. It will come in handy. I kind of feel better knowing that EIO suffers from this problem too. I was thinking I did it to myself by piping the WB through the AC. It is a little disheartening that it is not as easy as it seems. Makes me wonder how many other inputs are in error.

[jakebdb56]

I just got a LM-2 myself for mobile tuning, and I'll say that I really like the way everything is setup but it's not as fast as my PLX was. Still very accurate at WOT, but low speed air seems to make it go all over the place while the PLX would read out a pretty steady rate. Anyways, the spreadsheet helped getting my LM-2 setup right. Leaving the factory default voltage outputs for the LM-2 helped a lot in accuracy side by side with my PLX. I know my PLX is good since I've checked it against two other widebands recently

Once my voltage offset was corrected, everything was ready to rock.

[senthilmar]

thanks man for your tips

[Turbo-Geist]

Thanks for all of the information posted. I will be trying out the additional ground this week. I also have a question.

I was doing some reading and found this on wikipedia:

"Ground loops can form when equipment with floating grounds are connected to properly grounded equipment, and results in a current traveling between the two devices."

It made me wonder if a diode could be used on the signal wire and ground wire from the LM-2 to the EIO Interface to make sure that information is only traveling in one direction? Any thoughts on whether this is a solution?

You would think that there is one already within the LM-2, but maybe not. I'm just asking a question here and looking for input. Let me know if this has any merit.

[Ben Charles]

Any updates on this? My friend has an LM2 Wb and I am trying to get the AFR on the WB to match the scanner. I am about 1 point off from my WB to scanner reading(scanner leaner).

The equation I am using is .625 + 13.1 (seems to be just right at idle, but off 1 point at WOT?)


Also I have been used to working with my EGR WB and I can setup the Histogram for that no problem using this equation ([PID.2811]/.7095)+9.102

MY WB to AFR percent error is
100*((([PID.2811]/.7095)+9.102)-[SENS.121])/[SENS.121]

However, we are using the EIO Hp tuners pro version, what equation do I use? I tired for about an hour to figure it out and I couldn't get it to report percent error in the histogram.

[Ben Charles]

Ok as far as getting this WB to work with the EIO Pro version the equation is
100*([AUX.20121]-[SENS.121])/[SENS.121]
I believe this is correct if anyone cares to give any input. I will confirm next time I am working on this car

[5_Liter_Eater]

Ben we need to know how you have your LM2 set up to know if your equation is right. What AFR and voltage range are you using? Also what do you have stoich defined as and does that match what the tune has defined as stoich?

Also there is no reason to create your own AFR Error PID when using EIO/Pro. Once you have an air fuel ratio sensor defined (WB PID) it uses that for SENS.120 (AFR) and the built in AFR Error PID compares that to SENS.121 (commanded AFR) and give you the AFR error.