Computing stuff tied to the physical world

Hot spots on a JeeNode USB

In Hardware on Mar 13, 2013 at 00:01

Recently, a blower door test was done here at JeeLabs, as a way to measure – not guess – how air currents are affecting the house, and what sort of insulation measures would be most effective. This was a fascinating and very extensive test, combined with a thorough scan of the entire house – inside and outside – using a high-end thermal imaging camera.

While the guys from Bergenbos were doing their tests, I asked them to take some close-ups of a JeeNode which had been running RF12demo for a while (i.e. not in sleep mode):



Fantastic resolution, as you can see. No big surprises, but nice pictures nevertheless…

  1. Similar story, but this time our fuse box with an IR cam.

    You can see 3 “hot spots”: 2 of them are relays and the third one is a transformer for the door bell (4W standby!). The 3 digital electric meters are sitting in a row (I guess they also consume >1W each). Not very power efficient…

  2. Would be nice to know if the FTDI-chip will also warm up if it’s not used.

  3. JC, you made me do it…I’ve contributed:

    Thermal Camera kickstarter project

    Now let’s hope they will actually ship around may 2013…

  4. @Peter, the active relay consumption can easily be halved by one of the standard circuits that takes advantage of the holding current requirement being much lower than the “energize me” current. No side-effect and for most countries, no “Licensed Electrician” hurdle, since this is a control path, distinct from the rule-bound downstream power distribution. The simplest circuit uses an N/C contact pair on the relay – most have a spare, low-current contact set, meant for remote signalling of closure.

    Usual Caveat: if you don’t know what you are doing, keep your hands in your pockets when the cupboard door for the distribution board is open and leave the screwdriver in the toolbox.

    For the digital meters, most administrations allow ~0.5W – ~2.0W for the metering logic. These meters typically use an on-line capacitance dropper type supply (it’s all encased, so no safety issues with prying fingers and opto-couple any outputs). In samples that I’ve seen, 0.5W continuous (even under “noload”) is typical – there is a lot to do if you want accurate readings of real and imaginary power. There is an additional loss in the sampling shunt resistor(s) – but that climbs with the load current so remains a tiny percentage of the total energy consumed. Actually a fraction of what is lost in the skinny cables distributing power around the premises.

    Interesting low technology Green tip. Simply by doubling the cross-section of all the power cabling inside the house (sockets, kitchen, laundry etc) you can make a substantial reduction in energy losses. Done at construction time, it is only the incremental cost for the fatter cables, the labour costs are the same to first order. (Apart from your time lost persuading a truculent Electrical Engineer (i.e an electrician with an ego problem) that yes, it was permitted in the regulations, the cable sizes there are minimums, not mandatory.)

    It was popular for a while until copper prices went crazy again, stretching the payback period beyond most people’s time horizon. Personally, I think longer term: I’m still planning on leave a little bit of Earth in reasonable state for my grandchildren ;-)

    • That’s interesting! Your idea with thicker copper cables is good, but can’t be changed so easily in a ready built house as the cables are hidden in the wall. BTW: there is an easy rule of thumb selecting the right voltage to transmit power efficiently over some distance: 1V per meter length of the device. For example: bike 6V with a max length of some meters; car: 12V; truck 24V up to ~24m; 230V for some hundred meters, etc. (Otherwise the drop of voltage in the lines increases to much.)

      To get JeeNodes powered directly from the mains I found this nice AC/DC converter with only 30mW standby from Recom. It’s close to what JCW developed some time ago having 10mW but it also nicely decouples the output from the mains.

  5. @Peter – yes, a good find. Includes EMI filtering and has a good isolation spec. Still not quite right for the “sleeping node” case which has such unusual requirements: enough current available on demand to transmit a packet, then an almost negligible “keep alive” current for when the node goes back to sleep.

    This argues for a much smaller average current rating for the supply. Use this to top-up some local energy store and then a complete converter shutdown while there is enough energy saved in the store.

    Sounds easy? The problem is getting the boundary conditions right e.g. how to cold start reliably when the energy store has inadvertently faded to zero.

        Hmm… reaches for pen and paper…     ;-)

    • Hi Martyn, do you have any idea how to reach for this goal? Zero Watt consumption except when your module wakes up for a couple of ms and all powered from 230 Volts?

      On the other hand one could say “30mW is well below the point that it really can be seen on the electricity bill”.

  6. @Peter – agreed, an individual 30mW is below the horizon, but sprinkle them liberally around and it does start to count. Ironically, the energy I’ve consumed in the coffee maker trying to figure this out greatly exceeds a lifetime load for one household’s worth of devices.

    Yes, I have figured it out. But it is a long and tortuous path from the breadboard to a production unit that can pass CE etc., approvals. Maybe I should Kickstart it to push it up the priority list.

    As a guide, what is the functionality worth to you? (‘Zero Loss’, packaged to meet applicable ‘direct mains connect’ requirements, no ‘skilled tradesman’ requirement to install)

Comments are closed.