rusEfi: DIY ECU

rusefi · Post by **rusefi** » Tue Apr 08, 2014 9:22 am

DIY as in real Do-It-Yourself from scratch ECU.

I am Andrey, I am involved with www.rusefi.com - that's a project to develop an open-source DIY engine control unit. On the one hand, that's reinventing the wheel - on the other hand, there are not so many open-source ECUs (and no, MS is not open-source AFAIK), and by using a modern microcontroller and a higher-level modular code I think this new ECU can make a lot of new things possible.

But for now we have to develop it further :)

We have a board we have developed which is the ECU:
http://rusefi.com/forum/viewtopic.php?f=4&t=359

And we have a kickstarter project which has some details: https://www.kickstarter.com/projects/31 ... t-ecu-firm

and we have www.rusefi.com website which explains a bit more.

Why am I posting this here? Because we have just fired up a second test engine, and that engine is a Neon engine of one of the forum members:

[youtube][/youtube]

Mr Sabastian Half · Post by **Mr Sabastian Half** » Tue Apr 08, 2014 9:32 am

Dood! If you could make a computer for our cars that would read all Parameters ready & can pass emissions testing you friend will be well off cuz i personally would pay $500 for something like that just to keep my car inspected! A little change here & there on timing, fuel ect would be a great selling point also! Pls keep us posted very interested

Sent from my iPhone using Tapatalk

rusefi · Post by **rusefi** » Tue Apr 08, 2014 9:41 am

Mr Sabastian Half wrote:can pass emissions testing

U.S. Clean Air Act might stand in the way of this. While we already have some CAN support and we can eventually implement OBD via CAN, you would need to look into the legality of this in your specific region.

occasional demons · Post by **occasional demons** » Tue Apr 08, 2014 10:12 am

It would need to be C.A.R.B. certified to pass in a lot of states. But if you could make it appear on a scan tool exactly like the oem PCM, it would probably fool most inspectors, since they rarely look under the hood anymore. (at least here)

But once you go past certain physical parameters with valve timing, etc, the best electronics in the world aren't going to make it idle stock, or have 02 readings that would pass. You might be able to have the PCM give false readings, but when it smells pig rich, any inspector worth their weight will know something's up. If they do sniffer tests in your area, you would be done right there.

If all you want to do is mock a factory PCM, you might as well just use one. But where the PCM is hard to get, or you don't feel like waitng for an MTX PCM to come available for a swap, and don't have testing, this could be a nice thing to have around. It would be a decent option for the off public road use stuff too.

Post by **Vanquizor** » Tue Apr 08, 2014 10:47 am

Mr Sabastian Half wrote: A little change here & there on timing, fuel ect would be a great selling point also! Pls keep us posted very interested

Sent from my iPhone using Tapatalk

If by a little change here and there you mean complete customizability that's kinda where RusEFI is at. I'm the guy that fired up engine #2 and I keep putting my time and effort into this project because I see it's potential being FAR beyond what AEM, MS and others are doing...

belalegosi · Post by **belalegosi** » Tue Apr 08, 2014 11:02 am

So users would be given full access to the code?
Would their be a GUI for tuning or would adjustments need to be made in the raw code?

What about hardware customization? The board is all SMT. I don't like reworking SMT components.

95neonhybrid · Post by **95neonhybrid** » Tue Apr 08, 2014 11:08 am

belalegosi wrote:So users would be given full access to the code?
Would their be a GUI for tuning or would adjustments need to be made in the raw code?

What about hardware customization? The board is all SMT. I don't like reworking SMT components.

i totally agree with this, customization via sd card, usb, the ability to add or remove components in a proto area to allow better or faster upgrades. it would make other systems look under thought. live data and air fuel adjustability would be a nice add on

rusefi · Post by **rusefi** » Tue Apr 08, 2014 11:10 am

Users are already given full access to source code & documentation: http://rusefi.com/wiki/index.php?title= ... code_intro

rusEfi is already compatible with TunerStudio, there is also rusEfi own dev console which is a bit more suitable for current phase.

Looks like nicer components are only available in SMT these days, you would have to either force yourself into SMT (which is not that bad as long as you use a $5 syringe of soldering paste), or you would need to research through-hole options. Anyway, the hardware is also documented.

Sounds too good to be true? That's because some areas are not ready yet. A lot of stuff is still missing: still no knock detection, still no speed density. https://www.kickstarter.com/projects/31 ... t-ecu-firm is one of the ways to help a bit :)

SD already on the board. x2 USB already on the board :)

rusefi · Post by **rusefi** » Tue Apr 08, 2014 11:12 am

95neonhybrid wrote:live data and air fuel adjustability would be a nice add on

Data is already live, for now you just adjust timing and fuel amount, but we are working.

Ohh, did I mention https://www.kickstarter.com/projects/31 ... t-ecu-firm ? :)

belalegosi · Post by **belalegosi** » Tue Apr 08, 2014 11:21 am

rusefi wrote:Users are already given full access to source code & documentation: http://rusefi.com/wiki/index.php?title= ... code_intro

rusEfi is already compatible with TunerStudio, there is also rusEfi own dev console which is a bit more suitable for current phase.

Looks like nicer components are only available in SMT these days, you would have to either force yourself into SMT (which is not that bad as long as you use a $5 syringe of soldering paste), or you would need to research through-hole options. Anyway, the hardware is also documented.

Sounds too good to be true? That's because some areas are not ready yet. A lot of stuff is still missing: still no knock detection, still no speed density. https://www.kickstarter.com/projects/31 ... t-ecu-firm is one of the ways to help a bit :)

SD already on the board. x2 USB already on the board :)

this forum likes to eat links for breakfast.
[ u r l ] link [ / u r l ]

I'd be interested in the source code. Its been a while since I've tinkered in embedded systems... sometimes I miss it

I hear ya on the the availability of parts for through hole components.

What would be great though is to have a protoboard area. The main components can be SMT. Those don't/shouldn't get customized much. But "add ons" like launch control, boost, shift lights, etc can still remain through hole and be built in the proto area..

95neonhybrid · Post by **95neonhybrid** » Tue Apr 08, 2014 11:33 am

belalegosi wrote:
rusefi wrote:Users are already given full access to source code & documentation: http://rusefi.com/wiki/index.php?title= ... code_intro

rusEfi is already compatible with TunerStudio, there is also rusEfi own dev console which is a bit more suitable for current phase.

Looks like nicer components are only available in SMT these days, you would have to either force yourself into SMT (which is not that bad as long as you use a $5 syringe of soldering paste), or you would need to research through-hole options. Anyway, the hardware is also documented.

Sounds too good to be true? That's because some areas are not ready yet. A lot of stuff is still missing: still no knock detection, still no speed density. https://www.kickstarter.com/projects/31 ... t-ecu-firm is one of the ways to help a bit :)

SD already on the board. x2 USB already on the board :)
this forum likes to eat links for breakfast.
[ u r l ] link [ / u r l ]

I'd be interested in the source code. Its been a while since I've tinkered in embedded systems... sometimes I miss it

I hear ya on the the availability of parts for through hole components.
What would be great though is to have a protoboard area. The main components can be SMT. Those don't/shouldn't get customized much. But "add ons" like launch control, boost, shift lights, etc can still remain through hole and be built in the proto area..

thats what i meant to say for the protoboard area, ms does offer that and alot of people take advantage, the fic doesnt

rusefi · Post by **rusefi** » Tue Apr 08, 2014 11:44 am

Gotcha on the proto area, I guess no reason not to have it space permitting.

Fixed the urls in the previous messages.

Post by **Vanquizor** » Tue Apr 08, 2014 12:41 pm

belalegosi wrote:
this forum likes to eat links for breakfast.
[ u r l ] link [ / u r l ]

I'd be interested in the source code. Its been a while since I've tinkered in embedded systems... sometimes I miss it

I hear ya on the the availability of parts for through hole components.
What would be great though is to have a protoboard area. The main components can be SMT. Those don't/shouldn't get customized much. But "add ons" like launch control, boost, shift lights, etc can still remain through hole and be built in the proto area..

Why would you want to put that stuff on board rather than just using a module on the CANbus to gather your inputs or drive unique outputs?

belalegosi · Post by **belalegosi** » Tue Apr 08, 2014 1:06 pm

Vanquizor wrote: Why would you want to put that stuff on board rather than just using a module on the CANbus to gather your inputs or drive unique outputs?

That's another option. But then the user would have to source CAN supported modules to perform each operation they want.
How easy is it to find CAN supported shift light, boost control, launch control, etc modules?

My thoughts behind the protoboard is the flexibility to do what you want, when you want and how you want. Maybe I'm spoilt

Post by **Vanquizor** » Tue Apr 08, 2014 3:44 pm

Protoboard is nothing special- its just a bunch of holes. Why wouldn't you use a $3 chip like MCP25020 to put your bunch of holes outside the main case on a through hole board?

Either way RusEFI is really young so there are lots of opportunities to play with different ways of implementing the kind of add ons you are talking about.

I'm sure if you were to spec some hardware and write some code to play nice with neon wheel speed sensors then launch control, traction control and stability control are achievable. The discovery board we are using already contains a 3 axis accelerometer capable of +- 8g so I'm talking actual launch control rather than MS's simple 2 step rev limiter.

Picture this- you engage launch control in your SRT4 pushing the button on the steering wheel and mash the gas. As the revs come up the ignition is retarded and fuel mix goes richer to spool the ridiculously large turbo you have fitted. As you sidestep the clutch the ecu sees the clutch switch toggle and begins modulating a solenoid on the clutch line to control the line pressure in the clutch hydraulic circuit. This control provides the rate of clutch slippage required to keep the engine rpms in the power band and minimize drive line shock to keep the tires stuck and your axles in 1 piece. As the car is rolling out and the power is really getting applied the ecm is monitoring wheel slippage and correcting with slight applications of brake through the ABS module or engine power delivery for larger discrepancies. After the first few feet the ecu sees less intervention required to meet targeted wheel slippage and begins feeding in all the power it can by holding the waste gate tight and feeding in the timing advance. Several more feet down wind the ecu is delivering maximum power through conventional means and wheel slip again is sliding below target so it begins pulsing the nitrous solenoid. As the tach comes up to redline you pull the shifter and the ecu give a dab of the clutch matches revs and starts pulling hard in the next gear, again maintaining optimum wheel slippage for maximum acceleration as long as power is available.

Now on the other hand I'm a road racer, so I want this kind of wheelspeed control while respecting my desire to also maximize lateral acceleration...

belalegosi · Post by **belalegosi** » Tue Apr 08, 2014 4:12 pm

Vanquizor wrote:Protoboard is nothing special- its just a bunch of holes. Why wouldn't you use a $3 chip like MCP25020 to put your bunch of holes outside the main case on a through hole board?

Sure by all means use a CANbus system and yes a protobroad IS nothing special. But its not moving more to the non-DIYer side of things.
My comments are all based of this the first line of the first post

rusefi wrote:DIY as in real Do-It-Yourself from scratch ECU.

Vanquizor wrote:
I'm sure if you were to spec some hardware and write some code to play nice with neon wheel speed sensors then launch control, traction control and stability control are achievable. The discovery board we are using already contains a 3 axis accelerometer capable of +- 8g so I'm talking actual launch control rather than MS's simple 2 step rev limiter.

Clearly you're a bit biased.

Vanquizor wrote: Picture this- you engage launch control in your SRT4 pushing the button on the steering wheel and mash the gas. As the revs come up the ignition is retarded and fuel mix goes richer to spool the ridiculously large turbo you have fitted. As you sidestep the clutch the ecu sees the clutch switch toggle and begins modulating a solenoid on the clutch line to control the line pressure in the clutch hydraulic circuit. This control provides the rate of clutch slippage required to keep the engine rpms in the power band and minimize drive line shock to keep the tires stuck and your axles in 1 piece. As the car is rolling out and the power is really getting applied the ecm is monitoring wheel slippage and correcting with slight applications of brake through the ABS module or engine power delivery for larger discrepancies. After the first few feet the ecu sees less intervention required to meet targeted wheel slippage and begins feeding in all the power it can by holding the waste gate tight and feeding in the timing advance. Several more feet down wind the ecu is delivering maximum power through conventional means and wheel slip again is sliding below target so it begins pulsing the nitrous solenoid. As the tach comes up to redline you pull the shifter and the ecu give a dab of the clutch matches revs and starts pulling hard in the next gear, again maintaining optimum wheel slippage for maximum acceleration as long as power is available.

Now on the other hand I'm a road racer, so I want this kind of wheelspeed control while respecting my desire to also maximize lateral acceleration...

Sounds like your're talking more about traction control with a splash of launch control.

Post by **Vanquizor** » Tue Apr 08, 2014 4:56 pm

I agree we have different definitions of launch control- when I say launch control I'm talking F1 standing start to corner 1 launch, I'm not sure what your definition is.

To be fair to megasquirt its good for what it is, and I currently run 3 of the units. I just think its peaked and is now growing outside its market- at $1200 for the latest MS its not really a full feature entry level unit any more, and the fact that you have to use THEIR hardware prevents others from driving the cost down.

To clarify my bias- I have no financial or commercial interest in RusEFI, I have purchased all my components at full retail pricing and done my best to help support the project based on a request made on another race car forum. I fully support the initiative for what it is- a completely open and shared project that is likely to cost me a pile of cash, waste a bunch of my time, and maybe eventually give me and every "Johnny come lately" after me a pretty slick control unit at a reasonable price in the end.

95neonhybrid · Post by **95neonhybrid** » Tue Apr 08, 2014 5:16 pm

for me its matter of being able to tune without having to be a code wiz or or spend copious amounts of time trying to understand the fuel tables etc. i might not be the smartest in a group, but my efforts show a desire to learn. so a user selectable tune would be nice rather it be by level from 1-100 based (only way i can describe) kind of how you would install an operating system, automatic or custom

unacceptable · Post by **unacceptable** » Mon May 12, 2014 7:04 pm

This is pretty neat, any progress reports? I love that the "CPU" for this is a $15 unit, vs the $90-$200 that MegaSquirt's CPU goes for.

rusefi · Post by **rusefi** » Mon May 12, 2014 8:01 pm

unacceptable wrote:This is pretty neat, any progress reports?

There is definitely a lot of not really visible "under the hood" progress. On the hardware side, we are getting closer to a full self-contained PCB with the main chip & power supply right on the board. On the software side, we are moving towards MAP/speed density...

Any chance you would be interested to try all this & provide feedback? :)

belalegosi · Post by **belalegosi** » Mon May 12, 2014 8:53 pm

cost?

rusefi · Post by **rusefi** » Mon May 12, 2014 9:11 pm

belalegosi wrote:cost?

You can order your boards from any PCB shop or you can buy them @ https://www.tindie.com/stores/russian/ - there is an option for a fully assembled one there, too.

rusefi · Post by **rusefi** » Fri Jul 11, 2014 2:42 pm

new board - just ordered a first batch

Post by **joesf** » Mon Jul 14, 2014 5:01 am

So we would need something like this for the engine harness http://www.digikey.com/product-detail/e ... ND/1122079

rusefi · Post by **rusefi** » Mon Jul 14, 2014 5:08 am

joesf wrote:So we would need something like...

For the wire harness you need to either cut a piece of civic/camry harness in a junk yard or maybe one of these "civic harness adapter" from ebay:
http://i.imgur.com/HMaioSi.png

rusefi · Post by **rusefi** » Thu Aug 21, 2014 6:32 am

new board is mostly functional

but I am at the edge of my own car skills - I cannot tune a car from scratch properly, without that I cannot really make any meaningful progress. Looking for some gearheads to play with this, we can figure something out with the boards if needed.

Post by **boost junkie** » Thu Aug 21, 2014 9:37 am

Can we look at your tune and some data logs?

rusefi · Post by **rusefi** » Thu Aug 21, 2014 9:45 am

boost junkie wrote:Can we look at your tune and some data logs?

There is no tune - my test mule seems to consume both oil and coolant, I cannot find the timing mark on the pulley and the O2 sensor is located in a spot where I cannot take it out in order to put my WO2 in (it's not a neon).

I will probably get rid of this car and shop for another test mule, but the fact that I did buy this car in the first place tells you where I am with car skills in general.

pimp_daddy_patty · Post by **pimp_daddy_patty** » Wed Sep 10, 2014 1:42 pm

This is a very interesting project. I read over this thread, however I'm curious as to the advantages this would have over megasquirt. As in, are there specific reasons why I would want to replace my current megasquirt with the rusEFI unit?

Obviously, the cost compared to potential features appears to be an advantage.

I'm just asking in order to get more info out there.

rusefi · Post by **rusefi** » Wed Sep 10, 2014 1:55 pm

pimp_daddy_patty wrote:This is a very interesting project. I read over this thread, however I'm curious as to the advantages this would have over megasquirt. As in, are there specific reasons why I would want to replace my current megasquirt with the rusEFI unit?

The potential is the only difference. rusEfi is starting with modern microprocessor and ability to move to whatever would be the next microprocessor in the future. http://en.wikipedia.org/wiki/VEMS#Comparison I guess question would be what is missing and impossible in megasquirt?

It is not there yet, but rusEfi has the processing power to support dual wide-band O2 and four channels of EGT and a lot of other stuf. But in order for the cool stuff to happen, we need people supporting or testing or criticizing this :) What we have right now is a couple of running engines and an LCD screen which tells you that intake air sensor is disconnected instead of flashing a code in your face - I see this as an example of usability.

pimp_daddy_patty · Post by **pimp_daddy_patty** » Wed Sep 10, 2014 2:15 pm

Thank you for replying.

It seems to me that the rusEFI surface mount has cost as it's advantage at this time especially with the fancy LCD display and such. I have a 94 Ford Ranger that I intend to swap a turbocharged 5.0L V8 into over the winter/spring. What would I have to do/know in order to help with testing if I were to use this vehicle as a test mule?

Unfortunately, my neons have to pass OBD-2 emissions and would not be a fit at this time.

rusefi · Post by **rusefi** » Wed Sep 10, 2014 2:21 pm

pimp_daddy_patty wrote:I have a 94 Ford Ranger that I intend to swap a turbocharged 5.0L V8 into over the winter/spring. What would I have to do/know in order to help with testing if I were to use this vehicle as a test mule?

Unfortunately, my neons have to pass OBD-2 emissions and would not be a fit at this time.

You would probably need some hardware to run rusEfi logic. You can either invent your own hardware from scratch, or order the boards from your favorite PCB manufacturer using our open-source files, or buy a board/kit/assembled boards from us - see http://rusefi.com/wiki/index.php?title= ... e:For_Sale

For you neons you can make a direct plug-in ECU with the stock connector and swap it in place of the OEM ecu while inside the garage or on the road where it is legal :)

rusefi · Post by **rusefi** » Fri Sep 26, 2014 8:25 pm

A 2003 manual is the new official rusEfi test mule! Please do not hesitate if you want to donate a burned or non-burned stock ECU to this cause - I would need one ECU just for the connectors and another for bench-testing.

Post by **boost junkie** » Sat Sep 27, 2014 4:51 am

FYI most people here have 95-02 neons which use a completely different ECU. In 03-05 they switched to an "NGC" controller which has a completely different harness and cam/crank triggers.

fixit · Post by **fixit** » Sat Sep 27, 2014 6:11 am

Hi, seems your somewhat local -- pm if you want too meet for a beer.. would love to see the neon...

I've been thinking of porting the the MS code (http://megasquirt.free.fr/sources/MS/ma ... n_v2.890.c) to a raspberry pi w/ audrino

rusefi wrote:

A 2003 manual is the new official rusEfi test mule! Please do not hesitate if you want to donate a burned or non-burned stock ECU to this cause - I would need one ECU just for the connectors and another for bench-testing.

rusefi · Post by **rusefi** » Sat Sep 27, 2014 6:18 am

fixit wrote:I've been thinking of porting the the MS code

Be careful with MS jihad - would that work with their license? But that's a very different processor so it would be more a re-write then a port maybe you should be fine.

How far are you from 07302?

fixit · Post by **fixit** » Sat Sep 27, 2014 7:35 am

work about 15 min away, live in central jersey (use to live in boken)

http://www.reddit.com/r/arduino/comment ... _ignition/
So for a while I've been putting together a project for a DIY engine management system (EMS) designed to help people understand the concepts in these 'black boxes'. It came about because I seemed to talk to a lot of car people who were kinda interested in electronics and a lot of electronics people who were kinda interested in cars/engines, but few that overlapped both.
The main aims were to be in all places as simple as possible. No weird build environments, no knowledge of assembly needed, favour simplicity over performance where possible and make as low a barrier to entry as can possibly be achieved. To this end, I'm basing this on an arduino mega which is cheap and very readily available, but is really limited spec wise. It's not so limited that this isn't achievable, but it does mean using everything the hardware has to offer.
Some of the currently planned features are:
100% compilable and installable from the Arduino IDE
Little or no assembly and definitely none in the main algorithms
Fully programmable MAP (Speed density) or TPS (alpha-n) based injection and ignition
8x8 maps for the above with interpolation
Batch support up to 8 cylinders, full sequential up to 4 cylinders
Tuning across all platforms using Tuner Studio or Mega Tune . This includes things provided by these packages such as 'auto tuning' etc.
The software is still under development (Probably always will be) and the first round of shield PCBs have just been printed. It's complete enough to run an engine if anyone is interested in trying it out and has run a couple of single cylinder engines so far. I'm hoping to get a 4 cylinder engine up and running shortly.
If anyone would like to check it out, the code, schematic and gerbers are available at: https://github.com/noisymime/Kartduino Of course, I'm more than happy to answer questions about EMS' in general or this one specifically. :)

rusefi · Post by **rusefi** » Wed Oct 29, 2014 7:29 am

Lemons race season is over and I am now focusing on running my NGC Neon with rusEfi.

Couple of questions if maybe someone would have the info:

IAT and CLT thermistors curve? what is the OEM pull-down value?

What kind of idle valve do I have? Looks like a two wire. Just a solenoid? What kind of frequency?

Alternator field wire. Do I pull it to GND or +12 to turn alternator on?

what neon · Post by **what neon** » Wed Oct 29, 2014 7:42 am

Field wire is + on ngc motors.

Tapatalk Yo. NOTE II

Toomey200 · Post by **Toomey200** » Wed Oct 29, 2014 11:07 am

You're in NJ, I'm in NJ, come use my car as a test :)

Who is online