Viability of a Megasquirt application for FI on the renesis
#276
Humpin legs and takin nam
Join Date: Nov 2003
Location: Clearwater, Fl
Posts: 2,433
Likes: 0
Received 0 Likes
on
0 Posts
I tore apart my CZ unit and still have the harness/casing. Could I use this to build a MS unit?
Originally Posted by rotarygod
Since I spent several hours rewiring it so it wouldn't short out, I would suggest that anyone purchase the extension harness from Rotary Performance. It is around $200, all the wires are already labeled, and they are all finished very nicely. Pay the extra $75 and make it easy. Don't risk shorting out something on the car or the ecu.
#277
Registered
You can. Just figure out which wires go where and then use common sense if you think something could be done better. As long as you are thorough and careful you could easily use it. Someone brave enough could even just tap into the factory harness but I like the idea of being able to unplug it.
#278
Humpin legs and takin nam
Join Date: Nov 2003
Location: Clearwater, Fl
Posts: 2,433
Likes: 0
Received 0 Likes
on
0 Posts
I have a question..
When you initially hooked up the MS and turned the car on the first time, did you already pre-load a map to add fuel, allowing the car to start? Or did it somehow already have sufficient fuel to start?
When you initially hooked up the MS and turned the car on the first time, did you already pre-load a map to add fuel, allowing the car to start? Or did it somehow already have sufficient fuel to start?
#279
Registered
I had to load in a set of base parameters such as engine size and fuel injector size. Based on that alone it is easy to get started. The first time it started up, it idled on it's own. Of course I guessed at a setting as I already have experience with it on a rotary but it was still easy. I did not have a drivable map though so that we had to work out as the car was being driven. Brillo would drive the car and I would make real time changes. Still pretty easy and straight forward.
#281
Registered
Yeah we took the car out and drove around just fine. We were only running fuel only control though. The stock ecu still handled ignition. We did figure out that fuel only control isn't much good on this car. You also need ignition control. When we'd dial in a setting for fuel, the ecu would try to alter the timing curve to compensate. I know for a fact the ecu fights back which is why it is bs when people say that it doesn't. It did for us. It did when the Canzoomer was on the car too. If we also had full ignition control it couldn't have fought us and we would have gotten similar results ultimately that the Interceptor is getting as we too would have full control. The igniton code is written. I just don't have it yet as it is going through testing right now to check for errors. When I do get it, we'll work on a full map.
Last edited by rotarygod; 12-15-2005 at 10:11 PM.
#285
Go Texas Longhorns!
Wish I had an update, I'm still waiting for the ignition code to be finished and tested. I'll bug the MS folks and see how its coming.
They are in the process of rolling out some newer and faster hardware, that will allow folks to code in C, and as a result, there is a little bit of backup as far as these side projects while they get the main code ported over to MSII. Allowing coding in C will mean many more people can contribute, which is a good thing.
again, we have the fuel side working fine, but I need the MS to be able to read our stock trigger wheel for ignition. Its taken longer than I thought b/c the MS is starting to run low on memory(hence the need for MSII) so it took some time to find some spare room for the code (you have to realize how primative the original processor is)
I'll get Ken to post an update here if he can, but we haven't forgotten the project, we are just at the mercy of others for the moment.
They are in the process of rolling out some newer and faster hardware, that will allow folks to code in C, and as a result, there is a little bit of backup as far as these side projects while they get the main code ported over to MSII. Allowing coding in C will mean many more people can contribute, which is a good thing.
again, we have the fuel side working fine, but I need the MS to be able to read our stock trigger wheel for ignition. Its taken longer than I thought b/c the MS is starting to run low on memory(hence the need for MSII) so it took some time to find some spare room for the code (you have to realize how primative the original processor is)
I'll get Ken to post an update here if he can, but we haven't forgotten the project, we are just at the mercy of others for the moment.
#286
I'm the "Ken" that brillo speaks of in his last post.
Just an update. I finished the code back in Aug, but I never got to test it. Around October sometime I got brillo to send me an rx8 trigger wheel, but then none of the sensors I could get for cheap would work with that wheel (sensor heads were too big or too small), and the rx8 sensor is like $120 or something. The code got put on the back-burner at that point. About a month or two ago, I gave the code to jsmcortina over at msefi.com (he is the original creator of the msns-extra firmware) to look over, but he never got a chance to test either...
That's the history...
Where I'm at now:
jsmcortina wrote some code for an older, no-longer-for-sale version of the MS2 which simulates just about any trigger wheel imaginable... I need to port this to the newer version of the ms2. Once I've done that, I'll test the msns-extra version of the rx8 wheel decoder, fix any problems I find, and post it here, along with the code that makes the coil-on-plug setup on the rx8 work.
I can't give a good time estimate of how long it'll take... mainly because I have 2 project cars that I'm working to get back on the road, and those cars are my priority right now. Once I have those cars on the road again, (hopefully next week), I'll be able to look at my software projects again. I was going to rewrite the msns-extra generic wheel decoder to use all of a wheel's teeth to calculate timing (right now, it only uses 2 teeth for a 4 cyl/2 rotor, 3, for a 6 cyl, 4 for an 8 cyl, no matter how many teeth are actually on the wheel), but since that is just a pet project of my own, and the code that's there now works (I just wanted to create something better, with more accurate timing), I can put that on the back burner and try to get the rx8 stuff done.
If anyone has any questions about this, you can PM me here, on rx7club, or at msefi.com.
So we have enough memory to do what needs done, the only thing that's been stopping me this whole time is finding time and hardware to test with.
Ken
Just an update. I finished the code back in Aug, but I never got to test it. Around October sometime I got brillo to send me an rx8 trigger wheel, but then none of the sensors I could get for cheap would work with that wheel (sensor heads were too big or too small), and the rx8 sensor is like $120 or something. The code got put on the back-burner at that point. About a month or two ago, I gave the code to jsmcortina over at msefi.com (he is the original creator of the msns-extra firmware) to look over, but he never got a chance to test either...
That's the history...
Where I'm at now:
jsmcortina wrote some code for an older, no-longer-for-sale version of the MS2 which simulates just about any trigger wheel imaginable... I need to port this to the newer version of the ms2. Once I've done that, I'll test the msns-extra version of the rx8 wheel decoder, fix any problems I find, and post it here, along with the code that makes the coil-on-plug setup on the rx8 work.
I can't give a good time estimate of how long it'll take... mainly because I have 2 project cars that I'm working to get back on the road, and those cars are my priority right now. Once I have those cars on the road again, (hopefully next week), I'll be able to look at my software projects again. I was going to rewrite the msns-extra generic wheel decoder to use all of a wheel's teeth to calculate timing (right now, it only uses 2 teeth for a 4 cyl/2 rotor, 3, for a 6 cyl, 4 for an 8 cyl, no matter how many teeth are actually on the wheel), but since that is just a pet project of my own, and the code that's there now works (I just wanted to create something better, with more accurate timing), I can put that on the back burner and try to get the rx8 stuff done.
If anyone has any questions about this, you can PM me here, on rx7club, or at msefi.com.
So we have enough memory to do what needs done, the only thing that's been stopping me this whole time is finding time and hardware to test with.
Ken
#289
Go Texas Longhorns!
Many thanks Ken for the update, I know Richard and if you can carve out some time, I'm sure he can loan you the sensor.
Thanks Richard also for offering that up.
Thanks Richard also for offering that up.
#291
Alright, here's an update. I've got James's wheel simulator code ported to the newer version of the ms2, so all I need to do now is test it and hook it up to my ms1 running the rx8 code... and do some oscilloscope testing.
#293
another update:
We just had a really good meet this weekend in VA beach with Al, Bruce, James, and all the other megasquirt guys...
During this meet I was able to pick up a developer kit for the c32 processor, which means I should be able to test the rx8 decoder this week... maybe even tonight.
Ken
We just had a really good meet this weekend in VA beach with Al, Bruce, James, and all the other megasquirt guys...
During this meet I was able to pick up a developer kit for the c32 processor, which means I should be able to test the rx8 decoder this week... maybe even tonight.
Ken
#294
yet another update:
I hooked up the c32 dev kit, and got jsmcortina's wheel simulator running on it, and properly emulating the rx8 wheel last night. I checked it on my scope and it's definitely doing the 36-2-2-2 wheel.
Next step: I have to edit the msns-extra.ini file for the version of the megasquirt firmware that I originally added my decoder to, and make it so I can actually set up the firmware to use the rx8 code. Also, I have to set up my test bench to get the signal from the c32 dev kit to an actual megasquirt.... this shouldn't be too hard. I'll probably get to all this on thursday or sunday.
I hooked up the c32 dev kit, and got jsmcortina's wheel simulator running on it, and properly emulating the rx8 wheel last night. I checked it on my scope and it's definitely doing the 36-2-2-2 wheel.
Next step: I have to edit the msns-extra.ini file for the version of the megasquirt firmware that I originally added my decoder to, and make it so I can actually set up the firmware to use the rx8 code. Also, I have to set up my test bench to get the signal from the c32 dev kit to an actual megasquirt.... this shouldn't be too hard. I'll probably get to all this on thursday or sunday.
#296
OK, sorry for the lack of updates here. I have not had time to work on it as I planned to yesterday... I ended up spending almost the whole day tuning the ae86's 20v 4age. I'll do my best to get to it this week, but it's more likely that I'll get to it after my vacation (starting thu going to mar 26th).
#298
Registered
Because the project pretty much has too. We haven't pursued the MS anymore. The new code never came out as far as I know. It may have. I just haven't been in touch with the MS crowd in quite some time so I'm out of the loop when it comes to updates. Even if the new code did get done, it would be for a version of MS which I don't have. After Cobb tuning comes out with their device to retune the factory ecu, there won't be much need in this except for the few tinkerers out there who want to play around but the average person won't really find it useful on an RX-8. We originally did this just for the fun of it to see if it was possible since I already had an MS. I don't plan to pursue this anymore. It was fun at the time. I would use one on a Renesis conversion into another car though.
#299
Registered
OK it's been a long time coming but the RX-8 code is now done and it works through Megasquirt II. Remember I was using an MSI. The MSII has 100 times better resolution when it comes to injector pulse widths. The MSII can now read the stock RX-8 trigger wheel! It can also run negative split! Up to -50 degrees. 20 max conventional split. That's plenty!!! The onboard map sensor can go up to 21 psi of boost but if you need more a 4 bar map sensor is also available.
I do not have an MSII at this time but it's going to happen soon. We have also not actually used one on an RX-8. I'm just announcing that what we have asked for for over a year has finally come true! We know an MSII will control a Renesis engine now.
If you want to build your own, kits can be had for $250.
http://www.diyautotune.com/catalog/m...bled-p-59.html
If you need a wiring harness to go to the ecu, they have them available for $60.
http://www.diyautotune.com/catalog/m...eady-p-43.html
Obviously you won't want to cut into your stock wiring harness so you can get the adapter from Rotary Performance for $198
http://www.rx7.com/store/rx8/rx8harnessadaptor.html
This would be the minimum you'd need to make one work on an RX-8 assuming you don't want to splice into anything and you are cool with building one yourself. This adds up to $508!
If you don't want to build your own, you can buy one preassembled for $410.
http://www.diyautotune.com/catalog/m...unit-p-65.html
This still makes it $668 for a fully programmable ecu with negative split capability.
Keep in mind even a fully assembled MSII will still require a little bit of work on your part. You need a laptop so you can load the tuning software. Megatune is probably the easiest ecu software I've ever seen. Tuning is simple with it. Anyone can learn in a few minutes. You don't need to change back and forth between maps. All loads and rpms are on one screen in 3D mode. It's very easy.
You will need to wire each harness together. Fortuantely all the wires are labeled. If you can't do this I'm sure there is someone out there who you could pay to wire them together for you. It's not that hard though. The MSII really does need to be mounted inside the passenger compartment. It is not made to live in the harsh engine bay heat for long periods of time. Because of this you may even want to use the relay board. It's $88 fully assembled.
http://www.diyautotune.com/catalog/m...unit-p-32.html
This you can keep in the engine bay. Wire the MSII up to this through the use of the available cable (or make your own). It's $70.
http://www.diyautotune.com/catalog/m...able-p-47.html
Then you'll need to wire the relay box to the engine harness.
This all adds $158 to the above prices if you decided to go that route. I'm not using it but then again, I'm an idiot!
Just thought you guys would like to know. I know it's not a reflash and that ultimately that will probably be the nicest option but it is something for the tinkerers who wouldn't mind having control over their engine for relatively cheap.
I do not have an MSII at this time but it's going to happen soon. We have also not actually used one on an RX-8. I'm just announcing that what we have asked for for over a year has finally come true! We know an MSII will control a Renesis engine now.
If you want to build your own, kits can be had for $250.
http://www.diyautotune.com/catalog/m...bled-p-59.html
If you need a wiring harness to go to the ecu, they have them available for $60.
http://www.diyautotune.com/catalog/m...eady-p-43.html
Obviously you won't want to cut into your stock wiring harness so you can get the adapter from Rotary Performance for $198
http://www.rx7.com/store/rx8/rx8harnessadaptor.html
This would be the minimum you'd need to make one work on an RX-8 assuming you don't want to splice into anything and you are cool with building one yourself. This adds up to $508!
If you don't want to build your own, you can buy one preassembled for $410.
http://www.diyautotune.com/catalog/m...unit-p-65.html
This still makes it $668 for a fully programmable ecu with negative split capability.
Keep in mind even a fully assembled MSII will still require a little bit of work on your part. You need a laptop so you can load the tuning software. Megatune is probably the easiest ecu software I've ever seen. Tuning is simple with it. Anyone can learn in a few minutes. You don't need to change back and forth between maps. All loads and rpms are on one screen in 3D mode. It's very easy.
You will need to wire each harness together. Fortuantely all the wires are labeled. If you can't do this I'm sure there is someone out there who you could pay to wire them together for you. It's not that hard though. The MSII really does need to be mounted inside the passenger compartment. It is not made to live in the harsh engine bay heat for long periods of time. Because of this you may even want to use the relay board. It's $88 fully assembled.
http://www.diyautotune.com/catalog/m...unit-p-32.html
This you can keep in the engine bay. Wire the MSII up to this through the use of the available cable (or make your own). It's $70.
http://www.diyautotune.com/catalog/m...able-p-47.html
Then you'll need to wire the relay box to the engine harness.
This all adds $158 to the above prices if you decided to go that route. I'm not using it but then again, I'm an idiot!
Just thought you guys would like to know. I know it's not a reflash and that ultimately that will probably be the nicest option but it is something for the tinkerers who wouldn't mind having control over their engine for relatively cheap.