Yep.

 

sorry to bring back an old thread. but i have been contemplating this exact idea in my head over the past few days. it seems to me like it would work perfectly. I am a honda guy at heart and ive been chipping ecus for years using crome/ECtune/Hondata. I just recently got a miata and ive been looking into EMS's for it. Crome is FREE as long as you don't need the RTP. and its only $150 for the Pro version which includes the RTP and datalogging. I already have all the sensors, wiring harness, dizzy, ECU and a Crome Pro license. Now I just need to get to work on designing either a new cam gear, or a weld on extension piece like you have shown here.

has anyone put any effort into this? if so i would love to see what you have done. there are a few people that have done this with the SR-20.

http://www.youtube.com/results?searc...l550l2-1.1l2l0

 

I don't remember who (matthewsdesigns maybe?), but somebody here re-pinned an eeprom DSM ecu and is running their turbo miata on dsmlink. Nobody has done it with a honda ECU/Dizzy simply because odds are the cost and effort involved for the average Joe would be more than just installing a megasquirt and calling it good.

 

I believe his sn on here is "underdog". Running DSMlink though at some point I believe he was running (or is running now) a megasquirt.

I honestly don't see the point/need for something like is proposed here.
I mean if you're a honda guy at heart and want to do it "just to do it" then cool, knock yourself out, but if you're wiring/soldering savvy and can put together something like this, an MS is going to cost like 300-400 bux to put together yourself and is more than enough control for your average turbo miata. Maybe I'm missing something

 

my main reason for it is that i already have all the parts laying around. plus i would be using the same tuning software that i am already comfortable with and know how to operate correctly. really the only thing that would cost me anything is the coupler from the cam gear to the distributor. the rest is just labor.

 

The reason for doing it is because the Honda ECU is awesome and it costs next to nothing to work with.

Megasquirt is nice and all, but with a Honda ECU using FREE software, you can get resolution like this:

 

Full Tilt Boogie,

having 250 RPM and 2 psi jumps is useless unless you are willing to put in the dyno tuning time to tune each cell. ANd FWIW the AEM has 24 x 32 or something lots.
You really only need enough RPM and MAP jumps so that your timing error from simple interpolation is < 1° and your fueling error is <5%.
You can't really dyno tune a motor's timing map to < 1° of accuracy. Too many other variables - intake air temps, piston crown temps, exhaust manifold pressure.


Here's an interesting feature of the K series ECU. How much would it cost to buy a junkyard one and buy the software?

https://www.miataturbo.net/showthrea...t=62792&page=2

 

You know I honestly think the best feature for aftermarket ECUs that isn't there are autotune wizards for the stuff that's time-consuming to autotune, such as idle.

IOW they need features that save tuner time.

Only the (early) Adaptronic has a modicum of timing auto-tuning.

Only the Autronic has a VE x RPM 2D curve and whose fuel (inj) table changes when you change the AFR tables. This makes initial (and subsequent) fuel curve tuning easier.

 

hondata kpro is something like 1100-1600 iirc (if that's the one you're talking about.
a k20 ecu can be had between 150-300.

not cheap

And I was going to say the same thing about the resolution: I tried 250rpm increments on the adaptronic when initially tuning it and all it did was slow down tuning time and be more confusing. No difference in driveability/power/or smoothness between it and 500 or even 1000 increments. (though I ended up sticking w/ 500).

 

There are lots of plugins than can be used with Honda EMS that allow you to autotune, boost by gear, EBC, anti-lag, etc.

I know the resolution is not really needed for the most part, but its there if it is. Its not difficult to tune at that high resolution and I have seen many times where sometimes there WILL be places where you want to make very tight adjustments, especially at low RPM.

Have you guys ever seen the maps for a sportbike? The resolution is massive, they are using maps the size of the one above thand the engine is still NA! I can tell you that the resolution is definitely being used up too, its not like there are big expanses of cells with a linear change in cells.

There are people in the Honda world making well over 200 hp on 1.6L naturally aspirated engines, and people making 800hp Turbocharged and theyre doing it with a chipped Honda ECU. You cannot argue with results, especially considering the price.


---- K-pro.

Chipped p28 = easily under 100 bucks if you do it yourself
Ostritch = $175 new (easily found used for less)
Chip burner = $80 for a good one
Datalogger is literally just a USB to serial cable
Software = Free with Crome, not much more for licensed software


What I think would be cool would be modifying a Honda IM to fit on a BP, and just rocking a Honda throttle on it. You have the IACV, IAT, and MAP sensor right there on the IM, and the rest of the sensors in the distributor and youre done. Very clean install, all fully Honda, and works perfectly with the ECU.

 

I mean its cool but MS has nearly all those features already, and you can always define/scale map/rpm values closer together in those spots where you want more resolution.

I'm just providing my input, not arguing one vs the other. In fact I'd like to see your guys' ideas put to work: the more options we have for EM the better off we are.

 

I'm not sure if it's the same as this, but MSII has a feature whereby once the VE table is tuned to your target AFR's, you can turn on the feature (forget it's name and it's helpfully not documented yet), and adjustments in the target AFR table directly adjusts the VE table to hit your new target AFR.

 

Cool.

BTW adding a distributor is a step backwards. You add complexity, cost, length of spark cables, reduce spark energy, and add wear points.

 

Honda dizzy's have 3 signal rings, it'd be neat if you could make the honda ecu run off two, and then make a custom cas wheel via emachineshop.com or similar. Of course then you'd have to figure out a way to make it run 2 coils via wasted spark... or even hack a a hole in the firewall to run a FWD bp dizzy, lol.