Archived Forum

"I have clipped a CT sensor around the live main supply cable"

Have you clipped the CT around only the line cable? You must not clip it around both line and neutral. (Because the currents in the two cables should be equal but flow in opposite directions, they cancel each other, and you would read almost nothing).

"I cannot plug in the adapter very near to the board so it is on an extension lead totalling about a 5m round trip to the emontx."

That should make no measurable difference if it is the only thing on that extension cable.

"I don't really know whether it is using the voltage to measure power or the ct sensor or a combination of the two!"

It uses both (unlike some commercial units that measure only current and then estimate power). Power = voltage x current. This page http://openenergymonitor.org/emon/buildingblocks/ac-power-introduction tells you a bit more.

"the power rating [is] far too low"  - "should I change the 111 to 222?"

What is the error? Are you displaying voltage, and is that correct? Have you got the correct calibration constants for both your ac adapter and current transformer? Both need to be correct for the units that you have. This page http://openenergymonitor.org/emon/buildingblocks/ct-and-ac-power-adaptor... explains how to calculate the numbers for yourself (apologies if it is hard going), or you can look here: http://openenergymonitor.org/emon/modules/emontx/firmware/calibration if the parts you are using are listed.

"air gap in the sensor [is] causing accuracy errors"

The transformer core is the square grey metallic component that is in two halves within the blue plastic case. It is an air gap between the faces of the core that meet when the top is closed down and clipped that you must not have. An air gap around the main cable is perfectly OK.

Post again if you want some more explanation.

The voltage reading sent by the emontx is about 14500, what units is it supposed to be in? Volts?! If so, how do I correct the reading, I have changed the calibration number to match the AC adapter I am using.

I do not have a calibrated meter so I cannot accurately say how far out the power reading is, but when I turn my 2.75-3kw rated kettled on, it was only increasing by about 1.5kw.

The only other ting on the extension cable is the emontx itself, it appears to make no difference if i power this separately, should it?

"The voltage reading sent by the emontx is about 14500,"

That definitely looks wrong. The published sketches (as far as I know) all output in engineering units, i.e Volts and Amps.

[Edit]  ... but the value of voltage is multiplied by 100 and then transmitted over the RF link as an integer  - so you are seeing 145.00 V. So if you monitor the serial output of the emonTx, that is what you should see.  [/Edit]

(And no, having the emonTx on the same extension lead should not matter).

We need to do some careful and systematic checks.

1. What is your system voltage, i.e what should it be reading, in Volts?

2. Which sketch and library are you using? and where are you reading 14500 ?

3. Unplug the voltage adapter and unclip the CT from the cable, allow the emonTx a few minutes to settle, what do you read then?

1) I am not sure what the "system voltage" should be. Is it measuring what comes out of the transformer? In which case should it be 9V?

2) I am using the emontx_CT123_Voltage sketch with amendments to also read the optical pulses and to set the voltage calibration to 238.56 rather than 234.26 as I am using that model of adapter. I am reading the 14500 from the serial output of the emonbase

3) With no voltage adapter or CT the power drops to zero and the voltage fluctuates around 800.

If I read from the emontx board itself, the power reading is 0 and the Vrms fluctuates around 8

Only difference being that then it is being powered by the FTDI cable instead of the usb cable

There are 2 values of Vrms aren't there - there is ct1.Vrms which is the one shown on the serial port, and the one it sends to the emonbse (emontx.Vrms) and the 1srt value is multiplied by 100 in the emontx code to give the value to send to emon base - that's why I was reading a value so high from emonbase. However, should it be reading 8V when no connection is made and 145V when the adapter is connected?

By "system voltage" I meant the nominal mains supply voltage to your installation. In Europe, it is 230 V but in the UK although still nominally 230 V the centre value around which it fluctuates is 240 V.

From the calibration coefficient you give, I deduce that you are in the UK and using the Strontronics adapter.

Check the voltage divider resistors on the emonTx pcb. You might have a wrong value resistor for R13 / R14, or one could be faulty/damaged/cracked when assembled? All properly soldered? If all this looks ok, it is looking very much like your adapter is giving the wrong output. If you have a reasonable a.c. voltmeter, you should read about 11.2 V out of the Mascot adapter. The Sonotronics is likely to give a similar (i.e. higher than 9 V) output.

That looks right. Many people expect an accurate zero but in practice a small number is normal.

"Only difference being that then it is being powered by the FTDI cable instead of the usb cable"

That should not affect it.

"I do not have a calibrated meter"

I strongly advise you to get a reasonable digital multimeter if you are going to do significant electronics work. Something like this http://www.maplin.co.uk/ut50a-manual-ranging-digital-multimeter-46390 should be adequate. [Disclaimer: I don't get a commission - other makes and suppliers are available!].

I have a multimeter, I just didn't fancy putting it into my mains to see what the voltage reading was!!!! I meant I didn't have a meter where I could measure an accurate power reading for any appliances etc to give the emontx a more accurate calibration coefficient.

The adapter is giving a voltage measurement of 10.9 when switched on.

I checked the resistors and all was fine then I noticed one of my capacitors was incorrectly aligned - oops!! I have now rectified this and will see if it makes a difference to power readings. The Vrms number from the emonbase is now 21300-ish which means on the tx it must be 213V which is making more sense?

Methinks you've cracked it! Was it C10 the wrong way round? (If the embarrassed answer is yes, then the centre point of the bias chain was probably very near to 0 V, and the wave would have been clipping on the negative half cycle, hence reading low).

"I didn't have a [power] meter"

I think few do have - it's a fairly specialised piece of kit.

"The adapter is giving a voltage measurement of 10.9 when switched on."

That's totally believable.

"it must be 213V which is making more sense"

A lot more sense, though still somewhat lower than I would have expected. It's possible that you have resistors in the divider chain (R13 & R14) that are 5% out the wrong way, which would mean the reading would be about 10% low, and the correct(?) value was 233.4 V. The mains supply should be 230 V +10% -6%, which gives a low limit of 216.2 V. I think you need to take courage and measure the mains voltage (on the protected side of an RCD if you're really apprehensive - that would protect you but not the meter, just make sure the meter is on the highest ac voltage range first), then adjust the voltage calibration constant to give you that number ( x 100 if you read it as transmitted). When you get the voltage calibration correct, the power readings too should be correct. If not, we'll look for a problem in the current department!

Do we need a stable supply for the emonTx?

Like others, I checked the Tx calibration using a new pair of batteries for the power supply (3.25V).  But what happens when the battery voltage is down to 2.7V?  If the ADCs measure the inputs as a proportion of the supply voltage, they will start to read high as the battery voltage reduces.  At 2.7V the ADCs will be reading 20% high, so the measured power reading will be over 40% high (being voltage * current).

Have I missed something?

"Have I missed something?"

Indeed you have! (Sorry). The ADC reference voltage is measured internally against a stable reference. It emerges in two registers:  ADCL and ADCH, and you'll need to plunder the Atmel data sheet for more about those. These are read by readVcc() in EmonLib.cpp and from there it enters the main calculations as SUPPLYVOLTAGE.

Thanks!