FAST wb ecu onto a '95 240sx w/ custom turbo

TurboTR

sprayhead
Joined
May 25, 2001
I have a customer with a '95 240sx that we put a FAST batch wb O2 ecu onto last weekend. Engine is a 2.4L DOHC 4 banger straight out of the boneyard for a couple hundred bucks. They put a turbo on it with a custom exh manifold w/ Tial wg, some intake manifold mods and a custom front mount, but otherwise the engine is 100% bone stock. 84 lb inj and a simple low-end Aeromotive pump setup. We started the car for the 1st time Sat afternoon, he drove it to/from the track and raced it all day Sun without any misfiring or problems. Gotta love the FAST wb ecu :)

So it goes 118 mph already on its 2nd pass Sun at 17.5 psi boost, BUT it only put boost down for about 1/4-1/5th of the pass due to a mismatched turbo. It gained 32 mph from the 1/8th to the 1/4 markers once boost finally came up though ;-) Once it gets a better match and the boost cranked up I look for the car to run well over 130 mph in the 1/4 on turbo alone. Remember, this is a boneyard engine that hasn't even had the valvecover removed. The spray is installed and waiting after that :) The car is alot of fun, easy to work on and relatively cheap to do. May have to build one myself someday :) It's already being challenged Fri night by a high 10 sec/mid 120's turbo Miata for the title of fastest local import (other than all out, 150+ mph Supra of course :) We're gonna paddle 'em ;-)

I bought a 4 channel Tek TDS224 DSO so was able to scope the Nissan position and ignition signals and all, and after minor rework we were able to set it all up with the stock distributor with built in ignition and position sensing systems. No add-on crank trigger or ignition system required. It did require a different interrupter wheel from an earlier model 240 to get the ref angle and rotor position relationship exactly right though, but it dropped right in. We used the generic harness supplied and removed the stock ecu and some of the factory harness. Hooked into the stock TPS, replaced the Nissan air and coolant temp sensors with GM and added the 3 bar MAP. Haven't verified yet, but the Nissan temp sensors may be close enough to the GM that it may have been unnecessary to replace 'em. Anyway, we also hooked into and used the existing stock low speed fan relay system to run the aftermarket electric fan, and are using the other high speed fan relay system to run the nitrous solenoid. Found both trigger lines in the p.s. kick panel so easy hookup to the FAST ecu. Cool!

I hope to specialize in custom installs. That scope has to get paid for somehow ;-)

TurboTR
 
I guess I would call that "turning chicken $#!+ into chicken soup." Sounds like you did a good job with it Todd!

Craig
 
It went well, thanks to support from FAST, PT&E and a little luck :)

I think the import market potential is already big and growing. At the big IDRC race this past weekend at River City RW it was hard to find a car without a turbo, and often juice too. Many were young, but all were pretty friendly IME and the focus was on high output turbocharged racing, like it oughta be. Alot of these guys are struggling with Haltech or TEC2 tuning. What's exciting is that it's still in the early stages IMO and doesn't yet require $75k invested to run with the big boys, at least in the lower classes. I think we could build say a 240sx for $20K or less and dominate the street class for example. Anyway, I'd like to help jump start some of that potential business by doing custom installs, then use the info gained to maybe help make a plug 'n play harness for the more popular apps. We'll see how it goes I guess.

TurboTR
 
TONs of potential out there!

Go read the thread in the SupraForums about the upcoming AEM ECU. there are Miata, Mitsu, RX7, and other gorups DYING for that ECU! The reasoning is VERY simple - NOTHING exists for those cars right now except tons of piggyback crap things. Supra folks have to use a device to bypass their boost cuts which clamps the MAP sensor just shy of the boost cut. They then have to do other things to modify their injector pulses and still other things to dump their restrictive MAF sensors. Lastly they end up with electronic boost conrollers. Inthe end they've got a harness that looks like a pincushion (there's actually a market for an extension harness to support this crap) and a car that's nearly impossible to to tune. Which widget needs to be twiddled to do what?! None of it works together, what a mess! Some of the other imports aren't much better, they have o run alphabet soup devices too. Read the Greddy or HKS catalogs sometime for a list of potential customers. Mustang guys are stuck with chips and a programmable ECU plug-in board that darnnear requires HEX editing tables. I watched an LS1 Camaro dyno with a ton of electronic widgets on it (including a box to modify the MAF signal?!) that came up 20HP shorter than the other 10 LS1s :rolleyes: LS1-Edit isn't out the door yet and won't be as lexible - there's a market there too!

If someone were to make a plug-in harness for those cars people would beat a path to their door. The thread about the AEM drop-in ECU is 23pages long and their rep is promising everything from traction control to boost control to knock sensing. Their ONLY issue(s) right now is lack of support for domestic vehicles, crappy beta race only geared software on their WEB site, and a mid-January release date.

I've got a Mustang that badly could use a drop in box better than the older DFI I've got (bleah!). I've got an RX7 that I'm about to sell who's only support so far has been from PMS (think Crane Interceptor) and APEXi who won't release the software to properly tune their replacemen ECUs. Hondas are being turbo'ed all over the place and their ECUs weren't even designed to handle boost. Miatas have only been supported properly by one company's ECU named MiataLink thta has no wideband and no knock if I recall - their getting turbos and superchargers like mad too. Mitsu 3000GTs, Eagle Talons, Vipers, you name it - they all want ECUs. Did you know some of th eearly Vipers supposedly ran both a 4cyl and a 6cyl ECU? I know somone with one of these who is DYING for a way to reprogram the thing. I know someone with a later blown\nitroused Viper who has to have chips burned everytime he modifies the darned thing - there's money there :D

There's a market allright, it's all over the place. Trouble is no one has really recognized it until recently and the AEM ECU priced "under $2K" may be about to make like Godzilla through Tokyo in this market place. It's not out of the gate yet so there's still some time but trust me - there are people DYING to get their hands on something easy to program that will drop in place of their existing ECUs. Build it, you'll be swamped.

For that matter, do you have ANY idea how many 240SX turbo swaps are being done? Go find their mailing list, I pretty sure they've got one just for this, and mention a "drop in WB programmable ECU" - you will probably be swamped with inquiries!:eek:
 
I'm on 2 of their lists lately and basically it's been crickets chirping pretty much ;-) There does seem to be interest in tire pressures and such though ;-)

I did go read the AEM thread when you had posted about it. I'm afraid most Supra guys will want all the amenities to work too, like traction control, boost control, whatever, which the AEM seems to be promising. Maybe they have scammed access to the factory wb O2 control code, traction control algorithms, etc and are about to cut and paste them into their AEM ecu code, who knows :) I worked on reverse engineering a factory O2 heater control algorithm for example and it was over 40 pages by itself if I recall, where they were keeping track of airflow and load past and present, estimating temperatures past and present, monitoring and controlling heater element power, etc, etc :) If they can really get it all for under $2k then there's nothing to compare right now that Im aware of. I do have a Supra guy interested in going 150+ with the FAST ecu in central TX but "not right now" as the layoffs hit us very hard in this area. Soon I hope.

If anyone really wants a programmable, wb O2 ecu that works for reasonable bux it's here and we're willing to do what it takes to get it on there correctly :) However, getting people to see the light and trust enough to say "I'll take it" vs "That's cool" (or worse ;-) is a whole 'nother story IME...

TurboTR
 
WB control

I'm not sure that they're going for the factory code at all. It's a drop-in replacement. The traction control, at least on that car, is a seperate ECU if it was stock. I've not looked deeply enough into the beta demo code but apparently you can create triggers that will act upon the various programmable I\Os. Traction control would simply put in some sort of retard or cycle a dropped injection signal to lower power - perhaps boost too. Can the FAST system do a programmable boost very easily?

Frankly, their willingness to say "sure, can do!" for nearly every question with so much confidence makes me just a little nervous. Some of those requested features are probably going to require added sensors and some pretty decent programming to achieve but the response makes it sound as if it's trivial. I'm not quite convinced it's that easy yet :)

As for the market, I'm convinced it's there but I'll admit that may be a myopic point of view based upon MY desires :D However I've seen more than a couple of turbo Hondas in my area, more than a few turbo Miatas (at least a couple of which were running MiataLink ECUs), and TONS of blown Cobra and older 302 powered Mustangs. The Cobras are making sick power but having to have chips made is no fun and leaves little flexability. RX7 are having all sorts of "fun" too since without a MAF they just run leaner as you mod them until detonation trashes the motor - those guys have just now heard about the AEM promises. The RX7 APEXi EC has some issues, one of which is the manufacturer's attitude, and the PMS is being primarily distributed by a guy that few people in that community respect (including me). Surely that community would like to find something better! I'm actually to the point with my RX7 where I'm willing to sell it rather than go further and risk the engine. Good rotary mechanics are just too hard to find and the ECU support isn't where I'd like it to be even with the APEXi. If th eGN were newer and smaller I'd go for one but the Supra looks to be my new target. I still need an ECU for my SBC (in a 72 240Z) though and I'm going to have a hard time deciding this Winter.

I guess we'll see how the AEM box shakes out and if it will live up to it's promises. My fingers are crossed, it's feature list would certainly seem to give the Motec units a run for their money on the lower end. Is FAST going to be releasing an upgraded ECU anytime soon? Anything on the drawing board o consider waiting past January for?

Can you give more detail on the 240SX? I know some folks are going to be interested in it ;) Running an R200 rear in it? If so how's it holding up? Considered an R230 by any chance? What trans? Strictly race or is it going to be streetable for more than just to and from the track? If emisison weren't so nasty in my area that's one car I'd certainly consider playing with!;)
 
The 240 is a fast street/strip car, not race only. Same as my own turbo Buick basically, just a different brand.

Deciding on the ecu is not that tough really, once you decide you want the engine to run well and be easy to tune vs trying to find support or replacement for all the "alphabet soup" gadgets that are popular with the import crowd :) If you want solid boost control for example, a great, low cost solution is a regulator across the wg actuator diaphram in "delta" mode. Works for me.

TurboTR
 
Originally posted by TurboTR
The 240 is a fast street/strip car, not race only. Same as my own turbo Buick basically, just a different brand.

Deciding on the ecu is not that tough really, once you decide you want the engine to run well and be easy to tune vs trying to find support or replacement for all the "alphabet soup" gadgets that are popular with the import crowd :) If you want solid boost control for example, a great, low cost solution is a regulator across the wg actuator diaphram in "delta" mode. Works for me.

TurboTR

What is this "delta mode" that you speak of? I really like what the AEM computer has to offer, but they stuck themselves in the rear by not having WB 02 sensor capability. If they were WB friendly, I'd snatch on up quick cause they can control the trans. I guess I wil have to run a seperate harness to run the trans with the FAST. ERIC.
 
This is definetely something you should document! Write everything down in steps and post it! There will be others out there who would like to try the same thing.
 
Originally posted by Taffy
What is this "delta mode" that you speak of? I really like what the AEM computer has to offer, but they stuck themselves in the rear by not having WB 02 sensor capability. If they were WB friendly, I'd snatch on up quick cause they can control the trans. I guess I wil have to run a seperate harness to run the trans with the FAST. ERIC.

AEM can use a WB02 easily, it just needs a external controller. You can get one for ~$150 soon from the Australian DIY WB02. They are building them now.
 
I wouldn't get so excited about the AEM EMS' just yet. Right now the only available EMS (as of 02/18) were the Honda/Acura Plug & Plays. They have alot of promising features but I recently saw one installed on a stock 160 HP GSR converted Civic and their engineer had alot of difficulty setting it up. After he got it running it made a max of 161 HP. In on a modified vehicle I can see that this system will be valuable especially for the import guys who really have nothing to go to now. There is going to be alot of teething problems with this system and until AEM gets everything sorted out, the auto "ethuasist" is not going to be able to use this system. I do this for a living and trouble understanding what they were trying to accomplish with their system. And so did their head engineer.
 
Delta mode basically means using both ports of a dual ported wg actuator. The regulator is connected across the diaphragm. Regulator input in parallel with the opening side, output to the closing side. Makes a closed system, as opposed to a simple bleeder type hookup.

I hear that some people are keeping stock ecu's in place to run their trans. I don't know the details of each trans control app, but if it were just a case of controlling on/off shift solenoids then either something probably exists that could run these, or else something could pretty easily be made that would map shift points vs speed and load and control the solenoids.

As for documenting everything, most of it is written down at least ;-) We're also working on an Eclipse box that will allow plugging in a FAST, and maybe a Miata box too. Would like to work on other potential apps too as time permits :)

TurboTR
 
Sound Performance in bloomingdale, il (country wide known supra tuner) has done a few installs of FAST systems and still using the stock computer

both on supras and on some turbo lexus converts

they have made piggyback harnesses onto the stock computer and it has seemed to work well

dont know why others aren't trying it out

i've also downloaded and looked at the AEM software and it isn't very user friendly
 
Which head engineer was that?

Originally posted by JrTuner
I wouldn't get so excited about the AEM EMS' just yet. Right now the only available EMS (as of 02/18) were the Honda/Acura Plug & Plays. They have alot of promising features but I recently saw one installed on a stock 160 HP GSR converted Civic and their engineer had alot of difficulty setting it up. After he got it running it made a max of 161 HP. In on a modified vehicle I can see that this system will be valuable especially for the import guys who really have nothing to go to now. There is going to be alot of teething problems with this system and until AEM gets everything sorted out, the auto "ethuasist" is not going to be able to use this system. I do this for a living and trouble understanding what they were trying to accomplish with their system. And so did their head engineer.

Jason seems to be pretty knowledgable about the system. He's not given any indication that he doesn't understand it at all and I believe he's one of th ehead guys on it.

The AEM is attractive because of the great number of programmable I/Os. It has more than enough for most apps and unlike some others you don't lose things when going from bank\batch to say sequential. The box can also use a wide variety of sensors so you CAN use a Toyota MAF if you wanted to. While it looks like some other boxes allow you to modify some sensor ranges it looks like the AEM box opens these tables up completely. Want to run the oil metering pump on a rotary? Or control the guts of an electronic trans? This box has the I/Os that can do it. If the hype is to be believed we'll be able to program actions based upon software triggers on various I/Os. EGTs get too high and you want to cut boost? Should be able to do it! Lot's of potential there. Did you guys read Jason's post on the I/Os? I'd post it but the server is overloaded over there and I cannot get to it :rolleyes:

Anyway, this ECU is worth watching. It WILL accept WB O2s, it will do datalogging, it will allow all sorts of neat tricks with the I/Os. The software IS less than friendly, at least the beta stuff on their site is. The user manual had better be something special! We'll see - personally I want it because it's PnP for my Supra. I MIGHT use it on my V8 if I find that it's easy enough to tune. The price is right at about $1350 so I'll just have to see. My fingers are crossed - shoudl be out any day now supposedly...
 
One thing as of now AEM's EMS can't control low impedence injectors. I am not disagreeing that it does have a lot of good features but it is NOT proven yet. They really shouldn't have released it yet. There is alot more to a programable EMS then just saying it will do all of these things. Just ask Lance or Craig.
 
>Traction control would simply put in some sort of retard or cycle a dropped injection signal to lower power - perhaps boost too. Can the FAST system do a programmable boost very easily?

Just a few things I'd like to add BLKMGC. I agree the result of say, traction control is just retard the timing momentarily or whatever. BUT the algorithm to run this system, successfully, is more than just one or two lines of code. I would bet, safely :) that the factories have *alot* of time involved in getting these systems to work well :) For example- when would the system activate? How "strong" would it activate, and for how long? What would interrupt the process? From what sensors would the system gather wheel spin data? What interrupt handler would take care of measuring the time elapsed between input pulses from the wheelspin sensor(s)? How much space is available for the code? How much cpu time does the algorithm take up, and are we already near the limit of time available at say 7500 rpm?

See what I'm getting at here? The concept is simple, "just decrease the torque when the wheels are spinning". But the devil is in the details, and implementing them takes alot of development, code space and valuable cpu execution time. Again I refer to that factory O2 heater control code that I spent almost a year of my EE career in reverse engineering. This was 40 pages of code by itself, and the comments were in Japanese if you get my drift :) So I have to wonder, how exactly will AEM manage to deliver on everything they are promising? If a late T*y*** ecu has the knock sensor interface completely integrated into the ecu how will AEM deliver factory knock performance in that case, for another example? The code I saw was probably 50 pages in that case... Plus transmission control. Again, this was a for a late model ('00+) app, but the trans control became so large that we had to specify *dual* Mot 555 cpu's to handle the trans + engine control code.

They are doing a good job so far IMO and people seem lined up to buy already, but I just don't see how they can deliver all the promised stuff, at the cost promised too. I mean, deliver knock algorithms for example (not to mention the correct hardware interface!!) that work *properly* with a given factory knock sensor(s), etc... Plus don't forget, the wb interface is not integrated. Which means their ecu just accepts an already processed, linear input signal in that case basically. Heck I think even the Haltech has had that capabaility for ages.

TurboTR
 
Low impedance? Got it!

Originally posted by JrTuner
One thing as of now AEM's EMS can't control low impedence injectors. I am not disagreeing that it does have a lot of good features but it is NOT proven yet. They really shouldn't have released it yet. There is alot more to a programable EMS then just saying it will do all of these things. Just ask Lance or Craig.

Actually it can do this. Apparently the Supra have some sort of resistor pack stock and this box uses it just like the stock ECU. Stock injectors on this app are apparently low impedance and the AEM box is supposed to use them just fine (shrug).

As for not being proven this is true however not releasing it certainly won't solve that! How else will it be proven? they have made bold claims, no doubt. However I will not tell them it cannot be done simply because someone else hasn't done it. I'm optomistic but cautious. I wish I could find someplace where the Honda guys are talking about it...
 
Whew!

Originally posted by TurboTR
[B<SNIP>
Just a few things I'd like to add BLKMGK. I agree the result of say, traction control is just retard the timing momentarily or whatever. BUT the algorithm to run this system, successfully, is more than just one or two lines of code. I would bet, safely :) that the factories have *alot* of time involved in getting these systems to work well :) For example- when would the system activate? How "strong" would it activate, and for how long? What would interrupt the process? From what sensors would the system gather wheel spin data? <SNIP>

See what I'm getting at here? The concept is simple, "just decrease the torque when the wheels are spinning". But the devil is in the details, and implementing them takes alot of development, code space and valuable cpu execution time. Again I refer to that factory O2 heater control code that I spent almost a year of my EE career in reverse engineering. This was 40 pages of code by itself, and the comments were in Japanese if you get my drift :) So I have to wonder, how exactly will AEM manage to deliver on everything they are promising? If a late T*y*** ecu has the knock sensor interface completely integrated into the ecu how will AEM deliver factory knock performance in that case, for another example? <SNIP>
They are doing a good job so far IMO and people seem lined up to buy already, but I just don't see how they can deliver all the promised stuff, at the cost promised too. I mean, deliver knock algorithms for example (not to mention the correct hardware interface!!) that work *properly* with a given factory knock sensor(s), etc... Plus don't forget, the wb interface is not integrated. Which means their ecu just accepts an already processed, linear input signal in that case basically. Heck I think even the Haltech has had that capabaility for ages.

TurboTR [/B]

Whew, too much to go point by point but I'll try to address some of this from MY point of view...

First - I understand that they have written some pretty big checks here. Further I also understand why factory OEMs spend millions on this sort of research - it's not easy! However I would also point out that no one expects perfect OEM drive quality with an aftermarket ECU. Coming close would be nice but no one should expect it. DFI, FAST, Haltech etc. etc. simply don't have the I/Os to do it and the amount of programming involved if they did would be insane IMO. What I think everyone DOES expect is a car that starts, idles, drives, and goes WOT reliably and well. If it doesn't turn on the EGR or vent the gas tank "right" I doubt many people complain. Folks like Ford put code into the EEC to take into account the distance from the O2 sensor to the exhaust port - what aftermarket ECU is going to do THAT?!

So, when it comes to knock sensing I honestly expect something that will save my engine - not allow me to run on the hairy edge all the time and be all that saves me from toasting a headgasket or worse. The Toyota ECU is supposedly pretty aggressive in this aspect, I don't expect the AEM to act that way. Hell, look at the old DFI Gen6 knock sensing (cringe).

Traction control... This is supposed to use the stock ABS wheel sensors. My car has 4channel ABS, later models don't. This is supposed to be enough to do this and there ARE aftermarket traction control systems (dedicated) that are supposed to work VERY well - to include driver adjustment and launch control. IF you read that thread carefully you'll note that Jason mentions this being an add-on at a later date. It would seem that AEM is going to possibly use an external box to cheat this function :) Frankly, the OEM "traction control" on the Supra is pretty bad - most of us disable it by pulling the fuse! :D If AEM can provide an improvement over this terrific, if not I'll use the aftermarket box and be happy if I simply must have this feature.

O2 heater control - good points. Note that by using an external controller they don't have to do this IN their ECU. I'll bet that it's not as sophisticated as some OEMs but if you read the Aussie DIY WB postings that designer claims that his hardware circuit is more sophisticated than some OEMs. Do I really believe this? Nope! However see my first point - I do NOT need OEM-like performance. I simply need a reliable enough measurement to keep me from blowing my engine. If I sacrifice say 20RWHP because the sensor isn't accurate enough to tune to the bleeding edge I will NOT cry about it. I'm willing to give up HP I'd have in a perfect world simply to have the flexability to be able to tune. The OEM ECU is a black box, read some of the threads on SupraForums where guys are adding all sorts of expensive CRAP to their ECU harnesses in order to trick the OEM ECU. Given a choice between THAT and the AEM the choice is pretty simple, no?

Having said all that I AM wary of the claims that Jason and AEM are making. For instance he made a statement about MAF sensors that made it sound as if they were linear devices like a MAP might be (I'm not sure with regards to the MAP). Anyone who has read a Pro-M data sheet or flow sheet on a MAF knows that they aren't linear. Jason has also made complicated things like Traction Control sound trivial - you know it's not as easy as that and so do I. However I'm willing to give them a chance. NO ONE ELSE is making a PnP ECU for this car (or any of the others they've targetted) unless you count ApexI's crap so what choice have I\we got? Splice in a Gen7 DFI? I'm trying to avoid cutting things up and lashing sensors together. Trust me, I'm not truly stupid but I AM enthusiastic about what this box could do for me. If FAST or DFI wants to step up to the plate and build something too then by all means let's have some competition! Until then I'm looking for the best tool to do this job. No one else save MOTEC is making an ECU with all of these I/Os and MOTEC costs too much. If the AEM is half of what's claimed then it will be ahead of it's competition and I'd be stupid to dismiss it...

P.S. Now if they could just get their software to be as friendly as the GEN7 or even the FAST stuff. Their demo software on the WEB is pretty nasty!:eek: Oh and no flames here guys. I understand where you're coming from and I do actually share some of your concerns....
 
Todd,

I think the custom ecu installs is a great idea. With the complication and expense or inability of OE ecu reprogramming and with the price of aftermarket ecu's starting to drop I think you definately have a winner.

Good luck and I hope it works out well for you.


Chris S
 
Top