Copilot For Power Fc

[Kevin Beane]

The latest and greatest Power FC data logging, display, editing, tuning, and replay software is now available. It is loaded with new features. Some of the new features which were not in the 2.0.6 "Lite" version include:

Online Auto Tuning
Previously you had to do data logging while connected ("online"), run it in offline mode and click Analyze, then run it in online mode and write the new base map. Now you can set it up so it reads, logs, analyzes, and writes as you drive. This includes auto tuning the base map and/or ignition timing. Online ignition timing changes are limited to retarding only for safety reasons.

Improved Auto Tuning Algorithms
This is rather complicated, but basically samples are weighted and balanced, and you have a clearer picture of what is happening due to additional maps which can be displayed, some in higher resolution than the 20 by 20 Power FC maps. For ignition timing auto tuning, the program looks at what is happening for each knock sample obtained, for example whether or not you are shifting gears, and decides whether the knock is most likely true knock or false knock (false knock can be caused by shifting noise). Also, if you want it to, it will retard timing when there is knock (even a small amount, which you may think is just noise) until the point at which retarding it further does not reduce the knock. It will also advance timing, if you enable that feature. Advancing and retarding timing via "Find Insignificant Knock" works extremely well and can unlock power you did not know you were missing.

Street Dynamometer
The program looks for conditions that indicate a full throttle run in one gear and saves the data associated with it (not just HP and torque, but also AFR, knock, airflow, injector duty, boost, etc) in a Dyno file that can be opened later. If you want, the Dynamometer window will open automatically when there is new data. You can compare any two runs. Amplifying information, for example the car's weight and other factors affecting the HP calculation and notes, are saved and can be viewed as well.

Export to Google Earth
If you have a GPS antenna connected (example: DeLorme's LT20 -- around $60 and includes mapping software which can run at the same time), GPS data will be saved in the log, so you will be able to use the Export to Google Earth function. I've posted an example .kml file before (Google Earth input file).

Complete Replacement for FC-Edit
Well, almost. You cannot reinitialize the Power FC yet. I've added all the Fuel Injection Factors, Ignition Factors, Sensors, and so on. There are new windows for displaying, editing, saving, load, and writing all this new data.

Different Ignition Maps per Gear - REMOVED
This is actually described pretty well in the User's Guide. Unfortunately, I had to disable this code due to the complexity of the ignition timing auto tuning. It may be put back in later.

Better Communications Error Recovery
The Datalogit box is prone to communications dropouts. I have tried many things in software to work around this and the latest version seems to do a better job. I went for a two hour drive today and there were 6 momentary dropouts, all of which were handled automatically within a couple of seconds. The only reason I know it happed at all is there were messages in the log saying that it had handled them. UPDATE: Nope, that was too much to hope for. It still stops responding at times. Yet another update: It is working very well now after some more changes.

Rescaling Maps
A previous version added map rescaling, but I did not catch a bug that prevented the new map reference from being written. So this is just a bug fix, but it will not be in the Lite version... I am done updating the old code. It is just too much work to maintain two separate versions of the software.

Boost Gauge, Oil Pressure Gauge, other Gauges
These have not been tested, but should work with at least some of the sensors out there. Like the other virtual gauges, the data driving it is logged and replayed, so the gauges all work when replaying. Additionally, there is a Boost map, Boost trace on the chart, and Boost is included in the Dynamometer plot. The Boost map is 20 by 20 and can be displayed as average, min, and max, all per gear(s) (like HP and Torque maps). Many other types of gauges can be added, moved, and resized.

Export to Excel Workbook
Most of the data, for example all the maps, airflow, and map reference, can be exported as a complete Excel Workbook. In a later version I will expand upon this to include all the data. Having the Power FC data in Excel makes it easy for people with Excel talents to implement their own formulas. You can also print, of course. Whether or not this particular feature works for you depends on what version of Excel you have... no guarantees that it will work on your computer.

Expanded I/O Capabilities
With Datalogit, you get 4 Analog Inputs. Copilot also supports Phidgets 8/8/8 boards, which give you another 8 analog inputs, 8 digital inputs, and 8 digital outputs. The inputs are all logged and analog inputs can be used for a wideband O2 sensor (faster and more precise than Datalogit or even LC-1 digital), an oil temperature sender, oil pressure sender, or boost sensor. That's all I have defined at present. I have built a hand controller with 8 buttons that connect to the Phidgets 8/8/8 card and do dedicated funtions in Copilot. Let me know if you have any special I/O needs and I can work with you on the software. Copilot also works with Phidgets 2- and 3-axis accelerometers, and any GPS antenna and driver that outputs NMEA (standard protocol) GPS sentences on an actual or virtual comm port. Copilot can control a Phidgets 0/0/4 relay output board to close/open relays in response to user-defined conditions (examples: oil cooler fan control, warning light control).

Graphical User Interface Improvements
Many improvements here... too numerous to list.

Audio for Limit Check Alerts
You can now have it play any WAV file when a parameter (e.g., water temperature, knock, injector duty, ... up to 14 in all) exceeds a limit you define. A different WAV file can be used for each parameter, if you wish. A variety of WAV short files are installed.

2D Plot of Course and G Forces Display

This is an overhead view of where the car is with many features. A GPS antenna is required for this feature. The plot is a separate window and a miniature version showing just the track and vehicle position can be viewed on the Dashboard tab. Control of both plots center and scale is done via the 2D Plot of Course window, which you can set up to keep the vehicle on the plot (anywhere), keep the vehicle in the center, or keep it centered on any position. The Dashboard tab can now (as of 3.0.4) display a different style AFR gauge and a G forces plot (different from the G Forces window). Just click the different gauges/displays to make them change function. See the Jan 1, 2008 post on Page 17 of this topic for more information. There is also a Horizontal Situation Indicator (fake aircraft instrument) which is driven by GPS data.

Video

Now (3.0.5 and greater) you can replay an AVI file in synch with all the other data (Power FC, wideband, GPS, accelerometer). This will be expanded upon in the future.

Economy Display

Computes and displays fuel economy as you drive. Similar to a Prius "Consumption" display.

Platforms/Engines Supported:

1ZZ-FE, 2ZZ-GE, 3S-GTE, 3S-GE, 2JZ-GTE1, EJ20K (WRX), EJ207, RB25DET, RB26DETT, RB261PRO, RB26_2-D, 1JZ-GTE, 3S-GT1PRO, SR20DET1, 4G63 (EVO)

Note: platforms other that 1ZZ-FE and 2ZZ-GE have not been extensively and may still have some problems. I've been working on bug fixes.

12/16/07 EDIT: Disregard all references to licensing and cost. The software is free again. I'd like to encourage people who did not already donate or purchase the software to send some paypal to Tem for hosting the software (he's been doing it for over two years) and Little Rocket for providing valuable tuning advice. You know that it is the right thing to do!

How to donate:

I promised Little Rocket 20% of the profits, so he should get 20% of the donations. The other 80% should go to Tem for hosting it and all the help (tremendous!) he's given. So please make two small paypal payments to them, if you want.

Tem: temmeke @ hotmail.com
Little Rocket: dmlea @ earthlink.net

Edit: I've decided that it would be OK for me to accept donations, too, but only for work done to support platforms other than 1ZZ-FE and 2ZZ-GE. So here is my paypal information:
kevinbeane @ comcast.net

(no spaces in those paypal addresses... I just don't want it to show up as a link)

--------------------------------------------------------

If interested, first download and read at least the first two sections of the User's Guide here: http://www.mr-s.org/copilot/Copilot_2_0_8.doc. Right-click and save as. It will be updated after finishing all the preplanned development which has to be completed before making the software open source. Disregard the part of the document that talks about licensing and payment.

Next, download the program here: http://www.mr-s.org/copilot/ Look for the latest version of the Copilot msi file.


Demo

I've uploaded a folder with some 2ZZ-GE data and data logs. That folder is http://www.mr-s.org/copilot/Demo.zip which you should unzip after downloading and installing Copilot. When you unzip the demo folder, it must end up under your AutoTune folder (probably C:\Program Files\Copilot\AutoTune). It doesn't have to be named "Demo", but that's where Copilot will look for it when you click Open Log for the first time. So you probably want this folder, which contains Power FC data files and a subfolder called DataLogs, to be something like C:\Program Files\Copilot\AutoTune\Demo\. When running in Replay Mode, some buttons and features are not available because it cannot communicate with I/O devices, so don't think this is all there is to the program.


Jan 13, 2008: Added support for 3S-GE and added Injector Warning settings for 3S-GTE
Feb 10, 2008: Getting started on 13B-REW
Feb 19, 2008: Hopefully the Boost problem is fixed (converting from PFC format to PSI)
Feb 23, 2008: Minor changes. Boost is still wrong.
Feb 25, 2008: RB26DETT support added + minor changes
Aug 6, 2008: Bug fixes
Aug 26, 2008: Added Economy display
Sep 3, 2008: Improved Economy display, added some more error checking, removed preliminary 13B-REW support because of runtime errors (running out of memory)
Sep 10, 2008: Added another type of Economy display -- shows MPG for each gear at constant speeds to help you decide what gear to choose (if you want to save money)
Sep 14, 2008: Bug fixes, minor tweaks
Oct 4, 2008: Bug fixes mostly intended for EJ20K.
Oct 19, 2008: Bug fixes, added VVT-i logging for 1ZZ-FE
Nov 3, 2008: Added 1JZ-GTE support and a feature to Auto Detect the Datalogit comm port. Please don't try 13B-REW as it isn't ready to be tested yet.
Nov 18, 2008: Added more things that can be viewed in 3D: Logged AFR (filtered or unfiltered), Filtered Airflow %, and Filtered Injector %. These can be normal (20 x 20) resolution or Detailed (60 x 60) resolution and with or without Trivial data, just like on the 2D map display.
Nov 20, 2008: Included another background bitmap for the temperature guages in the setup package, so when you switch temp. units to Celsius, it will show the correct units.
Jan 2, 2009: Added support for RB261PRO plus other (minor) RB* platform changes.
Jan 10, 2009: Allow selection of RB26_2-D, treat like other RB26* platforms for starters.
Jan 11, 2009: Did more work on RB26_2-D and it is mostly functional now.
Jan 22, 2009: Trying to get to the botton of some run time problems and perhaps an installation issue.
Feb 16, 2009: Bug fixes
Feb 25, 2009: Bug fix for the calculation of "Throttle Accel (%/sec)", which is an auto tune filter
Mar 1, 2009: Fixed 2 RB26_2-D bugs
Mar 5, 2009: More work on RB26_2-D and other MAP platforms
Mar 8, 2009: More work on RB26_2-D and other MAP platforms
Mar 8, 2009: Fixed problem reading "PIM Scale, offset" strings for RB26_2-D and other MAP platforms, fixed display of minimum AFR on I/O Settings window
May 10, 2009: Added WRX STi (EJ207) support and enhancements
July 2009: Made numerous ease of use improvements, added 3SGT1PRO support, added Relay output support, new Limit Check Alerts, Advisor window, bug fixes
Aug 2009: Minor improvements
Sep-Nov 2009: Improvements to Auto Tune Ignition Timing (advancing timing), other enhancements, bug fixes, added SR20DET1 and 4G63 (Mitsubishi EVO I through IX) support.
Dec 2009: Virtual gauge additions & improvements, HSI added, bug fixes, hibernation

[Kevin Beane]

Sorry to report that the Datalogit comm dropouts are still a problem. On the way to work this morning I had quite a few, plus messages with bad checksum, and the software did not automatically reestablish communications some of the time. I hate Datalogit problems!

I uploaded a new version last night, still called 2.0.8. Minor bug fixes, plus I added a new feature that allows cells to be marked on the display, just for a visual reference. Say, for example you see something of interest, just click a cell or cells and then click the Mark button. Those cells in all maps will have a different colored border until you either clear the marks or exit the program. There was also an occaisional "permission denied" error at startup where it was trying to open a file with write lock -- fixed that. Also, whatever cells you have selected on any tab are still selected when you switch to a different tab. Fixed a problem with displaying two dyno runs at the same time.

[Kevin Beane]

Here's a Dynamometer example showing just one run. I guess I let off the throttle or something because this one only caught from about 7200 to 9000 RPM.



I need to make the scale numbers lighter (easier to read). It shows the HP and torque scale on the right and scales for AFR, percentages, and IAT on the left. The tooltip moves to the left or right corner, depending on which side of the plot the cursor is on.



This is what I posted earlier. The new version will have more information.

LordTakubon is helping me test and has asked for some more features, which I am putting in as we go. Next up is going to be a 20 by 20 map of boost (from an optional sensor). It will show min, max, or average boost per cell (indexed by load and RPM) per gear. Like the HP and torque maps, you can select which gear or range of gears to show the data for.

[aaronjb]

You've written some very, very special software there, Kevin.. it's top of my list for when I have the PFC

[Kevin Beane]

Thank you, Aaron. The extra sensors open up new doors. I want to do the video synch.

[Kevin Beane]

Well I guess the good news is I don't have to do any documentation, since there are no customers to support.

----------------


Heads up to all Copilot users: I found a problem this morning. I believe it has been in previous versions of the software. Copilot is not always reading the Function Select booleans properly. It happens more often that not, surprisingly. It reads them at startup. On the Function Select window, it has options for what to do at startup and exit. Mine is set to turn O2 Feedback off at startup and back on at exit. What I've discovered is that some of the other booleans in the same message (and maybe even O2 Feedback) can be read as being false/off, even though they are really true/on. If I click Read multiple times, it will eventually show them as being on. Examples: Knock Warning and Idle-IG Control. Since it turns O2 Feedback off at exit, it has to write the complete message. If it read the others wrong, it will end up turning Knock Warning and Idle-IG Control off at exit! That isn't good. Idle-IG Control really helps smooth out the idle and of course you want Knock Warning to be on. I will look into this and try to fix. It is pretty bizarre, though, since it is a very simple message and the code works the same way no matter how many times it is run... or when. I suspect this is a Datalogit problem, yet it works in FC Edit. Maybe FC Edit reads the message multiple times if it is a known hardware problem? That would make sense. For now, don't use the options on the Function Select window to turn O2 Feedback off or on. If you are not auto tuning, O2 Feedback should be ON and left ON anyway (use the Take No Action option). Then use FC Edit to see if Function Select is the way you want it to be and change it if it is not. Make sure you do a Read All in FC Edit before writing anything. I just did a Read All, changed Function Select booleans, and did an Update. Stupid Datalogit box. We really need a better interface than this POS.

[barkingspyder]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 12 2007, 04:01 AM) <{POST_SNAPBACK}></div>

Well I guess the good news is I don't have to do any documentation, since there are no customers to support.[/b]

You will as soon as I get my PFC installed.

[Kevin Beane]

Darn. Now I have to do documentation!

I'm going to try sending the fricking Read Function Select command three times in a row to work around that problem. And I'm going to beat this stupid interface not responding problem if it is the last thing I do. Stupid Datalogit! ARRRGH!!!

[Kevin Beane]

Well it is up to version 2.0.11 now. I put some changes in that appear to be working for Function Select and also Datalogit comm glitches (a response to glitches... workarounds that have worked perfectly since I put them in this morning). There are more user interface improvements including a Boost map which you can display as average, min, or max, all per gear(s) (like you can for HP and Torque maps). Boost PSI is on the Chart and Dynamometer, too.

[aaronjb]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 12 2007, 03:38 PM) <{POST_SNAPBACK}></div>

Darn. Now I have to do documentation! [/b]

Real men don't need documentation! (Or rather we do.. we just refuse to read it)

Ok, docs would be good - I'll be a buyer just as soon as the PFC turns up

[Kevin Beane]

OK, cool. I didn't make figures yet because I knew I'd be changing the windows. It will be so much work, but really needs to be done. I've always said that.



This is the main window as of right now showing the Fuel tab. I may have to stop adding features! So sad.

[Kevin Beane]

LordTakuban, I've got something new for you in 2.0.12 (just uploaded). I've opened up the gear detection parameters so you can modify, if necessary. On the Monitor tab is a new button. Click it to bring up this window:


These were hidden constants before and you could not disable further learning or clear learned data (well, the last one you could do by changing the Registry).

"Window" means the allowable range (+/-) for RPM/Speed to be in when in a particular gear.

Driving home today I cleared the data to make sure it relearned my gears. It did, although it is now harder to learn 1st gear than it was originally and then I remembered that I had made the "window" for high RPM/Speed ratios smaller because it was indicating 1st gear sometimes while I was shifting (RPM goes up, speed goes down, and it happens to fall into that range for long enough sometimes). It does that less often now. Anyway, I had to have it revved above 3K RPM with speed above 20 MPH (those are conditions required for it to have data good enough to learn). I think you had trouble because of the turbo... or didn't get 100 samples. So now you can make the Large Window bigger (enter a number > 0.1 for the size), but what I would try first is setting the number of consecutive RPM/Speed samples required (text box near the bottom) to a smaller number (maybe 50) and leave it that way for a while. After it has plenty of data for 1st gear, set it back to 100.

Notice on the right that Historical Data is not in order by gears like it is on the Monitor tab. That is the order in which it detected my gears. There is a button to Clear Learned Data (Restart Learning) so you don't have to mess with the Registry. I can't think of any reason to disable further learning other than it may speed up execution ever so slightly.

It might be interesting to note your current values for historical data (same thing is displayed on Monitor tab, but in order) after learning has been taking place for a few days of driving, then compare every few months. If you see different numbers, it could mean your clutch is slipping because the numbers are RPM divided by speed. Clearing learned data would make sure you got fresh data, so do that for this kind of test.

Any more problems, just let me know, but I'm burned out on this right now. Looking forward to seeing your dyno plot.

[LordTakuban]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 14 2007, 05:11 PM) <{POST_SNAPBACK}></div>

LordTakuban, I've got something new for you in 2.0.12 (just uploaded). I've opened up the gear detection parameters so you can modify, if necessary. On the Monitor tab is a new button. Click it to bring up this window:


These were hidden constants before and you could not disable further learning or clear learned data (well, the last one you could do by changing the Registry).

"Window" means the allowable range (+/-) for RPM/Speed to be in when in a particular gear.

Driving home today I cleared the data to make sure it relearned my gears. It did, although it is now harder to learn 1st gear than it was originally and then I remembered that I had made the "window" for high RPM/Speed ratios smaller because it was indicating 1st gear sometimes while I was shifting (RPM goes up, speed goes down, and it happens to fall into that range for long enough sometimes). It does that less often now. Anyway, I had to have it revved above 3K RPM with speed above 20 MPH (those are conditions required for it to have data good enough to learn). I think you had trouble because of the turbo... or didn't get 100 samples. So now you can make the Large Window bigger (enter a number > 0.1 for the size), but what I would try first is setting the number of consecutive RPM/Speed samples required (text box near the bottom) to a smaller number (maybe 50) and leave it that way for a while. After it has plenty of data for 1st gear, set it back to 100.

Notice on the right that Historical Data is not in order by gears like it is on the Monitor tab. That is the order in which it detected my gears. There is a button to Clear Learned Data (Restart Learning) so you don't have to mess with the Registry. I can't think of any reason to disable further learning other than it may speed up execution ever so slightly.

It might be interesting to note your current values for historical data (same thing is displayed on Monitor tab, but in order) after learning has been taking place for a few days of driving, then compare every few months. If you see different numbers, it could mean your clutch is slipping because the numbers are RPM divided by speed. Clearing learned data would make sure you got fresh data, so do that for this kind of test.

Any more problems, just let me know, but I'm burned out on this right now. Looking forward to seeing your dyno plot. [/b]

Nice. I got it and it's installed on the laptop. I'm having some issues logging out of my gauge to the FC Box. I may need to contact Autometer and see about getting a wire kit from them that has the pins for the plug with wires attached. CoPilot didn't read any values coming from my boost gauge, but I think that's due to the wiring. I've determined that I can only do a limited amount of testing and then I have to quit because my intake temps get too hot. I'll have that problem resolved within a week or so hopefully. Then I can really do some logging. I'll tweak the settings and then continue to log to and from work this week to get my gears all set up. I hope my clutch isn't slipping. It's an ACT HD pressure place and street disc.

Anyway, thanks for the additions. The color bar is in the right place for me now with the bigger window. I'm just so damn anxious to get a dyno plot. hahaha. But I'm patient enough to wait until it's safe to really push the car.

[Kevin Beane]

<div class='quotetop'>QUOTE (LordTakuban @ Sep 16 2007, 07:37 PM) <{POST_SNAPBACK}></div>

Nice. I got it and it's installed on the laptop. I'm having some issues logging out of my gauge to the FC Box. I may need to contact Autometer and see about getting a wire kit from them that has the pins for the plug with wires attached. CoPilot didn't read any values coming from my boost gauge, but I think that's due to the wiring. I've determined that I can only do a limited amount of testing and then I have to quit because my intake temps get too hot. I'll have that problem resolved within a week or so hopefully. Then I can really do some logging. I'll tweak the settings and then continue to log to and from work this week to get my gears all set up. I hope my clutch isn't slipping. It's an ACT HD pressure place and street disc.

Anyway, thanks for the additions. The color bar is in the right place for me now with the bigger window. I'm just so damn anxious to get a dyno plot. hahaha. But I'm patient enough to wait until it's safe to really push the car.[/b]

I'm glad you are blazing the trail for logging boost. Can you Y off before it goes to the boost guage? If you can't get it to work I will look into it because we need this to work. I have no way to test, but I can look through my code to see if anything jumps out at me. It is pretty simple and similar to the way other analog inputs are treated.

I have been working on making some bottlenecks more efficient and have had some success, but will hold of on 2.0.13 until all of the optimization is done. The small and large charts update much faster now and the jerkiness is gone. The Fuel tab updates quite a lot faster, too. All this will mean faster data logging and replay.

You should be able to see right away by looking at the chart of RPM/Speed if the clutch is slipping. The RPM/Speed line will curve down more than it should before becoming flat after a gear change because RPM is higher than normal for the speed. Here's one of my charts. The deep blue line is RPM/Speed. The dotted vertical line is the cursor and info on the left is for that cursor position.


The cursor is positioned right after a gear change. I'm not saying your clutch is slipping, but it would make sense in 1st to have clutch and/or wheel slipping in 1st. It could also be that the program didn't see 100 samples in a row or that the window for low gears isn't in the best place for a 5 speed.


You were only missing the detection of 1st gear, correct?


EDIT: The code looks good. If the Datalogit box's AN4 is connected to a 0-5V input, check the box on the I/O Settings window that says "A Boost Sensor is Connected" and don't check the box below that says it should be on IFKit (Phidgets I/O card). It should log boost and also update Boost map, Boost gauge, and Boost should show up in the Chart an Dynamometer.

[LordTakuban]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 16 2007, 07:54 PM) <{POST_SNAPBACK}></div>

I'm glad you are blazing the trail for logging boost. Can you Y off before it goes to the boost guage? If you can't get it to work I will look into it because we need this to work. I have no way to test, but I can look through my code to see if anything jumps out at me. It is pretty simple and similar to the way other analog inputs are treated.

...

You were only missing the detection of 1st gear, correct?


EDIT: The code looks good. If the Datalogit box's AN4 is connected to a 0-5V input, check the box on the I/O Settings window that says "A Boost Sensor is Connected" and don't check the box below that says it should be on IFKit (Phidgets I/O card). It should log boost and also update Boost map, Boost gauge, and Boost should show up in the Chart an Dynamometer.[/b]

The problem is that there are 4 wires coming from the control module that run to the gauge. So I don't know which one of those wires is the sending signal for boost. There are pin outs on the plug that are supposed to go to Autometer's logger, but I couldn't find a pin that would slid in there at radioshack. Another thing that may be of interest is that the gauge self-calibrates itself whenever it gets power. So I think that the signal may need to come from the gauge. I will contact Autometer or someone today and see how I can log out of the gauge/or controller.

I'm going to Y off of my oil temp sender wire too and run that to the FC box as well and start logging oil temps.

With the gear stuff, It doesn't seem to save my setting for reducing the number of samples to acquire a gear ratio. When I tried to get 1st gear before, i was on jack stands and ran the car at above 3K RPM for about 20 to 30 seconds. The speed may not have been high enough though. I wiped the numbers and will try to log on my way home today and see what happens. Once I get those ratio's set, then I should start seeing more goodness.

Hopefully, my meth injection kit comes in this week and I'll get that installed on Saturday and can finally start really logging some WOT runs without worrying.

[Kevin Beane]

<div class='quotetop'>QUOTE </div>

With the gear stuff, It doesn't seem to save my setting for reducing the number of samples to acquire a gear ratio. When I tried to get 1st gear before, i was on jack stands and ran the car at above 3K RPM for about 20 to 30 seconds. The speed may not have been high enough though.[/b]

Must be an oversight on my part... not saving your setting for # samples. It was a rush job... will fix soon. If you could email a log to me of when it was in 1st gear I will replay it and fiddle with those parameters until it finds the gear. Large Window size is what I think needs adjusting. Isn't 1st gear about the same on 5 speeds and 6 speeds?

I'd like it to find 1st gear automatically, but you can also right-click on RPM/Speed on the Chart tab when it is in 1st gear and select Designate blah blah blah as 1st gear from the popup menu. Then you will have to use the Use Override Values check box, though. It will work for now.

[CIN]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 17 2007, 02:55 PM) <{POST_SNAPBACK}></div>

<div class='quotetop'>QUOTE

With the gear stuff, It doesn't seem to save my setting for reducing the number of samples to acquire a gear ratio. When I tried to get 1st gear before, i was on jack stands and ran the car at above 3K RPM for about 20 to 30 seconds. The speed may not have been high enough though.[/b]

Must be an oversight on my part... not saving your setting for # samples. It was a rush job... will fix soon. If you could email a log to me of when it was in 1st gear I will replay it and fiddle with those parameters until it finds the gear. Large Window size is what I think needs adjusting. Isn't 1st gear about the same on 5 speeds and 6 speeds?

I'd like it to find 1st gear automatically, but you can also right-click on RPM/Speed on the Chart tab when it is in 1st gear and select Designate blah blah blah as 1st gear from the popup menu. Then you will have to use the Use Override Values check box, though. It will work for now.
[/b][/quote]

Kev I never donated but please let me be your first customer. Can you send me the payment details by PM please?

Kind regards
Joseph

[Kevin Beane]

Thanks, CIN. All you have to do is download and install. When you run the program before it being licensed it will guide you through the process of sending the Product ID and getting a Key. Part II of the document explains in more detail what to do.

http://www.mr-s.org/copilot/Copilot_2_0_8.doc

The current version of the program download is here:

http://www.mr-s.org/copilot/Copilot_2_0_13.zip

In case anyone was wondering about this, Tem and I are splitting whatever we make on this 50-50.

2.0.13 has some nice changes:
- Some code was optimized, making Online Mode and Replay faster to update with new data. Both charts were made smaller, the one on the Dashboard tab
can be resized by clicking inside the chart and dragging left or right. It redraws the chart when the mouse is released and the change in width is the distance
the mouse moved left or right. I tried redrawing it as the mouse moved, but it was taking too long and that would affect data logging rate. The small chart
on the Dashboard tab updates smoothly (not jerky), even Replay. The large on updates smoothly when the Replay speed is less than 16X (it cannot not keep
up at higher speeds and I don't think I can optimize it further).
- Check boxes were added on the big chart (Chart tab) to enable/disable display of each item.
- Changed the color key for Number of Samples to show a more useful color range (it isn't linear, unlike the others)
- Fixed a problem with LC-1 port not opening automatically on startup. (I caused this before 2.0.8, when I moved all the I/O controls to a different window.)
- When there is a PFC timeout and the last thing sent was a Write command (command with data), it will resend not just the command
(which would be a read), but also the data that goes with it (making it a write command like it should be). I thought this was important due to there being
more occasions when data is written (sometimes automatically)
- Bug fix: It was not saving and restoring what the person had set for "O2 Feedback Control Changes at Start and Exit" on the Function Select window.

[LordTakuban]

<div class='quotetop'>QUOTE (Kevin Beane @ Sep 17 2007, 08:55 AM) <{POST_SNAPBACK}></div>

Must be an oversight on my part... not saving your setting for # samples. It was a rush job... will fix soon.[/b]

Don't worry about getting the fix done too quickly. I just modified the registry to have it be 50 instead of 100.

I'll send a log of my drive home today after I load it up and see what all it has.

[Kevin Beane]

Justin, I got your files and was able to replay the log no problem. Replaying someone else's files is one of the nicer things you can do, but it makes the software a bit complex. I added code to save the Gear Detection Parameters in the Registry, but it was in with other code that isn't executed in Replay mode. (You don't want to save much of anything in Replay mode because if you are replaying someone else's data, you don't want it to affect anything on your computer.) So what I did to enable people to work out the gear detection parameters when replaying their own data is to have it save these and a few other things, as long as you do not exit the program via Quit Without Saving. I need to do some more work in this area and maybe come up with something easier, like a way to Copilot when starting up that you do not want anything saved on your machine.

Anyway, here is a chart showing just RPM, Speed, and RPM/Speed for 1st - 3rd gears. (The check boxes on the left are new, BTW.)



Here is one of my logs replaying, also showing 1st - 3rd gears.





Notice:
1. 1st gear on mine has RPM/Speed well above 190, which is why I made the setting that defines the end of the Medium Window 190. I see a min. of 232 there, so that is 1st gear. That number should also work for your car (and all 5 speeds).
2. On mine you can see that RPM/Speed varies quite a bit in 1st. This is to be expected because speed is a small number and only accurate to 1 k/hr. For some reason, on your plot, RPM/Speed has a downward trend. This means that RPM isn't increasing as much as speed. If the clutch was slipping later on, RPM/Speed would be rising. But what is happening is the clutch and/or wheels are slipping at first (RPM higher than it should be for the speed), then the ratio "gradually" (we aren't talking about much time... 2 1/2 - 3 seconds) drops down to where it should be for the gear. The default settings for Gear Detection Parameters have 0.1 for Large Window Size. This means that it will have to see a series of X (100 by default) samples all within +/- 10% (0.1 times the start value) of the start value. You may need to make Large Window Size something like 0.15. If you make it too big, though, it will think it is in 1st when you are really shifting.
3. After the shift into 2nd, there is a little hump on your graph and mine. That's just the clutch hooking up.
4. Notice in between shifts what happens to the blue line. It falls out of one window (where a particular gear would be selected) and into another one. Depending on how you shift, it could be all squirrely. Copilot looks for this and ignores knock that happens during a shift. Well, it doesn't completely ignore it.... knock goes into a False Knock map. Just a little FYI.

When I replayed your log I could not use the Designate Gear function because in Replay, the gear number comes from the log. The info I designated would be saved in my Registry, but it would not affect what is displayed for Gear during Replay. Thanks for your patience. We really had to get this working for everyone.

I noticed some other things while Replaying that you might want to change.

1. Your Deg from Knock map is all 0s. That is fine, actually, as long as you are not trying to auto tune ignition timing (online or offine). When it sees knock over the threshold you set (see Setup tab), it will ADD the number for the cell with knock and the number in the current ignition map cell and put the result in the new ignition map cell. The Deg from Knock map should have negative numbers.
2. Under Auto Tuning Required Conditions on the Setup tab, you have min water temp set to 120 or 140. I had mine set to 140 until the other day. I raised it to 160 because it seemed to me that the engine was still running rich at 140 (after starting it in the morning).
3. You Max Air Temp is 140 and that is going to be too low for your PE turbo. I was thinking NA before when I set the default max. Now you can set the limit up as high as 280 I think. With your PE turbo, I imagine intake air temp is going to be higher than 140. Look on the Monitor tab after replaying a file. It shows the min and max it saw.
4. Problem area: TPS % is off. The min TPS I saw was 15%. I did not notice the max and wasn't sure if you had it WOT anyway. TPS % is calcuated based on min VTA V and max VTA V. VTA V is an input to the Power FC. It varies from car to car and is self-calibrating. Copilot keeps track of the min and max it ever sees. On the Monitor tab, it shows these numbers near the bottom. If something gets messed up, for example the Datalogit sends bad data (there are better "sanity checks" now, so it is unlikely this will be a problem in the future), it will throw off the TPS calculation. So what you should do is open your log and find the min and max values of VTA V (this is logged in addition to TPS %). You'd want the throttle wide open at some point while logging. Then on the Monitor tab, enter those numbers in the override boxes and check the "Use these override values..." box. That will fix TPS %. This has to be right for the HP and Torque plots. Also the dyno. Under Settings --> HP/Torque and Dyno Settings, there is a parameter for "Min Throttle Percent for HP and Torque "maps" and Dyno". The default is 98%. This means it will not put data in the HP and Torque maps or generate a dyno file/plot unless the throttle is at least 98%. If it did not have this check, the maps would be filled up with useless data. To verify you have the right numbers for VTA V min and max, watch the speedometer. There is a bar graph showing TPS %. It should be 0 when your foot is off the gas and close to or at 100% when at WOT.

All that sounds like a PITA, but you only have to do it once. I think I will add some lines to the chart that are displayed when the Gear Detection Parameters window is open and show where the parameter are.

Thanks again for all the Beta Testing.