> I use the CA-42 clone cables, e.g. > > <http://cgi.ebay.com/CA-42-USB-DATA-CABLE-for-NOKIA-6015i-6016i-3120-5140_W0QQitemZ370153338441QQcmdZViewItemQQptZPDA_Accessories?hash=item370153338441&_trksid=p3286.c0.m14&_trkparms=72%3A1234|66%3A2|65%3A12|39%3A1|240%3A1318|301%3A1|293%3A1|294%3A50> > > USD$3.89, with free shipping in the US (I actually get them in Australia > for AUD$10 each). > > They have a Prolific PL2303 chip inside the long blue USB connector. >
Likewise. I am using a CA-42 clone that I purchased from bargaincell.com, via amazon.com for $.01 plus $2.98 shipping. It turned out that the local distributor was within walking distance, in oakland - small world. :-) I don't know about the ca-53 cable. I think that anything with the super-long USB hood (containing a serial UART) can be made to work, though you should check that it's 3.3v ttl and not 5v. A note on that - measuring the supposed 3v3 TTL outputs on both my router and ca-42 cable (unloaded) showed 5v on my digital scope. I haven't bothered to look at it when they're loading each other. Using linux, it's the pl2303 driver. The pl2303 driver that I apparently had installed in my dell windows xp laptop had an issue - it was a 'Tripp Lite' by 'Prolific', old date. When my laptop went into standby, the driver was not properly suspending/resuming, and I would have to perform a hard reset. The prolific reference driver did not work with my clone cable, but the driver distributed by belkin on their website for the Belkin-badged usb-serial device works smashingly well across suspends. On a side note, for anyone doing TTL interfacing - you should assume you're going to be coupling grounds, ie, no isolation. My house does not have ground on the circuit where my linux build machine resides (where I first tried to use the CA-42 cable). My computer's chassis tends to sit at a DC offset of 120V/2 = ~60V due to the switching power supply within. Therefore, anything that the computer interfaces to will tend to have a 60v DC potential. I had tried a tv tuner card in it, and ended up getting a nice shock when touching both the chassis and the (disconnected but with grounded shielding) coax line. What can I say? it's better than licking a ringing phone line. I saw such a large potential between the computer chassis and my asus wl500gpv2 - of course, I didn't realize it at first, but only noticed that I got a lot of garbage in the bit stream, and connecting the usb cable to my linux box would cause the router to flip out and reboot. Hopefully there's no permanent damage - it seems to work fine. I have similar results when I plug my laptop into its charger and try to talk to the router - they're on different circuits. I don't mind running off of battery; I rarely need a serial console. (I once spent some time tracking down a problem with a high bit error rate in a prototype IS-95 CDMA system I helped to develop - it turned out that the power supply for the RF transceiver board was on a different circuit than the data acquisition board in the cPCI chassis, even though the outlets were next to each other - there was a slight dc offset between them that was causing clipping when sampling the baseband output. Shame on us for calling chassis ground signal ground.) -g _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org http://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
