PDA

View Full Version : Problem Flashing VY Ute



johnv
January 20th, 2005, 09:22 PM
Just been playing around tuning a mates VY ute, read PCM no problem,changed desired calibrations then licienced the PCM and did a test upload without any problems in 30 seconds.

went ahead and uploaded new program and it wouldn't load leaving PCM in a zombie state,tryed again and same result 10 minutes and still zerrow %. :(
Thought I was going to have to hook up the bench harness , but decided to uncheck the high speed transfer box and try again.
This time started uploading imediatly and all done in 3 mins.

Any reason for this happening ?

Haven't tryed flashing my own car yet - it is a problem car to flash with Edit via the OBDII port - killed 2 PCMs and only tryed twice :evil: so always have to use bench harness which is a PITA.

cheers
John

Delco
January 20th, 2005, 09:37 PM
Just been playing around tuning a mates VY ute, read PCM no problem,changed desired calibrations then licienced the PCM and did a test upload without any problems in 30 seconds.

went ahead and uploaded new program and it wouldn't load leaving PCM in a zombie state,tryed again and same result 10 minutes and still zerrow %. :(
Thought I was going to have to hook up the bench harness , but decided to uncheck the high speed transfer box and try again.
This time started uploading imediatly and all done in 3 mins.

Any reason for this happening ?

Haven't tryed flashing my own car yet - it is a problem car to flash with Edit via the OBDII port - killed 2 PCMs and only tryed twice :evil: so always have to use bench harness which is a PITA.

cheers
John

John dont stress , you wont fry the pcm with flashscan ( I have tried to do everything in beta testing to kill one and havent been able to yet )

The issue with some of the utes is known and a fix is on the way.
in the meantime a rough workaround is to turn the key off momentaraly while it is hung then turn it back on , the noise on the bus will stop and then flashscan can continue.

Even if the pcm gets stuck in limbo as you found just power of then reprogram and it will reover the PCM no problems

note: this is not a efilive endorsed practice but I have used it sucessfully on over 300 flashes without a issue ( bench harness also works if you have one)

johnv
January 20th, 2005, 09:47 PM
Thanks Delco I'll give that a go next time.

Nice to know that flashscan can recover the pcm,
if that was edit the pcm would be winging its way to Ross :lol:

cheers
John

Blacky
January 20th, 2005, 09:56 PM
Thanks Delco I'll give that a go next time.

Nice to know that flashscan can recover the pcm,
if that was edit the pcm would be winging its way to Ross :lol:

cheers
John

Actually EFILive will recover most PCMs that have been "toasted" by edit. If you still have them just hook them up and reflash them with EFILive :)

We should be charging $50 a piece for recovering all those dead PCM's, right Delco? :)

Paul

Delco
January 20th, 2005, 10:13 PM
Thanks Delco I'll give that a go next time.

Nice to know that flashscan can recover the pcm,
if that was edit the pcm would be winging its way to Ross :lol:

cheers
John

Actually EFILive will recover most PCMs that have been "toasted" by edit. If you still have them just hook them up and reflash them with EFILive :)

We should be charging $50 a piece for recovering all those dead PCM's, right Delco? :)

Paul

Yes I have recovered 50 edit crashes PCM's - no there isnt a $2500 check in the mail Paul :lol:

GMPX
January 20th, 2005, 10:55 PM
Here is the real problem......and boring techie explanation, sorry if you get lost.

The data bus used on these cars specifies the 'first or master' node (which is always the PCM) must have a bus loading of 1.5K Ohms.
All other modules on the bus must have a bus loading of 10.5K so as not to load the bus down too much. These are specified values by the chip manufacturers (and what we use in flashscan).

In fact here is an 'official' design note:
"It is recommended that in J1850 VPW systems with less than 26 nodes, one node
should have a load resistance of 1.5K, with all other nodes having the nominal load of 10.6K,
as outlined in SAE STANDARD J1850 – CLASS B DATA COMMUNICATIONS NETWORK INTERFACE."

Now, to demonstrate with photo's off my CRO of the data bus I can show you why on a ute things can go astray.

Red shows the transition time from high to low.
Blue is the approx 'time' the signal is at a valid low point.

First two pohtos......
Data bus with 10.5K loading (as is recommended), but the voltage level in the 'blue' outline is not low for long enough, unlike the 2nd photo.
http://www.efilive.com/upload/public/GMPX/10K5.jpg

Data bus now with (non standard) 1.5K loading.
http://www.efilive.com/upload/public/GMPX/1K5.jpg

Also, look at these next two that demonstrate the amount of 'noise' on the data bus under reflash conditions.

Data bus with 10.5K loading (as is recommended).
http://www.efilive.com/upload/public/GMPX/10K5_Noise.jpg

Data bus now with (non standard) 1.5K loading.
http://www.efilive.com/upload/public/GMPX/1K5_Noise.jpg

So the easy answer would be to make Flashscan a 1.5K bus load, but this is NON standard and 'may' cause problems on some vehicles that have many other modules. (The Commodore only has the PCM and PIM).

There is a fix, and this was even sent out to Holden dealers, and that is to place a resistor across the Data line and GND causing heavier loading (like the pictures above).

What EFILive will soon offer is an inline module that has this installed, just incase you get one of these difficult cars that doesn't want to flash, you can simply plug this in-line, then when you are done it can be returned to the SAE standard bus loading.

For those who can't wait.....head on down to the local electronics parts store, grab a 1.5K resistor (1/4W is fine) and place it at the back of the DLC connector between pins 2 and 5.

Cheers,
Ross

GMPX
January 20th, 2005, 10:58 PM
Nice to know that flashscan can recover the pcm,
if that was edit the pcm would be winging its way to Ross :lol:

John

They still get sent to me John. From memory you got burnt twice didn't you?.
Try the resistor trick with your mates ute, we had some shockers (utes) over here recently that this fix cured 100%.

Cheers,
Ross

Blacky
January 20th, 2005, 11:13 PM
Here is the real problem......and boring techie explanation, sorry if you get lost.
Cheers,
Ross

I'd been pulling my hair out trying to find a "software" solution to the VY utes problem.
I was *impressed* (relieved actually) when Ross first showed me the cause of the problem on the CRO.
Puts everything into perspective. :shock: :shock: :shock:

Paul

GMPX
January 28th, 2005, 11:15 AM
:UPDATE:

We recently made some changes in the software to fix this issue.
One of our Beta testers had a car that simply would not flash at high speed no matter what (a VY).
Since the changes we made that car has been flashed at least 10 times without a problem at high speed (approx 30secs), just a bit more testing and this fix should find it's way into an update or patch early next week.
Oh, there is no hardware changes needed, it's just a software fix.

Cheers,
Ross

Steve Bryant
January 28th, 2005, 01:41 PM
Ross and Paul,

Good detective work! :D

Cheers,

Steve