Stand alone engine management (simple)

Dean924s · Post by **Dean924s** » Wed Jul 20, 2011 2:01 pm

I have been looking at all thing megasquirt for the last 2-3 years and quite honestly MJLJr was to be a stepping stone to going to a MS system on my car. However what I have come to realise is that the MS is trying to be all things to all people with all motors. They do it very well but it can be (and is) overwhelming to the average user.

Could a simple standalone engine control unit be made. When I say simple I mean just make it work one way with one set of parameters. You the end user have to make the necessary changes to the hardware to match the controller.

Some of the things I see are.

Only use one ignition system (edis of course)
Only have a pair of injector drivers that can Handel up to 4 high Ohm injectors (so you can use this on V8's)
Use MAP or TPS since the MJLJr has it already

The rest would be the sensors.
Intake air temp
Water temp.

And???? I think this is it.

Put two or possibly 4 outputs that you can set to trigger on or off that can be assigned to a trigger point at any of the 5 items noted above. This should give most everyone enough outputs to play with.

I think that is it

I have found hat for the most part it is easier to assemble the required parts (injectors, ignition system, sensors) to match a preset recipe that is dictated by the controller than to have to sort through the mountain of info and options that MS has.

The key here is keep it simple. This is what MJLJr a winning product.

Yes I know you have many other projects that still need to be completed but if you could put this type of unit out I bet it would sell really well.

luxseven · Post by **luxseven** » Thu Jul 21, 2011 7:23 am

Hi Dean,

Trouble is that everybody wants/needs a different set-up, has different skills both mechanical- and software-wise. In some point what you write is contradictory. Especially the EDIS point. If I got it right, the new generation megajolt will include its own ignition drivers and EDIS will be obsolete, i.e. less parts and a simpler wiring. So this might just go in your direction aiming at a standalone motor management system.
I totally agree with you for the rest. I want to convert to injection this summer and have that neat Emerald ECU lying here for some months. I read the manual maybe five times, three times in complete for sure. I'm tempted to leave ignition to megajolt and pick up the tach signal to trigger the Emerald (just as if I was still running a dizzy). But then again, this would be overkill with two different systems and softwares to tweak everything. Remember, changing maps with a simple rotary switch, both ignition and injection.
On the other hand, not a lot of people would need all those tweakings, that's where I join you again. A comprehensive GUI (Graphic User Interface) would also add to a better handling, especially for those not familiar with programming. Just imagine a simple fader going from slow to fast with all the tweakings behind staying opaque for the carefree user not wanting to know. When searching the internet, you may find similar systems for american V8's, quite simple installations to replace a four barrel carb. With just a small box with an LCD screen and a few push buttons and sockets to wire the sensors, it's all you need for a complete conversion.
But then again... if it's simple, you have finished early and you only get time to think about other upgrades you want to make

Cheers.

Jos

Post by **brentp** » Thu Jul 21, 2011 7:33 am

Hi Dean,

This is an excellent topic. Thank you for bringing it up.

In short, I think there's room for a simplified engine management system that has a decent set of features *and* can be implemented without immediately requiring hardware modifications (jumper wires, custom part installation, and so on).

The system I envision is Megajolt/C with the addition of injector drivers. Look carefully at the hardware discussion thread and consider the feature set as is- you'll see a full engine management system with just the addition of injector drivers. And, while I'm in the same breath, I'll mention that I've left PWM channels available for this kind of forward thinking.

EDIS should not be considered as the used parts are becoming more and more scarce; especially with EDIS-6 and -8 modules around the world. Megajolt/C will take care of this.

Thanks again for your contribution and your valuable thoughts!

Dean924s · Post by **Dean924s** » Thu Jul 21, 2011 12:12 pm

I will have to go read up on the MJ/C thread. I forgot one input obviously WBO2. If you want a guinea pig for this let me know. I have been gathering all the parts to convert the Porsche to MS but with the complexity of MS it really is a daunting task.

Regarding the availability of EDIS. Any local parts store in my area has them and they are still plentiful in the local Pick and Pulls around here. It sound like I should go get 10-20 of them and put them on the shelf for resale. No Kidding I counted that many EDIS cars the last time I was at the yard. Many were V6 and V8's

Dean924s · Post by **Dean924s** » Thu Jul 21, 2011 2:24 pm

I think I am an idiot I can not find the thread that you speak of Brent. Can you post a link? Also are you still interested in the other project we spoke of a while back?

Post by **brentp** » Thu Jul 21, 2011 3:37 pm

Dean,

Here's the thread:
http://www.autosportlabs.org/viewtopic.php?t=3034

Note that we will support the EDIS module in some form or another as a future variant of the Megajolt/E. Megajolt/C will simply be able to do it without the EDIS controller, gaining much of the flexibility long requested by people in addition to a simplified installation.

Spockie-Tech · Post by **Spockie-Tech** » Sun Jul 24, 2011 2:40 pm

I have done many (well,~ 6) Megajolt Installations and 2 Megasquirts now.

I agree that the hardest part with the Megasquirt is piecing together a coherent picture out of the mountain of options available to you. There are many different revision circuit boards, assembly options, Processors, Firmwares, Sensor Sets, and Configuration Software packages available. Its certainly a daunting lot to digest on your first one.

But, its a bit like "Linux". Potential Windows Refugees often follow something like "I hate Windows, I might try this Linux thing I hear about", only to discover that "Linux aint just Linux" and then be swamped in distributions (redHat, Mandriva, Ubuntu, Suse, Debian etc), versions (2009, Karmix Koala, Sargent, Corporate server), multiple desktop choices (KDE, Gnome, Unity, Ice) etc.

So they often throw their hands up in the air, head to the local "Best Buy" (or whatever), and buy a "Windows PC" thinking its much simpler.

The truth is, there are many versions of Windows too (Home, Professional, Ultimate, Corporate, Server, 64 or 32 bit) etc etc. Its just that when they buy from <BigBoxRetailer>, all those decisions are made for them..

Or for a more auto-related analog, its a bit like asking for "A Carburettor" - you would be expected to know the Brand, Model, CFM requirements, Jetting, Mounting Flanges etc etc.

So while a "Turn Key DIY EFI" system sounds easy in principle, in practice, its not, unless you head for an EFI shop and get them to make the decisions for you.

Someone will want Turbo capability, someone else will want MAP or TPS, Some will Want EDIS, others want to use their Dizzy, Some want a simple 8x8 Map, others want 16x16 to get maximum tweakage ability, some will want a software version of "Turn the Idle Mixture screw out", others will want "Hot Weather Idle Lag recovery if the airconditioning compressor is running" and so on.

The Megasquirt does a fairly good job of trying to be all things to all people.. The TunerStudio software has "Simple and Advanced" operating modes to hide the really confusing stuff for an ex-carb newcomer to EFI.

But there *is* still, unavoidably, a lot to learn. Throwing a spark at a particular time is a heck of a lot easier than feeding an engine fuel.

Ive watched a few "Simple EFI" projects grow over the years, and none of them reach a realistically driveable state (unless they are designed for 1 simple application like full throttle only salt-flat racing or something like that) without also becoming as complex as the thing they were trying to replace.

If it was easy, the first one-jet carburettors would still be on engines without air-correctors, accelerator pumps, progression holes, emulsion tubes and all the other hoo-haa needed to make a carb work.

Megajolt does a great job for those who want to stay with carbs.. I personally dont see any point in trying to reinvent the wheel and try to grow it into something its not. An EFI system is a *long* way from an Ignition controller.

Having said all that.. If you need a hand with a Megasquirt, I can offer some advice having waded through the initial learning curve. Its not *that* hard once you manage to piece the bits together into a picture.

Buy a pre-assembled MS2 (so you dont need to worry about the build options), configure it for an EDIS module (since youre already presumably comfy with them) which is just moving a few wires- It connects to the EDIS, just like a Megajolt does. 2 Wires (PIP and SAW) and youre away. Buy the premade wiring harness, get the $40 TunerStudio software (its much more userfriendly than the free Megatune app). Get an Innovate LC1 Wideband O2 sensor - and hey presto, the TunerStudio Autotune function does 95% of the hard work for you.

The hardest part for my installs has been hacking injector bungs into the manifolds, making up fuel rails, regulators, TPS sensors and so on.

Sensor wise, TPS, Water and AirTemp are all you need (plus the built in MAP, just like a Megajolt).

I dont want to sing Megasquirts Praise too much on Brents Megajolt Forum, but if you really want to go EFI, follow the basics I just listed and you shouldnt have to thrash about too much to get it going.

There are a couple of "Ex Carb head Simple EFI" system on the Market. (Holley Pro-jection is one name that comes to mind), that were just adjusted with Screws (No Laptop required). But why go to all that trouble just to make an Electronic Carburettor ?

Dean924s · Post by **Dean924s** » Sun Jul 24, 2011 3:23 pm

I here what you are saying. My MS build is here. >>http://clarks-garage.com/phpBB2/viewtop ... 38&start=0

I am almost done. I am waiting on a plug for the o2 sensor so I can use the stock wiring. One of the goals has been not to hack up the wiring harness that is in the car.

Looking back on things it would have been much easier (and faster) to just abandon the stock wiring harness in place and install a new/custom one for the MS conversion. (lessons learned). However once I am done I have paved the path for others to do what i am doing. It is the time it takes to figure things out. I don't think people realize that things that I cover in a single post can take weeks or even months to research and come up with a viable solution.

Post by **brentp** » Sun Jul 24, 2011 5:08 pm

Spockie-Tech,

Ah, the age-old question of "why make another X"? Why indeed? And what does history say about that in general?

You raise some excellent points, thank you so much for sharing. In my opinion the MS family of controllers are absolutely fantastic: they are incredibly flexible and have worked extremely well for so many people. Indeed, we are running 2 MS variants - A 1st gen MS1 in our 24 Hours of Lemons Race car + a modern MS-II in our turbocharged AW11 MR2.

The fact that MS exists with all of it's awesome flexibility (and complexity) *actually creates space* for an easy-to-use system that meets about 80% of the use-cases out there. Think about it. You want obscure feature X,Y or Z? Great! Go install a MS. Or, hack what is hopefully the open source design to meet your needs, if that's the itch you want to scratch.

It's also worth noting that we have *nothing scheduled* for developing an EFI system. I'm merely sharing opportunities I've identified based on my view of the landscape. And Dean stirred the pot.

Of the projects we've officially announced, RaceCapture and Megajolt/C are the top priorities for us right now, and we have some really exciting stuff to share soon!

Thanks again for the great discussion.

Post by **brentp** » Sun Jul 24, 2011 5:15 pm

Speaking of a new 'X' - have you seen our Google Plus profile? Lots of little bits of project activity here.
https://plus.google.com/#115841061102670630313/posts
I'm really liking the design and the engagement has been great!

Of course, you can check out our Facebook page here too:
http://www.facebook.com/pages/Autosport ... 4953614230

cng1 · Post by **cng1** » Sun Jul 24, 2011 11:02 pm

We see a log of people struggle with EFI systems. The megajolt is so successful because it is so simple to get to grips with. Despite the simplicity some people do still struggle. Many people struggle with Megasquirt because it is just so flexible.

The gap in the market therefore is for a something that does enough to satisfy the basic requirements that keep 90% of people happy without introducing any extra functionality that is likely to confuse things. The trick to doing that is to be prescriptive over how things have to be installed and not to give into temptation to add an extra feature because it is easy to do so and will help a few percent of potential users.

One thing we've considered doing is taking the Megasquirt/Microsquirt ECU which isn't far off the sort of hardware that we would be talking about and producing some custom "ini" files that simplify the user interface down to just a couple of screens. The beauty of doing things that way is that a user can start with the simplified interface and then if at a later date they want to use some of the more sophisticated functionality that the hardware can offer they simply swap the ini file over in the software and away they go.

Spockie-Tech · Post by **Spockie-Tech** » Mon Jul 25, 2011 1:44 am

There is certainly a market for making things as simple as possible for "Joe Average" - just look at Apple.. They turn complex computers into shiny consumer-toys and make a lot of money by doing so. But they do so by saying "You do things *our* way, or hit the highway and go elsewhere". Conformity makes it easy.

Megasquirt *is* very "Linux-ey" in concept and execution.. "You want triple overhead boost inputs with a variable PWM output to control your heater fan ? No Problem, just adjust this, tweak that, recompile the other and install a flange-whatsit and you're away"..

Which is great for "yes it *can* do it" factor, but as you've said, a little daunting for the I want a "Turn-Key Solution" Market.

Is there a gap in the market for another EFI computer ? Maybe.. it would all depend on your marketing I imagine. Put it in a white shiny box and call it an "I-nject" or something and you might find a niche.. but does that crowd ever get their hands dirty on DIY spannering anyway ? I suspect not.

I'm not sure that it would be feasible to make an EFI system that would appeal to the Budget-concious DIY mindset *and* still be flexible enough to handle all the different engine/car combos, factory wiring looms, plethora of different sensors, Injectors and so on. If everyone was running exactly the same hardware, maybe.

About the only way that sounds simply DIY do-able to me in the Holley Pro-jection Throttle Body Injection Style kit. Here is a Holley-4-Barrel Carb shaped widget that has all the temp, TPS sensors, Injectors, Regulators etc built into it with a single plug. Slap it on your manifold in place of your Holley Carb, Plug it in, feed it fuel, and twiddle these knobs. Jobs (sort of) Done.

Trying to do brain transplants on existing EFI cars, or retro fit Port Injection Systems with sensors and Injectors scrounged from all over the place is always going to be a bit of a Frankenstein Job and require lots of hacking about to get them all itegrated. Hence the ultra-tweakable MS.

@Dean: reading your build thread, it looks like you've already done the hard yards getting things sorted and are close to running, so my "how to do it simple" reccomendations are about a year or so late

@Brent. I have an Idea for you over in the RaceCapture thread.
(it was here, but didnt want to derail Deans Thread)

Dean924s · Post by **Dean924s** » Tue Aug 02, 2011 10:17 pm

@ Spockie-Tech

I here what you are saying BUT

I think that there may be a market for just such a "hacked" system.

If you simply say No you are not going to use the stock wiring harness and you have to use high impedance injectors and you have to run a MAP based system and you have to use GM IAT and water temp sensors and then specify GM coil packs and a wide band it would make setting up the system simple.

Yes it would take some junk yard hunting and posibly just ordering parts from your local parts store but it would make the setup of the "I-Inject" foolproof. No jumpers no messing with software and the menuing would become very simple. All you would have to do is imput the basics about your motor and start tuning. (ok over simplified I know)

I can see a market for this kind of product for kitcars and conversions to EFI. All the other fancy stuff like sequential injection and COP ignition just is not needed. Cool yes but not needed for about 90 percent these types of projects.

Spockie-Tech · Post by **Spockie-Tech** » Tue Aug 02, 2011 11:52 pm

I agree that there probably is a market for such a system..

In fact I may even have come across a similair product when I was surfing about sucking up Megasquirt information across the net. Problem is, I forget the name of it now.

I recall it looked like it was based on Megasquirt, but had a "kit" of parts they sold, plus possibly some custom tuning/gui software and was fairly pricey.. I remember their "tuning advice/tips page" was interesting reading..

And thats likely the biggest problem you would face.. Finding your Market.

Were you to make such a system, you have to get word out to right sort of people who want EFI conversions, but arent savvy/patient enough to hack through the Megasquirt Info-Jungle..

This would likely require some fairly heavy efforts in start-up Marketing to sing the benefits of your system whilst travelling the world/internet to get word out there so the Technically limited would hear about it. You would want to find the workshops that do custom cars, but dont have a resident geek to suss out the MS stuff themselves.

Do-able, but not much fun for a tech-head.. it would be more of a profit/business project for a salesman type to package&market such a system.

Dean924s · Post by **Dean924s** » Wed Aug 03, 2011 1:37 pm

What about just another "box" that is sold by some company that already sells "box's". (ya Brent I am picking on you

) Let it sell its self. I have done the marketing thing for another automotive related devise and it is alot of work. The better way is to let it grow slowly on its own.

What irritates me the most is when there is a good product out there and the designer / builder feels they have to "improve on it". This usually results in the cost increasing and a product that can do more but does not do any of the original stuff better. This in turn makes the product more complex and this is can actually hurt a product.

I think this is what has happened to MS. However they apparently gadget there market correctly and found that the added complexity gained them a larger market than what they were going to loose by adding complexity. In other words they had a net gain in there customer base.