Archived Forum

I am sorry to say that the CT you have chosen is not suitable for the Arduino. The input that the Arduino expects is approximately 1.6 V rms at maximum current, yours gives only 55 mV so instead of using the whole input range, you are only using about 3.3% of it, and instead of the ADC output swinging most of the way between 0 and 1023, it will swing only 11 counts from possibly 506 to 517. You can only expect an indication that you have a current flowing, you cannot expect to measure it.

I suggest you use a YHDC SCT-013-030 (30 A, 1 V output) to replace yours, then connect it without a burden resistor.

The point is to make the assembly as small as possible. It's for a license project.

I have seen some use ACS712 to monitor AC consumption. Yes, 230V. I thought this CT was for max of 30V.

Where are all of my posts disappearing? I've written 2 posts 4 hours ago and none of them are visible.

Anyway, there are some differences between the CT you mentioned and the one a datasheet on this site belongs to (YHDC SCT-013-000).

Here are the datasheets:

- mentioned above (YHDC SCT-013-030) - http://bit.ly/193F4AK

- in the guide  - http://bit.ly/1FQSM8e

When you look and the 2nd CT's datasheet you can clearlt see a max of 50 mV of output. That was one of the reasons I bought an ASM-030 because output is similar. Can the ASM be modified in any way in order to give me what I want?

What about this one instead of the SCTs? http://bit.ly/193ElQ3

"Where are all of my posts disappearing?"

Here

If you need to use your CT with the 55 mV output, then you have a choice.
1. You use an operational amplifier to increase the output voltage as seen by the analogue to digital converter, or
2. use the analogue to digital converter with the internal 1.1 V reference, or
3. a combination of both of these.

For (1), you need to source a single supply op.amp capable of giving an output that will go as close as possible to 0 V and to Vcc. You will need to set the gain to a little less than 1.6 V / 55 mV ( = × 29 ) and you might have to reduce the gain below that if dc offset and drift become a problem.

For (2), you need to program the Arduino to switch the ADC reference to the internal 1.1 V reference, then you need to alter the bias resistors that at present set the quiescent input to 2.5 V, so that the quiescent input becomes 0.55 V. Even when you do this, you will still be using only 14% of the input range.

The ECS1030-L72 gives you much more flexibility. The ratio is 60 A : 30 mA. The transfer curve indicates that it is good to at least 60 A with a 10 Ω burden, therefore it should be good to 30 A giving you 600 mV output with a 40 Ω burden resistor. That represents about 40% of the available range and is much better. You could, if you have the necessary instruments, measure the output voltage against current with a higher value of burden resistor (in order to give you more voltage), checking if and by how much the output waveform distorts as current increases.

You might find a similar ring-core CT that will be smaller than this one, but if you go too small, it will not be powerful enough to give the output voltage needed.

I was considering amplifying the output. What about the part involving the ACS-712?

That has a galvanic connection to the electricity supply, and means that you must observe proper safety precautions, e.g. it must be housed in a double-insulated or earthed metal enclosure, and you must have proper creepage distances between the current path and the output on any circuit board on which it is mounted.

If you can do that, it will give you a maximum current of about 9.5 A, which falls a long way short of the 30 A that your original CT gave you.

I guess it's safer and easier with the thoroydal CT.

There is a variety of CT amplifier schematics but I have no idea which one applies to my needs.

This http://bit.ly/193ElQ3 CT is similar to the ASM-030. Is there any slight difference?

That one appears to give a higher output voltage and has a higher maximum current before saturation sets in, so it is better for what I think you want (which you have not actually specified!)

It might give you a maximum current of 12 A with a burden resistor of 270 Ω, which would be suitable for use with an Arduino running at 5 V. But the data sheet is missing some important information so I cannot be certain that the output wave would not be distorted with that output voltage.

In essence, I am building an energy monitor with a Yun which will send data wirelessly to plotly.

Which amplifier would you suggest to use with the ASM-030

You need to do the maths on the burden resistor I suggested. Have you done that? Do you know how to do that?

Its datasheet has a schematic with a 220 ohm resistor at its terminals. I suppose that's what I'm looking for. It's not a wounded type. The wound is replaced with a ferrite core. The output is voltage.

I am looking at the ECS1030-L72, which I think is much more suitable.

That resistor (10 Ω) is the burden resistor, and it is an example only. The output is a current, not a voltage, because the ratio is 30 A / 15 mA.

What is the maximum current that you want to measure?

What will the CT output current be at that?

What value of resistor do you need for that CT output current to give you 1.66 V? That is the value of burden resistor that you need.

Next you need to check that the CT will not be asked to deliver too much power. If you look at the graph, it is a straight line so at the output it will give 300 mV @ 30 mA = 9000 μVA ( = 9 mVA). If your maximum output current × your maximum input voltage ( 1.66 V) is less than that, the CT is suitable. If it is slightly more, it is probably suitable but you might have a little distortion at maximum current.

What about if I replace the CT and schematic from the guide shown here with this one http://bit.ly/1DrA9re ? It's the ACS712 with it's own schematic.

My only concern is if the Emon library is tightly related to the guide shown here which involves the SCT-013-030.

Can the whole emon library be made into one ino file?

That is not a current transformer, it is a Hall effect transducer.

EmonLib does not care what the transducer is. All it requires is an input voltage that is a faithful representation of the current waveform, NOT a steady dc voltage proportional to the peak value that the diagram in that link will give you.

If you use that interface as published, you will not be able to use emonLib and you will not be able to calculate true rms current, nor real power, nor power factor, because the rectifier and smoothing capacitor on the op.amp output remove all the information that is needed to calculate those values.

If you wish, you can incorporate emonLib into your sketch. It is open-source, so you can do what you wish.

Updates:

I have passed the live wire through the ASM-030 4 times (4 loops) and the results look promising. The power factor varies between 6-7 units through readings. This happens when using a normal 9V transformer, with a primary and a secondary windings with a 9.3 V output. When using a PCB transformer like the HAHN BV201 results are going crazy no matter what calibrations I make. Maybe because the output is unusually high (12.6 V, more than 3V the voltage from the transformer mentioned above).

Did you not read this reply that I posted 3 weeks ago:

"...the CT you have chosen is not suitable for the Arduino. The input that the Arduino expects is approximately 1.6 V rms at maximum current, yours gives only 55 mV so instead of using the whole input range, you are only using about 3.3% of it, and instead of the ADC output swinging most of the way between 0 and 1023, it will swing only 11 counts from possibly 506 to 517. You can only expect an indication that you have a current flowing, you cannot expect to measure it."

Understood. But what about the SCT 013 000 from the guide? At max of 30 amps the CT will give max of 1 V. If I find a similar CT besides it's construction or type (clamp on, ferrite core, winded type) it should work?

I suppose this is what I'm looking for:http://www.tme.eu/ro/Document/70426a11254837900f0ea7ada5026da5/AZ%20Series.PDF

Those look as if they are much more suitable for an Arduino running on 5 V.

Remember that a current transformer is a current source, this CT appears to require an external burden resistor, just like the SCT-013-000, and that means you can adjust to value of the burden resistor to give you the correct voltage for your Arduino.

If you say what maximum current you want to measure, then it is easy to calculate the correct value for your burden resistor and the calibration coefficient so that you read the correct current.

The Yun will powered from the V in pin from a power supply, not a battery.

Please clarify the influence the pcb transformer mentioned above has on the results and why the normal one shows correct values.

Is the current supplied by the transformer important in this case?

I do not understand your questions. We are discussing a current transformer. It has two windings: the primary winding is the cable that you thread through the middle, normally it passes through once so there is one turn. The secondary winding is the fine wire inside and there are many turns, often hundreds or thousands. The secondary winding acts like a current source and it supplies a current proportional to and smaller than the current in the primary winding.

If you want a voltage, for example to feed into the analogue input of your Arduino, then you must convert the current into a voltage. You do that by feeding the current into a burden resistor. Then you connect the current transformer and burden resistor like this.

The output will be specified as so many mA at the maximum primary current.

Some current transformers are supplied with the burden resistor inside. When that is the case, the output is specified as so many V at the maximum primary current.

Every transformer has a maximum output. For a voltage transformer that you are probably familiar with, there is a maximum current that you can draw before it becomes overloaded.

For a current transformer, there is a maximum voltage that it can generate before it becomes overloaded.

Last posts were about the 220V - 9V transformer. One is a regular and the other is a pcb type, a compact one. Read the issues again.

The AX-1500 CT just arrived and I installed it. The problem now is the output. As you can see below, the output is very strange when a 40W light bulb is plugged. The real power output varies very much and that's because the power factor isn't calculated properly.

23.60 40.43 225.75 0.18 0.58
18.56 39.68 225.86 0.18 0.47
18.60 31.05 226.06 0.14 0.60
26.53 44.05 226.04 0.19 0.60
17.09 37.55 226.12 0.17 0.46
21.40 39.49 225.80 0.17 0.54
22.15 42.95 225.82 0.19 0.52
14.70 36.09 226.27 0.16 0.41
23.15 40.45 226.17 0.18 0.57
18.87 40.02 225.94 0.18 0.47
17.83 37.27 226.16 0.16 0.48
25.09 42.76 226.92 0.19 0.59

I don't know the source of the problem. I checked the EmonLib.cpp and I still can't figure it out. I have to mention I added abs() to the realpower and power factor in the EmonLib.cpp file. It kept reading negative.

And increased the sample rate to 1000.

If your maximum current is 30 A (I'm guessing as that's what you originally suggested you wanted), then 40 W is probably 0.5% of full scale, and you should not expect accurate readings at such low currents due to noise pickup. You should calibrate at a much higher current, say with a 1 kW heater (~ 4 A) and 5 primary turns to give the equivalent of 20 A primary current.

If real power is negative, it means that the power flow is in the opposite direction to that which you assumed. Reverse the CT connections or reverse the CT on its cable.

So, even if I calibrate it correctly, small consumers would not be read accurately? What would be the minimum?

That depends on the actual method of construction that you used, so I cannot give a figure. Great care was taken with the design of the emonTx V3 to minimise noise pickup, we seem to get more complaints about noise pickup from people who have constructed their own front-end circuitry and used an Arduino board for the processor. We believe the major noise contribution comes from the digital circuits inside the processor and makes its way into the analogue side via the analogue reference AVcc.

For now, the assembly is on a breadboard. I'm using a Talema AX-1500 CT (max. 15A read), which reminds me:

http://www.imagebam.com/image/f8b562405087177
 

In the attached image I've highlighted in yellow the load resistance and the 50 Ohm resistance in the graph on the right. What is with the load resistance and what's with the one in the graph? In the current setting there is no load resistance on my assembly, just the 50 Ohm resistor from the graph.

Those are suggested values for the burden resistor. If you look back over this thread, I think I've covered all the theory you need to be able to choose the correct value for your application. The sums are all in Building Blocks - see the article on calibration theory. It's only straightforward Ohms Law, but you need to allow a few percent for component tolerances. I aim for not more than 1.1 V rms across the burden at max current for the emonTx, or 1.66 V rms if you have an Arduino running at 5 V.

The AX 1500 has a reasonably higher output when using the 200 ohm resistor. Can that "feature" be used for a more accurate output and be applied to smaller consumers?

You can use whatever value burden resistor you like, subject to the maximum power that the CT can provide. You can estimate that from the graphs: it is in VA and it is the voltage × the secondary current.

As I have written before, if you try to generate too much voltage, you overload the CT and the waveform becomes distorted. You can see that at 5 V with 200 Ω burden the graph is starting to bend, that implies a current of 25 mA - that is 0.125 VA. Therefore, if you want to measure only to 20 A (13.33 mA secondary current), you would have a 124.5 Ω burden resistor, which would give you 1.66 V rms suitable for an Arduino running at 5 V. (You would actually use a 120 Ω resistor.)

The price you pay is increased phase error, as you can see from the data sheet.

I got access to my university's lab with precise equipment and a test load. The load was an 800 W heater with direct link to the socket. I calibrated the assembly as best as I could but the power factor is still a major problem. No matter what I did, it wouldn't get over 0.76. I noticed the current was oscillating constantly. Maybe that's the source of the problem. What is there to do?

Posting a complete circuit diagram and the sketch you're using might help.

Here it is:

Let's be frank here: I've spent a lot of time trying to give you some support. I asked for a circuit diagram and your sketch. You have responded with fuzzy photograph of a prototype board. Do you really expect me to try to figure out what the component values are and how they are connected?