Subaru Outback Forums banner

61 - 80 of 1047 Posts

·
Registered
Joined
·
1,072 Posts
Discussion Starter #61
According to the manual if the vehicle driven on stands, systems will not operate the same unless some brake drag is applied.
It can tell there is no load on stands vs. driven on-road.
This may have something to do with Franskers observations.
Makes sense to me, the ECU knows how much load is on the engine.


More tests for confirmation? :29::D
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #62
Made some edits to Page 4 post.

My key was a bit off for the adjustments, double checked that.
Every 1% adjustment = 1% duty cycle change.

I also revisited and retested the dynamic control values.
Those have been slightly adjusted to better reflect real world values.

Found a slippery hill on the property to accelerate up, was monitoring transfer duty cycle with FreeSSM.
The data stream is much more responsive than I would have expected. :29:
It recorded a max value of 95% for the first bit of momentum, then drops off as I mentioned before.
Also verified this same value with my DVOM.
The flat ground acceleration wasn't pushing the AWD system hard enough it seems.
I did notice and hear the front wheels spin for a moment (.25-.50 of a second maybe) before the rears caught up.
All in all the system works very well.

I still like my manual adjustment though :D
There is some difference from the standing start auto mode vs. manual 100% with the knob.
Gives the vehicle more of an initial push, since there is no delay.
Actually heard the rears spinning before the fronts.
Being able to hold the 50/50 is nice for:
1: Crawling up a hill/obstacle at whatever throttle input desired
2: Increased speed has no affect on torque distribution

On road, the typical driver will never notice a problem with the system, it simply works great.
I plan to do a lot more offroading and having crazy fun in the snow, so the benefits will be seen here.
 

·
Registered
Joined
·
2,105 Posts
In http://www.subaruoutback.org/forums/65-parts-accessories-performance/39426-freessm-complete-access-your-ecm-tcu.html#post367598, even when decelerating (as well as cruising) the duty cycle appears to remain significant.

As far as I can recall of threads I've read, with the car raised both front and rear drives were functioning (turning) when in gear, even at idle.

What is strange in that case is the rear VSS signal would have been at or close to zero, and the difference to the front VSS signal, one might think, should have triggered an increase in clutch activation (in the 2002, a reduction in the duty cycle).

Was there any further discussion in that thread about the rear wheels not turning? (Link)
There wasn't any further discussion about it. I was asking him about his experience with the disabled transfer clutch in 5th gear and he probably just thought I'd find it interesting.
 

·
Registered
Joined
·
2,105 Posts
According to the manual if the vehicle driven on stands, systems will not operate the same unless some brake drag is applied.
It can tell there is no load on stands vs. driven on-road.
This may have something to do with Franskers observations.
Makes sense to me, the ECU knows how much load is on the engine.


More tests for confirmation? :29::D
That explains it. I don't know if it was his car, but I'll see if he'll try it again.
 

·
Registered
Joined
·
16,408 Posts
On road, the typical driver will never notice a problem with the system, it simply works great.
Exactly. So why do so many here worry about 90/10, 80/20. 60/40 ad infinitem?

This is what I appreciate so much about where this thread is going. It is providing an unprecedented understanding about how the Subaru AWD works and, at the same time, is dispelling myths, such as "using 1 or 2 puts the car in full 50/50 AWD" or "it's normally 90/10 but increases . . . ." without the findings leading to anything other than it's a really effective system, when it's working properly.

As for the raised wheels situation, with the low engine load I can understand the transfer solenoid duty cycle being low. However, if the rear was not being driven (turned) to any extent, it would mean that there's a more complex relationship between throttle and differences in the VSS signals. It would suggest that regardless of the VSS difference, if the throttle/engine load is low, the duty cycle won't be advanced.

If that's the case, a question remaining is: What's the duty cycle at that stage? It would have to be low enough so that clutch friction doesn't overcome rear drive train friction. This might explain why all four wheels turn on some cars when raised -- their rear drives might have less drag, and so are able to turn even with the low and inhibited clutch engagement. More measurements? Perhaps.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #66
Because there are many people who don't understand how a basic vehicle works, not to mention something somewhat complicated system like this.
"Expectations" of vehicle performance come down to here say, rumors and marketing techniques, all of which I try to ignore at all costs.
Getting to the bottom of exactly how everything works is one thing I do best.
I say provide real unbiased data, and let people make the decision themselves.

Yeah I will have to throw it up on some stands and see what happens while monitoring the duty cycle.
I have a hard time believing the rear wheels wont turn even with 0% duty cycle.
The clutch plates are in very close contact, so I would expect at the least a fluid coupling effect.
I will be able to test automatic mode and the entire range with the dial 0-95%.
With and without brakes.

It is almost time to start up a .xlsx spreadsheet....
 

·
Registered
Joined
·
16,408 Posts
I have a hard time believing the rear wheels wont turn even with 0% duty cycle.
The clutch plates are in very close contact, so I would expect at the least a fluid coupling effect.
Good point. Then again, an older car with a stiff propeller shaft U-joint, or old gear oil in the differential, or a brake dragging ever so little, could be enough to overcome whatever (little) torque can be transmitted through the "disengaged" clutch. It might also depend on the engine speed and transmission gear which will determine the clutch's drive plate rotation speed.

Possible test: raise the car, engine running, tranny in N. Manually turn one front wheel while the other is held from turning or turn both front wheels in unison (in the same direction). Check if the rear propeller shaft turns.
 

·
Registered
Joined
·
16,408 Posts
Up and running

Cable arrived today. (For those north of the border, the US seller sent it in a 10 x 7 inch bubble pack envelope, with appropriate Customs Declaration on the label. Based on the USPS Office Date Stamp, it was delivered by our Canada Post carrier exactly seven days later -- far better than I expected.)

My experience so far . . .

I hooked up the USB cable to the laptop -- it recognized the cable, and the FT232R USB UART, but didn't have a driver (Windows 7 Starter). I opted to let Windows find a driver, and it did (via Internet). (Incidentally, before the system has a driver installed, the cable showed up as "Other" (with the FT232 etc. description) in the "Device Manager" list. Once the driver is in and connected, the cable shows up, as ntippet noted, under "Ports (COM& LPT)" as "USB Serial Port", in my case, followed by "(PORT3)".

Turned on FreeSSM, and went to preferences. It had already been set to Serial Port COM3.

I then connected the OBD cable, turned the ignition on, and selected "Test Diagnostic Interface". It failed. I tried again, and it failed again. I shut off the ignition and closed FreeSSM (which apparently got hung up and stopped responding. It took some time for it to close down).

The next time, I opened FreeSSM, clicked on Test Diagnostic Interface, and a window popped up asking to connect the OBD connector and turn the ignition to On, which I did. Then I clicked "Start" in the little window, and the test was completed successfully.

FreeSSM logged into the ECM and TCM without a hitch, and found no codes. (It also listed the Cruise Control, and Inhibitor, with a green check mark, at the left side.)

"Adjustments" and "System Tests" etc. all came up as in ntippet's screen shots. However, when I went to ECM "Measuring Blocks", I got the main screen (grid where Blocks are to be listed) but no list of blocks to select. Took a few moments of thought and I clicked on the green + sign at the buttom -- up popped the list. Same for the TCM. (Not exactly the same list of blocks as ntippet's; for example, instead of "throttle position voltage", it had "pedal position" in the TCM list, but it's for my 07 with the fly-by-wire throttle etc.)

So far so good. I didn't start the engine but selected a few TCM Measuring Blocks, such as "Pedal Position %", Line and AWD duty cycle, wheel speed etc. (When more than one is selected, the up/down buttons at the bottom come on, and a selected block can be moved up and down the list.)

With the gear in P or N, the AWD duty cycle was fixed at 5%, but in D or R was at 30%. This did not change when varying the Pedal Position %, although the Line duty cycle reading did respond. Not sure why the AWD % didn't change, perhaps because the engine was off.

The Measuring Blocks grid can be set to show a combination of min value, current value and max value. The min and max columns are selected by toggling the left two (of three) buttons at the center bottom. I didn't figure out what the third button does -- looks like max and min, but it didn't seem to do anything when I clicked on it -- anyone can confirm what it does?

To get the Measuring Blocks to read, the "Start" button at the lower left is used (same to stop). Till then, there's just blanks in the grid.

The refresh rate seems to be quite fast -- I found the Pedal Position % reading changing even when I thought I was holding the pedal steady, and it changed almost "instanteously" when I purposely moved the pedal.

I'll get back to it (and this thread) in a short while, when I can take the car out on the road. In the meantime I wanted to report on the cable's arrival and the setup.

Once again, thanks to ntippet and Seabass for the thread and support.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #69
Glad to hear its working. :29:
Its nice when something just works out of the box. (or bubble wrap, hah!)

Don't know what the 3rd button does either on the measuring blocks page.
The data measured does come in very fast, a good thing.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #71
Has anyone tried this software on Gen-4 Outback?
No one has chimed in yet...
Software documentation lists support up to 2010.

"The software supports the models Legacy®/Liberty®/Outback®/Baja®, Impreza® (incl. WRX + STi), Forester®, Exiga® and Tribeca® starting with model year 1999 up to model year 2009.
Model year 2010 is already supported, too, but some Measuring Blocks may be missing and some Diagnostic Codes may not be displayed as plaintext."
 

·
Registered
Joined
·
16,408 Posts
Had a netbook in the car today with FreeSM on it.

I was tracking a few transmission parameters, including accelerator pedal position and AWD transfer duty cycle, in order to get a better idea of the relationships with my 07 4EAT, but I was stymied by the small text. No way to see it in a quick glance at the screen.

I tried the Windows 7 "magnifier"; it does a good job of increasing size and visibility from a distance, but it splits the screen horizontally, limiting the number of rows of data (i.e., different parameters) that can be displayed in magnified form at the same time. To see the other rows of data, the scroll bar on the right can be used, but that's not something that will work easily while driving. (This might be due to the netbook screen size, which is rather small vertically.)

I'd like to find a way to simply increase the display size, as I can with other programs using the "zoom" level. This, of course, might cause some parts of the page to go off-screen, but that's okay -- I can move the page to display only the data fields (rows and columns) at the right, and recognize what they are from their order down the page and units (%, V, etc). But I couldn't find a way to do it.

I also noticed that FreeSSM doesn't retain Measuring Block parameters. If the current Measuring Block page is closed (e.g. to go from Engine measuring blocks to Transmission measuring blocks), or the key is turned to ACC or OFF, the list that was being monitored is lost, and has to be reselected again when going back.

Without any way to record the readings, at least so far, I'm thinking of setting up a camera and videoing the screen while driving. Then perhaps I can play back the video and plot the data. Sort of complicated, and will have to see if this will work.

These are really minor issues, especially in relations to what FreeSSM does provide, including ATF temperature, which some other scanners discussed in these forums apparently don't access. Also, the refresh rate (reported in the upper right of the measuring block grid) was consistently less than 0.1 seconds -- pretty good.

Some data observations (from memory). The transfer duty cycle was around 40% in R and D with no throttle (brake applied). There was no change when using the "manual" Sport selection in 1 or 2. When driving the duty cycle peaked (measuring block max reading) at almost 70% but I don't know when, although the throttle max reading for the same time period was less than 45%. (Never went near WOT.)
 

·
Registered
Joined
·
2,105 Posts
There are screen capture recorders that could make it a bit easier. Maybe there's a decent free one kut there.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #74
Hey now you can see how difficult it was for me to monitor parameters while driving!
Fun isn't it?
Terribly dangerous while on the road, I suggest having a passenger do the reading.
My DVOM screen is much much easier to read than in FreeSSM.
I could watch the road and take readings without squinting at the screen trying to figure out what I was looking at.

I was also thinking about using some screen capture software, then playing back over time and manually copying over the data to .xlsx and graphing it....
Downloaded CamStudio but haven't put it to action yet.
 

·
Registered
Joined
·
16,408 Posts
Hey now you can see how difficult it was for me to monitor parameters while driving! Fun isn't it?
And the screenshots above are so bright, clear and stable! But yes, I do appreciate the real life difficulty.

The video idea would certainly involve another person, either driving while I watched, or just operating the camera. I had no intention of trying to do it all. Today was just a "let's see what it (FreeSSM) does" trial.

The goal of being able to plot sets of readings might be another challenge (and another learning experience), especially with the numbers refreshing every 1/10-second. If I have this correct, a 15 minute test drive involving five parameters would be a 5 row by 9000 column spreadsheet (raw data). Using a video recording would require not only transcribing all the numbers to a spreadsheet, but also carefully advancing the video at small time increments, even if it's at one or ten second intervals, in order to read the numbers, provided their all clearly visible.

No question, FreeSSM is a quantum leap forward in measurement capability, at least for me, and will serve well in many instances. But it doesn't make studying parameters and their relationships a simple function.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #76
The data doesn't bother me.
I analyze 100MB+ raw files from data recorders I have installed for my job.
Well over 100,000 rows x 10+ parameters.

Having a way to record it from FreeSSM would be excellent, I might have to look into that somehow.
The video recording might work for a few seconds of data, but anything beyond that is far too tedious.
 

·
Registered
Joined
·
2,105 Posts
At least freessm is opensource. There has to be a way to implement some sort of basic logging to a spreadsheet. I don't mind sorting through huge amounts of data, but having to enter them into a spreadsheet is just nutso.
 

·
Registered
Joined
·
16,408 Posts
The video recording might work for a few seconds of data, but anything beyond that is far too tedious.
. . . but having to enter them into a spreadsheet is just nutso.
Indeed.

Having a way to record it from FreeSSM would be excellent, I might have to look into that somehow.
Had a look at the FreeSSM discussion forum (BerliOS Developer: Open Discussion). There's a question about logging the measuring blocks, and the answer, posted 2010-Jan-10 is

". . . curve-plots and data-logging are two of the major features which are planned for future versions. Work is already in progress (but will need some time)."

It's been almost two years (version 1.2.5 appears to have been released 2010-May-26, about a year after the previous version 1.1.1); perhaps we will see the next version with logging in the not-too-distant future.

I wonder if a question posted to the FreeSSM discussion forum would provide a better idea of the status of the logging project. (Seabass -- there's a question about ATF temperature data in the forum -- was that you?)
 

·
Registered
Joined
·
2,105 Posts
Yes. I was trying to find out whether there was any correlation between how freessm retrieved info and how a programmable obd2 scanner did it. Trying to find a code to access the atf temp sensor with a scan tool.
 

·
Registered
Joined
·
1,072 Posts
Discussion Starter #80
Doesn't the 4eat have 1 temp sensor and the 5eat has 2 or more?
 
61 - 80 of 1047 Posts
Top