EvoScan - Subaru Select monitor

[xbn377]

I did find that and got the update but it did not help still gives me the same message...Anyone have the actual evoscan cable they dont want anymore? Thanks for the idea though.

[log1call]

It's not the version that is the pronlem, it's because the drivers are not being found in the right place or some such thing.

You also need the microsoft Netframework, did that get installed or do you have that? You can check if you have the net framework by looking in add and remove programs.

I'd remove the driver using a tool thats availaible from the ftcd site, then install them again. Then look in hardware mahager and (if youv'e got a cable plugged in), you should have an extra usb listed as "usb serial converter", and under ports you should have one called "usb serial port". If those are not there you haven't got the drivers right... or your cable is defective?

[log1call]

Thanks for the file there Bsoc2600, I'll add it to my list.

Could you get the specific ECU number for me to record it as? That's the last two numbers of the 7232xx? They appear in most programs somewhere when they are connecting to the car. In evoscan it's in the info box that turns green when you have a connection. Then when we connect and get a funny reading we know it's a fault rather than having to go through a system of confirming that the reading is from the right place on the motor and not possibly a wrong bit of data.

[BSOD2600]

My ECU is 72.36.22 (as ID by b10scan years ago). Don't have a laptop atm to verify what evoscan sees.

Once I get my hands on a laptop for a weekend, I can do some better testing/comparison between EvoScan values and the Revtronix logging tool. Then hopefully set up evoscan data.xml to better match the correct values for legacys.

[BSOD2600]

After looking through my evoscan and revscan logs during last weekend, have a few questions...

• Pulse Width: What does the raw number mean or the conversion equation? What is the max for the stock injectors? 440cc injectors?
• Spark learn factor: Whats this used for?
• Idle speed: typically 650 rpm, when it drops down to 50-25 rpm, this is for fuel savings when coasting?
• Fuel Trim Short/long: How does this effect things? Trimmed based of the O2 or a lookup table?
• Air Flow: What is the max grams/second of the stock/JECS MAFs? Stock max voltage is 5.0V? JECS?
• Spark Base/Advance: These numbers are Degree BTDC? Spark base is what the ECU thinks it should run at and advance is the calculated?
• Boost DC: low number means its not trying to open the wastegate vs a high number is the % it's trying to open the wastegate?
• MAP: The max 2bar sensors can read is 200 kPa? When it gets maxed out, how does that effect/change things?
• O2 sensor: "0" readings are treated as ultra lean or just a lack of logged data?
• Boost DC: What are these for?
• Knock Reading: This takes the measurements from the knock senor and then retards the timing X degree? With EvoScan, I've noticed the knock parameter (1530h) some times logs 255 through 250 on some WOT pulls -- I assume this is bad?

[Legacy777]

This thread may be of some use

http://bbs.legacycentral.org/viewtopic.php?t=24376

I've got a few answers, but I'll quote your original thread and answer within.

[Legacy777]

After looking through my evoscan and revscan logs during last weekend, have a few questions...

[*] Pulse Width: What does the raw number mean or the conversion equation? What is the max for the stock injectors? 440cc injectors?

This number is in miliseconds. There isn't necessarily a max.

[*] Spark learn factor: Whats this used for?

This is probably tied in with the knock sensor. The ECU will advance timing until it hears knock on a regular basis, and then back's off timing.

[*] Idle speed: typically 650 rpm, when it drops down to 50-25 rpm, this is for fuel savings when coasting?

It shouldn't drop to 50-25 rpm. That's probably a bogus reading

[*] Fuel Trim Short/long: How does this effect things? Trimmed based of the O2 or a lookup table?

This is the a/f ratio learning control, and I believe it's based on O2 sensor readings

[*] Air Flow: What is the max grams/second of the stock/JECS MAFs? Stock max voltage is 5.0V? JECS?

Hitachi MAF is around 200 grams/second & JECS is around 250 grams/second. This should help
http://main.experiencetherave.com/subar ... ansfer.xls
Max voltage should be 5.0v

[*] Spark Base/Advance: These numbers are Degree BTDC? Spark base is what the ECU thinks it should run at and advance is the calculated?

Yes, I believe it's before TDC. Yes base is the physical timing, and advance is variable based on engine sensors/parameters

[*] Boost DC: low number means its not trying to open the wastegate vs a high number is the % it's trying to open the wastegate?

Without any signal/power applied to the wastegate solenoid, you can blow right through it. As you increase duty cycle, it closes off boost pressure to the wastegate solenoid and vents it back to intake. So it's backwards from what you posted. Low number means wastegate solenoid is seeing all boost pressure and trying to open actuator. High number means wastegate is seeing a portion of boost pressure.

[*] MAP: The max 2bar sensors can read is 200 kPa? When it gets maxed out, how does that effect/change things?

Yes, in theory. It maxes out and the voltage gets clamped at that high level

[*] O2 sensor: "0" readings are treated as ultra lean or just a lack of logged data?

Narrowband O2 sensor is supposed to cycle back and forth, but yeah, that's pretty much correct.

[*] Boost DC: What are these for?

Not sure what you're asking here.

[*] Knock Reading: This takes the measurements from the knock senor and then retards the timing X degree? With EvoScan, I've noticed the knock parameter (1530h) some times logs 255 through 250 on some WOT pulls -- I assume this is bad?

It can be, does it do it all the time?

I'd also suggest looking at the factory service manual in section 2.7 for more details on some of these values.

Code: Select all

http://www.main.experiencetherave.com/subaru_manual_scans/1992_FSM_(Engine_&_Electrical)

[BSOD2600]

Ah thanks, never ran across that thread/document before.

Not sure if RevScan is pulling additional info from the ECU that B10scan didn't find or if they fields are being calculated, but this is what it looks like (regarding some of my questions)

[*] Pulse Width: What does the raw number mean or the conversion equation? What is the max for the stock injectors? 440cc injectors?

This number is in milliseconds. There isn't necessarily a max.

There isn't? What about all the threads I read about stock injectors getting maxed out, running static, not in their 'ideal' 80% duty range? I can't find anywhere that states how people calculate that. For example, Revscan logs the raw IPW (where I see readings of 140-150) and EvoScan uses RAW * 0.128 (ok, I pulled that from b10 source code) which apparently is the milliseconds (which I see 18-19ms) all during WOT. Idling the RAW is around 20 and 2ms.

To me, it seems like the numbers are backwards. During WOT when more fuel is needed, the pulse width would need to be more frequent... hmm, so my 18ms reads mean what exactly? It's firing 18 out of every XX? milliseconds? There must be a way to calculate how much fuel is flowing with that injector pulse width to see if its close to the max capable.

[*] Idle speed: typically 650 rpm, when it drops down to 50-25 rpm, this is for fuel savings when coasting?

It shouldn't drop to 50-25 rpm. That's probably a bogus reading

At least in Revscan, it appears to be some sort of target idle speed -- because my the other parameter for rpms still doesn't even come close to it. Furthermore, revscan was logging idle speed numbers, which are multiples of 25 -- kind of odd for something to be a sensor reading. Wonder where they pulled it from...

[*] Fuel Trim Short/long: How does this effect things? Trimmed based of the O2 or a lookup table?

This is the a/f ratio learning control, and I believe it's based on O2 sensor readings

Alright, but what do the numbers actually mean? During WOT, I see the fuel trim short % often at 0. Other times (can't really find a pattern), I see the number fluctuate -20 to 5%.

[*] Spark Base/Advance: These numbers are Degree BTDC? Spark base is what the ECU thinks it should run at and advance is the calculated?

Yes, I believe it's before TDC. Yes base is the physical timing, and advance is variable based on engine sensors/parameters

What are ideal/acceptable values for this? I'm seeing ~25 when WOT and ~50 when idle.

[*] MAP: The max 2bar sensors can read is 200 kPa? When it gets maxed out, how does that effect/change things?

Yes, in theory. It maxes out and the voltage gets clamped at that high level

Clamping does what to other sensors? What does the ECU do then? Must be a reason why Revtronix is talking about a 3bar MAP sensor for higher boost levels...

[*] Knock Reading: This takes the measurements from the knock senor and then retards the timing X degree? With EvoScan, I've noticed the knock parameter (1530h) some times logs 255 through 250 on some WOT pulls -- I assume this is bad?

It can be, does it do it all the time?

RevScan never logged any, yet EvoScan did almost always during WOT. Not sure who to believe or how to interpret the results. I know the parameter is only 255/FF big -- so a value of 255 would be interpreted as knock detected, don't do anything. A value of 250 would be knock detected, retard timing by 5 degrees (255-250) ?

[log1call]

There are two documents you could be interested in reading here...

http://cid-4ca3c3459aaa7f7f.skydrive.li ... px/.Public

The one called diagnostic is the criteria for codes to be set and what readings should be within for sensors on the ssm ecus, and the 04 diagnostic criteria..... well that's what it says.

In my experiance, looking at all sorts of cars, not just subarus, it is more important that the readings behave in the right way than that they be exactly right.

[IronMonkeyL255]

Was anyone ever able to get this working on their USDM Legacy Turbo?

I finally got my laptop to speak with the cable, but the numbers are all screwy. It seems to communicate with the car, but gets incorrect numbers.

[log1call]

Are you using the right protocol in Evoscan's make selector panel. You can use either ssm or the subaru tick button and get readings in some models but only one set are right.

[log1call]

You might have to read around to verify which market your ecu was used in and which set of ecu addresses is going to be right for it.

Finally, if you are trying to verify readings or find new ones, I think it is a better idea to study their behaviours, under test conditions, rather than their specific readings.

[IronMonkeyL255]

I am just using the factory ECU, so it's nothing fancy.

I tried using just about every protocol and none of them were reading right.

I am assuming it would be either SSMI other or SSMI A30 113

[log1call]

The ssm protocol seems to work after you have made a conection on subaru, but it can loose it if you stop and restart. . It's best to stay on the subaru if it works at all.

First off, when you connect do you see what your ecu's romid number is... 7032XX or something? give us all those digits and it should identify it.

[log1call]

That model A30113 looks like OBD2 data addresses so I doubt that would be your model. You don't say but I'm assuming it's a 91 turbo and also presumed a two litre quad cam? I'd try addresses as per 7136XX or 7236XX. They might make more sense, then you might need to choose one to suit your transmission.

[BSOD2600]

EvoScan does not come with the correct parameters in the data.xml file for our Subarus. You need to use the data file a user posted previous in this thread and/or insert the fields I pasted a little into your data.xml.

[IronMonkeyL255]

I see......

I entered in the data from the VWRX site, and that cleared it up.

Still addressing the conversions, though. A few things I noticed. The vehicle speed doesn't need to be multiplied by 2 to get km/h. The coolant temp seems to be reading backwards. As my car warmed up, the temp was dropping. I swapped it to read 50-x, and at least the readings seem to be reading the right way. Not sure if they are accurate, but at least it isn't reading backwards.......

It looks like the information stored for the SSMI protocols are for the non-turbos and the jdm ecus, while our cars use different addresses.

[log1call]

When evoscan makes a connection succesfully there is a screen that turns from yellow to green, in that screen will be a ECU Id number, that's the way to identify your exact model.

[IronMonkeyL255]

Yeah. Mine is showing 7136XX

[log1call]

Ok, and do you have the data addresses for that ecu written into the evoscan's data file? So that the 7136XX ecu is an option?

If not then you need to go back and read through this thread to where I sent a modified data file in... it's being hosted by Josh somewhere or if you can't find it I will mail it to you.

[Legacy777]

There isn't? What about all the threads I read about stock injectors getting maxed out, running static, not in their 'ideal' 80% duty range? I can't find anywhere that states how people calculate that. For example, Revscan logs the raw IPW (where I see readings of 140-150) and EvoScan uses RAW * 0.128 (ok, I pulled that from b10 source code) which apparently is the milliseconds (which I see 18-19ms) all during WOT. Idling the RAW is around 20 and 2ms.

To me, it seems like the numbers are backwards. During WOT when more fuel is needed, the pulse width would need to be more frequent... hmm, so my 18ms reads mean what exactly? It's firing 18 out of every XX? milliseconds? There must be a way to calculate how much fuel is flowing with that injector pulse width to see if its close to the max capable.

See that's the thing, those ms numbers are just the amount of time the injector is staying open. To calculate duty cycle, you need to also look at engine rpm. This calculator is pretty nice.

http://www.stealth316.com/2-calc-idc.htm

At least in Revscan, it appears to be some sort of target idle speed -- because my the other parameter for rpms still doesn't even come close to it. Furthermore, revscan was logging idle speed numbers, which are multiples of 25 -- kind of odd for something to be a sensor reading. Wonder where they pulled it from...

Don't have a clue....

Alright, but what do the numbers actually mean? During WOT, I see the fuel trim short % often at 0. Other times (can't really find a pattern), I see the number fluctuate -20 to 5%.

It really doesn't mean too much. I would use it as a way to tell if you're running consistently lean or rich. if you see your fuel trim in the negative a lot, it means you're running very rich, and the ECU is trying to dial the fuel back. This pretty much only happens in closed loop to my knowledge.

What are ideal/acceptable values for this? I'm seeing ~25 when WOT and ~50 when idle.

50 seems a little high. FSM states 9-21 deg when idling I'm not sure how you can set these though. ECU is controlling this.

Clamping does what to other sensors? What does the ECU do then? Must be a reason why Revtronix is talking about a 3bar MAP sensor for higher boost levels...

It depends. I don't know the ECU code, and what the ECU looks for to adjust fuel & timing, etc. Revtronix probably wants to add a 3 bar MAP sensor so they can provide a raised fuel cut limit. To do this, you need a MAP sensor that can read higher boost levels accurately.

RevScan never logged any, yet EvoScan did almost always during WOT. Not sure who to believe or how to interpret the results. I know the parameter is only 255/FF big -- so a value of 255 would be interpreted as knock detected, don't do anything. A value of 250 would be knock detected, retard timing by 5 degrees (255-250) ?

I probably wouldn't be too worried about this. What you are seeing on the logging software may or may not be what is actually going on. These ECU's stop transmitting or don't transmit as well during high rpms. Or at least that's what Vikash had mentioned.

[impreza_GC8]

Where is the plug I connect my cable to? I only see the two plugs (green and gray) that are for reading trouble codes. Thanks.

[log1call]

Most are under the dash. Some, I have heard, are behind the kick panel. It depends what model and which market you are in.

[mlitz123]

ok fuel trims are a funny thing... simple but confusing let me see if i can explain them...

is there a long term fuel trim? i dont think there should be so i will give the short answer

short fuel trims are basicly readings strait from the o2 sensor, the closer to 0 the better, +% is adding fuel -% is subtracting fuel. this PID (oh sorry PID is Parameter Identification) is only useful if you are trying to diag a problem, like vacuum leak or a rich condition. so dont worrie about it to much unless it goes waaaaay out of spec (+/- 20 at idle even 10 is bad)

it gets more complicated when you add long term fuel trims in there... if you want me to explain it i will

[BSOD2600]

I found out some of the parameters that RevScan logs, which b10scan/EvoScan does not have, are easily accessible . Some things of possible interest: Different coolant temp source, Spark Learn factors, Idle target speed, Fuel trim (long/short term), MAF Airflow!, and knock retard (short/long term). Hopefully do more logging next week once I get the additional o-rings for the 440's and installed. Few more days after that, install the JECS and Revtronix chip Very Happy.

I'll post my new data.xml for SSMI 7236XX in another week or two, once I validate these additional parameters work and the conversion functions.