Archived Forum

Hi Peter,

I actually research the mechanics of the grid frequency with a view to developing dynamic demand management processes as well, so I'd be interested to see how you get on with this kit. I plan on doing a similar investigation in the summer to the one you wish to do.

I have not yet ordered or built any of the parts yet, but my understanding is that the emonTx is perfectly suitable for measuring grid frequency. You obviously can't plug the mains directly into the board but you can transform the voltage down to 9V using an adapter and just build the emonTx with the optional AC input. This page has a schematic including a 9V AC input and it links to this Google photo album showing pictures of the assembly (if you click on the individual pictures there are instructions in order).

I cannot give you any advice regarding the software side of things, but presumably the firmware on Github developed by the community here will allow you to measure the grid frequency. You'll have to get answers from someone else regarding the sample frequency ability of the emonTx, but presumably it is in the MHz region. For grid frequency a sample frequency of 1kHz has sufficed for myself in my research, it gives you plenty of accuracy for conducting fourier transforms.

I am just about to attempt to build a system which outputs a different signal depending on whether the frequency is above or below 50Hz, without having to resort to using fourier transforms, which are obviously very costly in terms of memory and processing power. If you're interested in the mechanics of how it works, send me a message.

Hi Sean,

Thanks for reply. I will start with emonTx as you suggested, any idea when we can buy it.

I had hoped to use Voltage to measure frequency and that this could be done relatively cheaply  and accurately- something around counting zeros or the like ( I'm a novice but hope to improve as I build project). 

Keep in touch.

Cheers

Peter

Apparently they will be available shortly to buy ready made, but you can assemble the kit yourself if you're happy to do a bit of soldering etc.

Counting zeros is a method I am currently investigating (alongside a band pass filter method). I'd bear in mind that this method might suffer from timing issues to do with the time it takes to execute each loop of the code, and in the 16MHz clock cycle not actually being exactly 16MHz. But this probably won't have a major impact on the measurements, and ultimately you could correct for it by using a frequency generator to generate an exact 50Hz frequency and seeing what your emonTx thinks it's measuring the frequency as.