Newbee questions on injected LPG for a 289 windsor

FreeEMS topics that aren't specific to hardware development or firmware development.
User avatar
oden
QFP80 - Contributor
Posts: 91
Joined: Sat Sep 24, 2011 12:27 pm
Location: Australia

Newbee questions on injected LPG for a 289 windsor

Post by oden »

Currently looking for an EMS for my 1965 Mustang which has a rather stock 289 windsor motor, this is a daily driver so I am looking for reliability and efficiency over power gain. Mostly interested in an injected LPG setup like this one http://www.cletusracin.com/index.php?pa ... megasquirt

Interested in this project because its open source and I have experience with embedded processors (PIC micros) and electronics so building the system will be possible. Although I am new to EMS technologies and only have basic engine knowledge, this will be my challenge.

So my questions are:
1) The configuration I will be looking for will be a v8 sequential injection and a distributor ignition ( ? based on above LTD link), is it correct that this configuration is not supported yet?
2) Been looking all over the forum and the general site and I can not find a feature list for FreeEMS. There is info scattered around but I dont have the understanding to make sense of it all, has a feature list been compiled? ( hardware specific I assume)
3) I found a post where Fred said knock sensing was low on the priority list, can I ask why is this so ? Is knock sensing not as useful as the car manufactures advertise?
4) How hard do you think this will be to achieve ( injected LPG ) ? Any suggestions welcome.

Love the work you guys are doing here :)
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Newbee questions on injected LPG for a 289 windsor

Post by Fred »

1) Not right now, however you can do semi sequential and distributor as is, and by the time you have hardware built Sean's XGATE code should be complete and tested enough to integrate into the main code. You could volunteer to test it in whatever state he has it, but that is at your own risk and you are strictly forbidden from badmouthing FreeEMS if you have any kind of bad experience with code that is not ready for general use such as that.

2) No, sadly no such feature list is available yet, however it would be best described as "less than the minimum required" or "not worth listing" at the current time. The priority is the rock solid stability of the core code, not features, for the time being, however, you should see this thread to realise that if you hook something up, your priority is my priority, assuming that it is a reasonable request and beneficial to the majority of other potential users. Additionally, you can check out the roadmap on the issue tracker to see some of the things that need to be done. Note, user requests and general features are in the ASAP section further down. They are generally easy and quick to implement, so it's not a big deal that they aren't there yet. Although there is a lot on that list, some of which look very important, the system is very usable now, I promise :-) BTW, "hardware dependent" would be a better term.

3) Correct, the OEMs have fuel quality concerns to deal with, and emissions standards to deal with, and very fine tolerances on everything. IE, they have to run right up into an unsafe state of tune and RELY on such systems to pull timing out if things aren't going well. This does NOT work well for very high output setups, ask any experienced tuner. It's also not necessary to get more than acceptable power and economy out of any given engine setup. If you tune correctly up to close to the limit and run fairly consistent fuel, you're going to be just fine. Once you've tuned an engine or two, this will become apparent :-) BTW, the newest hw design is planning to incorporate knock sensing, and I'm planning to use that as a development platform for it, so there will be support, but it MUST come after more important key things.

4) Very doable, just make sure you consult with me about what type of RPM pickup setup you're planning to run and make sure I have at least a week's warning to get it done for you.

5) Use the RavAGE pdf schematics combined with the TODO.md file to design your custom hardware. It is close to perfect and a great base for DIYing.

Welcome along! :-)

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Newbee questions on injected LPG for a 289 windsor

Post by Fred »

PS, I removed you from the noob group :-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
oden
QFP80 - Contributor
Posts: 91
Joined: Sat Sep 24, 2011 12:27 pm
Location: Australia

Re: Newbee questions on injected LPG for a 289 windsor

Post by oden »

Thanks for the status change :-)

Happy to test the new code. Worked as a software engineer for many years, so I know what it means to be running code that has not been heavily tested ;) Would be great to contribute.

Being an 1965 motor it has no crank / cam sensors. As much as I would love to install a crank sensor, not willing to take the engine out to do it. Would the best option be to install a laser cut wheel onto a modified distributor? Any designs available? Better suggestions ? Have been reading the MS documentation to get an understanding on it all.

The RavAGE thread is 281 posts, looks like it has been well thought out. Will start to read over it all tomorrow.

Exciting stuff..
User avatar
sim
LQFP112 - Up with the play
Posts: 112
Joined: Thu Jun 02, 2011 8:17 pm

Re: Newbee questions on injected LPG for a 289 windsor

Post by sim »

oden wrote:Thanks for the status change :-)

Happy to test the new code. Worked as a software engineer for many years, so I know what it means to be running code that has not been heavily tested ;) Would be great to contribute.

Being an 1965 motor it has no crank / cam sensors. As much as I would love to install a crank sensor, not willing to take the engine out to do it. Would the best option be to install a laser cut wheel onto a modified distributor? Any designs available? Better suggestions ? Have been reading the MS documentation to get an understanding on it all.

The RavAGE thread is 281 posts, looks like it has been well thought out. Will start to read over it all tomorrow.

Exciting stuff..
I use a modified distributor to drive a Mitsubishi Cam Angle
Sensor on my antique engine. I bought the adaptor, but if you
have a spare dizzy and access to machine tools, it should not
present much of a problem. It is a pretty elegant solution, and
good for full sequential.

I'm running a homemade board in my daily driver right now. The
build thread is only two pages, more digestible than the RavAGE
thread. There's lots of good stuff in the RavAGE thread though,
my build was pretty much cribbed straight from Dan's PDF.

Build thread is here: viewtopic.php?f=62&t=1263

I'm looking forward to a polished hardware solution, it looks
like it won't be all that long. A RavAGE board would be great for
my wife's antique daily driver, it needs a new engine harness
anyway...

I also appreciate the open base for the FreeEMS project. It's
important to me that I remain able to modify the firmware on my
ECU. I'm interested in following your experience with LPG. I'm
also interested in alternative fuels, though there is not much
available in my market.
<@TekniQue> but in the end, it's code that makes a computer useful
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Newbee questions on injected LPG for a 289 windsor

Post by Fred »

EDIT: Sim beat me to it! :-) His info is solid too! Top bloke, he is! :-)

Don't bother with the thread, just bother with the PDF and TODO.md which outline all required flaw fixes and additions etc. Obtain a copy with

git clone git://github.com/dvisser/Ravage.git

Then you can just run "git pull" to get updates.

Re the sensors, is it really that hard to pull the pulley off and do something with it? Not that you need a crank signal. In fact, if you know the dizzy drive has little or no slop and wobble, and you know someone with a laser or water jet cutter and are willing to do some testing of a brand new design, you could try this out:

http://status.freeems.org/firmware/decoders/FreeEMS

I have a DXF for it, if required. It needs a two sensor optical unit. You could adapt something from mitsi/mazda (not as good, so the design shown won't work) or a nissan or LT1. In fact, you could use an LT1 one as-is with existing tested code. It depends how easily you can mount it to the old dizzy location really.

A little more info on the FreeEMS disk effort here:

https://github.com/fredcooke/FreeEMS-CAS-HW/

You can go a lot simpler too, but these offer superior sync checking and bad ignition firing protection and faster starting. For the simple option, mount a 24 minus 1 or similar in there with a VR sensor or hall sensor. can also be 16 - 1 or 12 - 1 or any other number upto 60 or so, but you'll probably have issues over the 24ish mark.

Lots of options for you. Missing tooth code is yet to be written, but is on the short list of things to do. FreeEMS disk code is yet to be written, but I'm VERY excited to implement my idea and get it running something. You could use a mitsi/mazda dsm or 4/1 opto cas disk setup and it would be good enough for sure. That is written, tested, used, working. you could also retrofit a toyota dizzy with 24+2 in it and I would be happy to write code to properly support that, or you could run it as 24+1 on existing code.

Hope that helps! :-)

BTW, the best contributions right now are testing! I've got enough work cut out doing necessary firmware work without having to spend time with people helping them integrate their work at an acceptable standard. another good place, if your coding fingers are itchy, is OLV:

https://github.com/gufi/OpenLogViewer

Enjoy!

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Newbee questions on injected LPG for a 289 windsor

Post by jharvey »

DFH is a bit dated, but many parts are still good-ish like the wiki page where it shows the basic capabilities, and general layout. Check the wiki found here. http://wiki.freeems.org, also PUMA helps show the basic overview you might want to check http://puma.freeems.org.
User avatar
oden
QFP80 - Contributor
Posts: 91
Joined: Sat Sep 24, 2011 12:27 pm
Location: Australia

Re: Newbee questions on injected LPG for a 289 windsor

Post by oden »

Thanks everyone for your input, I am learning alot here :)

By the looks of things I wont be attempting contribute any code until I have a running LPG injected system running, that seems like a big enough mission at the moment.

Stage 1: Spark control
Stage 2: Injected LPG
who knows after that..

The motor has been rebuilt so there should not be (much) slop, from playing with a timing light it seems as good as you could get out of such an engine.

Dont have access to any special machining tools but I am willing to have the disk professional made if thats the best path to take. Looking at purchasing a second dizzy to modify for this purpose, mainly so I can restore the car to working condition in case something goes south with the install, remember this car is my (only) daily driver.

Just to make sure I understand, to use the FreeEMS CAS I will need to do the following:
1) purchase a second dizzy to modify
2) purchase a GM LT1 Opti-Spark distributor dizzy ( looks very wide, going to be an issue to fit)
3) make the FreeEMS CAS disk
4) modify the GM LT1 using parts from the windsor to fit onto the motor
5) install the CAS disk into the modified distributor

Then the optical outputs are feed directly into FreeEMS.

Went looking for who manufactures the optical sensor unit in the OptiSpark distributor, turns out its a Mitsubishi product. But you guys probably knew that..

I have to ask, why is the FreeEMS CAS Disk a big improvement over the LT1 disk?

Plan on confirming & ordering all the engine parts, then working on building the FreeEMS board.
User avatar
oden
QFP80 - Contributor
Posts: 91
Joined: Sat Sep 24, 2011 12:27 pm
Location: Australia

Re: Newbee questions on injected LPG for a 289 windsor

Post by oden »

Ok the Opti-Spark only has 4 slots on the second (inner) run, while the FreeEMS has 8. I assume this will make starting faster and generally be more accurate.

The first run of slots (outer) on the FreeEMS disc seems to have less slots then the 360 on the LT1, is that correct? Is this to support very high RPM which the LT1 apparently has issues with? Not that this will effect my application.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Newbee questions on injected LPG for a 289 windsor

Post by Fred »

No no, I was suggesting using the LT1 gear as-is. You can use the sensors from it with the displayed disk too, or any nissan optical sensor from any nissan engine up to the late 90's at the least, possibly much later. You could also use a mitsi/mazda sensor with a lower res freeems style disk, or a dsm/miata disk, as is.

The number of slots on the outside needs to be high to support lots of sync points, and low due to cam driven inaccuracy. For the code it's better for it to be mediumish, up to about 150 or so, perhaps? Right now lower as there is a performance issue that I need to work out. You should do what ever is easiest. You could use the mitsi/mazda sensor and a 24/1 disk from yoshifab to good effect. Starting will be a little slower with that, though.

If that's not clear, pm me your skype details or land line number and appropriate times and we'll have a quick chat to get you kick started :-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
Post Reply