$EE and V5, 94 caprice LT1

drunkenwhippet · Post by **drunkenwhippet** » Wed Jul 07, 2010 11:18 pm

Hey everyone, a bit of an odd situation I have here. long story short, data acquisition failed. unplug from aldl connector, cable test passes.

I'm running TunerproRT v5 on a laptop with no hard drive. I'm running puppy linux from a live CD with swap and storage partitions on a USB device. laptop cannot boot from USB, only cd. Tunerpro runs within WINE, seems to communicate with serial port enough to do the basic cable test. My cable is a 2 transistor, which when initially built wouldnt pass the simple test. The laptop's serial port needs the full -3v on the rxd, so i added a 9V battery to the circuit, so the transistor pulls rxd to -9 instead of ground. now cable passes simple test.

My next thought is to drag one of the desktops out there and give that a whirl running windows, my problem is that my inverter can't handle the load so it'll be on an extension cord, meaning no driving while logging on it.

the car is a 94 caprice classic wagon, LT1 and 4L60E. 365,000 miles, car runs like hell when in open loop, stumbles, backfires. when it warms up and hits closed loop it will set you back in the seat if you have your foot down, like someone flipped a magic switch. (being open / closed loop thing is my guessing, since i cant datalog yet). The PCM has been replaced, there is a "remanufactured..." sticker on it, I wonder if it's not the right rom on it? if anyone has any suggestions, please help me out. any extra info that might be able to help in the slightest of way, please ask away.

the cable also passes the test on an XP64 machine, with the +12 and gnd connected to the machine's power supply.

the cable was built with 3 7.5K ohm resistors, 2 darlington pair NPN transistors with a saturation voltage of 1.5V, so RS-232 should be driving them WELL into saturation... but is there enough currents floating back and forth within the circuit to open both transistors at once? I might try finding a couple transistors with a higher saturation voltage. the diode is a rectifier with a forward drop of .69V, should i try a germanium diode with like .3V forward drop? all of the components are salvaged from various electronics, mostly a power supply and I/O module from an 80's PLC, all tested good on multimeter.

I'm afraid it's my cable, as software side seems to be (somewhat) proven. (does anyone know of anyone running tunerpro under WINE ?)

I apologize if this post doesnt have much consistent direction, I've been at it for like 7 hours and ADD doesnt help. I've nearly exhausted all of my ideas, and it's a pain booting the laptop from CD every time I go outside, then have to install WINE and tunerpro, it takes around 20 minutes from power on to clicking the start acquisition. I am setting up the com in the settings, and COM2 is IRDA in the laptop.

I thank you SOOO much for reading this, and if you have any thoughts concerning this situation, please express them here.
Jim

Post by **robertisaar** » Thu Jul 08, 2010 6:11 am

http://tunerpro.net/forum/viewtopic.php ... light=wine

that may answer a few of your questions...

as for the cable being the issue, i've never had good luck making a 2 transistor... they always acted funny... the MAX232 on the other hand, works great.

drunkenwhippet · Post by **drunkenwhippet** » Thu Jul 08, 2010 8:18 am

I guess i didnt mention part of my situation. im dirt freakin poor. When I first started reading about tuning and everything, I was doing good, 50 hours a week, making good money. my plan was actually to buy a netbook for a couple hundred $, build a max232 cable... but a month later I go from doing great to being laid off. after bills, food and gas for my lady to drive the car to work every day, we have 0$ left at the end of the week. so much for the economy being "fixed" or even turning around, but lets save the politics for another day.

i didnt mention the car smells rich as all hell, runnin for 10 minutes in the driveway leaves wet spots under both of the tailpipes (factory dual exhausts) and the car likes to drink gas like a frat house drinks cheap beer. the car was fleet maintained, being a former taxi. It's the best running / driving car I've seen with 350K + miles on it. actually the closest I've personally seen was a cavalier with 310k, and it was pretty sorry to drive.

but anyway, back to the cable... I'm not sure what voltages the PCM is looking for on it's side, but the levels should be within spec on the rs232 side. I had an oscilloscope until the flood of 2007, if that damned thing still worked I'd be able to see just how ugly the interface is.

ok, just checked digikey. at $1.16 each for TI's clone of the maxim chip, i think i can go without tobacco for a few days and just order some to play with. If anyone has suggestions for the time between now and i get that built, i'd still love to hear them.

also, does anyone have an ads file for $EE on the V4? all i can find is the V5 only one. I'd like to at least try that, see if it says anything different. cable test passes on both v4 and v5, after adding the 9V to the circuit.

the link posted spans like 4 years, some of the wine config and commands seem to have changed since then (comparing current wine documentation to old posts including the on here on TP forum) but i havnt tried messing with any baud rates. I guess that tonight when i drag the desktop and a monitor out there I'll be able to either rule out the cable as being at fault or the software mess. I wont be able to do anything till the lady gets home in a few hours. thanks for the input, I'll keep working at it.

Jim

Post by **robertisaar** » Thu Jul 08, 2010 8:34 am

http://www.taydaelectronics.com/servlet ... MAX/Detail

i've actually ordered quite a few things from them, and considering where it's located, it's not bad at all... the $5 minimum should be easy enough to get to though if you do any kind of electronic stuff... i had to order enough parts to build 4 full MAX232 based ALDL cables to get to it... i'll probably build them and sell them off somewhere.... i could always use the extra cash in this day and age...

i MAY have the ADS i built when mark and i were setting everything up to see how well/if $EE would work with V5 after being imported... if you want it, i can email it.

drunkenwhippet · Post by **drunkenwhippet** » Thu Jul 08, 2010 8:57 am

OK, so i've seen a couple differend schematics for max232 based interfaces.

http://www.techedge.com.au/vehicle/aldl8192/8192hw.htm
and
http://www.mindspring.com/~amattei/aldlif.jpg

anyone have experience with both? the biggest difference I see is the one uses external transistors to do the pullup/pulldown, while the other just uses a diode and resistor on the max232's output pins. can the max232 sink/source adequate current to just use the R and D? or should it be the go big or go home mentality and just use the external transistors?

Post by **robertisaar** » Thu Jul 08, 2010 9:05 am

i use a modified(cleaned up for stuff i'll never use, like 160 baud, and the mode sense stuff) version of the techedge schematic. the other one is WAY overcomplicated.

drunkenwhippet · Post by **drunkenwhippet** » Thu Jul 08, 2010 9:12 am

so it's a pre-v5 ads you guys used to test the import function? i'd gladly give it a shot. my email is drunkenwhippet@gmail.com

and the 44 cents for the max232, that's at digikey's 10,000 unit price break. nice! how is shipping from thailand?

off topic, since being laid off, i decided to persue a formal education in electronics. I've applied to the local college, have an appointment with admissions person saturday. until this point, everything i know has been self taught, from books, internet readin, and fiddlin 'round with stuff. I figure a couple years in EE should expand my horizons a bit, I have a 2 year vocational school education in precision machining, and aiming to start a business initially doing fibreglass work (body kits, interior and such) and adding electronics (tuning, etc) and once i have the income to afford it, go to school for high performance motorsports (doing internal machine work, boring, balancing, etc.). I'm already well capable of milling heads, had the heads off my SBC400 in my truck on a bridgeport after blowing a head gasket when the water pump went.

so basically hopefully in the coming months, I hope to be able to offer an insight to the tuning community in a way similer to Mr. Moates has, all of the misc adapters and such he offers. somethin relatively simple I want to see is a bluetooth to aldl adapter. I have a BT GPS receiver that has an RS232 to BT chip in it, and once the connection is established in windows it appears as just another COM port, like a USB adapter does. with so many laptops having built in BT these days, it seems like it would be real handy.

Thanks for putting up with my ADD-inspired ramblings everyone, im sure I've got at least one logical thought in there somewhere...

Jim

drunkenwhippet · Post by **drunkenwhippet** » Thu Jul 08, 2010 9:14 am

robertisaar wrote:i use a modified(cleaned up for stuff i'll never use, like 160 baud, and the mode sense stuff) version of the techedge schematic. the other one is WAY overcomplicated.

so the max232 is fine with just the resistor and diode for pcm side... splendid, less soldering i'll have to do!

edit: is the CTS even needed? will the max232 function without it? i see the RTS is used for powering the circuit (i'll be going with a 12V source to the 7805) and i'll toss the 160 baud as well.

Post by **robertisaar** » Thu Jul 08, 2010 9:16 am

drunkenwhippet wrote:I have a 2 year vocational school education in precision machining

that makes two of us...

i was looking, but i cannot find the ADS files i made anywhere... since it's more or less for testing, a rather quick and incomplete ADS should work fine though, just to verify that software is not the issue...

Post by **robertisaar** » Thu Jul 08, 2010 9:23 am

drunkenwhippet wrote: edit: is the CTS even needed? will the max232 function without it? i see the RTS is used for powering the circuit (i'll be going with a 12V source to the 7805) and i'll toss the 160 baud as well.

by the time i was done, i only had three wires going to the DB9 port: RxD, TxD and Ground.

then again, i have mine going through a serial - USB convertor as well, and use a second usb cable to supply the +5V the MAX232 wants to see...

and thai shipping seems to be fairly quick, i think i got my parts in ~7 busniess days or so.

drunkenwhippet · Post by **drunkenwhippet** » Thu Jul 08, 2010 12:16 pm

what was cost on shipping? and i figure it wont be much to regulate 12v down to 5, my laptop only has one usb port, and im using it for the flash drive.

Post by **robertisaar** » Thu Jul 08, 2010 1:35 pm

i think shipping came out to like $2... which isn't bad at all.

drunkenwhippet · Post by **drunkenwhippet** » Fri Jul 09, 2010 5:23 am

the next step....

I'm getting together an old AMD K6-2 400 machine to run win2k, the comp should run on less than 100 watts. The plan is to run the computer on the inverter, monitor on extension cord. If all is well and I can actually communicate / log, I'll be able to boot computer, get everything logging, unplug monitor and go drive. done driving, i'll come back, plug back in, save everything and work from there.

question is, how well *should* tunerpro log (samples/sec?) on a k6-2 400 with no l2 cache? system has 384 megs of ram, and a SIIG high speed rs232 card (maybe it will offload the cpu some?) I'll kill every non-essential process and service in an effort to conserve cpu cycles, but the lack of L2 cache worries me. I guess i'll get to see just how CPU intensive tunerpro is when logging.

on a side note, after rent is paid i should be able to order some max232s and other misc components, if the 2 transistor cable seems to be the culprit. I didnt get to play last night, it was raining out and i didnt want to carry laptop back and forth in the rain.

edit: K6-2 is junk. Got a P4 running on the inverter, had to pull cd rom, a couple fans and a stick of ram and the network card to get it to where the inverter wouldnt scream (beeps when overload) when cpu load goes up. talk about pushing limits i guess..... so when the lady gets home, i should be all ready to go out and see what my cable does on a winxp desktop.

whatever happens, at least i've been occupied here lately instead of sittin round bored.

drunkenwhippet · Post by **drunkenwhippet** » Fri Jul 09, 2010 4:58 pm

the 2 transistor cable was a fail on the dell with winxp. So i'll be switching out some parts and see how it goes, gotta put a few $$ in the bank tomorrow before i order the parts for a max232 interface. thanks again for the help robertisaar, and massively huge thanks to mark for all he's done with tunerpro. I cant wait till i can give somethin back to him for his efforts and the tuning community as a whole.

Post by **robertisaar** » Fri Jul 09, 2010 5:16 pm

http://robertisaar.co.cc/web_images/max232_aldl.pdf

this may help you a bit... this is what i made for my stripped-down ALDL USB cable. should you ever need a serial-USB convertor, grab the directly from Parallax. the 28031 is the cheapest known convertor that will work at 8192 Baud.

drunkenwhippet · Post by **drunkenwhippet** » Fri Jul 09, 2010 10:12 pm

so now my interface(s!) wont even pass the little test in tunerpro. before i was at least getting that, from all 3 computers i have. I built another, from scratch. with, and without the 9V battery, im not getting DANG!. either i blew out the serial ports on all the computers or something. I guess my serial cable i've been using could be bad, i said fuckit and left all the DANG! in the car. i'll have to test it pin for pin tomorrow. I know somewhere out in the garage i have a bare db9f connector, if this cable is in fact bad i'll try that. even used 10k resistors on the second one. if the serial cable is good, i'll try going back to a silicon diode and see what that does. right now i feel like the little bastard has challenged me, seeing that it used to pass the test now i dont get crap. and I cant try one of my serial cards caus no (running) computers with an ISA slot. looking forward to placing that order, not looking forward to battling with these damned 2 transistor IFs till i get the max232s and everything in. i'll be really sad if ive managed to destroy the serial ports in every computer i have, but i thought the serial port was 12V tolerant, so there shouldnt be any way this comp's is dead, i've only been hookin the 12v directly to the comp's PSU. oh well, it's sleep time now. i'll save my frustration till tomorrow. once again, i'd like to express my appreciation for all the input. most people trying to make a few $$ on the side dont want to help you with squat unless you're handin them cash. but it's always been my philosophy as "i'll tell you how to do it yourself, or you can pay me to do it for you". or if i'm learning along the way, i'll probably do it for free, the first time at least.

Post by **robertisaar** » Sat Jul 10, 2010 4:21 am

the test isn't necessarily 100% concusive...NONE of my homemade cables pass the test, but they also read the datastream perfectly... odd little quirk , but oh well, they work better than the aldlcable.com one i had.

drunkenwhippet · Post by **drunkenwhippet** » Sun Jul 11, 2010 4:57 pm

just ordered enough parts to build 5x of the interface that robertisaar posted a schematic of. cost $7.94 after shipping, I also bought a few extra parts not needed for the interfaces but will still come in handy with things i'll be fiddling with. now i wait 7 to 21 days for shipping.

Post by **robertisaar** » Sun Jul 11, 2010 5:07 pm

it usually doesn't take that long for me...

they do seem to be quality pieces though, at least the quality of radioshack stuff... i also contected them about getting some DB9 connectors, and they should have them in stock next week. that's another 2+ dollars radioshack won't be getting from me!

drunkenwhippet · Post by **drunkenwhippet** » Mon Jul 12, 2010 1:21 am

I looked for DB9s, but they didnt have them.

the only REAL radioshack around here went out of business, it was a little franchise where they knew me by name for 11 years. and that was an 11 mile drive from where im living now. now all we have is the ones filled with overpriced RC cars, computers and other fully assembled out of the box junk they sell. not a single resistor or capacitor in the place. disappointing if you ask me. the cool thing about the RS that closed, my total never seemed to come to the total of all the price tags added up. I was probably one of the few people that kept them floating as long as they did.

drunkenwhippet · Post by **drunkenwhippet** » Fri Jul 23, 2010 5:34 am

WOOHOO! my parts just got here, so I have to finish my daily chores and then I can start putting the cable together. the car wont be home till round 3 or 4:00, so I've got a little bit of time to get it together.

They max232's were actually maxim, not ti or some other knockoff. I'm getting excited already, hopefully i'll finally have a working cable so I can get some datalogging and tuning done. have enough parts for 5x cables, so i should be able to get at least one built successfully haha.

drunkenwhippet · Post by **drunkenwhippet** » Fri Jul 23, 2010 5:09 pm

I got the max232 interface built, using robertisaar's schematic, and got SOME date from the car, the problem is that it doesnt seem to be right. in the item lists, the numbers jump around to crazy values (rpms from 1200 to 6000)

It seems like the serial port settings aren't right, as if the word length is wrong or something. It's sampling at around 3-4 hz, I am running tunerpro on puppylinux, under wine, on a dell c600 laptop without hard drive. using flash based USB drive for swap and storage partitions. Maybe I need to mess with the serial port settings within linux / wine? i've only made it as far as actually linking the serial port to com1 within wine, havnt figured out baud rate settings etc. I'll do a search to try and find what settings i need the port to be at, but if someone knows and wouldnt mind posting here i'd appreciate that too.

I wanna say thanks again to everyone, especially mark and robertisaar, for all the help and the software that I'm getting to use for free. Once my employment status changes i fully intend to buy a full license and help support mark.

drunkenwhippet · Post by **drunkenwhippet** » Sat Jul 24, 2010 8:15 am

a little update.. a winxp box does the same thing, totally erratic numbers and 3-4 hz sample rate, even with the car not running. So i dont think it's the laptop / linux config. only errors i get were when starting the car, about 7 errors then it started streaming right again.

if anyone has any ideas, i'd love to hear them. I thought other people were using the $EE_auto.adx and EE.xdf from the download page, without problems.

Is it a possibility that a bad solder joint on the interface would cause erratic data? I guess my next step is to reflow all of those and see where I get. I've got enough parts left to build 2 more, if I have to. I dont think it's crosstalk over the serial cable, it's only about 3-4 feet long.

there's a sticker on top of the pcm, with 8051U on it. and the reman sticker has boxes for an ID and vin number, both blank. If it were a goofy rom on there could it be sending a different data set than $EE is expecting? winflash doesnt want to run under wine, i didnt try it on the xp machine while i had it out there. thunder was starting and I didnt want to get it caught in the rain.

thanks for the help guys, hope to get it going sometime soon, feels like i've been messing with this for months now.

Post by **robertisaar** » Sat Jul 24, 2010 8:24 am

was this with the updated ADX i sent?

drunkenwhippet · Post by **drunkenwhippet** » Sat Jul 24, 2010 8:51 am

i must not have gotten it yet, no it was the old adx on the site. i'm gonna check the email right now.

to add to my headaches, the onboard LAN in my main computer here in the house took a dump yesterday, so i have to boot this laptop every time i want to get on the net. And i dont have any network cards that i can find 64 bit drivers for.

just seen the email, i'll get out to the car and try it when i can, looks like the rain is passing.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 2:59 am

hey robertisaar, im not dead, just had a bunch of "life" to deal with. living situation changed then my niece came into the world 3 months early. I havnt had much time to play.

I tried the new $EE you sent, and was still getting erratic data. I played with the delay from 20ms all the way up to 1000 and all over between. It appears that I'm getting good data for a short period of time, then everything goes erratic again. I think i managed to get one log saved that shows TPS% moving what seems to be accurately, from 0 to like 96%. then I noticed that the 7805 regulator on my interface is getting hotter than i'd like, so i wonder if its getting too hot and the power to the max232 is flaking out, causing the troubles on the datastream. I'll replace the regulator and add a heatsink and try again when i get a chance. if you want, i can send the log that I did recieve, i think the one i took was with like 400ms delay.

just wanted to say thanks for the help and i'm still trying at it when have the opportunity

Post by **robertisaar** » Sat Aug 21, 2010 7:33 am

the regulator flaking out is a possiblity. if the BCM/CCM was talking on the line, then that SHOULD trigger the error counter, not show erratic data.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 7:45 am

it would go fine for a few seconds, then count like 4 to 8 errors and then go to the erratic numbers.

i might have to drag a desktop out there again and see where it gets me, although power is a bit less of a convenience now

Post by **robertisaar** » Sat Aug 21, 2010 8:00 am

that is quite odd... not driving anything extra on the circuit that would be consuming the +5V from the regulator would you? i know some people have added in LEDs with success.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 8:25 am

i built it exactly to the schematic you posted, nothing more nothing less. power is fed on some 18 gauge speaker wire, the wire to the aldl plug is an old multimeter probe, and a serial cable.

do you use heatsinks on your regulators? i thought since it's so little current it shouldnt be needed.

Post by **robertisaar** » Sat Aug 21, 2010 8:26 am

my regulators NEVER get above room/air temp actually...

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 8:34 am

when (trying to) log mine heats up till the point of burning skin, but sitting idle it cools off

Post by **robertisaar** » Sat Aug 21, 2010 8:37 am

hmm.... got another vehicle to play with?

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 8:39 am

only things computerized i have access to is a 90 dodge with a TBI 318 swapped in (originally a v6) and the rest are OBD2. does TP do obd2? ive never looked into that...

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 8:43 am

another thought, an autoxray codescout 2500 is able to read codes from the car, if that means anything.

Post by **robertisaar** » Sat Aug 21, 2010 8:51 am

reading codes can be done with 1 sample of data... don't suppose it can show live data?

and the interface between OBD1 and OBD2 are very different, requiring an ELM327 IC and a much more complicated circuit for communication. i wouldn't know if TP can communicate with OBD2 though, but i imagine since the signals get converted through the ELM327, if you knew the right commands, you would get some action going on.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 2:16 pm

replaced 7805, added heatsink and couple caps before and after 7805. still same DANG!. gonna set up everything on desktop system and drag that out there. this whole thing is makin me saad.

could an adx be made to just read DTCs and that's it?

while not running, it seems to log decently untill i press the gas pedal past about half, then errors. last time i was messin with it it would go up to 96% on the data list. problem seems random and to be changing. going to redownload all files and try on the desktop and go from there

Post by **robertisaar** » Sat Aug 21, 2010 2:20 pm

you COULD chop out everything but the DTCs, but if you're thinking it will make a difference, i wouldn't count on it.

and it logs fine UNTIL excessive amounts of throttle is applied?

this is very odd.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 2:28 pm

i'm just hoping to read dtcs at this point, it started throwing a SES light, scanned it couple days ago and it was left o2 lean, want to see if it's still just that or any more, without having to drive down to oriellys to use the autoxray scanner.

and yeah it seems once it gets past 50% it starts error-ing. im bout to drag out this desktop to see what happens. the desktop i was dragging out there, i gave to my dad caus his comp called it quits.

Post by **robertisaar** » Sat Aug 21, 2010 2:35 pm

hmm... i guess either you or i could chop the hell out of the datastream and get it down to the bare minimum of error codes.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 2:45 pm

i think i've figured out the structure of the macros and commands and how they work, but what the individual commands are, i'm lost. i know they're hex, but dunno where i'd find a reference of sending what command nets what response from the pcm. but i'll try this pc before going through the chopping of the adx.

I seen someone here reference using the EE def, so im pretty sure it works, it's just something in my situation that isnt agreeing with it.

drunkenwhippet · Post by **drunkenwhippet** » Sat Aug 21, 2010 4:37 pm

tried logging with the desktop, dual core intel 2.8ghz, the EE def on TP site worked better than the one with 20ms pauses added, but still had flickers of erratic numbers. in the log you can see somewhat steady and appearing to be accurate numbers, but every so often something garbles the data (?) and they flash to waayy off stuff. in the logs i took i could clearly see the left o2 reading was much lower than right, the exhaust smells rich as hell, im pretty sure my O2 sensors are toast, especially the left, it's throwing the SES light. and the blms are often both pegged at 160.

now i just need to get the laptop logging perfectly so i can drive it and log, not just sit in the driveway idling...

but at least i think im getting info to work toward the right direction...

also, tunercats winflash utility to read the rom, it was unsuccessful. If i have to build another interface, i'll buy a socket for the max232 to eliminate the possibility of heat damaging the chip somehow while soldering.

I just realised that i know someone who has a caprice, i dunno if it's an lt1 or l99 (do they use the same pcm?) and i dont know if its a 94-95, but i'll try and get ahold of them tofind out the details, and if it is compatible i'll see if they'll let me plug into it and see if that has any better results. I'm sooo ready to get this working...

Post by **robertisaar** » Sat Aug 21, 2010 4:48 pm

AFAIK, the L99 and LT1 use both the same PCM and datastream, but it's been a while since i've looked at it.