Archived Forum

Hi, The most enlightening logs would come from emonHub, I would recommend setting the loglevel to "DEBUG" in emonhub.conf. 

Also which OEM pre-built image are you using?

How "customised" is the emonTx firmware? minor adjustments (to which sketch) or completely new?

Is the emonTx powered by AC/AC or AC/DC mains?

From what you say it would point to the emonTx as the emonTH continues to work and restarting that device previously corrected the issue. But that is far from conclusive.

When it stalls next time try and determine if the (tiny) led is flashing on the RFM2Pi to indicate packet recieved, How easy that is will depend on the update intervals you have used with each device, standard intervals are 10s for the emonTx and 60s emonTH so if no more than one flash is seen in ~30s the emonTx packets are not being recieved.   

Paul

Hi Paul, thanks for your reply.

I will look into setting the loglevel to "DEBUG", but how would that be helpful? I mean, do I need to use another command to retrieve information or the "tail -40 etc." would work fine?

As for the other questions:

• I am using the emonSD 13-03-15 build (I've just seen there is an update to it)
• I have edited the current calibration in the emonTX firmware, since I am using a CT that has its own burden resistor - I have removed the onboard resistor on the emonTX
• AC/AC powered

Yes, I haven't touched the update intervals, so I will check it out the next time.

Anyways, when the emonTX stopped working for the first time, I had to manually reboot it. I have to admit I did not pay much attention, but I think the red led that usually blinks when the device detects the AC wave did not blink, which made me think about a failure in the AC/AC adapter... but as I said a reboot solved.

Thanks again for your recommendations!

The later emonSD image is not a direct upgrade, it is a different version of both emoncms and emonhub.​

I would stick with the image you have for the moment as it is more stable and familiar for debuging purposes.

You will need to view the logs after the next occurrence, by pre-selecting "DEBUG" level you will capture much more info, when it happens. You can use tail directly or with that emonSD image you can use emonhub -dl (display-log) from the command line too.

I'm not sure the led flashes are linked to the ac detection, don't hold me to it but I'm pretty sure it just flashes with the transmitions. 

Paul 

"I'm not sure the led flashes..."

It is all written out here: http://wiki.openenergymonitor.org/index.php/EmonTx_V3.4#Power_Supply_Opt...

but note that in "Powering via AC (4)", the sequence of operations in detail is:

sample the CTs in use, in order
read the temperature sensors (if any)
transmit the data via RF,
delay for 10 s
flash the LED

so in effect, the LED is flashed before the sample is taken, and that could be nearly a second before the radio transmission.

Therefore, if ac has been detected at start-up or reset, the LED will flash and will continue to do so even if the ac supply disappears (provided of course it isn't using the ac supply as the sole power source!).

Oh! well that was a surprise, I was aware the led flashed during "detection" at start up but I had no idea the led was just then flashing every 10 secs regardless.

Paul

edit - if a ACAC was present at start-up of course, I don't actually recall ever running one without an ACAC so thats probably why I hadn't noticed.

None the less checking the led remains a valid test I guess, as if the led is no longer flashing and it was previously that would indicate the emonTx is no longer running, that may or may not mean a power failure, a temporary power failure should result in a successful restart and so either the power went down and was restarted by a "mains" power cycle or somrthing stopped it running and a power cycle or "reset" button cleared it.

How did you "reboot"?

Paul

The 'start-up' voltage detection isn't particularly onerous: it runs the calc_rms( ) method and looks for a voltage > 90 (out of a notional 240), and flashes the LED rapidly 10 times if that's found. Otherwise, it puts the LED on for 2 s.

Hi,

just a quick update in case anyone has been experiencing the same issue.

Long story short, the Ideal Power 9V AC-AC power adapter for the EmonTX proved to be the problem.

To explain a bit, during a power cycle after the n-th time the emonTX stopped transmitting, I disconnected and reconnected the 9V AC power adapter and saw the LED on the Emon TX going crazy, like blinking very quickly but without following a specific pattern.

I decided to get a new adapter, as it looked like I got to the root of the problem.

After setting up the new power adapter, I noticed one strange things that made me question my engineering skills! The mains voltage report in my feeds, which was around 250V with the old adapter, suddenly dropped to around 230V. 

You know what? The old power adapter was supplying 12V AC instead of 9V AC. I guess that explains it all, and I suppose I am lucky I did not fry my board!!

Anyways, it's all good now and my two emonTH and my emonTX haven't manifsted any strange behavior in the last 20 days.

Does the type number end in -09? But do remember that 9V is the output on full load, the open-circuit output voltage is specified to be 11.6 V ±3% with a 240 V input (for the UK version).

If the type number is wrong, or the output is significantly above (like 25% above) 12 V, then it would indeed appear to be a 12 V version.

But your voltage readings only show 8% above, so doesn't really fit with it being a 12 V adapter labelled as 9 V.

But in any case, it does appear to be faulty and you should probably let the shop know: support@openenergymonitor.zendesk.com

Thank you Robert, I did not know about the open-circuit voltage.

Mine is a EU version but there shouldn't be any differences, I suppose.

I will check the type number and load voltage when I can and report back.

The DE-06-09 data sheet gives a slightly lower open-circuit output of 11.5 V ±5%

(77DE-06-09-MI.pdf from http://www.idealpower.co.uk/external-power-supply/fixed-plug-top-ac-ac-linear/77DE-06-09)

mrcsrplln thanks for this post! I have the same issue and am about to place an order for a new AC-AC adapter. I can see that it is indeed the culprit. Looking a the EmonCMS Data Ican see when it had started to malfunction (~Sep10):

Hope this info helps other as well. Should put it as a sticky in the forum.

Which model is yours, flakattack?

Hi Robert,

I am using an EmonTx Arduino Shield V2 with an AHEAD JAA-0901000AS AC-AC 9V Adapter. Since I am located in Australia there is no available powerpack on the store.

I purchased mine from: http://www.ebay.com.au/itm/9V-Power-supply-Adapter-1000mA-AC-Adapter-Guitar-Effect-Pedal-Buy-4-Get-5-/181927112544?hash=item2a5bb36f60:g:jXoAAOSwpDdVOt6~

Also am running a HopeRF RFM69CW as the chip onboard.

Cheers,
Simon

OK, thanks for that.

I suspected that we had a quality issue as two adapters had gone down with apparently the same fault, but clearly that's not the case. It looks like an unfortunate coincidence.