Archived Forum

1. Yes, voltage measurement is necessary to know the direction of power flow, because the power is measured at the grid connection where the direction reverses. See here. Type 1 measures the current where the power flow direction is known (actually not true, it gives wrong answers because the inverter consumes a small amount of power when idle!)

2. Yes, they are redundant. But maybe not in the future, but it should be easy enough to add them later if required.

3.1. It's a simple matter of the designation of the various inputs. On the Shield, the voltage input is AI 0 and the c.t's are AI 1-4 in order. So all you need to change are the values for voltageSensorPin and currentSensorPin. But depending on whether and how many LEDs you have, you might have to change those as well!

3.2. As far as I know, all the versions do the same basic job, both Robin's (calypso_rae) and MartinR's. Robin's though have the remote load already included.

4. The last we heard, Richmc was poorly and in hospital, but hoping to fulfil orders when he's sufficiently recovered. It's possible, but not so neat, to build the controller without his kit. You can use the standard Shield and AC Adapter for the controller itself. Alternatively (and I appreciate you've already bought an emonTx Shield), for the remote load switches you could use the emonTx itself with a strip-board sub-assembly for the opto-trigger and triac.

5. I don't see why not (though Robin may well chip in with his thoughts).

6. In principle, again I don't see why not, but deciding your control algorithm and trimming it to operate within the various constraints could be trickier as the number of loads increases.

Hi Alan, delighted to hear you're intending to build a Mk2 Router.  All versions of the code will run on the same hardware so you may be best to start with a simple one such as Mk2_PV_Router_mini_3.ino.  It will heat your water just as well as any of the more complex later versions, and will be much easier to understand.  Looking quickly through your list:

1. Yes, you will indeed need to measure AC voltage, it's right at the core of the Mk2 Diverter's logic.

2. You will only need to populate one of the current sensors.  Any of them will do, you can then amend the pin allocation in the sketch to match.  On the emonTx, AIO1 corresponds with CT3.  On the emonTx Shield, I believe that AIO 1 is CT1.  But do please check in the OEM documentation for the Shield which I've never used.

3. All of my Mk2 sketches are intended for an Arduino, but they run just as well on an emonTx.  When using the Shield, the only lines that you may need to change are for the pin allocations for the voltage and current sensor.  But only one character on each line, so no worries!

4.  The last we heard of Rich was that he was unwell, we wish him a speedy return.  To build a basic Mk2 rig, you shouldn't need his front-end PCB because all the necessary circuits are on the shield.  Instructions for making the output stage using stripboard are available in Section 8 of the online guide.  Triacs and triggers etc are readily available from Spiratronix.  If you'd like a working strip-board version with triac etc, send me a PM, I have a few spare here.  If you were to become my first heatsink customer, I'll throw in an output stage for free!

5.  The rev5 version of Mk2i can control any number of loads, I just happen to have configured one of them to operate via an RF link.  Operating two loads via RJ45 cable or similar is no problem; operating two loads independantly via RF would take a bit more thought.  If required, it should be easy enough to include a loadNumber field in the RF payload.  Each message could then be processed by the intended load .  Then there would be no need to change any of the Node Id fields etc.

6. rev5 has always been able to control multiple loads, the only limitation is the number of pins on the processor.  This logic can be seen working in my  FlickerDemoMultiload_2 sketch

Hope this helps.

PS.  At risk of repetition, do please make use of the basic tools at the top of my Summary Page.   Running MinMaxAndRangeChecker.ino and RawSamplesTool_4ss_2.ino will confirm the state of your new hardware in just a few minutes.  If all is not well, you save a lot of time/effort by fixing any basic problems at this early stage.  Posted results from these tools should be relatively straightforward to diagnose.

Cheers Robert, you obviously type faster than me :)

Thanks for your help Robin & Robert.

just a couple of questions arising from your answers.

2) "You will only need to populate one of the current sensors" - does this mean I only need one current sensor?

Both type 1 & 2 show two current sensors.

Where is this sensor placed? Is it between utility meter &  junction box?

But you're not making a monitor! You need 2 c.t's when you want to know separately the imported power, the generated power and the consumed power (or energies of course). You have your other system for that. If you are only controlling the dump load, then you need only one c.t. that measures the nett power flow in the connection to your supply, so the c.t. goes on the supply side of the point where the PV feeds in. On the main meter tail is probably the convenient place.

Agreed.  A Mk2 diverter needs one voltage sensor, and one CT which needs to go next to your supply meter.  During periods of surplus power, the Mk2's logic can then stay one step ahead of your meter so that no surplus power is lost, nor are you charged for any diverted energy.

Are we measuring current on live or neutral line?

The names of the conductors are LINE and NEUTRAL, both are "live" when connected to the supply. Please yourself which you use; if they are not carrying the same current, there is something seriously wrong with your installation. You might get slightly more voltage pickup ( = unwanted interference) onto the analogue input with the c.t. on the line, so if there's a choice the neutral might give less inaccuracies at very low or no currents. Normally, I wouldn't expect you to be able to find any difference.