legacycentral bbs

Most of you guys are probably not familiar with EvoScan. It's a popular tool for datalogging mostly EVOs but it apparently works on newer Subarus too although I've never used it for that.

So why am I posting this here LOL? Well when I was looking to see if updates were available I noticed this:

EvoScan Subaru Select Monitor I (SSMI) 9pin/16pin DataLogger Cable (1990-1998) $69.00

HighSpeed USB Cable for Subaru 1990-1998 SSMI vehicles. Perfect for EvoScan. Specifically supports pre 1998 Subaru vehicles with a 9 pin diagnostic port for EFI Datalogging, or if your vehicle has a 16pin obd2 style plug with metal pins in pin12 and pin13, you will need this SSMI cable. Included is virtual comport software to use this USB cable with serial port based software.


You also need the sofware which is $20

http://www.limitless.co.nz/cables/

Granted I don't have this cable and probably won't ever (I've got standalone EM). Just wanted to post it for any of you who are looking for another option. The software is very good IF it really works for the early model Legacy.



The software normally supports the integration of various popular wideband units. I have no idea if it does that with with early model cable.

Very cool, good find.

Sent off a request to find out if it will work...

Anyone here know?

Yes. It works on the Legacy and also the SVX.
It's by far the best datalogging software available at the moment.

b3lha wrote:Yes. It works on the Legacy and also the SVX.
It's by far the best datalogging software available at the moment.

Good to hear.

This program also works with my RS232-type serial cable, so you don't need the EvoScan-specific one if you have something already.

I ordered the cable and software.

please post up when you get it working. I badly needs to log things to get a decent tune and see wtf is up with my wagon.

It isn't working correctly for me-

I know the EvoScan software works correctly because I use it to log my WRX, but the numbers it shows me on the SS are way off.

I bought the $60 evoscan cable for use with our obd1, and regardless of what ecu type I chose in the software the numbers are completely weird.

Is there a special driver needed for this cable? Vista automatically installed a driver when I plugged it in....

I have some XP laptops around I might experiment with those...

Anyone have an issue like this?

I'll be curious if anyone gets it to work. OBD1 communications is a lot more proprietary then OBD2, so there may be some issues.

You have to add in the proper RAM locations and conversions for our ECU. The default one is wrong for our cars.

TurboBeater wrote:You have to add in the proper RAM locations and conversions for our ECU. The default one is wrong for our cars.

And how does one obtain that information?

GD

Yes, lets get that info in this thread for those who have bought the cable/software...

What type of ECU are you using?

I'm currently working to find the parameters for a Hitachi "Z1" ecu, used by the Impreza CS EJ15. If your ECU is also Hitachi then you may find some helpful information in my thread at http://www.subiesmart.com/forum

The stock Hitachi ECU (93 SS 5spd)
I don't know much else about this particular ECU, just things I've read on this board etc.

Hoping to get a better idea of how it handles control of the motor via logging....

I need the ECU ID to send to the EcoScan people... I opened a support ticket, and was told if I send them this, they can start working on it....

Does Ecoscan show the ID when you connect?, I'll have to plug it in again and see...

I can get the info for you from Revtronix. Give me some time though.

That interface does work with older subarus and the only problem I had was that I needed to use a usb to serial adaptor as recommended with the download. I've modified the data file so it has more subaru models in it. If you go into the program, find the data file and right click to "edit", you can post all or part of this in it and it will read subarus...

Damn, that doesn't look right, I'll try again with just the subaru stuff. You just paste it in at the bottom with a <ecu> above and below it. If this doesn't work this time someone better tell me how to post the whole file on here someplace.

Damn, still didn't allow all of it. Someone better tell me how I can post the whole file somewhere.

If you want to email me the file, I can host it.

Hi Josh,
the file is on it's way. People need to check the figures they get because some of the ecu addresses are not correct I think but I have not got the cars to check them against and there are zillions of variations it seems. Perhaps we could have a thread where people can post their suspected ecu addresses as they find them and have them peer reviewed before we incorporate them into a new data file? Same thing applies to the S.S.M program from vwrx. Cheers, Brett.

Here's the files

http://main.experiencetherave.com/subar ... voscan.zip

Hi all,
if you are using those files please be aware that some of the addresses may not be correct and the information you get out man not be either.
i have used the 7232xx, which I modified because the O2 average was a meaningless number that never changed and was useless, so I have added a live figure, apart from that it all works I think.
I also made the 703315 file which I think is all correct except for the knock correction, which I didn't need at the time.
The three addresses that are the hardest to find if you are making a new file are the O2, A/F correction and the knock correction. They are awkward because they should all be hovering around a middle point which means that the actual figure that comes out of the ecu, before it has any correction made to it, should be around 128.
After correction factors have been applied the O2 should swing constantly from approximatly point one of a volt, up to approximatly point nine of a volt. If the system is working correctly though the voltage should swing within about point two of a volt, of four point five, that is to say, it should swing from about point three to point seven. To test your O2 either let an air leak in or pull one plug and the figures should go briefly down or up. Be aware though that if the O2 feedback is working the figures will soon centre back again around four point five volts. The other thing that effects this O2 reading is the A/F correction. It should normaly be reading zero, but when you let in a air leak, or pull a sparkplug wire, it should note the O2 reading and alter to a positive or negative number of about one or two. So, if everything is correct, warm motor idling, and you cause a missfire, the O2 should go up towards point seven and straight away the A/F should change from zero to a negative one or two. If instead of pulling a plug lead you cause an air leak then the O2 should drop towards point two and straight away the A/F correction should go one or two positive. As soon as the A/F has made a correction the O2 reading should go back to it's usual swing in the middle of it's scale. If either figure is slow to adjust, or not in the centre of it's travel, then you may have a problem
The other difficult address to find is knock because you have to load the motor and then the knock figure will only give a reading for a moment before it returns to zero. The give-away clue there is that as the knock figure goes up, the ignition timing goes down.
With all these tests and readings, it is best to turn off all the sensors you do not need to record, so as to get the fastest refresh rate possible.
Hope all that is understandable and helpful. Cheers, Brett.

Here are the parameters for my Legacy.

TomsSVX got the SVX addresses correct by eavesdropping a select monitor. Then I disassembled the program code on the Legacy and SVX and compared them to find the corresponding Legacy addresses. No guesswork involved.

1996 UK model Legacy EJ22
JECS "U1" ECU Rom ID: 7432A1

10a2 AFR Correction
109a Advance
1305 Load
1306 Injector Pulse Width
1307 Airflow
1310 O2 Sensor
1314 ISC Duty Cycle
1328 Knock Correction
1329 TPS
1335 Battery Voltage
1336 Speed
1337 Coolant Temp
1338 RPM
133a O2 Max
133b O2 Min