Enthusiastic Newcomer (Long but Organized)

[NathanS]

Who am I?

I'm 27 years old and a hot rodder for more of those years than not. By trade I'm a systems administrator in a predominately *nix environment, so I'm no stranger to open source, large and small. I've been turning wrenches long before I had my license though, and I just can't seem to ever stop tinkering, much to the dismay of the significant other.

Why am I here?

I was all geared up and ready to go to purchase MS3. Everything is budgeted out, down to the nuts and bolts. My parts list estimate was hovering around $900 for the first step of the road, with another $600 planned in future expenses (tablet for shadowdash, MS3X expansion for future dual-fuel capability). But honestly, there's so many things that bother me about megasquirt as a whole, and the not-so-open source nature really gets under my skin. There really should be a more flourishing community around megasquirt, but anyone who wants to develop for the platform in any significant way seems put off by the way many things get handled.

I recently found out about rusefi while searching for megasquirt news updates, curious what might be coming down the pipe, and instantly was interested. A REAL cpu at the core of the project, a FRESH start without the inhibitions of some of the early decisions megasquirt took, and people willing to just DO it, despite the small manpower.

My project.

So the engine/car in question is a Datsun/Nissan 280Z with a turbo engine. I have two blocks, one with dish pistons (OEM) and one with flat tops (OEM) that will eventually get forged slugs. The head is a harder to come by Nissan Maxima gas/petrol head with tiny peanut shaped chambers. Combined with the flat tops it will yield compression in the mid 11's, which would be perfect for my plan of running alcohol (e85 which is cheap, not methanol which is expensive). The original duel-fuel plan will probably have to get put on hold if I'm going rusefi, because I don't see myself doing that much coding anytime soon. This might force me to either choose to run the dish piston block for the time being, or to run e85-only which would be fine since it's a fair weather car anyway. I just wasn't planning on converting the original tank over to support alcohol. The original end goal was to be able to run gas/petrol majority of the time out of boost, which would help with engine start up, fuel economy, as well as washing the alcohol out before shutdown. I think I can give that up for a few years while this platform matures.

My Questions.

1. Is the current code base designed for MAF or MAP, I seem to be finding both answers depending on where I look.

2. If I order the following kit, am I missing anything other than sensors, wire and time?
https://www.tindie.com/products/russian/frankenso-01-full-bundle/

3. How can I get involved with documentation? I make a living off the very fact I can research and learn things faster than most people, and in the last 12 hours I feel like I've learned next to nothing about rusefi other than some basics. I want to find all the details and consolidate and it looks like that's what the wiki has been an attempt at, but there's got to be lots of information out there that hasn't made it way migrating. This is kind of a follow up question so I'll include it in line: Is the source anywhere other than sourceforge? I'd much rather see it on github simply from a ease of documentation/issue tracking perspective, but I also realize that's not even remotely my call.

Conclusions

If more of my initial research goes well, I'll be ordering a bundle asap. My engine is still in pieces, but I really was hoping to get to drive it at SOME point this spring/summer, so I'm going to progress as fast as I can manage (aside from my fill time work, wife, and kids of course). I really do hope to be a great alpha/beta tester as I'm not an "expert" when it comes to electronics, just a passionate researcher, which means I'm pretty decent at running into problems, solving them, and then telling others how to avoid them. Hopefully those skills can be utilized well here.

[AndreyB]

Good questions, and I wish we were same organized...

#1: both. MAP/speed density is more popular and proven, but MAF is either ready also or very close

#2: "Sold out since Jan 27, 2015". There is https://www.tindie.com/products/russian/frankenso-02-diy-kit-/ option
See also
I am hesitating between ordering a bunch of assembled boards sooner, or ordering better boards later. Also still labor intense from my end, also usability still sucks.

#3
https://www.google.com/?l#q=rusefi+github
http://rusefi.com/docs/html/
http://rusefi.com/forum/viewtopic.php?f=5&t=10
http://rusefi.com/wiki

[kb1gtt]

Welcome along.

Have you made a "system schematic" something that shows the leg bone and hip bone connections? Such a schematic would help identify potential issues.

#1, A key issue with MAF is knowing your signal to flow ratio. If you think you know that, you'll be much better off. The first engines done with rusEFI were MAF not MAP. However now that MAP exists it's becoming the more popular approach.

I agree the wiki is in pretty bad shape. I took a jab at it some time back. I mostly created a structure to help organize bits of information when people came to the wiki. However you are correct that the wiki is lacking information. It's very fragmented and doesn't flow very well. I'd like to see it offer a way to show the basics of how a project might go in like 15 minute chunks of time. Perhaps short videos would work well for that. I like this KICAD training video, the transcript could be run through google translate to make this multi-language friendly. I'd like to see short videos similar to this that show certain basics, and can be made multi-language friendly.

http://www.workinprogress.ca/kicad-video-tutorial/

[kb1gtt]

See example system schematics found here

This one uses DIA which is kind of nice but also lacking.
http://rusefi.com/forum/viewtopic.php?f=3&t=360&start=162

This one uses a pencil and paper, simple but less techy.
http://rusefi.com/wiki/index.php?title=Manual:Hardware

I'd like to find a better system schematic option. It would be handy to have some example system schematics which people can use as a starting point.

[NathanS]

Thank you for the links russian, I'm going to sit down with a big caffinated drink this weekend and start pouring over stuff.


kb1git, I could provide a schematic from my MS3 plans, but that's hardly relevant anymore. I currently have ZERO limitations to what sensors I can/need/have to use. The ONLY sensors I have ready to go are a GM IAT sensor (fast acting) that's certaily easy to just ditch as they were nearly free. The only thing semi set in stone is the injectors I have, because I don't want to bother (at this time) with milling my manifold for o-ring injectors.

I have NO harness. I plan to make one.

I have NO map sensor.

I have NO maf sensor.

I have a TPS(witch) that needs to get replaced by a TPS(ensor). I have one that should work, but need to made an adapter plate to mount it.

I have no clue how feasible it will be to have rusefi controlling spark at this moment, so even THAT is up in the air. MS3 could fire my existing coil directly, and also has good documention as to interpereting my optical dizzy signal to be able to run wasted spark, or even full sequential with the diyautotune trigger wheel that offers a secondary sync signal.

My original plan was to run two sets of injectors, one sequetial running gasoline, and a second set running in batch fire with ethanol. I don't expect to make that work with the existing firmware, nor any revision I could help build in the immediate future.


So all that to say that I'd have to draw up a new schematic, and it can be pretty much anything I want. If I go MAF I'll use a sensor that's already well documented as far as flow to voltage, so that tuning is at least in the ballpark to start with, and then follow up with rigorous tuning.

Also, currently the engine is still awaiting final assembly. Hopefully I'll get started on that in the next few weeks. Bottom is together, just need to slap the head and manifolds on and make sure everything gets buttoned up so I can plan my engine harness layout.

[kb1gtt]

What type of coil is it? Is it providing direct access to the primary side, or does it have an integrated ignitor? I'm expecting your has in integrated ignitor because last I recall MS didn't have the capabilities to do the high voltage stuff native-ly. However I could be wrong with that. It's been many years sense I was banned over there, and I haven't followed it sense.

Do you have some details about your optical, like how many pulses per rotation, and how many channels?

[NathanS]

My coil has an ignitor hanging on the same bracket, but that's only required for pre PCB 3.0 boards. You can see this article for an install guide that uses my exact coil and dizzy:

http://www.diyautotune.com/tech_articles/how_to_megasquirt_your_280zx_turbo.htm

As far as the dizzy signal, (this isn't an area of huge expertise for me) I think it's only sending a pulse per spark event. You can look at the tunerstudio settings in the article above and they're faily generic, and not giving you and wheel decoding settings, just a base offset to start with.

The trigger wheel I mentioned that I can run is this one:

http://www.diyautotune.com/tech_articles/how_to_megasquirt_your_280zx_turbo.htm

And then this article shows how to set it up (MS3 is at the bottom):

http://www.diyautotune.com/tech_articles/using_diyautotune_nissan_trigger_discs.htm

Now, if I was running the MS3X expansion board, in theory I could fire COPs directly, but odds are I'd end up using LS1/2 coils which have built in ignitors anyway. There's someone else on the Z boards who's running a setup like that but with Ford coils as well with success.

The only "issue" with that trigger wheel is that I know how much play the dizzy shaft has. It's far from "ideal" and I'd be cautious about being agressive on the spark map, but it's at least an easy way to get sequetial working. If I went that route it would be with a plan to mount a wheel on the crank eventually and have a sync wheel either on the dizzy shaft or on the cam itself.

[NathanS]

A little outdated now, but here's a rough schematic I drew up a few months about (geez, 6 months now) for the whole car, as I'm wiring from scratch.

https://slack-files.com/T02JGUC81-F02UATAM0-c669faaf8b

[kb1gtt]

What did you draw that with? Looks similar to KICAD, but probably some other package. It seems a bit different than what I've seen.

I'd like to find a good system schematic tool. I haven't found anything that I'm really happy with.

[NathanS]

I honestly can't remember. It was definitely geared towards the arduino crowd as it had a lot of those chips ready to plop in, and it allowed you to draw your PCB, Schematic, and Diagrams all in one program and as you dropped components in they went into every view. It was extremely cumbersome to create custom components to drop into a schematic, and lacked what I thought were some pretty basic symbols (like a lamp).

The best program I've used to date is Eagle, which is what I plan to go back to for my final chassis diagram, which will get laminated and mounted in the car.

Edit: Hey look, the logo is on the bottom right of my schematic

http://fritzing.org/home/

[kb1gtt]

Looks like an interesting option
[video][/video]

[NathanS]

Here's a rough location diagram for my own purposes. Not ALL wires listed will actually bundle to my bulkhead, but most will. Also, COP wiring is proposed, not sure when it will be implemented. At least initially I'll be firing off the stock coil.

[NathanS]

Bah! I'd written up a big post, but it required re-login once submitting, and now it's gone. As a sysadmin I should know better and have copied it to a clipboard/buffer of some form first.

Since I'm not going to write it all up again I'll try to make this quicker (last post was a solid 500 word essay).

Status Overview: Still going, but slowly. Other things are getting in the way, and I'm sorting those out. I'll likely be putting a white board in the house AND in the garage with projects and sub to-do's for projects. I'm more and more convinced that rusEFI is more appealing to me than other EFI options.

Engine: The cylinder head I've always wanted became available via trade, so I jumped on that. It had two exhaust manifold studs needing to be removed, one of which broke. After failed attemps with an easy-out I'm considering buying a used drill press, and rigging it up with an abrasive and a sand bag on the lever and let it go for a while.

Chassis Wiring: Coming along but I need to put in more hours. The bulkhead is 90% done, and entirely self-contained, which is what I wanted. If I need to fix wiring on ANY system in this car, I want to be able to easy remove it and fix it on the bench. I find wiring things in-car a huge pita. You'll notice in the following picture that all the other systems interconnect around the perimeter to make things easy to route. For example, all the switches will wire to the bottom strips to run down to a switch panel closer to the driver. Though I have all the wiring for each section run, I need to start wiring devices up. Once complete I can test system functionality minus engine.


(side note: my last gauge panel in this location was stainless, this is alu. Saved a good bit of weight!)

The Future: Not sure what it holds yet. It's tempting to just buy a running engine and plop it in and drive the thing. It's been sitting for 5 years, and it's getting to me. For my original budget of MS3X I could easily buy a running engine + trans + wiring + ecu as a running assembly from a local JDM importer and then learn to be happy with what I get out of an all-oem setup. Or, just get the other projects that are in the way done and gone, and stick with the original goal - a dual fuel L28ET in a bare-bones chassis. My opinions about ideals have changed a lot in the last few years, and the "beast" I once wanted to build with this seems less interesting. Might just table any dual-fuel plans for now and just see where things go, which would mean shooting for sequential out of the gate.

[kb1gtt]

DIY EDM machine could be of interest for drilling a hole in a hardened steal rusty bolt. See this one build out of a door bell. http://reprap.org/wiki/Electrical_Discharge_Machining

[AndreyB]

For my original budget of MS3X I could easily buy a running engine + trans + wiring + ecu as a running assembly from a local JDM importer and then learn to be happy with what I get out of an all-oem setup.

I am considering buying somebodies else turbo miata project just to save myself a year of actually building one. Get a running turbo miata and play with my efi at least on a running platform.

BTW the image does not show up, probably not shared publicly?

[NathanS]

Image should be fixed.

kb1gtt: thanks for the link. Will look into it. I've seen profession versions of those, hadn't thought about making a home-size version.

russian: Yeah I hear ya, loud and clear. It's not always fun doing these things yourself when you don't have a ton of time on your hands.