Tightmopedman9

Tuners Rejoice! Free Tuning For M4.4!

Recommended Posts

Hello guys, back to stress !

I bumped my brains against the wall. I just ordered the MPX450AP boost sensor, and I don't just want to copy-paste things. I understood the formula in TunerPro for logging in PSI should be ((((X/255) + 0.04)/0.004)*0.145) - 14.7

Now I searched high and low through the thread and found out :

X - Binary value returned by the ECU

255 = number of steps in the ADC

0.04 = one "extra" step to make 256, and it's actually 1/255

14.7 = substracted from the result to read out relative pressure.

Now I don't know what the heck 0.004 and 0.145 factors are for ?!

Thanks !

Later edit - Found out 0.145PSI = 1kPa, and looking at the datasheet, MPX450AP measures in kPa. So where's the 0.004 from ?

Edited by Midnight Caller

Share this post


Link to post
Share on other sites

Oh, it came from the RAM scaling ! I have that document, yes, but never thought looking in it, considering this was somehow a "hack" and didn't take into account the stock multiplication factors.

Thank you for enlightening me.

  • Upvote 1

Share this post


Link to post
Share on other sites

For the ones who like to have their TCV Pilot control map set up properly in a fast way, I have created an XLS:

http://www.paerl.it/volvo/TCVPilotcalculator.xlsx

It works like this:

Create a history table in TunerPro.

set Output (Z) Object to current load. Set the amount of samples to a high value.

2015030601.jpg

Give it the same RPM rows as in the TCV pilot table.

2015030602.jpg

give it columns als in the throttle values in the TCV pilot table.

2015030603.jpg

Drive while trying to "touch" as much cells of the TCV pilot table.

If you play back your log in the history table, please select history average.

Paste this data into my Excel sheet. As well as the other requested values. This is documented in the sheet.

Of course their will be some pollution in the history table. You have to use your intelligence to correct these values.

The sheet calculates the new TCV pilot table.

This pollution is caused by situations where the turbo is spinning down.

Another way to avoid this pollution is to only play back the parts where the turbo is spinning up. The history table only remembers the parts you play back.

To my experience you can have a good TCV Pilot map in three iterations.

The reason for such a fast result is that the algorithm I used is converging.

Edited by razorx
  • Upvote 1

Share this post


Link to post
Share on other sites

Razorx method works really good.

It's absolutely worth trying, it wiil make your boost/load curve noticeably smoother.

Edited by Piet

Share this post


Link to post
Share on other sites

The memory bank switching speed can be observed by changing one of the maf tables with the faulty maf tool. You will then see the kg/h number oscillate with the bank switching frequency.

Share this post


Link to post
Share on other sites

The memory bank switching speed can be observed by changing one of the maf tables with the faulty maf tool. You will then see the kg/h number oscillate with the bank switching frequency.

Isn't that influenced by the sample speed of the logging? It seems logical to me that if you don't log every iteration of the main loop you will miss readouts and thus bank switches..

Is there a reason other than to satisfy curiosity to want to know how often the bank switching is going on?

Share this post


Link to post
Share on other sites

Curiosity is the main force behind most of my Motronic and communication research.

Understanding the bank swithing could be useful. I don't know yet.

My own car runs M1.8, so all the VST, MT-Pro, VCT2000, DiCE, M4.x, ME7, EDC, CEM, REM, UEM, ... research is not really of any use to me yet.

But if I buy a newer car, step 1 would be to make a full backup of all controllers. Backup of the tools software can also be useful.

Share this post


Link to post
Share on other sites

Like once per revolution, once per tooth, or even faster?

Same speed always, or depending on running/not running, idle, part load, WOT ...?

Share this post


Link to post
Share on other sites

Like once per revolution, once per tooth, or even faster?

Same speed always, or depending on running/not running, idle, part load, WOT ...?

No. Whenever it needs to address this higher bank.

The CPU, well you now the story ;), cannot address the amount of memory. So Bosch needed a dirty or clever trick.

Obviously you are thinking of something. I am curious.

Share this post


Link to post
Share on other sites

If we forget the interrupts for a moment.

In the main loop there are about 160 (give or take) lcalls to subroutines.

Half (about 80) of them are to the same subroutine in the upperbank.

The rest of them (all different routines) are mainly in the lower bank, they are called inbetween the 80 lcalls to the subroutine in the upperbank

So there is a constant switching of banks.

On top of that comes the interrupts.

Every tooth for instance there is an external triggered interrupt to a routine in the upperbank.

The cpu switches bank by setting or clearing CCM7 (Port P5.7).

Bank switching isn't really time consuming for the microcontroller.

Edit:

In one of the "other" subroutines lcalled in the main loop a logging frame is made and send.

Edited by Piet

Share this post


Link to post
Share on other sites

I think the logging speed is high enough. The sawtooth I saw back when we tried the faulty maftool looked like it hadn't any teeth missing. I'll take a look if I can find an old log made with a faulty maf table.

Best way to measure it would be monitoring p5.7 with an oscilloscope I guess.

Edited by venderbroeck

Share this post


Link to post
Share on other sites

For the ones who like to have their TCV Pilot control map set up properly in a fast way, I have created an XLS:

http://www.paerl.it/volvo/TCVPilotcalculator.xlsx

Nice Work!

I added this to the M4.4Wikia/Tuning page, seemed like the relevant place to add it - feel free to move/ edit. I only inserted your text and referenced your posting.

Should the XLS be added to the packet Venderbroeck is working on?

EDIT: I saw that there is also a MAF Linearization ADX, that is not qualified. If it is the one from >tjwasiak, shouldn't it be vetted first?

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.