Getting more resolution using a v.3 meter

Forum is open to all questions, answers, and discussions related to electric meters, ours or others.
Post Reply
eventhorizons
Posts: 5
Joined: Mon Jul 23, 2018 1:08 pm

Getting more resolution using a v.3 meter

Post by eventhorizons »

I have two v.3 meters on my building (model "021017"), one to read the "consumption" of the building and one to read the solar "production". I have been running this setup for about 4 years now.

Originally I had a custom PHP script that would collect the data and display some graphs on my web page. It was basic, crude and not flexible at all but it worked very reliably.

Recently I implemented a "Cacti" server to monitor my network and I decided to move the meter data to it. I modified my PHP script and implemented a "Data Source" for Cacti.

The annoyance I am experiencing is that the kWh reading are in 0.1 increments. This causes an amount of "spikiness" on the graphs, caused by the difference between the 5-minute polls from Cacti and the rate at which I'm moving energy. It works fine, but it looks ugly.
cacti_graph_133.png
cacti_graph_133.png (21.41 KiB) Viewed 43744 times
I would like to increase the resolution at which I read the data.

My first thought was that I could access an internal pulse-count register, but when I reviewed the protocol I do not see any such animal. There are three pulse count registers in the data packet, but I don't see any references to being able to use them on the V.3 meter. I'm assuming they were used on some other models of meters?

In theory I could upgrade from v.3 meters to v.4 meters. This would actually be my preference, but I really can't justify the expense.

Increasing the turns through the CT's is not an option in my situation.

What would be the downside to simply telling the meters that they had, for example, 1000A CT's, instead of the 200A ones?

Thanks for your input.
Jameson
Posts: 860
Joined: Fri Nov 04, 2011 7:42 pm
Location: Santa Cruz, CA
Contact:

Re: Getting more resolution using a v.3 meter

Post by Jameson »

Hello Eventhorizons,

This is very cool! What kind of a machine are you running this on? Are you reading your Omnimeters with a USB converter?

Before we move forward, can you clarify something? You mention that you would like higher resolution than the 0.1 kWh that the Omnimeter I v.3 provides. But then you show a graph of watts. Are you wanting higher resolution of kWh or watts?

Yes, you are correct that the Omnimeter I v.3 does not have pulse counting abilities. Only the Omnimeter Pulse v.4 and the Omnimeter Pulse v.4 UL can pulse count at this time: http://documents.ekmmetering.com/Omnime ... -Chart.pdf

If you want higher resolution of kWh, then yes, I would upgrade to the Omnimeter Pulse v.4. When the v.4 meters are set to 100, 200, and 400 amp CT ratios, then they have a kWh resolution of hundredths of a kWh ( 999999.99 kWh for example)

If you just set your v.3 meter CT ratio to 1000 instead of 200 (and your v3 meter is using 200 amp CTs), then the meter will multiply all amps based values (amps, watts, kWh) times 5. In order to get the correct value you would then divide by 5. This may give you what you want.

Keep us posted on how your project works out.
Jameson
EKM METERING
http://www.ekmmetering.com
831.425.7371
eventhorizons
Posts: 5
Joined: Mon Jul 23, 2018 1:08 pm

Re: Getting more resolution using a v.3 meter

Post by eventhorizons »

My setup has the two Omnimeters connected to a Ethernet-RS485 bridge (terminal server) I can't remember the brand, just something I picked up on ebay.
My Cacti server polls the meters every 5 minutes using a PHP script that I wrote. (I just finished another script to let it poll my old WX-200D weather station, so now I can see temperature/humidity/wind/rain data.)

Cacti (https://www.cacti.net/) is very customizable, but was originally intended for monitoring network performance. It likes to read "counters" that are constantly increasing in value. The PHP script reads the kWh value every 5 minutes and passes that to Cacti (as watt-hours) which then graphs the "rate" as "watts vs time". Since I was leveraging the built-in abilities of Cacti, it required little effort on my part to set up and works very well. (I did not want to use the "power" value because intermittent loads could be lost in the 5-minute polls.)

If I were to tell my meters that they are connected to 1000A CT's then all power-related readings should be multiplied by 5. I could then divide it back down either in the PHP script or in Cacti which should change the resolution from 0.1kWh to 0.02kWh. Unless I did something fancy this would cause a glitch in the graphs across the transition but on either side it would be correct.

However, nothing is ever free. What would be lost by such a maneuver? Precision? Accuracy? Range?

I am starting to write some scripts to process the collected readings and display cumulative production-vs-consumption for specified time periods. I had this before I switched to Cacti but it was very clunky and required editing files to change the time periods.
Jameson
Posts: 860
Joined: Fri Nov 04, 2011 7:42 pm
Location: Santa Cruz, CA
Contact:

Re: Getting more resolution using a v.3 meter

Post by Jameson »

Cool to hear about your system and Cacti.

Im not sure what you would lose by changing the CT ratio to 1000 amps, then dividing by 5. I would just give it a shot.

You should have the same accuracy and range. Not sure about precision.

Is your weather data also based on a count? We have a simple wind anemometer and rain gauge connected to the pulse inputs of an Omnimeter Pulse v.4 for a little experiment, it gives some interesting results:

Wind Anemometer Widget:
http://widget.ekmmetering.com/EKM2/?rea ... _reads=100

Rain Gauge Widget:
http://widget.ekmmetering.com/EKM2/?rea ... _reads=335
Jameson
EKM METERING
http://www.ekmmetering.com
831.425.7371
eventhorizons
Posts: 5
Joined: Mon Jul 23, 2018 1:08 pm

Re: Getting more resolution using a v.3 meter

Post by eventhorizons »

Late Breaking Update!

It's been 4 years since I posted my original question and I never got around to fiddling with the setting of the meter. You know...life!

However, my power consumption meter started acting up recently. It behaved as though it was not taking power readings as frequently and so it would use the same reading for long periods of time, sometimes many hours! This caused wildly erroneous power reporting.

A power cycle of the meter resolved the problem and I improved my scripts to send me an alarm if it happened again.

However, I was worried that my meter was failing and I had always wanted the higher resolution, so I pulled the trigger for a v.4 meter.

Got the new meter installed last week, amended my scripts for allow for the slight protocol difference with the v.4 meter and it was up and running.

This is the new appearance of the graph. So much cleaner and easier to read! (Compare with the graph in the first post)
20220721_ShopMonitoring.png
20220721_ShopMonitoring.png (18.44 KiB) Viewed 35066 times

The 2kW "baseline" load is my building's summertime swamp cooler, the server closet and whatever other loads are always on. You clearly see the power consumption from when I worked in the shop last night, running the mill and with the high-bay lights on.

The next step will probably be to install the old v.3 meter on my server closet. I will use smaller CT's but leave the meter set to 200A to increase the effective resolution. I'll make the corrections in my polling scripts.
Post Reply