A Voyager 205 revival.. loadfi.exe : loadfi [-i ] [-p ] [-c#] [-a[v][s]] [-f ] [-s#] [-h|/h|/?|--help] -i Image file name, default is "TEImage.bin". -p Parallel port address, default is 378. -c# Loadfi cable version, default cable is 1 (-c1). To run with new loadfi (v2.x) cable, set it to 2 (-c2). -a[v][s] Run loadfi in automatic mode, e.g. -a, -av, -as, -avs. 'v': Verifies the flash after programming. 's': Serializes the flash after programming. -f SNum/MACs input format, e.g. "-f sm1-m2" means sn & mac1 in 1st line, and mac2 in 2nd line, default is "-f m1-s". -s# Serial clock mode (-s1,-s2,-s3,-s4), default is -s4. clock modes: s1 : Inactive Clock Polarity Low s2 : Inactive Clock Polarity High s3 : SPI Mode 0 s4 : SPI Mode 3 (default) -h|/h|/?|--help Displays help message. cables: c1 cable: D5->cs, D4->clk, D3->si, D2->rst, ~S7<-(100ohm)<-so, GND25->gnd. c2 cable: D7->si1,D6->rst1,D5->cs1,D4->clk1,~S7<-(100ohm)<-so1,GND25->gnd1, D3->si2,D2->rst2,D1->cs2,D0->clk2, S6<-(100ohm)<-so2,GND24->gnd2. hmmm. wtf is ~S7 on a Parallel cable? Googling tells me that it's pin 11.. http://www.geocities.com/gear996/sub/parallel.html Thanks m8! Here's what works for some other clones.. JP1 ......... ATMEL AT45DB161 ------------------------------------- 1 ........... 13 [si] 2 ........... 14 [so] 3 ........... Exception: see below (*) 4 ........... 11 [cs] 5 ........... 12 [sck] 6 ........... 7 [gnd] -------------------------------------- (*) Pin 3 of the connector is not wired to the Atmel chip, it is connected to pin 4 of the LPT cable. PC LPT ................. JP1 -------------------------------------- 4 ---------------------- 3 5 ---------------------- 1 6 ---------------------- 5 7 ---------------------- 4 11 ---------------------- 2 (via 100 ohms resistor, 0.25w) 18 ---------------------- 6 -------------------------------------- Firstly, the Voyager 205 has a *seven* pin connector. Bummer. Some simple investigations with my multimeter lead me to conclude that I haven't a clue what I'm doing, but pins 1-6 look likely, pin 3 seems to be indirectly connected to the reset on the Atmel, and pin 7 isn't directly connected to anything, so I follow the instructions as writ.. Instructions.. -------------------------------------- Make the parallel cable. Connect modem to PC with the cable. Boot into pure DOS (like from a floppy boot disk). You'll need to have TEImage.bin on a partition that pure DOS can recognise, i.e. FAT or FAT32. Or else you might try to use ntfsdos to access an NTFS partition from DOS. Alternatively, you might want to try "UserPort", to give an XP DOS Session raw access to the parallel port. Run Loadfi.exe. Apply the TEImage.bin Reboot the router. Well, that didn't work! Something was happening. loadfi would recognise a device, but it was never the same device twice! It would go through the motions when pushed, mostly ending up with "manufacturing cycle complete. replace boards and try again" or words to that effect. reading the flash gets you a file of gibberish, and writing the flash has no effect, though it does produce a pretty display of X's. Interestingly, I had to reverse my 6-pin connector to get it to do anything. Even though my cable is assembled exactly like the chinese example, round the correct way it's always "no connection" from loadfi. This is not encouraging. It looks like the interface will need some tweaking for the Voyager models. Fortunately I have a one-puter-one-bricked-voyager network setup to test all this. If the whole lot goes up in a puff of smoke, nothing is lost. And that's exactly what happened! I made another cable, this time following the wiring diagram in loadfi, which is similar, but not exactly the same as the Chinese example, and plugged that in. Switch on. Nada. The router doesn't even power up. The power pack also seems to be kaput now (putting out a variable 0.5-0.9V instead of 16), and I wonder if it was the cable, coincedence, or both. And now I have no test router. damn! A 13V pack that used to at least make it power up also has no effect. I'll try feeding it some power directly (from my variable power supply) and see if I can get it to come on. Another time!