WBO behaviour

[Philip Lochner]

Installed the Innovate LC-1 in my Disco.
 
When I look at the AFR log using the Logworks sw, the AFR seems to vary much more dynamically than what it does when I look at the MS2 log using Excel.
 
Also, at idle (I set my param so EGO will work even at idle), I certainly do not get the readings on Megatune that I would get with the Logworks sw.
 
This leads me to distrust the MS2 readings.  I intend getting a second serial port going on my PC so I can log both simultaneously and hopefully superimpose the one on the other to assess the correlation.
 
Am I being paranoid?
 
Kind regards
Philip Lochner
 

Posted by email.

[grippo]

Several things can be going on here. First I have recently found a possible error in how the EGO table in MS II is burned in flash. I will fix it this weekend, so that's one possibility. The other is that there is a lag factor in MS II. If that is set too hi(or low I can't remeber which ) you could have excessive averging. Another possibility is that the analog ego signal from the WB board is an averaged signal, but the signal captured by Logworks is a raw signal. As far as data rates, MSII updates its ego readings as fast as they are needed to provide control - which can be as fast as every spark event. MT then samples 10x a second. Your plan of side by side comparison is a good one. If you suspect error in the MSII result you need to compare it with the voltage going into the MSII EGO ADC line using a scope. MSII should be capturing this quite well. If there is an offset there are inputs that allow you to correct for this.

[Bernard Fife]

If that is set too hi (or low I can't remember which ) you could have excessive averging.

Al,

Just for others info, the lag factors can range from 0% to 100%.

100% means no averaging at all, use the current value.

0% means never use the current value, so the indicated value never changes.

A value of 30% smooths the response more than a value of 60%, for example. There's more here: http://www.megasquirt.info/ms2/configure.htm#lag

Lance.

[efahl]

Lance,

I've set the upper limit on all the lag factors to 100%, this will appear sometime soon in the MT 2.25 patch with a bunch of other little tweaks.

Eric

[Philip Lochner]

Hi Guys
 
Thanks very much for the supporting input.  I did manage to solve my own "problem".
 
I found the "problem" to indeed be my averaging setting on the "Settings/general" page where I had 10%.  Increasing this to 50% (close to default....) now results in a more "lively" AFR.  What threw me off a bit was the fact that it is called  LAMBDA average etc (BTW, the online megatune manual does not refer to this specific menu item??). 
 
Using the word "LAMBDA" , led me to conclude that its is the RESULT of the division of AFR(stoich) div by AFR (Actual) that is averaged - not the actual AFR. Deciding to change this anyway resulted in a more  lively gauge so I concluded it was merely a terminology issue.
 
How about changing it to "LAMDA/AFR" ?
 

Kind regards
Philip Lochner

Posted by email.

[Philip Lochner]

Having installed a second serial port on my PC I have found the AFR reading on Innovate Logworks to consistently differ by about 0.9 to that displayed on MT 2.25 by displaying both simultaneously on the screen.
 

Grippo: "If there is an offset there are inputs that allow you to correct for this."

Where are these inputs and how do I go about adjusting them, or just a pointer to the relevant info? (I made sure that my LC-1 is grounded at the exact same point as the MS2)
 
Kind regards
Philip Lochner
MS2/V3/2.35
 

Posted by email.

[FoundSoul]

I’m currently using the Innovate LC-1 with the MS-I and MSnS-E and it reads dead-on accurate with the grounds shared between the EMS and LC-1. This is comparing the digital readout of the XD-1 (same numbers you see in logworks) to MegaTune’s numbers which are D/A converted by the LC-1, and then A/D converted by the MS and interpreted by MegaTune. My LC-1 analog outputs are configured for 0-5v 10-20:1 and I didn’t change any other LC-1 settings from the defaults, nor did I tweak any of the lag settings in MegaTune. I have done the same thing with an LM-1 and MS-II in the past, shared grounds, same config, and dead on accuracy between the digital readout provided by the LM-1, and the D/A and then A/D converted readout provided by MegaTune.

I’m also using the resistor in between the analog output of the LC-1 and the o2 input of the MS, 470 ohm ¼ watt is what should be used here.


All that just says I know it can and will work properly, I know it doesn’t help you very much yet. I’ll try to do that now. ;)

Are your outputs programmed for 0-5v representing 10-20:1 AFR on the LC-1? Were any other settings changed from the defaults at all? If so I’d change them back.

On the Megatune side of things—in the MegaTune Configurator, is the Innovate 0-5v 10:20 LAMBDA sensor selected?

Then inside MegaTune, calibrate the AFR table using ‘Tools > Calibrate AFR Table > Select Sensor Type > OK’

Lastly in MegaTune, Settings > EGO Control > Single Wideband > Burn to ECU > Close



You may have done all of these things already, and if so it’s hard to imagine why you’d still have a differential in your readings—I see no reason why you should as mine are rock solid in all cases I’ve seen where these steps have been followed. Let us know what you find out.

I’ll let Al comment on the input offset correction, though from my experience with this combination of components you shouldn’t need it at all…

[Philip Lochner]

Hi Jerry
 
I recon you’ve nailed the problem!  I’m using output 2 of the LC-1 which outputs 0-5V by default.  Somehow I missed the fact that I need to change the default settings of 0V = 7.35AFR and 5V = 19.36AFR.  So clearly I’ll be doing that NOW!
 
Will try to give feedback later today.
 
Thanks, yet again!
Philip
 
-----Original Message-----
From: FoundSoul [mailto:websales@diyautotune.com]
Sent: 22 March 2006 03:29 PM
To: megasquirt2@msefi.com
Subject: WBO behaviour
 
I’m currently using the Innovate LC-1 with the MS-I and MSnS-E and it reads dead-on accurate with the grounds shared between the EMS and LC-1. This is comparing the digital readout of the XD-1 (same numbers you see in logworks) to MegaTune’s numbers which are D/A converted by the LC-1, and then A/D converted by the MS and interpreted by MegaTune. My LC-1 analog outputs are configured for 0-5v 10-20:1 and I didn’t change any other LC-1 settings from the defaults, nor did I tweak any of the lag settings in MegaTune. I have done the same thing with an LM-1 and MS-II in the past, shared grounds, same config, and dead on accuracy between the digital readout provided by the LM-1, and the D/A and then A/D converted readout provided by MegaTune.

I’m also using the resistor in between the analog output of the LC-1 and the o2 input of the MS, 470 ohm ¼ watt is what should be used here.


All that just says I know it can and will work properly, I know it doesn’t help you very much yet. I’ll try to do that now. ;)

Are your outputs programmed for 0-5v representing 10-20:1 AFR on the LC-1? Were any other settings changed from the defaults at all? If so I’d change them back.

On the Megatune side of things—in the MegaTune Configurator, is the Innovate 0-5v 10:20 LAMBDA sensor selected?

Then inside MegaTune, calibrate the AFR table using ‘Tools > Calibrate AFR Table > Select Sensor Type > OK’

Lastly in MegaTune, Settings > EGO Control > Single Wideband > Burn to ECU > Close



You may have done all of these things already, and if so it’s hard to imagine why you’d still have a differential in your readings—I see no reason why you should as mine are rock solid in all cases I’ve seen where these steps have been followed. Let us know what you find out.

I’ll let Al comment on the input offset correction, though from my experience with this combination of components you shouldn’t need it at all… <![if !supportLineBreakNewLine]> <![endif]>


Jerry a.k.a. 'FoundSoul'
86 MR2 NA w/MSnS-E

Now Available:
USB to Serial Adapter -Tested and works with MegaTune!




This post is at: http://www.msefi.com/viewtopic.php?p=109289#109289

Posted by email.