NewKleer

thanks for that. looks like what you would normally change the K value by is what you can apply via consult. but this bit im not sure if you could do anything about: that "blas-off" is what i was referring to as "opening time" - i guess it depends on how different they are as to whether its significant (and what the value is used for). perhaps 370->440 wouldnt be too bad, but issues with higher (eg 550)?

with upgrading injectors, could you get away with, say going from 370cc->440cc injectors with just a leaning out of whole rev range via consult? eg set fuel injection time to 84% (370/440). or the "opening time" of injectors may play a bit of a role here? but if the injectors behaviour is linear across rev range, then the linear change to fuel injection time via consult might be able to compensate for higher flow rate injectors? you probably wouldnt want to make too big a jump, but interesting nonetheless by the way, i got myself a consult capable sr20det pulsar over the weekend, which will come handy

group buy will finish at 8, so thats one left at the moment.

id be interested in one im pretty sure (well i want an afr meter, just not sure which yet) edit: depends on price....the pre-assembled ones seem to approach the cost of a lm-1 unit after a quick check of the order page

thats them, thanks! jaycar categories are so rooted, hard to find anything

the ecu wont learn around the open loop stuff (ie, power region) so thats not a problem, but u probably do want it to learn around the closed loop region (idle, cruising). the "learning" aspect of the ECU is limited to the closed loop operationg only, so theres not much chance that your settings will be learned around by ECU (except idle/cruise where u want to be at 14.7:1). so in the end, i guess its not much of an issue after all.

any idea on price? also anyone got a cheap source for a exhaust clamp/venturi pipe?

doubt id ever release source code, just cause theres a chance that if i do a good enough job, someone could just make (and sell) their own program from it. not ready to pass on actual executable code yet either. i have a sr20 ecu ive used for testing, but the active test things, while they should work in theory, wont be able to test unless i have a car (although my sister has a pulsar which i can use) doing it in .net, and doing it such that a lot of code between pocketpc version and desktop version would be common, hence similar things would be available to both (just a matter of working out how to structure it on a pda screen given how small it is)

a simple thing like that would just be leaning/richening out over whole rev range (and likewise for ign timing) so u would have to limit yourself to an adjustment that wont stuff up any part of the rev range (ie, if ur car is leaning out top end, and you use that top end in driving, then u cant really do much). however if u have a solid 10/11:1 over top end of rev range all the way, and can live with leaned mixtures bottom end it might be ok. caz said 18rwkw from it, so presumably the leaning out lower/very top end wasnt an issue, but essentially depends on the AFR graph and how varying it is. eg going from 12:1 to 10:1 from 5k->6k means if u want to lean out the 6k to 12, ur 5k will be upwards of 13-14. obviously wouldnt be much use if your running a dfa (though u could zero ur dfa for purposes of testing it out) active tests is next on my software after fixing the gauges a bit (might leave the gauges looking a bit dodgy for time being so i can play with the active tests). i dont have a consult capable car yet so cant really test anything i do out too well till i do. do u have nissan datascan? if so u can have a play with the active tests yourself

no changes (or active tests) are permanent, thus the software would apply them each time the vehicle is turned on. theres no way to get permanent changes in a plug and play manner (pt 5.) and as such any software or device would have to reapply the settings on startup. one thing that may cause a problem with a change to fuel injection width (ie what ecuxtend does atm, likewise running active tests with any consult software) is that the ecu will have to (attempt to) learn around the change (it might not be successful, depends how far they can learn around things) to get back to stoich. guess this isnt much different to a AFM interceptor, but that can be mapped so that it doesnt change idle/low loads.

.2 of a volt is about 10% of your average usable AFM range (eg load points 30->90 = 2.5v range) so that could have something to do with it. and maybe since you dont seem to think its much of a problem, perhaps you were a bit too lenient with other calibration steps? maybe retry it all again?

all the calibration turned out fine as per the manual and so on? tried setting a load point to 0 and make sure that in = out?

did u do it in course mode or fine mode? is "untuned" meaning the -4 or -6 u did urself just for fun? or all zeroed? definitely something stuffed up there. if u did it in coarse mode and -4 to -6 without a dyno, im not surprised if u damaged some part of the engine, that is a pretty major change (30%?)

last time i was at jaycar i saw these: http://www.jaycar.com.au/productView.asp?ID=ZD0480 http://www.jaycar.com.au/productView.asp?ID=ZD0482 im pretty sure they had replacement globes for other things (like the ones ur trying to make in this thread) so might be worth to check out before trying the diy way

consult supports active tests where you can adjust fuel injection time (a percentage of normal, eg 95%, 110% across entire rev range) and static ignition timing (essentially same as changing cas, but no effort required ), which should be supported on most ecu's that support consult. check out nissan datascan webpage (screenshots) as it lists most of the common active tests there (http://home.curl.aunz.net/tpkolo/DataScan.htm). i think maybe scantech nissan (or some version of it?) can do active tests also, but not many other free software does. after i get gauges done, active tests would be next on the list for my software. as part of my last year uni project next year im aiming to make a monitoring/tuning device that will let you map out adjustments to fuel injection time as a function of some inputs (not sure whats best, combination of RPM and AFM voltage i guess?) and ign timing if i have the time. essentially a plug and play AFC, but with the advantage of not messing up your ign timing when u change fuel injection time. i assume this is similar to what caz is referring to by "in blocks". id probably write similar stuff into my desktop/pocketpc app first in order to test before doing it on a hardware/lcd etc level. however theres a few problems - serial comms to the ecu could potentially be interuppted, which could mean you get "stuck" with a certain value (eg leaning out a rich top end) which is then applied to whole rev range. likewise for timing adjustments. theory behind 100% safe tuning to deal with that would then be another page or so, and i need to get back to work but certainly, exciting to think of the possibilities

i think if it fully fails (eg power wise) then it will pass input off as output (ie, a short circuit between in and out? i remember with it switched off there seemed to be a short circuit between them. if u are a bit hesistant, then wire up a switch (which is what i did) to get it to bypass the DFA.

ferni - no it doesnt. google fclogit (not sure on spelling). not sure whether u need a special interface for it or not. sorry caz, my comment was based on the developers decision to sell it for that amount, not you as a reseller. "workshop quality, highly developed" is just marketing spin. if id decided to make interfaces for the purposes of assembling them to sell, rather than the diy kit for 50, then i would have gone smd and could add "workshop quality" and "highly developed" in blinking lights, but that was never my intention, which was from the start to make an easy to assemble interface to make it easy for those who couldnt justify the cost or afford a pre-made one. of course it will look "home made", as most people will make it at home! the only difference is that it had to be made with normal sized, thru-hole components otherwise it wouldnt be too easy to solder if it was all smd. (btw you got the numbers wrong. 70 inc postage vs 105 inc postage, but like you say they fit different purposes). i have no problems giving credit where its due - im quite pleased to see your group buy for the usb ones at a good price, theres no point in getting one of mine assembled if you have a usb only laptop now you have your group buy, and ive said as much in my assembled interface thread. do the same for the serial ones, and ill be able to spend all my time on the software (unless you want to convince the developers theyd make more money if they priced it at a level more people would be willing to pay) anyway i think theres been enough crapping on this thread Peter

having a bit of a play around tonight, and have a reasonably primitive gauge happening (this one was for battery voltage) still a fair way off (been working on pc version first as its easier) but getting there.

only its price makes it not for the casual/non serious user - its using the same consult functions that all other consult programs do/can sure it might be worth $500 in it's usefulness, but i guess thats the distinction between doing something with the focus on making money, vs the focus on helping others out

what we're "wishing for" isnt something with a $600 price tag if ecuextend wasnt so expensive then i'd have no need to even attempt to write my own - it looks like a great piece of software. but the bottom line is its priced such that its out of reach of casual or less than serious users. little bit of competition never hurt anyone!

excellent to hear! seems i underestimated everyone here, as i guessed that maybe 1/3 to 1/2 of the people who get a kit would have some issue with construction they wouldnt be able to figure out, but overall its gone quite well

bluetooth isnt something i know much about, and would at the least require some kind of extra hardware or something. bluetooth would be useful, but low priority atm. if pocketpc's can have an emulated com port over bluetooth (so no need to change the program) then it would just need some serial->bluetooth converter (if one exists) attached to the consult interface

this is the null modem thing i got: http://www.expansys.com.au/product.asp?code=3011213 essentially just switches over the send/receive data lines (pins 2/3) i think. what model ipaq do u have?

most of them have some method of connecting up - ive got a hp jornada 568 (or something) and all it takes for me is to put a null modem adaptor on the serial sync cable (and this appears as com1) google it (typically u will find gps sites that tell u how to hook up a gps serial unit to ur pda - however they hook it up is how you would) and u should be able to find out. i think a lot of ipaqs are the same, and it may just be a null modem adaptor on the serial sync cable (which u may have to buy if u dont have one). as for pricing etc, i dont really know, thats something to work out later. obviously on one side it would take a lot of time to do so, so id hope to get some money from it, but on the other hand i know everyone are tightasses (despite owning skylines....). maybe donation or something may be the way to go, so that those who'd appreciate the time/effort that itd take to do can send along some encouragement

sent a couple more out today, only a few left now. note that after 33., there may be up to 2 more, but then thats it - ive got no plans at this stage of extending the group buy any further.