AVR-DB: emonTx V4, new hardware in progress

Looking forward to Emonpi2… I am ready for a hardware refresh…

1 Like

Probably a little late in the design cycle,… but has any consideration been given to some forced cooling,.My ‘emonpi v1’ can get a little warm,… particularly when the ambient temperature reaches 40C… …although chips may be running within their rated temperature spec,… its always nice for their extended life to run them as cool as possible,…

Thanks @garyjuergens

@diyhouse no forced cooling, it’s not something we’ve looked at yet. Really the Pi4 is way over powered for what we need. It would be good to look at under-clocking, and other options to reduce power consumption / chip temperatures. Would prefer Pi3’s and the pi zero 2w, the latter we plan on making available as one of the options with the emonpi2.

The emonPi2 is now available to order:

See also new emonPi2 thread here: emonPi2 Pictures & Spec

1 Like

@TrystanLea , I have eventually got round to trying the 12 channel firmware, following the documentation on the wiki, but the sketch never returns any readings over the serial connection. If I add an “else” to print something at the end of the “if (EmonLibDB_Ready())” condition it is running, so it seems not to be returning “true”.
I have tried modifying a couple of other sketches as well and all work fine with just 6 channels, but adding the full 12 has the same effect. This is with the CM library so “if (EmonLibCM_Ready())” has the same problem.
Is there any plan to release an “official” firmware that is programmable from the EmonPi ?
If you wish to see my modified sketch, let me know.

I suggest you download the emonLibDB library zip package from this forum EmonLibDB - Version 1.0.2 and therein you’ll find working demonstration sketches for up to 3 current and 3 voltage channels (i.e. single phase, split phase or three phase but no temperatures), which are briefly described in the accompanying documentation.

You’ll need to add “trimmings” as necessary - i.e. I’m waiting for Trystan to specify what he wants by way of on-line calibration, then I can produce a “fully engineered” sketch which will fulfil your requirement of “programmable from the EmonPi”.

@Robert.Wall , Yes done that already with no success. The EmonTx4DB_rf sketch with the expansion board in compiles and uploads but then just sits there with
“EmonTx4 + EmonLibDB Continuous Monitoring Minimal Demo with rf
Starting reports every 9.80 seconds approx.”
on the screen.
Comment out the “#define EXPANSION_BOARD” line and the same happens.
The other two sketches for 6 channel only in the zip do the same.
I did re-download the zip and start again.
I have one sketch modified by me that I can switch between 6 and 12 channels, and when it is 6 it works fine, but not 12.

Have you got a voltage source connected?

@Robert.Wall , No not at the moment. I was programming on my desktop, but my voltage source is in use on my other TX4 in my “cupboard”.

That could well explain it. If it can’t count zero crossings, it can’t work out when to report. Put your finger on the bottom of the power connector socket (where the ribbon cable from the emonVs plugs in) so that it gets a bit of voltage pick-up from your body. It might get enough to run.

Because the concept all the time was an integrated power source and voltage reference - the emonVs - emonLibDB was never intended to run without a mains voltage to measure and derive its timing from.

OK, just tried the EmonTx4DB_rf sketch with my emonVs and it works…
Thanks for that.
Might be wise to add that to the documentation?

I didn’t think it was possible to accurately measure energy without knowing the voltage.

I’m working on V1.0.3 so I can add a note in the documentation for emonLibDB - I’ve no control over what appears in Github.

It was more just to set up and test the 12 channel emonTX, but when it didn’t return any data I was thinking there was something amiss!

The question was discussed with Trystan at a very early stage, and we decided the complication and additional code necessary to run with and without a mains voltage, i.e. to be able to run measuring current only and outputting a coarse estimate of apparent power, was not justified, again as I said there was not at the time any intention to run without the emonVs.

It would be a relatively simple matter for the sketch to “time out” if results didn’t become available after (say) two and a half times the reporting period - but what to do then? It’s only when a human is looking at the serial output, or at the on-board LED, that the sketch can do anything which might elicit a reaction.

Might be worth adding that as an error message at startup (i.e. no Voltage detected) if it is required to generate values (if not already there).

Is the LED a fixed colour? Could it be set to stay on constantly?

From the documentation:

bool EmonLibDB_acPresent(int8_t channel = 1)
Returns true when the a.c. voltage on input 1 has been detected (greater than approx.
17 V, well below the voltage at which the emonVs can operate). Voltage Input 1 is used for
timing purposes and this voltage is required for proper operation of the library.
N.B. Only works on channel 1.

It’s the sketch’s responsibility to use the library to achieve the desired result.

I need to add the expansion board now but the link to the shop is giving 404?

Also cant find under development boards, hopefully just out of stock?

Hello Paul, I believe @Gwil sent out an emonTx4 expansion board as part of your order for the other bits yesterday.

For anyone else interested in the emonTx4 6CT expansion board we do have a few of these expansion boards left and can order more if there is demand. Please send us an email at [email protected] if you would like one.

Hi Trystan, I did email the shop and Gwil sorted me out and I ordered more CT clamps as well - glad you had some left available and I can now expand as I originally intended.

Arrived today but not had a chance to fire up the soldering iron yet to attach the header and install the upgrade.

Just a tip, drill 2 holes in the modified bezel and fit 2 long m3 bolts through to with washers on the inside back edge to hold the new expansion board tight against the bezel so you are not relying on the header to hold the board in place…