Voltage discrepancy fix just deployed - let us know if you still see issues

Update on voltage discrepancy issue: we just deployed a fix. Please read on for more details.

We did indeed find a voltage calibration problem in our production process. We conduct a careful calibration and QA step for every monitor we produce, but had an error in the reference for the voltage part of the calibration. We first suspected something was wrong with the reference itself, but found it was actually correct. The problem was that in our factory setup we had an unexpected voltage change between the point where the reference was measured and where the Sense monitors were being powered for the test.

We are making changes to our calibration and testing in the factory for future monitors (and adding more post-production QA steps). And, the reference error was constant so we’re now deploying a software update for existing monitors to correct this.

The release should correct any places in the app that display wattage or voltage both in any real-time and historical views. Please let us know if you are still seeing discrepancies, so we can work together to resolve them.

6 Likes

I am seeing 122 volts and 121 volts on two UPS units that are on different legs far from the breaker box and am seeing Sense report 121.9 and 121.8 on the legs so I think you have it now. Previously Sense was showing in the range of 117-118 volts at the panel. Nicely done, thanks!

5 Likes

Great! I’m now reporting very close to my actual readings.

Just looked before reading this post and thought it looked higher than it has in the past :slight_smile: :slight_smile: :slight_smile: :slight_smile:

I just wanted to add that you guys have completely restored my faith in the company and the device. For a while, I was getting disheartened with the data it was providing. I’d now have no reservations recommending it to energy-conscious family, friends, and local businesses. You’ve earned a customer for life :slight_smile:

This wasn’t an easy thing to do - admit there was an error, describe what it was and how you were going to fix it with so many units delivered already. Yet it needed to be done and I’m happy you did. All to often tech companies lose their touch with their consumers these days. I think you guys have put yourselves in a good position to handle any problems that come your way by putting your best foot forward here.

Cheers to you guys, this was a great Xmas present. Happy Holidays to everyone there.

5 Likes

Thank you! We really appreciate the feedback. We are a small company (~20 people now) and we are working very hard to bring a brand new, very hard, and groundbreaking technology to market. There is a long list of improvements and issues we are working through, but we think we are on a great path and with your help we are working hard to make Sense truly great. We really appreciate your help in bringing issues like this to our attention and helping us get to the bottom of them. We’re committed to delivering the best possible product and input like this really helps!

5 Likes

Hi Rhoda,
Just preliminarily, and with the caveat that I’m waiting on replacement components yet, but it still looks like my solar watt reports are off by about 6.5%. I’ll wait for more conclusive evidence, but just wanted to post up in case others are still seeing a discrepancy. The volts seem more appropriate, but the watts haven’t budged much. Thanks for checking!

6.5% under again today (31.2/33.4):

Ugh. Thanks for the heads up, Nicholas. I’m passing this on to the Engineering team to have a look.

Maybe the fix wasn’t hitched to solar? I’m still limping with one main ct, so can’t tell yet if usage is accurate.

The error and fix was in the voltage, which would have affected solar in the same way (the main and the solar CTs measure current and are multiplied by voltage (point by point throughout the cycle) to get wattage.

I’ll reserve any further judgement until I get my replacement items in. After I get things set up I’ll collect cumulative data again for another week and post up. Perhaps there are two issues as someone suggested elsewhere, and the voltage fix only corrected some percentage of that error. Alternatively, maybe I’m the only one and everyone else is getting perfectly satisfying data. It’d be nice if they’d drag those outside conversations here to the community forum so we can get a consensus :slight_smile:

I was collecting data for a while, but stopped. Guess I’ll have to start doing it again and see if the wattage has changed based on the voltage changes.

@Howard do you have solar? I think @NJHaley was flagging this specifically for solar, and as he is waiting for new sensors it could be particular to his home. If you do see an issue w solar calibration, of course, please let us know! Thank you!!

Yes I have solar, I have a 10KW array - that is the main reason I bought Sense.

3 Likes

We just finished our first full electric billing cycle, and adding up our daily totals, and accounting for the first day being off, due to being installed around noon, it is perfectly matched with the electric meter. Great job Sense! :smiley:

4 Likes

Hmm. Wondering if my wonky CT threw off my calibration. I’m all hooked up with my new equipment, just waiting for an account reset so I can log in and get back to sensing :slight_smile:

I installed my monitor over the weekend and it has been under reporting. Voltmeter says 120-121 and sense says 117-118.

What version of the firmware is on your unit, perhaps it has not been on long enough to receive the patch. I believe I was the person that first noticed this error and once they patched it mine immediately started reading correctly.