Landis+Gyr E470 iec1107/62056-21 protocol

I have a British Gas installed Landis+Gyr E470 smart meter. This meter has a IEC 62056-21 optical interface. I am experimenting with an Onmina IR probe connected to Raspberry pi GPIO serial port. I have hacked up a bit of python code to send the IEC sign on command /?! followed by CR LF.

I would expect to see the meter respond with an identification code, but I get nothing back.

If I loop back the TX and RX pins at the GPIO I see my command reflected, so the Pi end is working. If I look at the probe through my phone camera I can see the IR TX led flashing on the probe. So either the meter is not responding or the RX side of the probe is fried.

Has anyone managed to get anything back from one of these meters?

I was thinking of using lirc to test the RX side of the probe.

Ideas anyone?

Do these meters need a meter ID in the sign on?

Roger

Could you use any remote control to check the receiving side for signs of life? Is it being saturated by daylight?

I don’t know anything about that meter, so I can’t help any more.

Hi Robert,

I connected oscilloscope to the receive side of the sensor. I can see a
bit of crosstalk coming back when the transmit side is active. But
nothing when I just shine a tv remote into it. However if I connect up
the send side and run my test program which sends the init command every
few seconds and I press the remote I see something else going on, weird!
I know that the remote is sending irda frames not ordinary asynchronous
ascii characters, and is working at a different IR frequency than the
IEC1107 sensor. So this is probably some random side effect.

Or of course it also means the rx side is fried. I did initially connect
the rx and tx the wrong way round due to poor documentation. So 3.3v
being sent back up the incoming data line might have done for it :slight_smile:

Roger

I can’t identify your probe for sure, so I can’t really help. But if there’s absolutely nothing coming back, and you fed 3.3 V into (presumably) an op.amp output, then I suspect the worst.

Robert,

These probes are sold on ebay
http://www.ebay.co.uk/itm/Optical-Probe-IEC1107-IEC61107-62056-21/201996001267?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649
as slight seconds for £25 by Omnima Limited http://www.omnima.co.uk/.
They are based in Oxford. The probes are not mentioned on their web site.

They do a usb version and a “linux” version with a little pcb with a 4
way socket header. The markings on the pcb are misleading, the line
marked TX on the header is actually connected to IR transmitter and the
line marked RX is connected to the IR receiver. Because this pcb was on
the other end of a meter of cable from the actual sensor I initially
assumed that the markings were relative to the cable end not the
device. But I agree with you I have probably fried the receive side of
the sensor.

That said, I still think these are good value for money seeing as I can
see various other companies selling versions which look identical for
£150+ and probably came from the same Chinese factory. I would recommend
paying the extra 4 quid for the USB version and avoiding the mistake I made.

Roger

For information. Here is an official reply I received (finally) received from British Gas about the optical probe port on the Landis + Gyr E470 Zigbee meter that is being installed by them as part of the smart meter rollout.

For reasons of security the optical probe system is closed after the point of manufacture. It doesn’t pass readings via the internal mechanism and therefore no information can be gained from it. The optical port can’t be opened remotely so unfortunately this is not an option. Also each meter manufacturer creates their own internal software and are unlikely to make it accessible to the general population. We know that Landis and Gyr won’t allow their version of software out and British Gas don’t have access to it.