legacycentral bbs

Yeah, there's no relation to RS232 at all.


The fastest time I've seen between state changes, that I can recall off hand has been on the order of a few hundred microseconds; but I'm sure it can change a little faster than that (down to less than 100 microseconds would be possible)

I did notice that during a stream, sometimes the seperation interval would get regular; ~1000 microseconds followed by ~500-600 microseconds, then back up to about a thousand.

there's no reference point that I'm aware of; the only way to differentiate between a set of state changes is timed seperation.

I've seen it take as long as 10 to 30 thousand microseconds before the state changes again - but it's always very rapid when looking at it, never longer than half a second.

from what I counted, there was an average of 150 to 200 state changes in a second - sometimes more, sometimes less.

Basically, the ECU is trying to send something to subaru's select monitor, be it real time diagnostics like trouble codes or MAF voltage or what have you; so unless you have the select monitor, you'll play hell figuring out what the data is all about.


Even if vrg3 was to release his source code, it would confuse the hell out of most of us; almost gaurenteed.

Anyone having issues using the scantool with a 92 EJ22T Wagon? I cant get it past the "Reading Rom Id . . ." Screen it just sits there.

help!

all cables connected correctly, parallel port enabled, everything is in the right spot. <---frustrated

thanks for any help

Luigi

Found out that it was a computer error. Parallel port outputting the wrong voltage. program works great!

Have been reading through this, cant see if anyone has tried this with a jdm spec ecu... Has anyone tried this or are those ecu id's just not defined in the program...

I know very little about logic circuits, but in theory wouldnt it be fairly simple to build a cricuit that would read the signal put out by the ECU, and retransmit it with timing info, so a multitasking OS (such a windows) could use it?

From a troubleshooting point of view, would it be possible to select and display two real time parameters on the screen? Four? Just tossing it out to help diagnose a problem I'm having.

Not straightforwardly and probably not very effectively.

Vikash, thanks for your quick reply. I was hoping. I love the scantool software you wrote. I'm having an intermittant problem with high idle on my '91 N/A and was looking at the IAC %-age and trying to see if another parameter was out of spec at the same time. Normally, the high idle returns back to normal slowly in 15 to 20 seconds without a CEL, but once in a while I get a CEL with "normal" but rough idle when the IAC reading was around 55% which should have been giving me a high idle condition. The only codes I'm pulling are all four injectors, IAC, and purge solenoid. Luckily the CELs are very rare; once or twice for a few seconds every 3 or 4 days.

Just tried out the cable and software tonight on my 92 AWD turbo. Works GREAT.

It appears that I'm only getting 6.5psi boost MAX - seems VERY low. Also appears that my boost control duty cycle is stuck at about 1.65 (by my memory) and I would think that it should be a changing number. All other data seemed normal as I don't have any history built up yet.

Any thoughts on max 6.5 PSI boost on a lecacy turbo with less than 50K miles on it?

Anybody played with the Boost Control Duty Cycle before?

Any suggestions would be appreciated.

As far as connectors go - there are other connectors of that type elsewhere in the car. I have lots of corrosion in the connectors on my car, so I stripped SEVERAL from a junker and keep them in a box - anyway, one of them matched the diagnostic port. So if you have a junker you regularly visit - look for the connector instead of buying one...

Are you still out there, Vakesh??? I realize this is an old thread. Thanks very much for sharing your work.

Are there any downloadable assembler programming tools out there that I might use to modify your program. I'd like to convert atmospheric pressure to psi and make some tweaks (maybe make the text white and background black etc. ) If you have any suggestions, I'm all ears.

Thanks so much,

I am still out here, 'cept my name is Vikash.

You're getting 6.5 psi because the ECU has stopped driving the boost control solenoid -- the 1.65% reading is effectively zero duty cycle.

I don't know exactly what prompts the ECU to do this, but I have seen it happen when the dynamics of boost control change too much (like if an intercooler that presents significant pressure drop is installed), and I think it also happens if the engine is knocking excessively.

Less than 50,000 miles? Good lord, man.

I will be publishing the source at some point, as I've already said.

Black text on a white background is the most readable color scheme in daylight. At night it's suboptimal, though.

What did you guys use for the connector?

I couldnt find the stereo harness locally. I had some paperclips jammed in, but they fell out after a day.

I used the stereo connector that was mentioned in VRG's post I believe. I wish I could remember the number. Magnolia Hi-Fi or Car Toys carries it; should be one of those in Victoria.

How much did you pay for it?

You can find the harness easily on eBay or in a bunch of online stores.

Or you could splice directly into the wires... Have a DB25 connector hanging right out under the dash.

Well the cable has to run all the way to my trunk, so Im using a USB cable spliced into a DB25, since I only need 3 conductors.

I just dont want to fuck up the stock wiring.

Splinter wrote:How much did you pay for it?

Less than $20 as I remember. I used an old serial cable for the DB25 and wired it as per VRG's notes. Works great on my old laptop with freedos as the OS.

I have it wired up to my car PC

When I want to use the engine diagnostics, I reboot, hit F9 and it boots off the USB drive instead of the Windows install

Probably what Ill do is do the paperclip thing again, and just ducttape the whole mess together.

Im getting an Autronics SM4 in a few months, so Ill be hooking that into the PC then.

VRG, if you're planning on doing any updates to the program, here's what I'd like to see:

Splinter wrote:VRG, if you're planning on doing any updates to the program, here's what I'd like to see:

be careful not to get your hopes up. the programming format that the code is written in is sevrely limited by size (100k max size IIRC)

Factoring in the high/low/mean would mean pointer and memory management, adding considerable size and complexity to the code.

You'd be lucky to get the text blown up to fit the screen.

I doubt this program has any future development. It's a great diagnostics tool, but that's about as far as the stock ECU is useful.

Does anybody know if this bcbfscan program will work if I use a usb to parallel adapter and a dell inspiron b120. I know this is a long shot but I am hoping to able able to use the softwafe with what I have.

Alright, everyone... finally, a new version. Highlights:

Open source! That's right, download the source code and do what you will (within version 2 of the GNU General Public License, that is). If you want to try to make it work on your non-US model, or make it run under XP, or make it work with other types of hardware interfaces, or show multiple parameters simultaneously, or whatever, give it a shot.

There's experimental support for a new serial port interface. It's more complicated than the parallel port interface, and it might not work with all serial ports, but it should work for many. Hopefully this will increase the range of laptops that the scantool can work with.

It's self-booting. It no longer uses FreeDOS or whatever unless you want it to.

Named it the "B10 Scan Tool," since "BC/BF" was incorrect as it omitted BJs.

...and some bug fixes.

My hero!


Ill try out the serial interface this weekend, hopefully Ill be able to determine whether I fried the parrallel port on my PC, or the diagnostic port on the ECU, and give me a bit of insight into what precisely is causing my poor mileage

Cool. I'd like to hear if it works for you... It's worked on all my computers, but I usually make a point to pick hardware with good I/O ports.

I picked up all the parts a few mins ago, $4

I'll get it soldered together when I get home.

The more I think about what happened, the more I think it MUST be the parallel port that got fried.

Was there any noticable pattern to which serial ports didnt work? chipset manufacturor, computer age?

Four bucks! What a rip-off!

I haven't tried any serial ports that don't work.

The RS232 standard specifies that a logic "0" should be transmitted as a voltage between +5 and +25 volts. For this interface to work, it has to be at least somewhere around +6 volts.

So it's conceivable that some RS232-compliant serial ports might not work if they're close to the limits of the spec. Luckily, most use 10 or 12 volts. But some laptops could conceivably hug the lower end of the spec to save power, or even not actually comply with the spec.