EmoxTx v3 Accuracy

Hardware emonTx
1

Hi All
Newbie here - from Tasmania, Australia.

I have an EmonTx v3 (bought as a ready to use assembly) talking to an EmonPi.
I am using an ac power supply for the EmonTx.
I have 3 x CTs connected to the EmonTx, as follows:
1 - Light & Power feed (into property - everything except hot water service and heat pumps)
2 - Hydroheat feed (into property - hot water service & heat pump loads).
3 - Solar array output

Everything is communicating just fine and I have feeds set up with kWh logging enabled on all the above. I have an email report set up to provide me with daily kWh data for the three feeds.
Unfortunately I’m not convinced that the data from the EmonTx is correct, as all my kWh numbers appear to be approximately 50% lower than what I’m expecting (based upon the daily kWh numbers on my solar inverter and also my quarterly energy bill).

I didn’t perform any calibration on the EmonTx - just installed out of the box.
Is there something I’m missing? I’m using the correct CTs and the AC power supply is in spec (the measured voltage coming out of the EmonTx corresponds to measured supply voltage).
I’d appreciate any advice / others experiences.

Cheers

2

Are you on a single phase or a 3-phase supply?

3

Hi Robert
I’m on single phase supply.

4

Good. That’s OK, and one idea shot down.

Obviously there’s something drastically wrong for you to be seeing only 50% of what you expect. I take it you’re looking at the numbers on the emonPi.

Are you using CT Inputs 1 - 3?

Are all three the same?

What are you seeing for the power on each input – the actual powers in watts/kW?
Does that correspond roughly with what you expect, or is it about a half? i.e. Is it reading 50% or accumulating energy at 50%?

How long have you been recording the data for?

5

Hi Robert
Yep - using CT inputs 1-3.
All three CTs identical - I have to confess at this point that I bought my CTs from ‘Banggood’. They are ‘yhdc.com’ devices, part number SCT013, 5-100A.

I believe it’s the actual watts numbers that are low - I’ve been suspecting it for a while and when I look at output watts on Solar inverter LCD display compared to measured watts from the EmonTx the EmonTx watts are a lot lower.

6

What is the exact part number of the C.T?

Because there are two types, the SCT-013-000 which is a 100 A : 50 mA genuine current output current transformer, and there’s the SCT-013-000V which has an internal burden resistor and is 100 A : 1 V output.

The emonTx and emonPi both require the former, though with a small hardware and calibration change, you can use the voltage output one.

7

Hi Robert
I believe it is the SCT-013-000.
The link to the CTs I bought:
https://www.banggood.com/5-100A-SCT-013-000-Non-invasive-AC-Current-Sensor-Split-Core-Current-Transformer-p-1193260.html

8

You need to look at the c.t. itself - that page lists 9 of the 10 types within the SCT-013 family that YHDC make.

It seems very likely that you do have the ‘V’ version, because by having both an internal and an external burden, the secondary current is splitting two ways and developing half the voltage it should (approximately) hence you are seeing half the reading that you should.

9

Hi Robert
I’ve just read the order detail more closely - looks like I bought:
#2: SCT013060 (60A/1V). Darn it.

This would explain why they are reading incorrectly!
Now - how on earth do I fix?

10

Three ways:

  1. The “fudge” that will reduce the accuracy, because you’re using only half of the input range: Change the calibration. If you have a programmer, you can do that in the sketch in the emonTx, the best place; else in emonhub.conf you change the "scales = " coefficient for the powers. (If you do that, I can supply details.)
  2. The preferred: Remove the internal burden resistor in your emonTx. It’s the larger SMT component immediately behind the socket, labelled “22R0” (22 Ω) with two holes for a wire-ended resistor, one at each end. You’ll still need to recalibrate.
  3. Money!

I’m going to struggle a bit to work out the new calibration for you for option 1, as I believe they use slightly different numbers of secondary turns for the “V” versions, and I don’t know the values they use for their internal burdens.
For option two, the calibration coefficient in the sketch will be 60 for you, in emonHub it will be 0.6

11

Hi Robert
I think I’m going to take the ‘chicken’ way out and buy the correct CTs.
Thank you so much for guiding me through this - for the past 10 months I’ve been thinking my solar array has been seriously under performing!

All the best!