pcmhacking.net

Tazzi wrote:Really does make me worry about how they compensate using the ELM for instances where the ECU fires off a "7F 78" response when its being slow. I bet they probably just put a default delay between messages when receiving it since realistically there is no other way. Then they run the risk of missing the next response ect. Absolute nuttas using an ELM for flashing!

I imagine its all single thread blocking with hard coded delays which makes me shudder. Probably works 95% of the time though.

In other news..

my newly rewired test bench has two fully functional obd2 ports and I thought... while I'm here why not see if I can do a read with the VXDIAG VCX send set FEPS manually.

It started reading fine, but then after a while it suddenly failed the security unlock again.. in mid read.. very odd. I left FEPS on and at 14.6 volts for the whole read.

1:03:27 PM Loaded J2534 Library Via: VXDIAG
1:03:27 PM Successfully created OBD2 connection
1:03:27 PM Unlocking Controller...
1:03:27 PM Enabling FEPS
1:03:27 PM User requested to cycle the ignition
1:03:46 PM User cycled the ignition
1:03:48 PM Starting TESTER_PRESENT message pump
1:03:48 PM Unlocked Controller
Updated Flash Info (used for next read or write)
PCMType: Black Oak 1472K
Strategy: HAAO7
Year: 2004
Vin: 6FPAAAJGCM4T9708
1:03:48 PM Reading Flash from Bank 1
1:03:48 PM Read Address: 0x0 Size: 0x800
1:03:49 PM Read Address: 0x800 Size: 0x800
SKIP A BUNCH.....
1:04:29 PM Read Address: 0x1F000 Size: 0x800
1:04:29 PM Read Address: 0x1F800 Size: 0x800
1:04:30 PM Read Address: 0x20000 Size: 0x800
1:04:31 PM Read Address: 0x20800 Size: 0x800
1:04:31 PM Error: Sent Cmd:0 Recieved: CONDITIONS_NOT_CORRECT
This error is generally raised when trying to enter a security mode and the ignition was not cycled with the correct timing
1:05:15 PM Stopping TESTER_PRESENT message pump
1:05:15 PM Disconnect

Oh, and if you want me to sniff anything else.. like module programming in IDS or whatever. Let me know.

14.6v might be borderline. You could test by disconnecting FEPS on purpose and if it immediately triggers CONDITIONS_NOT_CORRECT

MeZmeriZe wrote:Oh, and if you want me to sniff anything else.. like module programming in IDS or whatever. Let me know.

It is ok I don't plan to add anymore features to the tool other than possibly adding some more read/write methods (eg green ok etc) other than that just bug fixes now as I want to concentrate on the editor.

rolls wrote:14.6v might be borderline. You could test by disconnecting FEPS on purpose and if it immediately triggers CONDITIONS_NOT_CORRECT

Iv tested from smack bang 12v to 15v.. no dramas with constant supply.

Also have a whirl with varying it between 12v and 15v while reading, seems fine.

Wasnt till the 8v mark where it wasnt too happy.

*Edit
Thats power to ECU and VCM as its all tied together.. so could be either one not being a happy camper at those kinda voltages.

Tazzi wrote:

rolls wrote:14.6v might be borderline. You could test by disconnecting FEPS on purpose and if it immediately triggers CONDITIONS_NOT_CORRECT

Iv tested from smack bang 12v to 15v.. no dramas with constant supply.

Also have a whirl with varying it between 12v and 15v while reading, seems fine.

Wasnt till the 8v mark where it wasnt too happy.

*Edit
Thats power to ECU and VCM as its all tied together.. so could be either one not being a happy camper at those kinda voltages.

What are you using as a power supply Mr Tazzi?

MeZmeriZe wrote: What are you using as a power supply Mr Tazzi?

Using this little mungral: https://www.jaycar.com.au/compact-switc ... y/p/MP3800

Iv got a multimeter hooked up to be able to accurately monitor the voltage.

Tazzi wrote:

MeZmeriZe wrote: What are you using as a power supply Mr Tazzi?

Using this little mungral: https://www.jaycar.com.au/compact-switc ... y/p/MP3800

Iv got a multimeter hooked up to be able to accurately monitor the voltage.

nice.. I just noticed that my big old lab powersupply is outputting fairly low current.. especially going from the giant heatsink on the back I hadn't expect that but perhaps I'm using the wrong outputs.. it has about a dozen of the damn things.

not been an issue because I put a 12v UPS battery in parallel anyway to provide extra kick when needed but still.

I have a piece of shit old charger I hacked up, it is meant to be 12v but its dropped to about 10.8 now, everything still seems to work though. Will need to buy a decent bench supply at some point.

rolls wrote:14.6v might be borderline. You could test by disconnecting FEPS on purpose and if it immediately triggers CONDITIONS_NOT_CORRECT

MeZmeriZe wrote:Oh, and if you want me to sniff anything else.. like module programming in IDS or whatever. Let me know.

It is ok I don't plan to add anymore features to the tool other than possibly adding some more read/write methods (eg green ok etc) other than that just bug fixes now as I want to concentrate on the editor.

Well the biz school at the uni where I work just had a sell off of academic laptops and I just picked up a really nice quad core HP with 8gig of ram and a 250gig SSD for 100 bucks. so I have free drive space again. I have already installed VMware on it, so soon I'll be able to tell you if having hardware shared though a VM makes any difference to the timing.

TBH I prefer the way redhat KVM works as a hypervisor.. you can actually share the hardware device exclusively with a VM.. and the parent OS need not even have a driver available. We have licensing dongles for academic software that only works with the KVM hardware passthough because KVM passes though the USB port, not the device plugged into it.. so it works in places where HyperV and sometime Vmware too, won't. anyway, off topic.. will let you know if it makes any difference.. I also plan to test the VCX again on a clean windows 7 laptop with no IDS or anything installed.. plus this one is 64 bit instead of 32 so who knows.. it might be that FEPS works on this one.. I can hope right? If it doesn't I might well get an openport anyway, just to have a fully functional up to date J2534. The VCX guys either didn't believe me or don't care, but I'll send them a video of it not working if it doesn't work on this laptop either.. Other than FEPS its a really nice unit that is compact, works great with IDS (except presumably for anything that writes.)


i did have one thing I wanted to discuss.. On of the things I thought about with my wideband, and something that is a limitation of the open port too.. is with datalogging.. presumably openport can only log params that are on one of the busses it can read... with other solutions you can use a laptop or whatever to get other info.. but that wastes the benefit of the openports SDcard slot.

I had an idea, to basically turn my wideband signal into a canbus de vice.. have it respond on the bus and be logable as such.. the question is.. how would you go about addressing it? at first I thought of using one of fords default PIDs for widebands. (which i have written here somewhere.) but as I understand it, anything requesting that pid from a PCM that doesn't support it would get a not supported message. So I guess my question is.. can the PCM be setup to accept the output of the wideband and then pass it though as the wideband? clearly it would as things like this already happen with the speed from ABS etc.. but I'm guessing you'd need pretty low level access to the RTOS code to actually impliment it.

what other alternatives would there be? The BA and BF's both have J1850, but not to the PCM..
BA PCM plug schematic: http://www.fordforums.com.au/vbportal/v ... icleid=686
BF PCM plug schematic: http://www.fordforums.com.au/vbportal/v ... icleid=687
(you probably need a ford forums login to see those)

but that might not necessarily matter. If an arduino similiar could be tasked with providing a bus interface you could basically put anything on the cars bus and log it with whatever you want.. would be handly as it makes things like my 100 dollar wideband more valuable because without the context of all the other logging parameters, a wideband is just a readout on your pillar.

Anyone have any ideas how that might best be achieved? can would obviously be the best option.. so the question becomes.. how best to address it so normal datalogging kit can read it. Perhaps that is another software project for me to look into. (teaching myself Visual studio and C#, so I'm going to make a network interface to J2534 driver. so any language that can connect to a localhost TCP port can easily run with J2534 so perhaps I'll add create your own canbus devices and datalogging to the list)