I actually had a recent interesting experience with an older HP Jornada 720 PDA, its the one that looks like a tiny computer.

I installed something called JLIME which is a linux OS to replace the windows CE on the unit as there is no longer any form of upgrade path for the unit. Anyway, the device has both a CF slot and a PCMCIA slot, and I had a USB to CF adapter card that I plugged in and it detected right away. When I plugged the Tactrix cable into it, It detected it as an FTDI device.

Here's a few little sniggles I have come across... Logging works, but its unlikely you'd be able to flash as part of the flashing process uses power from the PC and adds it to the 12v from the OBDII port, this voltage is applied to the boot pin on the ECU (the green wire to the white connector) The added voltage should be about 17 volts) Unfortunately its not, because the USB interface is passive and provides very little voltage, which means that unless the battery is charged to over 13 volts, the voltage may never be enough to trigger bootstrap.
(FWIW this is also the likely reason some laptops and combinations won't work to flash someones ECU, while others will)

This may be a problem with reflashing on any PDA, so be aware of it.

 

Jack,

Crap, that is a major hurdle to flashing. What wiring a small battery (9 volt?) into that boot pin with a switch on it?

donour,

Colby won't let me see the source despite repeated requests, so unless he is motivated to redesign and compile for mobile devices, I don't see it happening unless someone else steps in.

 

The most reliable way is a DC to DC converter in the cable, but depending on the current required they can be bulky (although I expect the current should be small).

So the cables would need to be altered or redone. Since the boot voltage facility is the main addition from a normal KKL cable then it isn't a bar to getting the Pocket PC working, just not the ideal of simply plugging in the Tactrix cable to your USB host device perhaps.

With realtime facilities I don't think full reflash capability will be such an issue. It is done and it works fine, just needs rolling out to lots of different ECUs.

 

Thanks for the info. You got PM.

 

That's weird. I had to be persistent, but I've had the source for at least at year. I have 2 or 3 different snapshots, not really the most recent, but close. Whenever folks ask me for it, I send them back to colby. However, the tarballs I have are licensed with the GPL so redistribution is legally at my discretion.

That's how I did the OS X native build last fall.

As for your PDA app, it looks pretty rad. I'm willing to help with whatever you need as well as build an analysis program. I can also host the dev repository if you want (either on sourceforge or here at my university research group).

d

 

Although I havent tried this, an Active USB hub with a proper power supply should "fix" the problem as it provides enough power to the ports. In some cases the voltage is too low to even initialize the tactrix cable at all, this could explain some people's problems coding on PDA's that have USB Host controllers.

 

That's actually a good idea about the hub, Jack. Even in a car, you could have a powered hub. Maybe even battery powered.

Hmm, I wonder if that's my problem with using devices that are supposed to have a host controller but won't recognize the cable. I will retry the Sprint phone with a hub...