T-leg Piggyback ECU R&D

[free5ty1e]

sounds like a good feature to add. Moreover, sounds completely do-able without too much additional trouble.

Please anyone with more information feel free to add to the antilag functionality. I likes it so far.

I envision just one of the buttons being an anti-lag toggle button. While it's on, when the car idles it pulls the timing properly for the antilag functionality. It'll keep doing that at idle until ya toggle it off. Sounds simple enough. It can easily be an adjustable effect, as well.

[azn2nr]

im wondering, how big of injectors will this be able to control if it scales back maf signal. id imagine a large say 3inch housing will help a little but since yellows aren exactly easy to come by i was planing on using modded n/a injectors which are suposed to be something around 740cc?

[THAWA]

Yes two is fine. I only have Na ECU's they have all the pins on the ecu but apparently they're missing pins in the wiring harness.

I was under the impression that anti lag was basically the extra air produced by a turbo that was stored somewhere, and released back into the intake when you start to accellerate again so you dont have turbo lag. I've heard it's hell on turbos and engines.

[azn2nr]

my knowlege of it is verry minimal adn i had only heard of it in a few discussions. after reading this ( http://www.rallycars.com/Cars/bangbang.html ) i think it may be a little more complicated to run.

your right in the fact that it is hell on turbos. it is also hell on manifolds espcialy stock ones. but it isnt stored air that is reintroduced as you can see from reading the website above

[THAWA]

interesting

[vrg3]

Turbo and NA ECUs have the same connectors. Turbo harnesses do have more pins in them though. I've made a few ECU extension harnesses and in each case had to use more than one NA cars' worth of wiring.

If you think you can handle reading the crank angle sensor all inside the PSoC that'd be great. Remember it's really important for the input impedance to be very high; these are very high-impedance signals and the ECU's already loading them a lot.

You could try advancing spark just from the timing you get from the ignition or tach pulses, but I think it'd be weird trying to predict when the spark would happen based on what happened 180 degrees ago. But maybe it'd be good enough. Maybe you could try to estimate crankshaft acceleration over the last three samples or something.

Yes, Jason, your description is a pretty good explanation of this particular type of antilag. It doesn't necessarily have to result in combustion in the exhaust, though; it could just be timing that's so retarded that it prevents the engine from increasing speed above idle but still allows heat to spool the turbo. Either way, you do need more than just timing control; at the least you need a means of blocking the throttle plate open when at idle.

If you have the money to replace all the parts that anti-lag is gonna break, though, you probably have the money for direct (rather than piggyback) engine control.

We won't know for sure how big an injector this setup will be able to handle until we can test it out. I'd guess 740cc/min is probably close to the upper limit though.

[azn2nr]

well if you could rev the motor and push the antilag button to engage it it could possibly work but your right in saying that it would break things easily. i think if you go garret this would be easier just because they are easy/cheaper rebuilds.

also if i rember corectly rsstiboy had a antilag system in his car. maybe we could get him to chime in.

[vrg3]

What you're describing there sounds a lot like launch control. Launch control could easily be put in. Just randomly cut out spark on one cylinder at a time, often enough to keep engine speed below the preset level.

The MSD DIS-2 has this feature. I tried it once. I think I set it to 4000 RPM, and found that I could build about 5 psi in the manifold when out of gear.

[azn2nr]

now was that with a stick or a auto. i have an auto which is why id want something like that. maybe you could incorporate a launch control and a antilag together???

[vrg3]

Mine's a stick.

With an auto, why don't you just brake torque it? You can make boost without moving the car just by pressing both pedals.

[azn2nr]

actualy when i break torque my gauge reads lower than normal vac when it stalls out. that and even if i rev it up to red in neutral it still doesnt build boost.

i wonder when this gets stikkied

[vrg3]

Really? You stand there with your foot firmly on the brake pedal and then you rev the motor, and it stalls out? That's not how every automatic I've ever driven worked.

Even revving to redline in neutral still involves an almost fully closed throttle, so you can't expect to build boost that way. If you did open the throttle up to bounce off the rev limiter you might be able to get some boost, but your engine wouldn't like it.

I don't think there's a reason to sticky it until there's something important to announce.

[evolutionmovement]

The stored air antilag is indeed a real system. There was a thread on this a while ago. I had come up with this very idea and read a year later or so that Subaru and Ford had used a similar system, but the FIA quickly banned it. Basically a compressed air tank would store excess boost and reintroduce it when necessary to reduce lag. The traditional system of burning fuel in the exhaust is horrible, and could not be used with cats. Hate to think what it would do to o2's also. Rally cars can use it since they replace everything all the time anyway. On a street car, you'd be wrecking turbos and manifolds constantly.

Steve

[free5ty1e]

lol... yes it looks like antilag is definately possible with some mechanical additions to the turbo's air passageways, but after reading that article and thinking about it its probably not the best of ideas for us. Launch control sounds like its possible too, we could go into detail on that if its a feature that generates enough interest...

Besides, even with the 16g on our cars... do we really have that much of a lag problem? Go get a bus turbo and maybe we've got a need...

Oh and - as long as the N/A ECU connectors/harness are physically the same shape as the turbo (connectors, not charger), I will be able to work with it. I'll have to be getting in there and messing with the connections and pins anyway so I'm not worried about that.

So anyway, on that website that explains antilag systems, they say something about an EGR-based antilag system but go into no further detail. I know we don't have an EGR system but I'm still interested if anyone knows more about this one to find out more. The boost storage tank is a good idea too...

[azn2nr]

if the na ecus harness is physicaly similar enought to use, maybe if the people that want plug and play can find their own at junkyards and butcher the ones there and send them to you for modification. this would make the option much more easily accessable.

[evolutionmovement]

http://www.bbs.legacycentral.org/viewto ... ht=antilag

Apparently, nobody was interested as I got no replies .

I think I remember discussing it with someone at some point (ciper and Vikash maybe?), but search didn't turn up another thread.

Steve

[vrg3]

Yeah, we talked about it in person, I think. The same discussion as the air-conditioner intercooler.

I think the "EGR-based" anti-lag system basically uses a repurposed EGR valve to dump pressurized air from the compressor outlet directly into the exhaust. Kind of like a blowoff valve that discharges into the exhaust instead of the intake or atmosphere. Then the ECU just needs to make sure enough unburnt fuel ends up in the exhaust to ignite and keep the turbine spinning.

[Matsumoto]

I'm not sure if anybody's seen Initial D here, but they actually talk about this in the series and an EVO has something similar equipped. They refer to it as mis-firing system though. Not sure but it sounds the same as ALS.

Just thought that was kinda cool.

[azn2nr]

yep. its like a built in launch control. the thread where i picked up the link above was on a guy with an evo that was trying to figure his out

edit. i was just looking back at the charts on page 9 or something and saw that your adjustments would be at every 500 rpm's and that they would extend up to 8000 rpm's. i was thinkin, what if you used up the last colums and instead made the adjustments every 250 rpms instead. this would make its tuning capeabilties alomst identical to most piggybacks on the market. a highest resoultion posible would be best though because even with 500 or 20 rpm increments you may be able to feel the load changes thorughout th erev band.

ok just so were clear on exactly whats been devloped. will there be control over fuel and timing? what about boost?

earlier there was a mention of the back to stock feature. i think that would be reall cool except that it should be optional as in a lock button on the controler that you just push to turn the controler on or off.

i was also reading in the em tuning that (at least for the wrx's) the stock ecu does a better job at less than 60% throtle and that in piggybacks like the utec (see here:http://www.turboxs.com/stiutec.shtml) the tuning only activates above that line. though i dont know why you would want that its just something that i noticed.

[free5ty1e]

Well the charts on page 9 are quite outdated. This thread would be many pages longer if I were to keep my code updates here. Here's the RPM range as of now, and yes it would be a big deal to have tables larger than 16x16 as the size in SRAM required increases exponentially:

Code: Select all

// 50, 90,130,170,210,250,290,330,370,410,450,490,530,570,610,650<--------RPMx10, stepsize=400, range=500-6500

This product will control airflow and timing, and if you'd like it can control boost via the stock boost solenoid.

[azn2nr]

isnt sram fairly inexpensive now?? or am i thinking of something else.

when you say ariflow you mean the maf sensor signal to the ecu right?

also is there a minium or how small is the minimum amount of signal the ecu needs to see to even operate the injectors??

[free5ty1e]

SRAM itself is yes fairly inexpensive right now but in order to increase the amount I have available to me I need to add a memory chip as well as interface to it, which eats up valuable i/o which can be used for much better purposes...

This is the most capable chip in the family of microcontrollers that I use, it has the maximum amount of sram and flash memory they can squeeze into the silicon with the current technology. 2K of SRAM and 32K of flash is nothing to sneeze at in the embedded world.

a 16x16 table of single-byte values (which is how the trim data will be stored) occupies 256 bytes of sram. For contrast, the lookup tabkes vary from one byte values to four-byte values (for floating-point values I'd rather lookup than calculate) and a 16x16 table of floating-point lookup values takes 1024 bytes of flash memory! If I were to attempt to go up to the next sensible table size for higher tuning resolution, 32x32, one trim table would then occupy 1024 bytes of SRAM (half of it!) and thats if we stuck with 8-bit values for each trim! Believe me, I've considered the alternatives, and 256 bytes is the largest SRAM table I want to deal with for many reasons, among which are that SRAM is only accessible in 256 byte pages at a time. The flash tables can be considerably larger but any tables we have to adjust should be limited to 16x16.

Yes airflow sensor signal = MAF sensor signal. No idea what the voltage threshold is for the ECU to see enough air to open the injectors. I'd imagine it's always injecting something as it knows it must be idling if the car's started....right?

[azn2nr]

i guess we'll see when you test it out. man im sooo excited

[THAWA]

Look at this!



Does that look familiar to anyone? That particular ecu is one made by APEX-i for the FD RX-7. There is also a pl ug in connector that has both sides that is sold. So what does that mean? If APEX-i can still get the connectors someone has to still be selling them. More googling revealed that there are many more cars that used the same connector as ours. There are also other cars that used three of the four connectors and two of the four connectors aswell. I believe a few years of the supra did along with the celica and many others. There has to be someone still making these connectors. We just need to find them.

here are pics of the ecu pinouts from toyotas: http://translate.google.com/translate?h ... f%26sa%3DG

it's not the actual ecu but it shows a drawing of what the connectors look like and how many pins there are.

[vrg3]

Yeah, my brother used BC/BF harness connectors for his JDM 1JZ-GTE's ECU. I didn't know any US-spec cars other than ours used them though!

The connectors are made by AMP, owned by Tyco Electronics. If we could get AMP part numbers we should be able to purchase the stuff from an AMP dealer, as long as we buy in sufficient quantity. I remember looking long ago through their online catalog trying to find them and having no luck though.

]Yes airflow sensor signal = MAF sensor signal. No idea what the voltage threshold is for the ECU to see enough air to open the injectors. I'd imagine it's always injecting something as it knows it must be idling if the car's started....right?

I think we'll have to determine this empirically. The ECU does cut fuel on extremely low load, but not at very low engine speed. And it might also require the throttle plate to be closed.