Maximizing Precision

For more advanced meter data users. Learn from others and ask questions about how to read meters, how to use the EKM Push Data, and how to display the meter data in useful ways.
Post Reply
cmh
Posts: 6
Joined: Sat Oct 14, 2017 9:59 am

Maximizing Precision

Post by cmh »

Hi there, I am trying to monitor a circult several times per hour that has relatively low current draw for research purposes. I have the OmniMeter I v.3, but it only seems to report in 0.1 kWh increments, and my circuit can take several hours to increase by 0.1 kWh. This leads to several questions.


Question 1
I see the OmniMeter I v.3 also has a pulse output at 1/800 of a kWh. Is it possible to somehow read the meter with this greater precision so I can get better resolution on the power usage?

Question 2
Alternatively, I am thinking about returning the OmniMeter I v.3 and instead purchasing the OmniMeter Pulse v.4. Would this be the better approach?

If I do that, I am not clear from the API documentation if can I read the meter:
A) in 1/800 kWh increments (can I read the pulse count directly? or would I have to loop the meter's own pulse output into one of its pulse inputs?)
B) in 0.01 kWh increments by somehow setting kWh_scale to 0x32. (The RS-485 documentation says to do this in the A request, but the A request documentation does not list kWh_scale as a request parameter)

Appreciate any feedback.
Jameson
Posts: 860
Joined: Fri Nov 04, 2011 7:42 pm
Location: Santa Cruz, CA
Contact:

Re: Maximizing Precision

Post by Jameson »

Hello cmh,



There are a few approaches you could try.

1. Get a v.4 meter: The Omnimeter Pulse v.4 has a kWh resolution with 2 decimal places when using 100amp, 200amp, or 400 amp CTs. For example it will report kWh as 999999.99 kWh. The v.3 meter only reads kWh to one decimal place (no matter what CT ratio it is set to).

2. Count the pulse outputs from a v.3 meter or v.4 meter: The v.3 meters pulse at a rate of 800 pulses per kWh. The v.4 does the same from its normal pulse output, but it also has a separate pulse output called the "Settable Pulse Output" that can pulse as much as 1600 times per kWh, when set to a 100 amp CT ratio.

See fig 6 here for "Settable impulse constant range": http://documents.ekmmetering.com/EKM-Om ... -Sheet.pdf

3. You can wrap the wire you are measuring multiple times through a CT to get a higher resolution using math:
Wrap the wire you are measuring through your CT 10 time for example. You meter will read 10 times too fast, but if you divide all of your amp based values (amps, watts, and kWh) by 10 you should have a higher granularity end result that will be accurate.
Jameson
EKM METERING
http://www.ekmmetering.com
831.425.7371
cmh
Posts: 6
Joined: Sat Oct 14, 2017 9:59 am

Re: Maximizing Precision

Post by cmh »

Thanks, this is very helpful. Multiple wraps through the CT won't work here because its 6 AWG wire and there's not enough space in the panel or CT for wrapping. It sounds like the v.4 is the way to go.

Do you envision any problems if I use the v.4 to count its own pulses? I assume I can wire the settable pulse output running at kWh/1600 to one of the pulse-counting inputs and then query that input over RS-485?

Given I'll be counting a lot of pulses if I run the setup for a long time, is there a maximum number of pulses that the meter can count up to? And what is the behaviour when that number of pulses is exceeded (overflow back to 0?)
Jameson
Posts: 860
Joined: Fri Nov 04, 2011 7:42 pm
Location: Santa Cruz, CA
Contact:

Re: Maximizing Precision

Post by Jameson »

Yes, you can use the V.4 meter to count its own pulse output. You would want to use the "Settable Pulse Output". The v.4 meter can count as fast as 20 counts per second.

The v4 meter has 3 separate pulse inputs. Each can count up to 99,999,999 (8 nines) before rolling over to zero.
Jameson
EKM METERING
http://www.ekmmetering.com
831.425.7371
cmh
Posts: 6
Joined: Sat Oct 14, 2017 9:59 am

Re: Maximizing Precision

Post by cmh »

Hi Jameson,

I've got the v.4 meter hooked up, and have jumpered terminals 15 and 11 to connect the variable rate output of the meter to its own pulse 1 input. I left all settings at default (including ratio of 1/800 for the pulse output because I didn't replace the CTs).

I am getting unexpected results comparing watts, pulses and change in kWh. With a 60s interval, and a roughly constant load, the meter reported 4336 Watts, an increase of 0.08 kWh on the total kWh and the pulse counter increased by 25. The Watts and kWh make sense (allowing for some minor variation in load and rounding error) but the pulses are not right. 25 pulses @ 800 pulses/kWh = 0.03 kWh, nowhere near 0.08 kWh. It should have generated/counted 64 pulses in that minute.

Can you provide any assistance/ideas to get the pulse counting working?

For info, here is a dump of the meter parameters for both the A and B responses.

Code: Select all

Connected to /dev/cuaU1
Sending request with CRC = 7b5e. Message = 2f 3f 30 30 30 33 35 30 30 30 31 34 33 38 30 30 21 0d 0a 7b 5e.
Meter response CRC success. Calculated CRC = 7759. Received CRC = 7759. Response = 02 10 24 1c 30 30 30 33 35 30 30 30 31 34 33 38 30 30 30 30 32 32 35 34 30 30 30 30 31 32 39 34 30 30 30 30 30 30 30 30 30 30 30 30 31 31 34 36 30 30 30 30 31 31 30 37 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 32 32 35 34 30 30 30 30 30 30 30 30 31 31 37 32 31 31 38 36 30 30 30 30 30 30 30 34 32 30 30 30 34 32 30 30 30 30 30 30 30 30 30 32 34 38 30 30 30 30 32 34 34 30 30 30 30 30 30 30 30 30 30 30 34 39 32 4c 30 34 39 4c 30 34 38 43 30 30 30 30 30 30 30 34 34 38 30 30 30 30 34 34 38 30 30 30 30 30 30 30 30 30 30 30 38 39 38 36 30 30 35 30 30 30 30 37 39 35 35 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 31 31 32 30 30 31 37 30 39 33 30 30 31 30 30 30 39 31 38 30 30 21 0d 0a 03 77 59.
	type:	02
	model:	1024
	firmware:	28
	address:	000350001438
	kwh:	000022.54
	kvarh:	000012.94
	kwh_rev:	000000.00
	kwh_l1:	000011.46
	kwh_l2:	000011.07
	kwh_l3:	000000.00
	kwh_l1_rev:	000000.00
	kwh_l2_rev:	000000.00
	kwh_l3_rev:	000000.00
	kwh_reset:	000022.54
	kwh_reset_rev:	000000.00
	volts_l1:	117.2
	volts_l2:	118.6
	volts_l3:	000.0
	amps_l1:	0004.2
	amps_l2:	0004.2
	amps_l3:	0000.0
	power_l1:	0000248
	power_l2:	0000244
	power_l3:	0000000
	power:	0000492
	powerfactor_l1:	L049
	powerfactor_l2:	L048
	powerfactor_l3:	C000
	var_l1:	0000448
	var_l2:	0000448
	var_l3:	0000000
	var:	0000898
	frequency:	60.05
	pulse1:	00007955
	pulse2:	00000000
	pulse3:	00000000
	pulse_state:	0
	direction_state:	1
	output_state:	1
	kwh_decimals:	2
	reserved:	00
	time:	17093001000918
	resptype:	00
power=0000492
pulses=00007955
kwh=000022.54
Sending request with CRC = 7a22. Message = 2f 3f 30 30 30 33 35 30 30 30 31 34 33 38 30 31 21 0d 0a 7a 22.
Meter response CRC success. Calculated CRC = 4c16. Received CRC = 4c16. Response = 02 10 24 1c 30 30 30 33 35 30 30 30 31 34 33 38 30 30 30 30 31 37 30 38 30 30 30 30 30 35 34 36 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 31 31 37 32 31 31 38 36 30 30 30 30 30 30 30 34 32 30 30 30 34 32 30 30 30 30 30 30 30 30 30 32 34 38 30 30 30 30 32 34 34 30 30 30 30 30 30 30 30 30 30 30 34 39 32 4c 30 34 39 4c 30 34 38 43 30 30 30 30 30 30 35 32 37 30 30 31 30 30 30 31 30 30 30 31 30 30 30 31 30 32 30 30 30 30 38 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 00 00 00 31 37 30 39 33 30 30 31 30 30 30 39 31 38 30 31 21 0d 0a 03 4c 16.
	type:	02
	model:	1024
	firmware:	28
	address:	000350001438
	kwh_tariff1:	00001708
	kwh_tariff2:	00000546
	kwh_tariff3:	00000000
	kwh_tariff4:	00000000
	kwh_tariff1_rev:	00000000
	kwh_tariff2_rev:	00000000
	kwh_tariff3_rev:	00000000
	kwh_tariff4_rev:	00000000
	volts_l1:	117.2
	volts_l2:	118.6
	volts_l3:	000.0
	amps_l1:	0004.2
	amps_l2:	0004.2
	amps_l3:	0000.0
	power_l1:	0000248
	power_l2:	0000244
	power_l3:	0000000
	power:	0000492
	powerfactor_adj_l1:	L049
	powerfactor_adj_l2:	L048
	powerfactor_adj_l3:	C000
	max_demand:	00052700
	max_demand_period:	1
	pulse_ratio1:	0001
	pulse_ratio2:	0001
	pulse_ratio3:	0001
	ctratio:	0200
	reserved1:	0
	pulse_output_ratio:	0800
	reserved2:	00000000000000000000000000000000000000000000000000000
	time:	17093001000918
Sending request with CRC = 5544. Message = 01 42 30 03 75 55 44.
Jameson
Posts: 860
Joined: Fri Nov 04, 2011 7:42 pm
Location: Santa Cruz, CA
Contact:

Re: Maximizing Precision

Post by Jameson »

Will you run the test again? This time manually count the red flashing light on the face of the Omnimeter. The red flashing light indicates a pulse. So your test results for the time period of the test will have Total kWh Incremented, the Pulse Counts read by the Omnimeter Pulse v.4, the visual count on of pulses from the face of the meter.

It would also be helpful to run a separate test for about an hour showing the Total kWh Incremented and the Pulse Counts read by the Omnimeter Pulse v.4.
Jameson
EKM METERING
http://www.ekmmetering.com
831.425.7371
Doug
Posts: 1
Joined: Wed Aug 13, 2014 11:21 am

Re: Maximizing Precision

Post by Doug »

I tried a similar test here and have an observation.

I'm jumpering 16 with 13, and 15 with 12.

The pulse count on the settable pulse output is about half of the fixed 800 imp/kWh output. The 800 imp/kWh is tracking quite well with the reported kWh.

CMH--I suggest you use the 800 imp/kWh for your project, not the settable pulse output.

Doug
cmh
Posts: 6
Joined: Sat Oct 14, 2017 9:59 am

Re: Maximizing Precision

Post by cmh »

Thanks Doug, very helpful. I will switch over to the fixed output as I am not using the right CTs to get the 1600 pulses/kWh working anyway.

Jameson, here is the test result you requested:

Over a 60 second interval, I counted 59 flashes of the LED. According to RS-485 data for that same 60 second period, the meter counted 29 pulses (9520 to 9549) and recorded 0.07 kWh (26.93 to 27.00). The wattage reported during the period was approximately 4370 W. Here the flashes and kWh agree with the wattage, but once again the pulses are not working.

Over a 1h interval, the load is not constant, but roughly averaged 2246 W. Over the hour I recorded 2.11 kWh (24.89-27.00) and 755 pulses (8794-9549).


p.s. Jameson, if you want me to run any other tests against the settable pulse output I am happy to do so. -Colin
Post Reply