On Friday 06 Feb 2015 19:04:45 Ralph Corderoy wrote: > All looks fine. The USB device uses a common Scottish UART-to-USB > interface chip. https://en.wikipedia.org/wiki/FTDI
I had also assumed that. > I just take this to mean a couple of things have a poke of the new > serial device to see if it's something that understand. Neither do. > `lsusb' should list the `Bus 005 Device 009: ID 0403:6001' device. Yes that's what I thought too eventually. lsusb says: Bus 005 Device 003: ID 0403:6001 Future Technology Devices International, Ltd FT232 USB-Serial (UART) IC When the device is plugged in. > Have you tried Googling for your efergy's device name and `linux' to see > what progress others have made? I think the normal route they take, if > this kind of interface is undocumented, is run it under Windows and > snoop on the USB traffic. I found > http://jdesbonnet.blogspot.co.uk/2010/09/smart-electricity-meter-based-on-ef > ergy.html which looks interesting, though probably not directly relevant. You are right; it is interesting, but not at all relevant :-) The annoying thing is that this used to work. It's always been a faf to re-establish the USB serial connections to my meter (and Satnav) after an upgrade, but it has always worked in the past. This time it has defeated me. As suggested I searched for the Vendor and Product IDs on Google and found a few things. It would seem that several things have changed if I understand the sources correctly. At http://www.picaxeforum.co.uk/archive/index.php/t-25488.html[1] I discovered that after kernel 3.12, the handling of Serial USBs changed. This and all of the other sites I visited appear to be advocating a number of solutions to fix this; scripts, using modprobe to add the ftdi_sio module, changes to the rules in udev to add kernel=ttyUSB* or all three. I'm a bit uncomfortable with this because these fixes are all related to devices other than my efergy power meter. Since what we are really talking about is support for the serial chip inside the meter which seems to be the common factor it should make no difference; even so I am reluctant to b*****r around with something I don't understand (the kernel. not the chip). I then had another idea. I installed a Serial Comm program (mocom) and was able to connect and disconnect from the device without error. Of course I currently have no idea what to send to it to get a response and it doesn't seem to send an error if I send a random string, so I can only assume that it's working. So it could be that my machine is correctly identifying the device and connecting to it and the problem lies with VirtualBox, as I originally thought. Any further ideas? -- Terry Coles -------- [1] http://www.picaxeforum.co.uk/archive/index.php/t-25488.html -- Next meeting: Bournemouth, Tuesday, 2015-03-03 20:00 Meets, Mailing list, IRC, LinkedIn, ... http://dorset.lug.org.uk/ New thread on mailing list: mailto:[email protected] How to Report Bugs Effectively: http://goo.gl/4Xue
