nAst1: Progress and Concepts Thread

[robertisaar]

take a look at how i setup the low RPM BLM histogram, i made a " RPM too high trap cell" to filter out that exact problem. i also base my 2000RPM adjustments off of the low BLM table, which is why i never made a trap cell on the low side of the mid BLM histogram. probably not ideal, now that i think about it.

[pocket-rocket]

maybe you're running a stock BIN?

You mean Ben said he tuned my car when he really didn't?

I started thinking about (and boy, did it hurt ) it while enjoying the 100+ heat index with high humidity at work today... Why would I want to start with an already tuned chip for my car as a basis for tuning? I mean, Ben said my fueling and such was spot on. The only way he could have tuned my car even better was with a wideband and strapping it down to the rollers. So, I've got a backup of my Ben tune, he has a copy, I'm going to make another copy or two for safe keeping, and I'm thinking I'm going to dive face first into tuning using the BCFA port of nAst1

also: hero = guy with way too much time on his hands to do anything else productive?

But this is productive, and useful

EDIT: Lol, I'm retarded when I'm exhausted from work. Ben can attest to this if he remembers me timing my car after work once, lol. Let's just say if you skip the step where you rotate the crank one revolution in between locking banks of cams back in (all flats up or all down vs the intended 1 set up, one set down), instead of 6 individual cylinders running, you get a double 3 cylinder since two cylinders fire at once instead of alternating... at any rate, I mislabeled the chip dump of the stock chip. The .bin I have labeled as a stock chip is actually my modded chip...

On a side note, the facepalm smiley is one of the best recent additions to this site.

[pocket-rocket]

So I'm going about entering into the nAst1 tune the differences between my tuned chip and stock and I'm finding some differences. Some things I'm ignoring for now, like 009A and 009B in DF.

Very first table in DF:

Address
0000 AIR FLOW CALC - I'm not finding it in nAst1
0863 LEAN CRUISE = Highway fuel in nAst1, I assume.

There are more, but those are what stuck out initially.

Besides the other stuff that I knew would be different, I've gone through the BCFA port of nAst1 and set things to my liking. I've left the speed limiter alone at 120mph. I don't plan on hitting that any time soon. I've set the minimum temp to enable EGR at 304.25 F, lol, and I set a low rev limiter just to play around with it.

Wasn't there a file on the site somewhere that was like a tuning guide? The book I read a while back (Performance Fuel Injection Systems) talked all about different setups, ECMs, what you need to do to tune stock ECMs, ect., but I don't recall a section that said "Start at point A and tune your idle- this is how you do it...".

[SappySE107]

I think I started a few tuning guides over the years but that was about it. You have to drive around and then tune idle with the engine warm and the O2 hot. Then you need a wideband to tune the cold start up and warm up idle.

[bszopi]

This is probably the closest to what we have... http://60degreev6.com/content.php/15...g-OBD1-Systems

[pocket-rocket]

Hmm, I thought there was a PDF or a series of PDFs that Robert had written. Oh well, time to burn this puppy and see how the 92 reacts to it at idle with a barely off idle rev limiter for kicks. Even though this engine isn't staying in the car, I don't feel the need to rev to 7k on a cold start, lol.\

Edit: As a side note, this laptop couldn't run for it's life if I shot it out of a cannon going downhill

[robertisaar]

So I'm going about entering into the nAst1 tune the differences between my tuned chip and stock and I'm finding some differences. Some things I'm ignoring for now, like 009A and 009B in DF.
0000 AIR FLOW CALC - I'm not finding it in nAst1
0863 LEAN CRUISE = Highway fuel in nAst1, I assume.

9A and 9B are highway mode spark timers
0000 airflow calc: nothing, since 0000 is actually the MSB of the PROM ID, the "airflow calc" is actually a seperator between different sections of the calibration.
yes, lean cruise = highway fuel.

[robertisaar]

okay, decided to go through the thread and find all of the things i haven't done yet, but mentioned. trying to categorize them as to how simple they are to integrate, which combined with their popularity will determine when they are integrated. should no one be interested in a specific item, i may just strike it from the list entirely.




uber simple:

oil pressure tracking in the datastream. obviously you'll need an oil pressure sensor to use this, for +12V ones, a resistor may need to be connected in-series to allow for a more acceptable range, otherwise a relatively low pressure (40 or so) could peak the A/D converter in the ECM and you wouldn't be able to read above that. and since we may or may not all be using the same sensor, you'll have to come up with your own conversions to make an accurate PSI number. basically, you would watch a mechanical guage(or use one in your cluster if you have one and trust it) with the datastream running, keep track of the A/D counts at various pressures, then input that data into a table in the ADX and you'll have a custom oil pressure reading.

low oil pressure/high coolant temp fuel cut. obviously, this would need a pressure sensor to work for the oil pressure cut, but the coolant temp can be integrated right now if need be.

fuel pressure sensor. not really sure if anyone would need this unless they might want the SES light to fire off when the pressure drops below a certain range, maybe MAP compensated as well?

fuel composition sensor. expensive, i'll just leave it at that. would be nice though to automatically change the stoich and PE target AFRs though.

full time barometric sensor. extremely easy to deal with, just grab a MAP sensor and don't hook up to a vacuum source.

wideband in datastream. also easy to just show up in the stream, later on, i'm planning on allowing closed loop operation, with the option to use or disable the constant rich/lean driving normally used for cat operation should you just not care about the cat or emissions.

TCC PWM operation. i may be needing this soon as well for a 95 4T60E to make some butter smooth TCC lock operations. not sure if PWM unlock would be necessary.

customizable indicators. lean cruise, torque limiting, knock retard, cat overtemp, DFCO active or any other number of situations that you might want to know about, can choose which one you want to keep track of. assuming i have any outputs leftover from other stuff.

valet switch. spark starts getting cut above a certain amount of throttle to limit torque and discourage abuse.

newer style PRNDL decoding. also something i'll be needing soon for the 95 4T60E. i'm not sure if the 4L60E is setup the same as a 4T60E or 4T65, but most definitely allow 4T60E support.

shift points above 6375RPM. not sure how many people are running a -E trans and need shift points above that and have a trans that will stand up to it.

transaxle temp sensor. for both general tracking purposes, and for disabling enabling certain things, like TCC enable/disable, gear lockouts, maybe other stuff?

alternate rev limiter. cut spark down to 0* instead of cutting fuel, much less shock and would prevent a spray motor from going far too lean and eating pistons.

3BAR fuel and spark modifications. simple yet effective tables for both. almost done with the code for it, just need someone with a 3BAR to check the A/D result of 1 atmosphere and 3 atmospheres to get the table calibrated.

adjust histograms. Ben came across a potential issue with the way the tables are setup currently, i'll need to think about it more to confirm it being an issue.





slightly more complicated:

VSS based traction control/torque limiting. essentially, you'll need to find the longitudal grip limit of the tires you're running, then set an acceptable "acceleration" limit, should the limit be exceeded, spark gets reduced until it gets back within acceptable limits.

MAF. probably only for fueling, though i might be persuaded to allow the spark tables to be run off of a "grams/cyl" type value seen in OBD2 as well, assuming it's not too processor extensive to impliment.

oil level sensor/light: somewhat complicated, since all of the calculations done on this are done after the key is in the off position in the original code. basically, a timer starts when the key goes off and counts how long it takes for the oil level switch found in the newer motors to go from the "low" to normal position. takes too long? oil level light(or maybe SES, to save an output) gets triggered on next startup.

water/meth injection during knock event. to cool down the combustion chambers when a knock count happens, maybe have an adjustable on-period when a knock happens?

3 step rev limiter/no lift shift. only difficulty is not having the testbench i would like to test this on.

BLM Cell 2. BLM doesn't like to update in cell 2... no idea why yet. i have it disabled in my tune.

explaining PID tuning. it's not as easy as i would have hoped, but it's not too bad, you'll just spend a lot of time making BIN revisions.

knock adaptive spark. whenever a(or multiple, your choice) knock happens, the main spark lookup result has a small amount of it's value removed on every lookup, essentially to keep you out of a cycle of knock, retard spark for a while, dial back in, only to repeat the process.

wheel speed sensor based traction control module. this MAY be simple/cheap to design and integrate as 5V signal to feed into the PCM to reduce torque, but i'm not entirely certain yet.





long term:

VVT/VI control. this is only difficult to deal with due to me wanting to do it closed-loop style using the cam sensor to track how far cam is advanced/retarded and to adjust the control signal to match a desired advance/retard. VI... nothing terribly difficult about it, other than finding out where it's useful.

single fire. getting it to work correctly with properly sized injectors has been fruitless and frustrating and for potentially no benefit.







and that's nAst1 in a nutshell. things may or may not be added in the future beyond this list, if someone can come up with something useful that is within the capability of a 16149396.

[bszopi]

I was just thinking last night about ability to control a 4L60-E, so perhaps I need to do some investigating for you...

Sent from my DROID2 GLOBAL using Tapatalk

[robertisaar]

i would suggest using a well hacked 94-95 truck PCM or even the 94-95 3.4RWD PCM as a TCM for anything equipped with a pressure control solenoid.... the code and tables required to run one of those properly is ridiculous. there's a reason why none of the PCMs that are setup to run one from the factory clock in at anything less than 64KB BINs. 128KB for the 94-95 N* and LT1.

[robertisaar]

was thinking about the ABS wheel speed sensor based traction control for a while.

http://forum.allaboutcircuits.com/showthread.php?t=22222

now, confirming one of my suspcions, using a low-pass filter, a PWM signal can be converted to an analog voltage that the PCM can read and interpret. that's the simple part of the operation.

i don't know if there is any combination of discrete components that will "compare" the wheel speed sensor's frequencies from the front and rear and create a PWM signal, however, i know a MCU can... as long as it can track the frequency of 4 speed sensors and output a PWM signal when necessary. or if PWM isn't possible, just a simple on/off signal would work as well, but with PWM it would be possible to base retard off of the amplitude of the signal, not just a "signal or no signal" comparison.

tough part about that is that the only MCU i really know is the 68HC11, which are getting tough to find, not to mention none of them are in DIP packages, which makes hand-soldering a no-go. after that, learning a new programming language would be quite a learning curve.

i need to look into finding some chips that run on an instruction set similar to the HC11s.

[robertisaar]

did some more calculating....

to use the 2E style of MAF tables and extend them to reach the ranges of 976 to 12,240Hz would cost 132 bytes.
ro make them all a single 16 bit table from 976-12,240 would cost 178 bytes.

$EE stuff(94-95 LT1) uses 4 calibration tables, supposedly from 1,488Hz to 11,216Hz, all of which are 16 bit values. i may need to steal it's 2D 16-bit lookup subroutine and see if it can be used with a single 89 entry table. if so, that will allow up to a 512 grams/sec max table value with increments of 1/128th of a gram/sec.

[pocket-rocket]

was thinking about the ABS wheel speed sensor based traction control for a while.

i don't know if there is any combination of discrete components that will "compare" the wheel speed sensor's frequencies from the front and rear and create a PWM signal,

Differential amplifiers to compare wheel sensor speeds, then a converter to change it to PWM.

[robertisaar]

i've only ever heard the term before, but those work off of frequency?

[pocket-rocket]

i've only ever heard the term before, but those work off of frequency?

Sadly no. They amplify the difference in two voltages.

A differential amplifier is a type of electronic amplifier that amplifies the difference between two voltages but does not amplify the particular voltages.

http://en.wikipedia.org/wiki/Differential_amplifier

I thought they amplify the difference in frequency before I posted, but they amplify the difference in voltage, apparently. We need to get Dave involved in this It's been far too long for me to remember everything I learned in high school, but I'm willing to give it a shot. In fact, I may just pass on my Black Ops for tonight and go dig my old electronics books out of the garage since somehow I ended up with a pair of the same textbooks I learned from. I'll start in the digital electronics book since that's more of the scope of what we are after. I remember differential amplifiers and such around the same section as the logic gates.