T-leg Piggyback ECU R&D

azn2nr · Post by **azn2nr** » Wed Feb 23, 2005 6:43 pm

is power fc a piggyback or a standalone??

evolutionmovement · Post by **evolutionmovement** » Thu Feb 24, 2005 2:32 am

Whoa! We buy direct from an AMP truck at work. I'll see if I can find out about these connectors. How should I describe them best?

Steve

vrg3 · Post by **vrg3** » Thu Feb 24, 2005 2:39 am

Woah. Hand them an ECU -- or, failing that, pictures of the connectors -- and ask them!

THAWA · Post by **THAWA** » Thu Feb 24, 2005 3:49 am

muwhahah!

After a few hours of searching I found the part number for the ecu connector. It is 2-177771-7 (76 pins). The part numbers for the individual connectors are, from largest to smallest; 174516-7 (26 pins), 174515-7 (22 pins), 174514-7 (16 pins), and 174913-7 (12 pins). You can verify this with AMP's site just to make sure I didn't mix something up. As far as ordering them. It looks like the indiviual connectors can be ordered through some place but the ecu connector doesnt look like that's possible. Anyway it's a start.

vrg3 · Post by **vrg3** » Thu Feb 24, 2005 4:31 am

Woah, nice job, Hardy! A couple years ago I tried looking for weeks and failed.

So it seems that the large contacts are part number 173631-1 and the small contacts are part number 173716-1. Cool. The contacts are the only part that you can't scavenge from a car.

THAWA · Post by **THAWA** » Thu Feb 24, 2005 5:17 am

Oh, the plugs don't come with contacts? Weak. At least we've got all the part numbers we need

. OOOOOOO I can just use this to create a plug n play thing for the engine swap and just change and add wires with the new connectors. AWESOME! Now to find the 76 pinner

THAWA · Post by **THAWA** » Thu Feb 24, 2005 7:22 am

Just to add more to this: http://kaele.com/~kashima/car/evo/cover.jpg

EVO 8's use the same connectors!

But that's not all. Since EVO owners get screwed if they go in the dealer with piggybacks and snipped wires, companys have started selling an adaptor harness which you can splice in whatever you want. For example: http://www.machevo.com/cocoecupaha.html or http://www.buschurracing.com/cgi-bin/ez ... ?20X376545

So we do have a harness we could buy and splice into without doing any cutting. It just costs 150 bucks to do so. :\

vrg3 · Post by **vrg3** » Thu Feb 24, 2005 7:36 am

Woah! Awesome! Judging by those prices the connectors themselves must be pretty cheap.

THAWA · Post by **THAWA** » Thu Feb 24, 2005 7:55 am

Wow so there's a company that does just this sort of thing: http://www.quicksilverperformance.com/pages/3/index.htm

Look at how many different cars they make harnesses for! See the EJ20G there.

ooo here's another: http://www.autosportwiring.com/products.htm

Whoa, Adam sells the EVO 8 harness for less than others: http://z1auto.com/prodmore.asp?model=ev ... prodid=241 $110 shipped! I'm going to have to support this

free5ty1e · Post by **free5ty1e** » Thu Feb 24, 2005 3:00 pm

WOW..... kickass you guys - I go work on something else for a day, come back and we've almost got the quick-connect worked out! That's great news.... those connectors must be available somewhere inexpensively....

Just wanted to let everyone know, by the end of this weekend I should have the unit reading MAF, MAP, and tach signals as well as converting into selectable units. So at the very least, so far, it's a digital airflow, boost/vaccuum, and tach gauge. Depending on how far ahead I get with my other work, I should be able to start miming the airflow signal out from a DAC to show unscaled passthrough, and then start utilizing adjustments from the trim table...

free5ty1e · Post by **free5ty1e** » Thu Feb 24, 2005 3:05 pm

Actually... there's another way to scale airflow voltages....

Instead of sampling with an ADC and repeating out through a DAC... another possibility is feeding the airflow signal into a PGA and then having the trim tables be gain levels below unity to set the PGA to - the output of the PGA would then be the scaled signal.

Just wanted to throw that out there. I may try both methods to see which seems to work smoother and have better end resolution for trimming, but its definately a possibility. And uses less resources in the chip, leaving more room for future expansion and capabilities.

vrg3 · Post by **vrg3** » Thu Feb 24, 2005 4:13 pm

But you don't want to scale the airflow signal linearly. You'd want to push it through the transfer function, scale it, and then push it backwards through the transfer function. Right?

azn2nr · Post by **azn2nr** » Thu Feb 24, 2005 10:21 pm

if there are harnesses that plug into other cars ecu's doesnt that open up more possibilites on the standalone market for us???

vrg3 · Post by **vrg3** » Thu Feb 24, 2005 10:27 pm

The other cars' pinouts aren't similar to ours.

THAWA · Post by **THAWA** » Fri Feb 25, 2005 5:55 am

Well no not stock. But couldn't you change the pins to which ones the new ecu uses. The problem I see with this though is what kind of sensors the new ecu needs to use and if the ecu has control over other things like VVT or whatever.

free5ty1e · Post by **free5ty1e** » Sat Feb 26, 2005 2:06 pm

V - I didn't realize we wanted the MAF signal to be processed; either method I described simply pulls a trim from a table and changes the read MAF signal by that amount and ouputs it. There's no transfer function... the PGA method would more allow scaling, there are many gain levels below unity, and it would be a continuous time function. Hmm. Either way, haven't had a chance to test it out yet. Damn paid work.

The ADC/DAC method would be the best if the table trims were to be added or subtracted from the MAF reading. The PGA method would be the best if the table trims were to multiply/scale the MAF signal. How did you envision this functionality? Perhaps I need to rethink it.

vrg3 · Post by **vrg3** » Sat Feb 26, 2005 7:08 pm

Well, it's your call... But I think making it a table of multiplicative airflow correction factors is the way to go. Then you can go like, "well, my new injectors are a third bigger, so at low loads I can just make all the values -25% and I'll get about the right amount of fuel." Or, "my air/fuel ratio is 12.5 but I want 10.0, so I'll set it to +25%." Or something like that. My arithmetic might be off but you get the idea.

A multiplicative correction applied to the voltage rather than the airflow wouldn't achieve this. Nor would an additive correction applied to either voltage or airflow.

Does that make sense? Do I have that right?

It doesn't really matter whether the time function is continuous; the ECU still samples the voltage relatively rarely. So performing the mangling of the signal, say, five hundred times a second ought to be just fine.

NICO · Post by **NICO** » Sat Feb 26, 2005 10:45 pm

can i ask you guys one qwestion what stand alone are you guys trying to use

stipro · Post by **stipro** » Sun Feb 27, 2005 1:40 am

NICO I WRX U wrote:can i ask you guys one qwestion what stand alone are you guys trying to use

free5ty1e is building his own piggy-back for our cars, and trying to sort out the input/output tables.

THAWA · Post by **THAWA** » Sun Feb 27, 2005 1:48 am

Noone's using a standalone. free5tyle is making a piggyback.

Vikash, I asked you before about what would be needed to make a Hitachi ECU understand a JECS MAFS. You mentioned the transfer function table. If I understand that right it's just a table where x voltage means y airflow. Since we're intercepting and modifing the signal anyway, do you think it would it be possible to take the signal from the JECS MAFS and modify it to match what the stock Hitachi MAFS would've sent to the ecu?

vrg3 · Post by **vrg3** » Sun Feb 27, 2005 2:26 am

It would be possible, but I don't see what the point would be. You'd be throwing away a third of the sensor's range anyhow, wouldn't you?

NICO · Post by **NICO** » Sun Feb 27, 2005 9:12 pm

holy shit thank god free5ty1e got a turbo legacy car, man if you build one i will buy one, good luck free5ty1e on building one i wish i could help but i dont no shit about that

u no what if you need a ecu and clips that clip in the ecu i can help you out.

THAWA · Post by **THAWA** » Tue Mar 01, 2005 1:10 am

Hmm, I'm not sure if I understand. If we're going to be modify the signal from the MAFS it's still going to top out at 200 g/s right? Sure you're taking other variables into account to come up with a guesstimate at what is reall flowing. But if you used a sensor with a higher range you wouldn't need to guess. Though then I suppose you wouldn't really need this afc anyway. Hmmm

anyway my harness came from z1auto:

That's my old stock ECU

vrg3 · Post by **vrg3** » Tue Mar 01, 2005 1:21 am

So your suggestion is to make a piggyback that uses a big MAF sensor and uses the airflow signal as an actual input? Sure, that could be done. Then index the tables by airflow per revolution and by engine speed. You just have to make sure your MAF sensor reads right and handle whatever reversion issues there may be. But those problems can be solved. And you would get a little more precision at high airflows.

It just seems cooler and cheaper to use the sensors that are already there.

Cooool, is the construction of the harness good? It looks nice. Much more uniform than mine.

THAWA · Post by **THAWA** » Tue Mar 01, 2005 4:52 am

Yeah it looks solid. I'll try to take a close up tomorrow or something.

legacycentral bbs

A forum for 89-94 BC-BF(BJ) Legacy Owners and Fans

T-leg Piggyback ECU R&D