Copilot download!!

[greenlabel]

Damn it features auto-tuning, etc.. check out spyderchat.com

[kortik]

nice

[Jesse IL]

El reposto:

http://www.newcelica.org/forums/showthread.php?t=199164

http://www.newcelica.org/forums/showthread.php?t=204670

I've spoken with the guy writing that program. By his own admission, he is a programmer, not a tuner. He has another guy trying to help him with the tuning aspect. From what I have seen of the program, it seems like it does a good job making sense of things for people who are completely confused by FC Edit. If you're an advanced FC Edit user, this program won't help you much as it doesn't really do anything that FC Edit doesn't also do (albeit in a more cryptic way). Think of it like DOS -> Windows 3.x You're still running DOS is the background, and people who were good with DOS continued to use it for a long time.

About the auto-tuning, it may help you but I would say at this point it is rough at best. I've offered suggestions about tuning methods but really got limited feedback. IMO, the auto-tuning requires that you start from a pretty good map. It won't turn garbage into a fully tuned car, so don't think about it like that. The other problem is that as development continues, it becomes more advanced and has all sorts of extra settings like selectively ignoring inputs, which if you don't know anything about tuning, you'll be at a loss to set.

Bottom line is, the auto-tuning is far from a foolproof system at this point, and is not something that your average guy could turn on and magically have tune his car. Programmable ECU's are very advanced pieces of equipment, and I fear that in the last two years, people who have no business using them have been lulled into a false sense of empowerment by programs like this. I still feel the best use of that auto-tuning would be for part-throttle cruising, to get maximum fuel economy. WOT, I'd still do the old fashioned way.

Also, if you're interested in that program, there is another one called FC Tune that is being developed by a guy in Germany. Its pretty slick, but at this point has no Toyota applications. There's a link to it in a sticky above. Perhaps I should link this program as well.

[blinding-gts]

I see Beanie has been busy. As for tuning the best way I found to get a good map is start with the idle zone and then work up from part throttle to full throttle logs. You're working on getting the AFR right for each cell. Once you have that around where you want it you want to make sure you watch for knock and adjust IGN timing accordingly. After that I finish off with tuning the VVT map which the stickies provide a nice walkthrough.

[Jesse IL]

I see Beanie has been busy. As for tuning the best way I found to get a good map is start with the idle zone and then work up from part throttle to full throttle logs. You're working on getting the AFR right for each cell. Once you have that around where you want it you want to make sure you watch for knock and adjust IGN timing accordingly. After that I finish off with tuning the VVT map which the stickies provide a nice walkthrough.

I actually have a new method that is much better than that. I'd agree on doing the idle first and doing a rough setup of your map. But then I ignore part-throttle and focus on full throttle. You need at least a working VVT map, but I like to tune to a rich/safe AFR, get to two degrees off knock, then go through a round of VVT tuning, incrementing by 3 each run and then making a rough map based on logged PIM values.

From there, you work down to your target AFR, work to your best power IGN and then revise your VVT map. Your VVT map will be very close and will probably only require slight revision. That will be your final VVT map.

From there, you can tweak fuel in case you changed VVT and AFR was affected. At that point you'll be at idealized fuel and VVT, and you might ply with your IGN by a degree or two to look for any extra power.

Once you have maximum power under control, now start looking at making it cruise nice and get good gas mileage. The reason I say do this last is because you don't want to mess with your max power settings, so you do as much as you can at cruise for fuel economy without upsetting your max power fueling.

If you want to get real advanced, here is a trick: actually use the INJ vs. TPS setting, don't leave it at 1.000. To explain this better, you have a certain range of load cells that get used at idle and cruise. You have a certain range of load cells that get used at WOT. Then you have a crossover range where you have part throttle acceleration. What I've found is that I can be on the highway and roll into part throttle acceleration, I'll shoot up to fairly high load rows. If you try to keep that lean since you're just slightly accelerating in traffic, you risk leaning out the car at WOT. By using the INJ vs. TPS correctly, you can keep it lean under part-throttle acceleration and enrich it only when you're really on it.

[blinding-gts]

If you want to get real advanced, here is a trick: actually use the INJ vs. TPS setting, don't leave it at 1.000. To explain this better, you have a certain range of load cells that get used at idle and cruise. You have a certain range of load cells that get used at WOT. Then you have a crossover range where you have part throttle acceleration. What I've found is that I can be on the highway and roll into part throttle acceleration, I'll shoot up to fairly high load rows. If you try to keep that lean since you're just slightly accelerating in traffic, you risk leaning out the car at WOT. By using the INJ vs. TPS correctly, you can keep it lean under part-throttle acceleration and enrich it only when you're really on it.


I played around with that around September last year and ended up enriching the AFR a lot. :sadpace: It's a really sensative setting to play around with but I never gave it much more then a few trial runs. I think I'll tweak with it again once I have the turbo all installed and running with a safe map. Thanks for the input Jesse.

[Jesse IL]

Well the trick to the INJ vs. TPS (now keep in mind that I've never messed with it yet) is that you have to define the percentages at certain throttle inputs. FC Edit would define 0v as 0% and 5v as 100%, but in reality you never see those values.

On the GT-S I've been working on tuning, I get 0.520-0.525v at idle (10.4-10.5%) and 3.98-3.99v at WOT (79.6-79.8%).

What you would want to do is take your car as it is today, get the TPS voltage and idle and WOT, then do some driving and figure out what kind of TPS voltages you're getting at part-throttle acceleration. Of course, this is extremely subjective, but I sort of know from running on the stock ECU with my wideband installed that you can roll into the throttle on the highway to modulate your speed in 6th, and the ECU will keep the car right at stoich. At a certain throttle percentage, it will start to richen things up. The best way to figure this out would be with the stock ECU and an OBD scanner. You could sit on the highway and at certain speed, look at AFR vs. TPS voltage.

In fact, I was going to try and offer jlitman some money to borrow his scanner, as my car is currently converted to NA and I'm trying to do some troubleshooting.

[Beanie]

Make sure you only use version 1.4.1 for rescaling maps. There was a bad bug in 1.4.0.

Nope, it's not magic. It is just doing some calcuations for you. It is up to you how you use that information.

Hope you all get some benefit from it and although I will have to take another break from it to work on some other projects, keep an eye out for updates and don't hesitate to suggest improvements.

Thanks.

[Beanie]

One correction to what JesseIL said: I have two people helping me with the tuning aspect of it: LittleRocket and Tem (both of Spyderchat).

I don't want to get into a pissing contest over this program or give the impression that I don't respect someone like JesseIL. I'll take all the input I can get. There is a LONG list of things planned for it. I've decided to put another project on hold and work on this some more. There are more auto tuning features (such as dealing with knock as you drive) that will be coming at some point in time. I have some of that working (including responding to knock and knowing when it is false knock or real knock), but it is not in the "Lite" software because we plan to put that and some other things in hardware/firmware.

As I said over on Spyderchat, there is more to this program than tuning, for example gauges, lateral G meter, GPS, replaying your logs (graphically... just like it happened when it was recorded) including replaying at different speeds, alarms, graphs, etc. I have many ideas for things to add and there are actually a lot of new features "waiting in the wings" (mostly coded, but not integrated yet).

[Jesse IL]

I don't want to get into a pissing contest over this program or give the impression that I don't respect someone like JesseIL. I'll take all the input I can get.
No offense taken. I've for the most part given up on trying to explain tuning to people, unless they are pretty proficient to begin with. I tried to explain VVT tuning to someone the other day in person and it was a huge challenge. Without him being able to constantly stop me and ask for clarification, it probably wouldn't have gotten anywhere.
As I said over on Spyderchat, there is more to this program than tuning, for example gauges, lateral G meter, GPS, replaying your logs (graphically... just like it happened when it was recorded) including replaying at different speeds, alarms, graphs, etc.
Which is why I made the DOS vs. Windows analogy. For me, gauges, alarms, playback of logs...it doesn't help me in any way. To be quite honest, I think watching logs in realtime feedback would hurt my understanding of what's going on. I think you'll get more out of looking at a static chart where you can zoom into the area of interest and spend as long as you want scrutinizing. For me to find a program like Copilot useful, it would have to automate all my labor-intensive processes, which it does not. But that's me, hopefully for other people, incorporating all the displays will help them to better understand.

[Beanie]

I'm putting some more enhancements over the next week or so. I don't know how helpful you will find them, Jesse. Graphic curve editing for one thing.

Boy, you guys have quite the collection of smilies over here. What is this one for? :monkey: Monkeywrench Racing?

[Jesse IL]

Actually, its a miniaturized version of a really old smiley. Its for stupid people who are "monkeys".

[Beanie]

There's a new version uploaded (2.0.1)

http://www.mr-s.org/copilot/Copilot_2_0_1_Lite.zip

Always look here for the latest stuff:

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

[Beanie]

I just updated the link above to the latest version. It has many new features, including horsepower and torque plots that, like other plots, update as you drive (or later, as you replay a log). Celica owners should create their own vehicle parameters file (go to Settings-->HP/Torque, enter Cd, frontal area, and mass, then save as MyCelica.CSV or whatever). It uses these constants when doing the HP and torque calculations. There are quite a few other improvements in displays and data logging (logging is 2-3 times better than before in terms of samples per second). More things were planned for this release, like much improved smoothing... and much less time to achieve target air/fuel ratios, but I felt like that code needed more testing.

Examples:

From a test session of my NA 2ZZ-GE powered MR2 Spyder. I think I've seen a little higher than this. It will be better after incorporating JesseIL's changes in with mine, I'm sure.

http://www.mr-s.org/copilot/HPgraph2.jpg
Notice:
- dotted vertical white line is the cursor. Data on left is for that point in time.
- Air temperature is flat. This was on a warm day and I'd been driving pretty hard. Why so flat? That's my super duper insulated, dual inlet air box. Never gets hot.
- Injector % is just a tad under the level that would trigger a CEL (90%). So I don't need bigger injectors yet. It is also quite rich at those points. I think it needs more air. May be a candidate for a throttle body bore.
- The first big hump (HP) is 3rd gear and the second one is 4th. Makes good power over a decent RPM range in 4th.

http://www.mr-s.org/copilot/HPmap.jpg
Here's the same data as a "map". That's showing just 3rd gear (see drop down list at bottom... you can select any gear or things like 3-4, 3-5, 4-5, etc)

It shows WHP, by the way. There is a window for HP/Torque settings where you can enter a Fudge Factor (any percentage to add, like your drivetrain loss or whatever you want). This is with nothing added, so it is wheel HP. I hope. ;)

[bp_me]

such as dealing with knock as you drive) that will be coming at some point in time. I have some of that working (including responding to knock and knowing when it is false knock or real knock), but it is not in the "Lite" software because we plan to put that and some other things in hardware/firmware.


A newbie question prehaps, but I haven't seen mention anywhere before of an ability to change the PFC maps in realtime? Can this actually be done?

[Jesse IL]

such as dealing with knock as you drive) that will be coming at some point in time. I have some of that working (including responding to knock and knowing when it is false knock or real knock), but it is not in the "Lite" software because we plan to put that and some other things in hardware/firmware.


A newbie question prehaps, but I haven't seen mention anywhere before of an ability to change the PFC maps in realtime? Can this actually be done?

I've read through all of Beanie's Copilot threads on Spyderchat. I'm 99% sure it can't do this.

[bp_me]

I've read through all of Beanie's Copilot threads on Spyderchat. I'm 99% sure it can't do this.

If thats the case, how can it deal with knock on the fly? Unless the PFC can be programmed to retard timing when it detects knock above a certain threshold, but it was my understanding that this was not the case.

[enjoi4586]

If you want something thats going to deal with knock in real time, all the time, then I believe you would be talking about Beanie's 'gadget' (for lack of a better term at this point). Check it out

http://www.spyderchat.com/phpBB2/viewtopic.php?t=35157

[blinding-gts]

I've read through all of Beanie's Copilot threads on Spyderchat. I'm 99% sure it can't do this.

If thats the case, how can it deal with knock on the fly? Unless the PFC can be programmed to retard timing when it detects knock above a certain threshold, but it was my understanding that this was not the case.

You CAN retard knock via the commander in real-time. You can program software to detect the knock levels and retard timing for the cells where the knock is occurring. I don't know if Beanie's software does it using a similar routine but I don't see why it wouldn't be possible. Also, that's normally a bandaid solution since you could have a bad AFR causing the knocking, VVT setting. etc which the software would probably not be trying to isolate.

[enjoi4586]

I've read through all of Beanie's Copilot threads on Spyderchat. I'm 99% sure it can't do this.

If thats the case, how can it deal with knock on the fly? Unless the PFC can be programmed to retard timing when it detects knock above a certain threshold, but it was my understanding that this was not the case.

You CAN retard knock via the commander in real-time. You can program software to detect the knock levels and retard timing for the cells where the knock is occurring. I don't know if Beanie's software does it using a similar routine but I don't see why it wouldn't be possible. Also, that's normally a bandaid solution since you could have a bad AFR causing the knocking, VVT setting. etc which the software would probably not be trying to isolate.


Beanies Gadget could do that, but I'm not even sure if he's still working on it anymore. The last thing he said about it was back in January or something like that. Seemed like a really good idea, but thats after he shows a price tag. But I suppose if your budget permits it. :shrugs:

[david.mcmahon1]

holy ****- i feel like a caveman reading this thread..all i can say is WHAT????

http://img87.imageshack.us/img87/2438/picturestoyota008yw1.jpg

[Beanie]

There is a new version already. 2.0.1 had a graphics bug. Oops.

LATEST VERSION: http://www.mr-s.org/copilot/Copilot_2_0_4_Lite.zip


Copilot did (before I disabled it) adjust maps as you drive. It would adjust the base map to hit target AFRs -OR- adjust timing (advancing and retarding) on the fly, all within safe limits and parameters you set. You had to tell it which you wanted it to deal with, fuel or timing. The idea was to set up a target AFR map and have it auto tune to hit that, then let it adjust timing. You could also have it do the timing first, if you wanted. I put a different intake in my car (different diameter, length, different type of filter) and drove for 2 hours, getting samples in most of the cells I would ever see, with the online auto tuning (AFR only) enabled, and it zeroed in on the target AFR map again. One problem was that it didn't deal with knock (other than log it) as you drove while it was busy dealing with AFRs. For one thing, the coding got complicated. For another thing, I didn't think it was all that safe to adjust timing on the fly (without getting "approval" for changes from the user, as would be the case when doing the auto tuning offline) when using a Datalogit, either, because Datalogit spits out bad packets once in a while. Finally, I did not have a good algorithm for doing both at the same time, since one affects the other.

Now about the gadgets (plural): The gadget idea started out as something ultracheap, which would not have cost much to make, but I decided I wanted something a whole lot better with tons of I/O capability like USB, Ethernet, CAN 2.0B (for an Elise), an LCD, a alphanumeric dot matrix LED, buttons, etc for my gadget. So I started building the ultimate gadget, which is still in progress. At the same time, my partner and I were working on the less fancy solution (he did some, I did some), but I passed that part of it off to him to work on so I could a) continue working on my fancy one (except for the Power FC interface part) and b) try out some new algorithms in Copilot that would go into both types of gadget. I also wanted to put some gee whiz features in that people (like enjoi###) had asked for. My partner is hard at work on the less expensive gadget, by the way. My fancy gadget (the white box) will not be for sale unless I can get the company to make them, which would be a generic microcontroller development system for engineers like me who want to get started in microcontroller development or prototyping. It would probably sell for about $900 because the parts alone are about $500. I doubt many people would want to pony up 900 bucks for it in a car like a Celica or MR2 (maybe an Elise, though), whereas there is a market for it as a microcontroller development system. Nevermind that white box gadget, in other words.

My partner and I have the Power FC interface mostly hacked. The less expensive gadget will do everything a Datalogit does, plus have auto tuning (taking care of fuel and timing) and some other neat features (I don't want to say exactly what yet). It will be more reliable (communications dropouts nonexistent) and faster... quite a lot faster. I think we can sell them for quite a bit less, too. Don't ask how much because we have not decided.

Meanwhile, I figured out how to get better auto tuning results (much smoother maps and much less time to do it.. and have implemented that in Copilot (did that a few weeks ago), but we can't give that away in software that works with Datalogit because that would compromise sales of the inexpensive gadget (we still don't have a name for it, but it might just be "Power FC Gadget"). We have money invested in making the gadget, so it just would not make sense to subvert that. Make sense? So the Copilot that is released for anyone to use free doesn't have those auto tuning features. What you can download free... and always free... is a "Lite" version of Copilot.

[Jesse IL]

If you want to retard ignition timing on the fly due to knock, I would HIGHLY recommend you go with a J&S Safeguard knock sensor. I wouldn't trust the factory knock sensor with the task.

[Beanie]

J&S Safeguard with a different knock sensor is over $500, right?

We wouldn't be trying to advance timing to the bleeding edge anyway. We'd be using the knock sensor as feedback... to know how far timing can be advanced safely. Advancing timing is also something that could be disabled... if you don't want it to advance timing at all, don't do it. You could just have it retard timing when it thinks there is a potential problem. Think of it as a Datalogit with more features. Why so negative???

[enjoi4586]

Beanie, have you tryed FC-Edit's new universal software ? If not, you should really give it a try. And If you can't get it, just PM me.

[Beanie]

Yes, I have it. They started incorporating some of the things I did. I've noticed that people keep asking for more things Copilot already has, too. ;)

[Jesse IL]

J&S Safeguard with a different knock sensor is over $500, right?

It also works about 1000X better than the OEM equipment.

[Beanie]

I know about the individual cylinder retard in the J&S. You'd want that to get the most power because you wouldn't have to retard ALL cylinders by the same amount. I really don't think anything the works with the Power FC could do anything per cylinder. I would want a J&S if I had a turbo (might actually do that someday) to keep it safe while getting the max power. But we are talking about two different things, in a sense. I visualize the gadget as something that would find out how when it is too advanced per cell and correct that, just as a person would do when tuning their car completely manually, yet having it happen very quickly and automatically. The gadget would respond very quickly. We aren't sure how quickly yet... would have to test (and we will give the test results). I've let Copilot retard my timing before and the knock I had... just in certain cells... went away. There is a setting for what knock level you want it to start retarding and another for how many degrees when it sees that. Doing it in hardware makes a lot of sense because software, particularly Windows software, doesn't have a guaranteed response time, plus there is a delay getting the data back and forth, plus the danger of bad, missing, or late packets. The gadget will be much better than what already did the job for me (normally aspirated 2ZZ) in software.

[Beanie]

Spyderchat is down, so I'm posting over here. The next release is going to knock your socks off. I had a three really good ideas and got two of them coded and tested. The third one I'm going to have to put on hold because it will take quite a bit of work and I got an even better idea today, something that won't even be that hard to do. I intended to do it earlier, but couldn't figure out an easy way to do it. Anyway, I will be expanding up the Curve Editor to allow you to see the map (and curves) from a different angle or in 3D. Right now the X axis is RPM, so I will add the capability to make the X axis be Load, plus have a 3D view. As before, changes you make by dragging points up and down will go into effect (be written to the Power FC) as you drag (if that's what you want it to do). I will probably get rid of the old 3D viewer because it isn't interactive and I would be hard pressed to make it that way. Oh, one of the ideas I already have working is resizing the main window. I use an 800 by 600 tablet PC in my car, but when I run Copilot in my house (to replay and edit offline), I have a much higher resolution monitor. Boy, seeing the maps and graphs in a larger size is great. There will probably be a new version out there within a week with all this and more.

[Beanie]

LATEST VERSION: http://www.mr-s.org/copilot/Copilot_2_0_4_Lite.zip

It has a smoothing function now. See the topic on spyderchat mentioned in the first post here for more information.

[Beanie]

It has been quite a while since I've posted here and a lot has happened. Copilot for Power FC went commercial and we made a few sales, but we decided to make it free again. Go here (http://spyderchat.com/forums/index.php?showtopic=45059) for more information. Disregard the fact that this is in the Commercial Sales forum.

The auto tuning works really well and there are all kinds of other new features. I'll be adding some more things in the coming weeks and then the software will become open source, except for of the components, possibly. It's all in VB6.

[blinding-gts]

Thanks for the update Kevin. I'll give it a test run as soon as this snow storm here dies down. :gap:

[shadytunerGTS]

This looks somewhat promising. Was wondering if anyone on newcelica.org has tried it?

[Beanie]

No, they havn't tried anything after it went commercial because they would have had to have purchased a license. It is very much improved compared to the older versions. The User's Guide is way behind but people are better off just reading that and then reading through the topic on Spyderchat to see what other people did than ask questions (especially here because I don't sign on here very often). I'm also going to be quite busy finishing up some things, including the document.

By the way, it can adjust the base map to hit target AFRs and retard timing in response to knock as you drive, plus a whole lot of other things "at the same time".

[Jesse IL]

Couple questions:

1. Is knock threshold used as a single value like in FC Edit or are you able to define a knock calibration curve a la the AEM Pro software? I've found the single threshold problematic because the knock sensor seems to basically respond in an exponential manner with rpm (single-wire sensor on the 2000-01 cars).

2. Does the auto-tuning consider the displayed values from the 20x20 grid as absolutes? I've found that treating these values as absolutes becomes problematic when you are attempting to fine tune the car. Because FC Edit has to choose a cell to display a value in and the load and rpm spacing is fairly large, the results will often be misleading. This is exacerbated by the fact that the AFR log will be fairly jagged, so the values shown in the grid will often not be representative of what's going on. I've found that when I'm attempting to fine tune a car, I like to look at the actual log curve and guestimate an average AFR over a range, sort of like a mental smoothing factor.

3. Is there any functionality that considers keeping fueling proportional to load? The biggest danger I've found of running constantly through the same cells and tuning them is that you can a strange fueling map, where you can actually have more fuel at less load. I typically work around this by changing an entire column of fuel values by a percentage while rough tuning and then making smaller tweaks to individual cells when fine tuning.

[Beanie]

Couple questions:

1. Is knock threshold used as a single value like in FC Edit or are you able to define a knock calibration curve a la the AEM Pro software? I've found the single threshold problematic because the knock sensor seems to basically respond in an exponential manner with rpm (single-wire sensor on the 2000-01 cars).

2. Does the auto-tuning consider the displayed values from the 20x20 grid as absolutes? I've found that treating these values as absolutes becomes problematic when you are attempting to fine tune the car. Because FC Edit has to choose a cell to display a value in and the load and rpm spacing is fairly large, the results will often be misleading. This is exacerbated by the fact that the AFR log will be fairly jagged, so the values shown in the grid will often not be representative of what's going on. I've found that when I'm attempting to fine tune a car, I like to look at the actual log curve and guestimate an average AFR over a range, sort of like a mental smoothing factor.

3. Is there any functionality that considers keeping fueling proportional to load? The biggest danger I've found of running constantly through the same cells and tuning them is that you can a strange fueling map, where you can actually have more fuel at less load. I typically work around this by changing an entire column of fuel values by a percentage while rough tuning and then making smaller tweaks to individual cells when fine tuning.


Hi, Jesse.

1. There are 3 different knock thresholds you can define in Copilot. One, which should be the highest, will cause timing to be retarded according to how high the knock is. It does not depend on RPM or load, but implementing that wouldn't be a problem (and I may do that). Anyway, when this high threshold is exceeded, Copilot ignores cell protection (cell protection is something inside Copilot... you can protect cells from being changed by yourself and the tuning algorithm). Then there is another threshold that works the same way, but cell protection is observed. Finally, there is a low threshold you set to something like 6 and if the knock is below that, it will just assume that any knock there is insignificant (noise). The Find Insignificant Knock code is pretty neat. It will retard timing in cells above this lowest threshold (if you enable it) until retarding it does not cause the knock to be reduced. That's explained in more detail in the topic.

2.

a.) As far as the base map goes, it looks at the RPM and Load values and treats every sample's "address" that way, not as a 20 by 20 cell index. It is quite complex. The program considers when and where the AFR sample was obtained (very precisely) and uses a map of logged AFR that has 9 times the resolution of a Power FC map PLUS a Logged AFR map that is the normal 20 by 20 resolution. Each 20 by 20 cell is subdivided into 9 subcells of different sizes/shapes. Each AFR sample affects a subcell and the "strength" of the sample (in terms of averaging with others in the same subcell and then later on with other subcell averages) is a real number based on distance from the center of the real (20 by 20) Power FC cell. Also, the program estimates what the AFR would be if it were obtained right in the center of the real cell and compensates for the fact that it was not... and was influenced by neighboring cells. It knows how each neighbor affected it because of the Power FC's interpolation (it does the same thing, in a way). You can see this higher resolution logged AFR map if you want by checking a box named "Detail". (See Lord Takuban's post a few pages back from the latest one.) Whenever it calculates the effect of an AFR sample on this imaginary 9X resolution Logged AFR map, it recalculates the normal cell in the 20 by 20 Logged AFR map based on balancing the subcells. So it ends up being very much like there are no cells at all (not a discrete method), but rather a continuous 3D surface. If you look at what Lord Takuban (silly username) posted, you will see what I'm talking about. It looks kind of like a topographic map and clearly is not just some blocky 20 by 20 estimate. Here's another analogy: it looks like a higher res image than one made with a cheaper camera or the same camera set to a lower quality. But when you switch to the 20 by 20 representation ("Detail" not checked) you can see that the edges of the cells don't have very abrupt and/or random changes (unless your Target AFR map did) as they would if your samples were not obtained right in the middle of each cell, so it is handling AFR samples obtained towards the edges in a sensible way. It took quite a long time to work all that out.

b.) For knock, that information is associated with a particular cell and it will retard timing just in that one cell, not neighboring cells. That's something that could be changed/improved, too, but I just got the whole Find Insignificant Knock code working with this somewhat simplified approach. Affecting other cells surrounding it, too, would be better. If you get the knock taken care of while driving a particular way (not increasing RPM or load extremely quickly), the code I have now is fine. I have not tried this during "spirited driving". There's always room for improvement, you know?

3. There is a "Base from Formula" map (just an imaginary map inside Copilot) that you can build using 3 sliders to create a smooth map for reference purposes. Before I put that feature in, I tried all kinds of things (different formulas) to come up with a prediction of what the AFR would be, based on the maps and some samples. The formula could also be used to predict what the AFR would be based on the maps. That was the idea, anyway. I gave up on it... and found that it wasn't necessary anyway. Copilot doesn't force the fueling (base map) to be proportional to load, though, and there is no need to. Provided it has enough data (samples) to go on, it can adjust the whole base map to conform to the target. If the target AFR map is smooth, the base map will be smooth. Note: It depends on how you drive, how smooth it is. I am still trying to work out how to compensate for how you drive (throttle being a big influence), but I'll tell you, what you end up with if you just drive the way you normally do is a base map that gives you the AFRs you want to have driving that way, if you know what I mean. And you can always go in manually and use the smooth button (it does progressive smoothing of an area of cells, each time you click the smooth button it gets smoother) and/or drag the points in the curves up and down. You can also select any area you want and bump it up/down with the keyboard arrow keys or use the Edit button and add/subtract/multilpy/divide all by some number. Finally, you can temporarily change the fuel injection factors like Inj vs Accel TPS so they don't have any effect while auto tuning. I haven't done that, but I'm not as particular as some people about smoothness. The automatic process, where all it is trying to do is adjust each base map cell so the logged AFR map converges on the Target AFR map, works very well. It has been greatly improved over what I had before. You don't even have to use a Holdoff value. It is able to utilize every single AFR sample obtained.

[Jesse IL]

1. I did some research into how the knock sensor behaves and found that it is highly dependant on rpm and should be calibrated for every single car. If you've ever experienced the situation where you hear knock and the knock warning doesn't flash, this will explain it. I basically found that the knock threshold is different for each rpm.

http://www.newcelica.org/forums/showthread.php?t=247596

If you look at the AEM Pro software (what you use to tune their EMS), you'll find that you define a baseline knock curve and then a certain level above that that can be considered normal operation of the sensor. You can clearly see this noise from my data. Basically the problem is that setting any knock warning that doesn't flash constantly will only warn of knock at high rpm. If you were looking to make a further enhancement of your program, this would be the very first thing I'd do. Anyone can download the AEM Pro software, so you may look to DL it and see how their knock control works.

2. That is very impressive. As I'm sure you're aware, the PFC pulls from four cells to calculate fueling/ignition/VVT for any given event. Using the 20x20 grid as an absolute for tuning is misleading at best.

3a. The reality of the situation is that fueling is directly proportional to load. The torque curve of the engine is basically representative of the volumetric efficiency of the engine. Since the VE goes up and down over the rev range, that's why the fueling doesn't simply increase as a straight line as revs rise. What I've found in my own experience is that if you're commanding AFR in a separate map and have fueling basically proportional to load, making a percentage change in the entire column when rough tuning actually works pretty well to get AFR where you want it. I understand what you're saying about the program eventually reaching the target with enough samples, my concern is always starting with something truly awful and it not being safe. I've always found that its tough to get many data points in the region between full throttle and cruising.

b. One thing I noticed you didn't mention was INJ vs. TPS. Most people simply defeat this (set all correction factors to 1.000) but in reality, AFR is almost directly tied to throttle position. You pretty much want stoich under cruising and something else at WOT, with maybe a few other points in between. For the most part, this is very subtle, because you can cover most of this by simply commanding AFR. But what you'll find is that there are a number of cells you'll hit while cruising (where you want stoich) that you'll also hit under some acceleration conditions, where you want things richer. If you use the INJ vs. TPS feature, you can keep the fueling lean for cruising but then run through the same cell while accelerating and get the richer condition you want. Not sure how you would incorporate that into your program, but its an important feature to have, especially on a boosted engine where you really don't want to see boost at stoich.

In any event, I'll have to look more into your program when I have some time. It sounds like its come quite a long way.

[Beanie]

So earlier cars had different (not as good) knock sensors? Hmmm. Mine seems fine. The Find Insignificant Knock feature will still work because if knock goes away by retarding timing, it was real knock. At any rate, there is a Knock Calibration feature already, but it doesn't have multiple points. I'm going to add the capability to inputs some points at different RPMs. I think I'll do that right away.

Since it makes changes to the base map as you drive, it rapidly converges on the right values, provided you drive so that you are hitting the same cells. I'm not sure I want to put any code in that adjusts other cells based on a trend or known dependence on load. It could do this, but it might be a little dangerous. Will think about that some more... wouldn't want to cause a problem. It would probably be better the way it is, having the user make adjustments in places there isn't any (or enough) data.

There is a filter (Auto Tuning Required Condition) on the Setup tab for TPS Accel that you can set up so it disregards AFR samples obtained when the throttle position is changing rapidly (you set the min and max rate of TPS change per second). It is your choice whether to tune for cruise or whatever. And later on, after the cruising area of the map is taken care of, you can protect those cells and change the TPS Accel filter.

[Jesse IL]

I'm not talking about TPS Accel, I'm talking about absolute throttle position. On all my newer maps, I base enrichment off of throttle position, which is a better way of doing things. Again, I need time to look at your program and try it out.

The 2000-01 single wire knock sensor isn't bad per se, it just has a totally non-linear response over the rpm range. Basically, a knock reading of 15 is definitely knock at 4500 rpm but is definitely not knock at 7500 rpm. Seriously, download the AEM Pro software and look at how they set up knock control. Its really very sweet.

[Beanie]

Boy, our knock sensors must be very different. My highest knock anywhere in the map is 9. I've just about finished putting in a 20-point knock sensor response curve, though. You'll be able to open and save your curves to files -- for tuning cars with different types of knock sensors.

[Jesse IL]

Boy, our knock sensors must be very different. My highest knock anywhere in the map is 9.
You have a 2002+ 2-wire knock sensor. They have a very flat response curve, typically reading from about 0-10. The earlier knock sensors read from about 0-41 during normal situations.
I've just about finished putting in a 20-point knock sensor response curve, though. You'll be able to open and save your curves to files -- for tuning cars with different types of knock sensors.

Awesome, that's excellent news.

[Beanie]

The knock sensitivity curve feature is in there now. Source code has been posted. I'm going to do at least 2 more versions, one of which is almost done. It will have some more graphical "toys", like a CG Circle window and some more displays for the Dashboard tab. One is shown here, which shows G forces. A Phidgets USB accelerometer is needed for that to work. These things cost about $60 as I recall and all you'd have to do is mount it and plug it in. The software detects when it is plugged in.

http://www.mr-s.org/copilot/SmallCGCircle.jpg
Small circle above center is a display of lateral/longitudinal Gs and can replace the AFR gauge. A small track display will be below center. Click one of the small displays and they all switch to the other functions (like they were before this).

[shadytunerGTS]

This is coming together really well. Beanie, let's say I just installed a turbo. If I set up an AFR table and go do some driving with the copilot, it will adjust to the desired AFR and ignition? What I mean is, will I have to frequently stop. Adjust what copilot has accomplished then go do a part throttle and stop, etc?

[Beanie]

You have to:

1. Set up the Comm Port and I/O Settings, including the wideband input... one time.
2. Make sure the wideband is calibrated and working.
3. Make sure the knock sensor is working and the sensitivity curve is correct.
4. Define a Minimum Base map... this tells Copilot how low it can go with the base map when it makes changes. You also have to define a Maximum Base map. For ignition timing, there is just a Minimum Ignition map (actually, for 1ZZ-FE, there are two, even though Leading and Trailing don't BOTH appear to do anything... I think the Trailing one does nothing).
5. On the Setup tab, define the conditions you want it to use when filtering out AFR samples. It is completely up to you how to do that.
6. Enable the auto tuning (fuel (base map) and/or ignition timing).

Basically, you can do these things in any order and at any time. What I suggested earlier was for someone who wanted part throttle to be "just so" and not have it influenced at all by data acquired and analyzed during WOT. I didn't do that. I just defined an AFR map, set my limits, set the conditions to be very permissive (although I set the number of samples/cell to be about 100 most of the time), and let it do its thing. What you suggested sounds like a good way to proceed, though.

Note that it will never advance the timing. Maybe you'd want to get it to hit the desired AFRs first with timing retarded a little more than necessary, then manually advance the timing to where you think it should be and let it continue auto tuning the fuel. It is really up to each person how to use the program. If it were me, I'd have the auto tuning of ignition timing enabled all the time with a knock limit of about 20 until I had a nice safe tune. The Find Insignificant Knock function should probably be done last. That's the fine-tuning part. It will leave the timing at the point where knock just went away or possibly retarded 1 more degree. It stops changing the timing in each cell when knock stays below the threshold you entered for the number of samples/cell you entered, so it may not retard a cell at all if it was fine to begin with. I've heard people say that 2 degrees from knock was a good setting, so after it is all done with that, retard 1 more degree. I'm not a "tuner" like JesseIL or LittleRocket, though. I don't give tuning advice, really. It is just advice on how to use the program (what it can do, how to get there).

You can smooth things out anytime you want, although you have to click Stop Acquisition to make manual changes to the maps. I need to do some smoothing in my timing map. The engine runs very nicely, though. :) I don't have a turbo, just a 2ZZ.

[Beanie]

Check out the improvements since January in the Copilot topic over on www.spyderchat.com. The latest version adds nifty MPG-related things including an Economy display similar to the one in the Toyota Prius... only better.

http://www.mr-s.org/copilot/Copilot_3_2_3.zip

[shadytunerGTS]

Damn Beanie. I guess I don't know how to use your program. I have it downloaded on my laptop and it just won't connect. Does the PFC edit version and your program have to be the same? Maybe you could give me a quick lesson. Thanks.

[Beanie]

There's a quick and dirty on setting it up a page or two back from the latest page over on spyderchat. If it doesn't start reading from your PFC, most likely you have not set the Comm port. Go to Comm --> Settings and set the port number, save that as the default, click OK. It will remember the port number from then on and try to connect. If you have a serial to USB converter with the Datalogit like most people do, always use the same USB port, otherwise Windows will assign it a different port number and you'll have to go back the the Comm Setting to change it.

[shadytunerGTS]

I really want to try out this program. I finally got the comm port set up so it doesn't get anymore errors and it still won't let me connect. It says at the bottom "Read Bad PFC packet length" or something to that effect. I've tried 1-11 on the comm port with no improvement. I don't have the USB converter. I have an older laptop so I can directly plug into the laptop. Also, I've got an AEM UEGO wideband O2. How do I setup for that WBO2? My comm port setting for FC edit is 3 and the setting for Copilot is 4. Thanks

-Charlie

[Jesse IL]

Still haven't had a chance to check this out as my car has been down 12 months now.

[zero2toy]

After I get my afr dialed in as close to be good enough and not knocking, I will be testing this new version of copilot this weekend, Ill report back with the results.

[Beanie]

I really want to try out this program. I finally got the comm port set up so it doesn't get anymore errors and it still won't let me connect. It says at the bottom "Read Bad PFC packet length" or something to that effect. I've tried 1-11 on the comm port with no improvement. I don't have the USB converter. I have an older laptop so I can directly plug into the laptop. Also, I've got an AEM UEGO wideband O2. How do I setup for that WBO2? My comm port setting for FC edit is 3 and the setting for Copilot is 4. Thanks

-Charlie

Use the same comm port in Copilot as you do in FC Edit. Go to Comm --> Settings, put in the port number, click Apply, Save As Default, OK. You can exit and restart or go back to Comm menu and select Connect.

For the wideband, if it is nonlinear (don't think that one is, but I could be wrong) go to Settings --> I/O Settings. Make sure "WBO2 and AN1-AN2 is checked, which I think it should be by default. Click the "poly.." button, put your coefficients in there (just like you do in FC Edit), check the "Use Polynomial" box, click OK. If it is linear, make sure that box is NOT checked and instead just enter the Values for 0V and 5V like you do in FC Edit.

Communication problems sometimes occur. I've improved it some in the ability to detect bad data, bad packets, no packet received, and so on, but the bottom line is the the interface (could be inside Power FCs and/or Datalogit box) is just flaky.

[Entranced]

just downloaded this program and wow! looks great, haven't used FC edit or this program, but i'm probably going to be using this one instead!

[shadytunerGTS]

It just does not seem to want to work. Is there anything else I'm missing? My comm port for FC edit is 3, and the only comm port I could use without getting an error on CoPilot is 4. It says connected, but also in the bottom status box says "Read bad PFC packet length". It won't let me Start Aquisition, read base or inj maps, VVTi map, etc.:shrugs:

[Beanie]

You'll have to close FC Edit and then Copilot should be able to work with Comm 3, but you must be doing that, so you may have a newer (black) Datalogit box... the one with 8 analog inputs. :( I've yet to hear from anyone who has the black box working. Supposedly they did not change anything in the PC/Datalogit interface, but I am conviced they did. I'm hoping someone will mail me their black box so I can do some port monitoring and figure out what's different. Anyway, if you have a beige box and it doesn't work, I just don't know what it could be. You should see the platform (1ZZ-FE or 2ZZ-GE) and version at the upper right corner if it is reading anything, then see "Reading maps" in the status panel at the bottom of the window, then "Ready".

If anyone out there has a black box they'd be willing to let me borrow for testing, send me a PM. I promise to return it when done. I just don't want to buy one, since there's really nothing in it for me besides work.

[shadytunerGTS]

Oh ok, I have FC edit open while Copilot is running, that's one thing. Another thing, yep I do have the black Datalogit box, dammit. Well there's the reason, at least we figured it out and I can stop while I'm ahead.

[Beanie]

Please try it without FC Edit running at the same time. Might work.

[shadytunerGTS]

All right I'll try that next time. I never would of thought that would screw with the Copilot, though it does make sense to only have one program running for the PFC.

[Beanie]

Yay... someone told me that it works with the black box. :)

[shadytunerGTS]

I still haven't tried it. Someone else has gotten it to work with the black box? It should work for me then.

[zero2toy]

My black box works with co-pilot as well :)

[shadytunerGTS]

I got to play around with it today. I like the launch control setting. I was browsing through the different settings and came across buttons, then enabled the auto tune to the fuel. I don't think it would auto the fuel. I stopped the aquisition and hit write base and Inj map to write the map to the PFC. After that, my fuel was a lot leaner, too lean for me. Another thing my cam changeover was screwy after the revision. I'm sure it's due to the lean A/F reading at WOT. It would be sweet if the Copilot auto tuned VVT, although I probably didn't have everything set up correct on Copilot and that is why it leaned out way too much.

[Beanie]

The Buttons menu pops up when I squeeze the trigger button on my hand controller... a hacked Microsoft Sidewinder joystick connected to a USB I/O board. I go up and down the menu using the HAT switch. Other buttons (8 in all) do dedicated functions like switch tabs. http://spyderchat.com/forums/index.php?showtopic=37622&hl=sidewinder&st=20

You set the conditions for auto tuning on the Setup tab... things like water temp (so it doesn't use the AFRs read while the engine is warming up) and TPS %. It works, but WOT causes RPM and Load to change rapidly (so there are fewer AFR samples per cell in a given time period) and therefore takes longer. I started out at a slow speed in each gear and gradually gave it gas for about 3 seconds so the TPS Accel factor didn't come into play. Then I just let the revs climb to redline. Repeat several times for each gear (I only did 2-6 and didn't take 5 or 6 to redline because it was too fast for a public road). It does best when you are just commuting or something and the RPM and Load don't change drastically. Sorry it is so complicated, but it has to be in order to work and have some flexibility.

When you are doing the auto tuning online (connected), it will automatically write whenever it has enough to go on (see Setup tab... number of cells with changed data, number of (AFR) samples required per cell). I doesn't change the New Base map (what gets written) unless it has enough to go on, so if you stopped acquisition, then wrote the New Base and/or New Inj map, it would not really change anything. It puts its changes in the New Base map (a copy of the current Base map), then writes it and reads it back. It doesn't change the Inj map unless you have a platform (engine) like EJ20K, RB25DET, or RB26DETT, which has no Base map (just Inj). Read through the Copilot topic in Commercial Sales (there are some other, older topics) on www.spyderchat.com -- where I walked different people through it. Note to others: Even though that topic is in the Commercial Sales forum, it is free.

[shadytunerGTS]

I'm so exicited and giddy inside like a little girl at a Miley Cyrus concert. I really like this program. It's probably something I could never create in my whole life. I give you props Kevin. I want to play with this program some more, I only got a feel of what it can really do. The only thing is I don't think it adjusted anything as I drove around for over 30 min. That is when I stopped it, pressed 'write base map' and drove some more and noticed it was lean at WOT and cam changeover was shifty and unsmoothe. I definitely need to set this program up better in order to get it to work to it's potential. It does nothing to the VVT map, right?

That is quite a nifty hand controller!!

[h8eternal]

does anyone have before and after dyno results using this program and not?

I got to play around with it today. I like the launch control setting. I was browsing through the different settings and came across buttons, then enabled the auto tune to the fuel. I don't think it would auto the fuel. I stopped the aquisition and hit write base and Inj map to write the map to the PFC. After that, my fuel was a lot leaner, too lean for me. Another thing my cam changeover was screwy after the revision. I'm sure it's due to the lean A/F reading at WOT. It would be sweet if the Copilot auto tuned VVT, although I probably didn't have everything set up correct on Copilot and that is why it leaned out way too much.

do you have any videos of your celica using the launch control feature.
im really curious to see what its like on a celica as silly as that may seem.

[Beanie]

shadytunerGTS, thanks.

As JesseIL has said, people shouldn't get the idea that this is "magic" or something. It isn't going to magically give you more horsepower. The auto tuning of the base map is basically a calculator that works under the conditions you set up. It looks at the average AFR in each cell and compares that to your Target AFR map. The goal is for the actual, logged AFR to get as close as possible to the Target AFR under whatever conditions you specifiy. This works, I've done it for years now and others have, too. Once the base map is correct, you can then just modify the Inj map to change the AFR. So to the extent that your Target AFR map (which can be tied to the Inj map, as in you modify the Target AFR map and it changes the Inj map for you) would give more power, you'll get more power. You can also get better economy, if that's your goal. Since you, not the program, specify the AFR, before/after dyno results wouldn't be a way of evaluating the program... just your Target AFR map.

The auto tuning of ignition timing is limited to retarding the timing in response to knock. I've used it to get my knock down into single digits everywhere. I had knock in the 20s in some places and Little Rocket told me that if you can make that knock go down by retarding the timing, it is real knock and should be eliminated, otherwise over the long term it will harm the engine.

Nothing is done to VVT-i.

Some people gave me a hard time about this not being what they thought it would be... some kind of magic auto tuning program that would find the max power somehow. Sorry, that's just too hard to do. There's already about 100,000 lines of code and that would be a whole lot more, plus there'd be no way you could do it unless you were on a dyno or very controlled conditions on the street. It is what it is, which is a tool. Learn how to use it and you can do all kinds of things. I personally like the new Economy display (to see where I am just plain wasting gas by being in too low a gear), gauges, replay, and the things that having GPS and an accelerometer add to the program. It is kind of a PITA to start up my tablet PC (mounted in the dash) every time I go for a drive, but worth it. The Limit Check alerts have given me peace of mind, too. I don't have to watch the coolant temp (I had a leak a few months ago) because the program does that for me. When the temp goes over 190 (what I have it set for), an alert pops up and a wav file is played.

So anyway, enjoy, but read all the information on it before doing any auto tuning. I just don't have time right now to do what I should, which is update the User's Guide. (I would have done that if this had turned out to be a more successful commercial product. I gave a "lite" version away for too long, I guess, plus people don't seem to want to pay much for software. There's this mentality that software and music and so forth should be free. I did have some people buy it for $50 (one guy even gave me $100 because I implemented the audio warning for him... he had an air bubble and couldn't watch the coolant temp all the time) but then I had "customers" and felt like a "vendor", which just wasn't worth it to me. I have a good paying real job.)

[zero2toy]

Copilot 3.3.2 will not connect for me. I have it on the same comport as the FC-Edit, and in the lower left corner keeps switching from Connected to Disconnected, is this happens when im not even connected to the PFC and when im connected. Ive tried all comports but still the same result, as if i wasnt even plugged into my computer... Any suggestions,

I have a black datalogic box, i thought i was connecting before but this weekend could not get it to work.

[Beanie]

Is that still happening? I don't visit this forum much, so I'm sorry I didn't notice this until now. If it worked before, you may have plugged the USB cable into a different port, causing the port number to change. Find out what port it is via the Control Panel. Maybe I'll put an option in the next version to find the port with the Datalogit.

[zero2toy]

Right on. Im going to be messing with the car tonight, and see if i can get copilot set up to auto tune and get this ball rolling. Ill check the comport via control panel see if i can get it to connect, if not, back to tuning with FC - Edit, booooooooooo lol. Lets hope i can get copilot working.

Thanks for responding, better late then never, thanks again.

[shadytunerGTS]

do you have any videos of your celica using the launch control feature.
im really curious to see what its like on a celica as silly as that may seem.

Ask and you shall receive. I used the program again today. It was being stubborn. It didn't want to read my inj % and water temp forever. Then, I just closed up the program and restarted and it read it all. I love that launch control!! This is also a sh!tty cell phone video, but the sound quality is decent.
http://s378.photobucket.com/albums/oo224/ShadyTunerGTS/?action=view&current=launchcontrol.flv

[zero2toy]

^ nice vid, do you have a white datalogic box or a black one shady? Im having connection troubles, wondering what you might beable to tellme to get my black box to connect.

[shadytunerGTS]

I've got the black box. I was having trouble before and the only things I did was use the same port as FC edit and close FC edit with Copilot being the only program open. I don't have the current copilot which is 3.2.3, I have the one before I do believe.

[h8eternal]

Ask and you shall receive. I used the program again today. It was being stubborn. It didn't want to read my inj % and water temp forever. Then, I just closed up the program and restarted and it read it all. I love that launch control!! This is also a sh!tty cell phone video, but the sound quality is decent.
http://s378.photobucket.com/albums/oo224/ShadyTunerGTS/?action=view&current=launchcontrol.flv

that is awesome man
thanks for the heads up
i appreciate it
BADASS!

[TRlPPlN]

Ask and you shall receive. I used the program again today. It was being stubborn. It didn't want to read my inj % and water temp forever. Then, I just closed up the program and restarted and it read it all. I love that launch control!! This is also a sh!tty cell phone video, but the sound quality is decent.
http://s378.photobucket.com/albums/o...nchcontrol.flv

WOW...i need to catch up with this program once i get my car back. thats b@d@ss

[Beanie]

A new version is available that has an Auto Detect option on the Comm Settings window. You have to check the box to enable Auto Detect. I tested it today and it worked for me. No other new features other than support for 1JZ-GTE engines. Trying to add 13B-REW (RX-7 support, but that'll take some time to finish).

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

Latest version is 3.3.4, so look for Copilot_3_3_4.zip in that directory.

[zero2toy]

Thanks Beanie, ill try that tonight.

[Pegaso]

A new version is available

Beanie, for temp and pressure gauges you get an evolution or not?

I need to change units and scale.

Gauges like speed and RPM it's impossible?

water temp is too large scale for example.

If I pay you for changes?

[v3bahumut]

Been playing with the program, looks interesting.

[zero2toy]

^ nice vid, do you have a white datalogic box or a black one shady? Im having connection troubles, wondering what you might beable to tellme to get my black box to connect.

Im an idiot, I was not hitting APPLY, just ok in the comm settings, thats why it was not connecting. :bang:

Now I'm having 2 problems,

1. The afr im seeing on my AEM Uego is like .4 lower than what im seeing in FC - Edit, and CoPilot cells during logging/map trace.
a. I have the I/0 settings right for the Uego, 0v=10 5v=20, i also checked the documentation that came with the Uego and confirmed it was right. The Uego is Linear.
b. I think I may have this floating ground issue I've read about previously on the forums here, unless someone else has a suggestion on why I might be seeing this difference in AFR from gauge to FC-Edit/Co-pilot.

2. Copilot is reading the active cell in the map like 500 to 1000 rmp higher than it really is, but FC edit reads it right... IDK what thats about, but my car idles around 1000 and CoPilot is showing im at 1500-1600??????

Thanks for any help anyone could give me on these issues :confused:

[darkninja697]

so has anyone tried this out personally? i'm thinking about using it once i get my car back but didn't see any testimonials from it..

[v3bahumut]

It's a pretty cool program, lots of toys the FC edit doesn't have.

[zero2toy]

so has anyone tried this out personally? i'm thinking about using it once i get my car back but didn't see any testimonials from it..

Im working on some kind of setup turtorial accompanied by a help file for this program, but work has been extremely busy, so i havent even had time to figure out my problems with the program, let alone how to use it/set it up correctly, or different ways to set it up to do different things.

[Boosted2.0]

Im an idiot, I was not hitting APPLY, just ok in the comm settings, thats why it was not connecting. :bang:

Now I'm having 2 problems,

1. The afr im seeing on my AEM Uego is like .4 lower than what im seeing in FC - Edit, and CoPilot cells during logging/map trace.
a. I have the I/0 settings right for the Uego, 0v=10 5v=20, i also checked the documentation that came with the Uego and confirmed it was right. The Uego is Linear.
b. I think I may have this floating ground issue I've read about previously on the forums here, unless someone else has a suggestion on why I might be seeing this difference in AFR from gauge to FC-Edit/Co-pilot.

2. Copilot is reading the active cell in the map like 500 to 1000 rmp higher than it really is, but FC edit reads it right... IDK what thats about, but my car idles around 1000 and CoPilot is showing im at 1500-1600??????

Thanks for any help anyone could give me on these issues :confused:

check the voltage logged versus the AEM stated afr vs voltage curve. If there is a discrepance then its simlpy a setup issue (ie: you need to adjust the scale on the FC edit software)

[Freakn Pwned]

Ask and you shall receive. I used the program again today. It was being stubborn. It didn't want to read my inj % and water temp forever. Then, I just closed up the program and restarted and it read it all. I love that launch control!! This is also a sh!tty cell phone video, but the sound quality is decent.
http://s378.photobucket.com/albums/o...nchcontrol.flv
Sick as hell...i wish i had to balls to do that to my car but being its my only car and i couldnt afford breaking anything big i guess it will just remain a dream for now...

[shadytunerGTS]

It is pretty sick. I should've kept recording through a couple gears though. All it does is sets the fuel cut lower and stays there until moving over a desired speed, ex. 1 mph. No damage besides the launch itself, tire spin, drivetrain, ect. Oh well, I want a new clutch anyway. Light em' up!!

[Jesse IL]

1. The afr im seeing on my AEM Uego is like .4 lower than what im seeing in FC - Edit, and CoPilot cells during logging/map trace.
a. I have the I/0 settings right for the Uego, 0v=10 5v=20, i also checked the documentation that came with the Uego and confirmed it was right. The Uego is Linear.
b. I think I may have this floating ground issue I've read about previously on the forums here, unless someone else has a suggestion on why I might be seeing this difference in AFR from gauge to FC-Edit/Co-pilot.
On the AEM Pro software, in addition to the linear slope of the sensor output, there is a correction coefficient that you need to play with. I had to play with it a bit on my friend's MR2 to get the log to match the gauge.

[Entranced]

^ curious to see if it's the same way for my LC-1. I never bothered to install the software on it, yet I seem to be getting the same issue. I double grounded the wires as well straight to the battery, 0.0v drop across the wire. Either that, or my times that set for recording lag are off... hmm. This is in FC edit. I still have to try this software out. It's alot more involved than FC Edit, haha.

[blinding-gts]

^ curious to see if it's the same way for my LC-1.

Copilot was designed around the LC-1 from what I remember when discussing it with Beanie in the past; doubt it'll have an issue.

I wonder if anyone has tried the auto-tune feature on a boosted car?

[zero2toy]

On the AEM Pro software, in addition to the linear slope of the sensor output, there is a correction coefficient that you need to play with. I had to play with it a bit on my friend's MR2 to get the log to match the gauge.

Thanks Jesse, Ill check that out.
I did play with the scale in copilot and readjusted it to match what i was seeing at idle on the gauge. I got it dialed in to about .1 difference and have left it there for now. Seems to be all i needed, but a patch not a defenite fix, but it will work the way i have it now.

Thanks Again

[payatz]

this software for free?

[blinding-gts]

this software for free?

Yes, it's free. You can go to http://www.mr-s.org/copilot/ to download the latest one.

[v3bahumut]

this software for free?

Paypal me $42 and I can get it for you.

[payatz]

the other person said it's free then the other said 42 bucks.. Who's telling the truth haha!!

[vegeta4ss]

oh wow I had forgotten all about this program. Can't wait to give it a run.

[Beanie]

Version 3.7.2 is the latest. It is improved in terms of ease of use and how long it takes to auto tune the base map. Still free.

[Jesse IL]

I'm still using the old vehicle-specific FC Edit :laugh:

Maybe I should check this out finally.

[v3bahumut]

I check in over at spyderchat every once in awhile to see if you have the boost gauge input worked out yet.

Still waiting :(

[Beanie]

I'll do that one of these days. Lacking data, I will have to make it so each user has to enter numbers defining how the boost input relates to the gauge.

[Magusownz]

Jesse your not alone. Many of us use the original FC Edit that is vehicle specific. For me I could not be happier...

[Magusownz]

Beanie btw where in florida are you located?

[Beanie]

Jacksonville

[blinding-gts]

I like the color scheme on the new FC edit which makes checking the cells a lot easier. It also detects the COM port with less issues.
Copilot has a lot of neat features. It's a shame an auto-tune feature isn't available in FC Edit.

[Beanie]

There have been many improvements since I last posted, including auto tuning ignition and lots of virtual gauges (including an XD-16 AFR gauge, knock, airflow, speedo/tach, clock, and more).

[v3bahumut]

yay, new toys...too bad my laptop died :(

[justynsi]

Has anyone ever heard of anyone having problems with the engine after using the copilot's fuel cut launch control?

[bajan_speed]

Has anyone ever heard of anyone having problems with the engine after using the copilot's fuel cut launch control?

I really like that feature from this software and I am interested in the answer to this ^^ question

[Beanie]

What sort of problems did you have? All it does is temporarily lower the rev limit. I've hit the rev limit on my car plenty of times. Back when I autocrossed, I'd "bounce off" the rev limit for several seconds at a time, multiple times during a run, and we're talking about 7300 RPM (1ZZ engine). I've bounced off my 2ZZ's 9000 RPM rev limit, too. If you don't want to do this, even at a low RPM, then please don't use that feature.

[justynsi]

None.. Is there any way it could do an ignition cut?

[bajan_speed]

What sort of problems did you have? All it does is temporarily lower the rev limit. I've hit the rev limit on my car plenty of times. Back when I autocrossed, I'd "bounce off" the rev limit for several seconds at a time, multiple times during a run, and we're talking about 7300 RPM (1ZZ engine). I've bounced off my 2ZZ's 9000 RPM rev limit, too. If you don't want to do this, even at a low RPM, then please don't use that feature.

Beanie what do I have to do to use the software on my pfc? I just downloaded your 5_1_9 version, do I just install it on my pc and use it instead of the dataloggit software?

[Beanie]

You need the Datalogit box.

[bajan_speed]

I have everything I need then. My tuner looked at the program and he said it is great we will use it next week. We normally use fc edit

[Celicasaur]

Beanie, any way you can get the launch control to work via the commander, ie without a laptop running your software when racing?

Also, is there any downside to having a fuel cut as the launch control limiter as opposed to an ignition cut? I know different models of car have different ways, so i guess both have their pro's/con's :shrugs:

[SuperDave]

Rana, might get a faster reply on Kevin's new website. http://www.mr2spyder.net/spyderweb/index.php?/forum/16-engine-management/

But I doubt the launch control would work from the Commander since it is an extra function he has added via Copilot; the software provides the launch control, not the PFC itself. I'd be curious to know what others use for the settings though.

[Celicasaur]

Thanks Dave :thumbup: yeah i guess it is wishful thinking to have it as a commander function, but I guess it's better than nothing to have it via the laptop. I'm going to start racing some 4wd guys soon and it'll be nice to have a bit of help on that al important launch.