Archived Forum

Broadly speaking - yes, you are right. The power take-off from the mains voltage monitor was only ever intended to supply the absolute minimum current necessary to run the processor, the RFM12B radio module and one temperature sensor, at the lowest permissible mains voltage. You can try running multiple sensors, whether you get away with it depends very much on the lowest mains voltage you see. You will probably see two effects: firstly, the 3.3 V regulator will fail because the resistor that minimises the distortion of the voltage wave will drop too much voltage (so you will see dips in the 3.3 V supply), and the mains voltage measured will be wrong because you are loading the adapter but only over part of the cycle and so distorting the voltage waveform. I have not done the sums to check what the actual effect might be.

What size should the power supply be in terms of Amps for this application?  I know that they do vary widely.

It is not the VA rating of the adapter that matters, it is its regulation - which appears as an impedance in series with a perfect voltage source. When the reservoir capacitor charges near the positive peak of the mains wave, there is a short, heavy, current pulse that causes a voltage drop across the transformer impedance, and it's that which is the problem. The series resistor R29 that has been put there to limit the current (and hence broaden the pulse) and that minimises the effect, though at the same time it naturally limits the current that can be taken from the 3.3 V rail.

The circuit schematic is available on Solderpad if you want to redesign the power supply to suit your needs.

That all assumes a level of knowledge I do not have.  All I want is a simple answer as to what power supply do I need to plug into the micro USB to power at least 5 sensors.

This one or this one should be fine.

Ok, great thanks.

Hi,

I am evaluating to buy EMontx V3 (prebuilt).

I plan to connect 4 TA to measure power + 1 temperature sensor. Everything feed by AC-AC power supply (available on the shop). AC-AC also as input for real voltage measure (and to measure real electrical power).

Is this configuration supported ? Should I change anything ?

Thanks.

Regards,

Emiliano.

By "TA" do you mean current transformer? That should be OK.

Yes I meant CT (Current Transformer).

I apologize, I used italian term TA thinking it was also english expression.

Thanks.

I'm pretty sure the Wiki says one temp sensor is OK. More than that and you need a DC supply.

That might not be true. I did the original calculations as an add-on circuit for the emonTx V2 and the Mascot adapter and only one temperature sensor. A couple of weeks ago I built a new Spice model for the 3.3 V regulator that the V3 uses, but I can't remember whether I have checked everything for the emonTx V3, which uses the RFu and the different regulator, and of course the TDC/Ideal adapter. I need to do those checks for another reason too.

Edit: I have now checked the numbers and it appears that with the recommended UK ac adapter, at the minimum permissible supply voltage of 216 V, it may just be possible to run 5 temperature sensors. This is from a circuit simulation and has not been verified in practice.

So: with 4 CT + Temp sensor should I buy also DC power source or not? Thanks!

For this application, an AC source will be required anyway for measuring real power.  The question is whether this AC supply  can also provide sufficient DC for the processor and several additional temp sensors.  

Most people can lay their hands on a suitable DC supply which could be temporarily connected to the V3.  If this additional DC supply makes any noticeable difference, that implies that the AC supply on its own is not sufficient. 

One could start with a simple system and keep adding extra temp sensors while checking the value of the 3.3V rail and monitoring a known source of energy flow.  If anything starts to change when the temporary DC supply is not present, but normality returns when the DC supply is present, then that would show that the extra supply is needed; otherwise, it probably isn't.

Has anyone already tried to use in reality emonTx V3 with 4 CT ? Any recommendation ?

There is some confusion here. The emonTx V3 is fine with 4 current transformers and one temperature sensor. This is what it was designed for. There is no problem. The number of current transformers does not affect the 3.3 V power supply.

So, Emiliano, you do not need a separate 5 V power supply. Even if you have two temperature sensors (your other thread!) you do not need a separate 5 V power supply.

It was Borpin who asked about 5 temperature sensors.

Each temperature sensor consumes current from the 3.3 V supply and that is where the problem lies. It was not designed for, but it should work with, 4 temperature sensors; and it might work with 5. If you want 5 temperature sensors, you might need to have a separate 5 V power supply. If you want more than 5 temperature sensors, you must have a separate 5 V power supply.

And if you have more than one temperature sensor, then you must change the sketch to send all the temperatures.

Thanks Robert !

Ordered :-)

Ok I now have the additional DC supply, how should I start the EmonTX?

I know I need to remove the jumper but should I switch on the DC or AC power first?  One then the other immediately, one then wait then the other?

Is switching off the same or reverse?

You need the AC to be on when the processor starts up (or when you reset it), else the sketch will assume it is reading current only. You can power both adapters at the same time because even though the d.c. will probably come up fractionally later than the a.c. (which will be instantaneous) and the sequence will be correct, there's a 1 s test sequence for the RFM12B before it looks for the c.t's and another 10 s delay before it looks for the a.c. voltage. So you have a little less than 11 s in hand.

Switching off doesn't matter, but if you don't want to risk a zero value being set, switch off the d.c first.

Great - many thanks.