Eyedro monitoring system for a wall connector circuit

[Mike]

I installed the Eyedro (wireless) energy monitoring system for my (Tesla Wall Connector) EVSE 60 amp 240 volt circuit:

http://eyedro.com/home-electricity-monitors/

I'm starting this thread in the hopes that some other folks that are using this brand of energy monitor can provide some insight into setting up the software interface side of things.

The system is installed and (as per the Eyedro instructions) only one sensor is being used because this is a balanced circuit with two leads and a ground (no neutral).

I did a 20 second test with the stove circuit esterday and when I turned the oven on, the system responded instantly so I know the raw data works:

My circuit breaker is on and my Tesla Wall Connector is live and has been so for almost 30 hours now, but in this snapshot of "today", no consumption is shown (unless the vampire drain of the Tesla Wall Connector is too small). And it does not list any money having been spent (even though I did set up all the TOU information):

Also, the "bill estimate" dates seen in the lower right corner are fixed for a standard calandar month, but our billing cycle does not end on the last dy of a month, but early morning of the 27th of each month as that is when our smart meter pushes the formal reading to Hrdro One Networks .

Anyhow, if anyone out there uses the Eyedro monitor system, I have some (minutia) questions regarding how to set things up so the monthly harvest of data does not turn into an epic 20 minute activity......

Signed,

Guy who knows nothing about computer software and hates software that expects you to do something that is not intuitive.

 

[Glade_EV]

The Tesla wall connector probably draws less than a watt in idle mode and I suspect the resolution of the eyedro is more than that. I had my mobile connector idling for 24 hours and also didn't register any power usage.

On the Rates page, there is an "effective date" under "Rate Structure Configuration"/"General". Set it to something like Apr 27, 2018 (or any month, year) and it should look better to you. (It appears that with the monthly "Reports" you don't have a choice - you're stuck with first of month)

 

[Mike]

The Tesla wall connector probably draws less than a watt in idle mode and I suspect the resolution of the eyedro is more than that. I had my mobile connector idling for 24 hours and also didn't register any power usage.

@Glade_EV, thanks!

I'm thinking the draw must be super low.

Since your unit is not registering any power usage either, I feel better about my setup now.

On the Rates page, there is an "effective date" under "Rate Structure Configuration"/"General". Set it to something like Apr 27, 2018 (or any month, year) and it should look better to you.

Okay, here is what I have now:

I also set up my TOU rates to capture all taxes, fees and distribution surcharges into the TOU rates page:

The reason I did that is if I simply entered our advertised rates (13.2 cents/9.5 cents/6.5 cents) and tried to play around with the "other fees" entry page, I found the available entry fields not sufficient to deal with our setup here in Ontario:

I end up running my monthly hydro bills into a simple spread sheet to figure actual retail price per TOU period:

 

[Mike]

Update:

I arrived home from the delivery event last night and after putting my car to bed, it had 50% battery remaining.

I set to charge to 80% and this morning my car reads 80%.

So I injected 30% last night.

Back of napkin math: 10% of a LR Model 3 battery should be "around" 7 kWh.

So I expected my Eyedro setup to announce that yesterday I used "about" 21 kWhs.

My Eyedro results are about 1/2 (half) of my expectations:

My Tesla Wall Connector is considered a "balanced" load (two line wires plus ground, NO neutral).

As per Eyedro's instructions, I only have one sensor being used on one of the TWC line wires and have the sensor set to 240 volts:

I'm going to add the second sensor and see if that helps things.

Will report back tomorrow......

 

[Mike]

Update: charging from 80% to 90% with both sensors introduced an error of summing:

So, configuring with only sensor "b" I get real world live use:

But.....non of the live raw data is being captured into my billing info, etc....i.e. it is tracking the disconnected "A" sensor and not the real live working "B" sensor:

So, live feed of actual power is good (sensor B), but no data is being captured......

My e-mail to Eyedro support has been sent.

Back on Monday when I hear back from them.......

 

[Glade_EV]

EyeDro's user interface isn't the friendliest - I seem to recall something I was trying not showing up properly until I exited the website & re-entered (or maybe signed out and re-logged in?). - There is also a distinct lag on the Highlights screen (> 5 minutes)

I know I tested the configuration that you are using (it seemed to work fine), but since I don't currently need to monitor anything else in my house I'm using both sensors (one on each line) ; set to 122.5 volts each (my measured voltage is averaging 245 V) and monitoring the aggregate power.

You're getting closer...

 

[Mike]

Right now I'm charging from 80% to 90%.

The car says 11 kW

This is my setup:

So, as we can see, I've pulled the sensor input out of the box labeled "B" and set sensor "A" as above........but sensor "A" was physically moved onto the other actual line....(?)

So, raw data now shows 11.29 kW of power, which agrees with the IU, which agrees with my back of napkin math.

The problem comes to the info actually reaching the other pages.

Here is the live demand, after me moving input leads around:

So, raw sensor data and "live demand" are showing the true data right now.

Problems begin when I go to "highlights" page:

The raw data in the upper right corner is showing the correct power right now, on the speed-o meter showing 11.29 kW.

But look at the circled info in the lower right. It's showing "0" cents per kWh as the current rate (all my rate data is inserted and accounted for). That $2.75 and $2.76 are static.....not moving at all.

Sidebar: You'll also notice in the above screen shot the time at 1200 showing use of 12.307 kWh to raise the car from 80% to 90% after lunch today prior to a 50 km drive....at 15 cents all in for power at off rate (weekend), that 50 km drive cost me $1.84 and one liter of fuel here today is $1.35. So over 50 km, it cost the same as 1.37L worth of fuel for that 50km.......or 2.73L/100 km.

Almost twice as efficient as my 08 Prius used to be (4.8L/100 km)


Taking your advise, I just logged out and logged back in and now look at the highlights page:

Bill to date went from $2.75 to $3.04.

Maybe I got this sorted out (?).

I'm only using one sensor actually plugged into the kit.

Said sensor seems to be satisfied on the different line than I was using last night.

Sop for balanced load, you need one sensor, have this system recognise both sensors, but only plug in one sensor.....but you have to experiment which line that sensor has to go on.

It seems to not be "omni" in its approach, but is for some reaason sensitive to which of the line leads is selected and the used as the host for the sensor....

 

[Mike]

Update: charging session from 80% to 90% complete.

Car UI shows 8 kWh's of energy added.

Updated "highlights" (with session total added for clarity because screen shot would not capture figure):

So, to recap, a 50 km drive took 8.396 kWh's (at my circuit breaker panel). All in my energy costs me .15 cents per kWh right now.

.15 multiplied by 8.396 = $1.25 for 50 km's.

I hope some Canuks (with our $1.35 gas) are viewing these numbers........

 

[Mike]

Another update: with about 36 hours of data since bringing the car home from Toronto:

My back of napkin math was $5.18 so far and Eyedro has $5.27.

So my goof-ups have cancelled each other out.

 

[Mike]

Just a heads up for any other Eyedro users.

As of yesterday (about 0630), I keep getting this:

I've sent an e-mail to these folks, but it's a long weekend up here so I figure I won't hear back until at least Wednesday.

If I try to log on via any devise, using any browser, I get the same issue.

On Explorer, some error code that involved "AJAX" showed up yesterday.......?

 

[Mike]

For whatever reason, I'm able to get my Eyedro info back from their server again .

Some numbers for me to ponder:

First, this is what Eyedro says was pushed to my car via my EVSE circuit for June 2018:

343.615 kWhs.

Then, (according to the readout on the car itself for each session), I had a total of four supercharger sessions in June that added up to 67 kWhs (+/- 1).

Thus, my total June 2018 energy use into my car is (343.615 + 67) 410.615 kWhs.

My archive shot of the end of month data, as per my car screen shows 349 kWhs:

My data says 410 kWhs was pushed into this car for June 2018.

The car says 349 kWhs was "used" by this car for June 2018.

I expected an 8% loss due to the AC to DC converter in the car.....but this looks to be about double that (17%).

Anyone else tracking actual grid enegy used versus distance traveled?

410 kWhs over 2533 kms equals 162 Wh/km, not the indicated 138.

Any thoughts on this discrepancy is welcome.

Note: raw data taken directly from the smart meter (single day example shown below) adds up to within 1% of Eyedro reading, so I am confident of the Eyedro figures....

 

[Glade_EV]

You don't seem to be taking into account "phantom" losses. I believe the trip meters are only measuring the energy to the drive system.
25 kWh over a period of a month for electronics and climate control wouldn't be unreasonable.

-I would have thought that the trip meters would take into account climate control (at least when the drive system in engaged), but I just sat in my car in the garage with the heater running, gear select in Drive with my foot on the brake until the available range dropped a couple of miles and the trip meter energy usage never budged. -Then again, maybe a longer test is called for that uses more than a kWh (too bad it's not winter).

 

[PNWmisty]

My data says 410 kWhs was pushed into this car for June 2018.

The car says 349 kWhs was "used" by this car for June 2018.

I expected an 8% loss due to the AC to DC converter in the car.....but this looks to be about double that (17%).

The system you are using tracks all electricity used after the sensor. But it only measures current, not voltage, it makes the calculation based upon the default of 120V (unless you have changed the default). If your circuit is getting voltage drop, I think it will measure a higher amount. So, you may need to measure voltage drop when the car is charging and change the default voltage to a lower value. It also measures losses in the WC connections, the WC charge cable and the cars charge port. You might use a non-contact infra-red thermometer to look for warm spots in this path.

Also, do the sensors have a built-in self-centering system to ensure the wire being measured is centered? The wire should be straight for a short distance on either side of the loop and go right down the middle of the sensor loop.

 

[Mike]

I would have thought that the trip meters would take into account climate control (at least when the drive system in engaged), but I just sat in my car in the garage with the heater running, gear select in Drive with my foot on the brake until the available range dropped a couple of miles and the trip meter energy usage never budged.

This is what I am starting to think is the case, the numbers on the trip odometers only show actual energy used to drive the wheels....

 

[Mike]

The system you are using tracks all electricity used after the sensor. But it only measures current, not voltage, it makes the calculation based upon the default of 120V (unless you have changed the default). If your circuit is getting voltage drop, I think it will measure a higher amount. So, you may need to measure voltage drop when the car is charging and change the default voltage to a lower value.

Understood. The setting I have is based on what I actually see at the car screen "most often", which is 240 volts (it always shows between 238 and 242 at the car screen itself).

It also measures losses in the WC connections, the WC charge cable and the cars charge port.

Agreed. This is why I have this set-up, to include all losses as I have to pay for them upstream at the actual hydro meter.

Also, do the sensors have a built-in self-centering system to ensure the wire being measured is centered? The wire should be straight for a short distance on either side of the loop and go right down the middle of the sensor loop.

My setup:

For this first month, I have taken data directly from (Ontario) Hydro One Networks and compared to my Eyedro data (screenshots of daily raw data):

Acknowledging all the recommendations made here, my "raw hydro data" versus my "Eyedro" data is within 1% of each other.

So, although the voltage setting may not be 100% perfect, or the sensor setup not 100% perfect, my only care is what Hydro One Networks says is my (car) use.....and I am satisfied that my Eyedro setup is reflecting that.

Which goes to my core observation: I use about 17% more juice than the car says I do.

 

[PNWmisty]

Understood. The setting I have is based on what I actually see at the car screen "most often", which is 240 volts (it always shows between 238 and 242 at the car screen itself).

If it shows 238 when charging and 242 when idle, there's almost 1% error right there. Does it ever read 236V when charging?

Agreed. This is why I have this set-up, to include all losses as I have to pay for them upstream at the actual hydro meter.

So, although the voltage setting may not be 100% perfect, or the sensor setup not 100% perfect, my only care is what Hydro One Networks says is my (car) use.....and I am satisfied that my Eyedro setup is reflecting that.

I'm not following how you are able to compare the Hydro meter readings to the Eyedro readings. Are you saying that the electric car charger is the only consumption on that Hydro meter? I admit I didn't try to review all your raw data. Also, I'm a little surprised I didn't find any recommendations in the Eyedro Owner's Manual to position the wire centered in the inductive loop. My experience has been that current measurements can vary quite widely depending upon the shape of the wire and the position of it within the inductive loop.

I know you feel that the similarity between the Hydro and Eyedro numbers guarantee accuracy but I would be more inclined to consider there may be inaccuracies that cancel each other out.

 

[Mike]

Does it ever read 236V when charging?

Most of my charging is in the middle of the night, so I can't say for certain......but I have never seen it that low during the day when I recharge.

I'm not following how you are able to compare the Hydro meter readings to the Eyedro readings.

My routine is to take a screen shot (for archival purposes) of what Eyedro says it pumped into the car and what the smart meter said.

Here is a random (for June 2018) example, which shows a large uplift:

23 June 2018.

A screenshot of what Eyedro says went through the (dedicated EVSE) circuit:

It shows 34.324 kWh's flowed through my EVSE circuit on 23 Jun 2018.

My whole home, smart meter, 23 Jun 2018:

In this example, we see that the charging was a three hour event, starting at 0200 and ending at or before 0500.

The total home uplift for those three hours was (11.8752+11.9964+12.0432) 35.9148 kWhs.

One must subtract what the actual house use was between 0200 and 0500 on 23 Jun 2018.

We can quibble about the +/- 1kWh accuracy in my technique to arrive at actual house use between 0200 and 0500, but for the record I subtracted (0.4 + 0.5 + 0.6) 1.5 kWh's to represent the house use between 0200 and 0500.

Hydro meter 35.9148 kWhs minus "best guess house use" 1.5 kWhs equals 34.4148 kWhs.

Hydro One Networks smart meter reading of car re-charge for 23 Jun 2018 says 34.4148 kWhs.

Eyedro says 34.324 kWhs.

34.4148 versus 34.324 equals splitting hairs.

This example is typical (repeatable) of all days in the month of June 2018 (first month of car ownership and therefor data.

I know you feel that the similarity between the Hydro and Eyedro numbers guarantee accuracy but I would be more inclined to consider there may be inaccuracies that cancel each other out.

That may be true, but from the data I have, what Eyedro says was uplifted to my car is accurate enough (within 1%) for me.

 

[PNWmisty]

Thanks for explaining that, now I understand how you are guestimating the usage via the Hydro meter.

That may be true, but from the data I have, what Eyedro says was uplifted to my car is accurate enough (within 1%) for me.

This doesn't make sense to me. If there is the possibility of self-cancelling errors (between Hydro derived numbers and Eyedro) then the fact that they are within 1% doesn't show accuracy, it could simply reflect that the errors self-cancel to within 1%.

 

[Mike]

This doesn't make sense to me. If there is the possibility of self-cancelling errors (between Hydro derived numbers and Eyedro) then the fact that they are within 1% doesn't show accuracy, it could simply reflect that the errors self-cancel to within 1%.

(I'm no longer on my laptop, with all my data at hand)

I had about 21 home charging events in June.

All 21 of them, using the technique to compare the home smart meter to EVSE Eydro meter (as explained above), yielded results within 1% of each other.

If anyone else uses Eyedro for their EVSE circuit AND has a home smart meter, I'd like to see more anecdotal examples of 1% accuracy.

 

[PNWmisty]

I had about 21 home charging events in June.

All 21 of them, using the technique to compare the home smart meter to EVSE Eydro meter (as explained above), yielded results within 1% of each other.

I understand that. But self-cancelling errors can be very consistent in nature. Your Hydro meter could be out of calibration for whatever reason and the Eyedro could be reading high due to the way the wires enter/exit the induction loops. If this is the case, you could be getting billed for more electricity than you're using.