Power Quality Chart - not up-dating - Nov 2022

[Kathy]

yeah, I’ve experienced similar.

I tried to engage early on about such weak links in the system but didn’t feel it was effecting anything.
So instead of experience frustration while trying to help pursue solutions … I have learned to stay clear of that realm unless necessary.

I felt it was time to give it another go to try to solve the gaps in the power quality data but it quickly turned into the loss of the entire power quality lab service.

And, it doesn’t seem like we know what is up yet. Only a few monitors are effected… how is that possible?

[Kathy]

@Dcdyer,

I do not know what is involved in producing the power quality / voltage graph but that being said I agree with you and want to reinforce your comment about the power quality lab /voltage info being just as important as the wattage info.

It is actually MORE important to me as I am off-grid with my own remote microgrid.

The power quality graph and the reports of unusual dips and spikes, as well as dropped data/gaps in the graph have alerted me to problems like a short-cycling water pump and put me on a search for possible floating neutrals or open grounds… scary shit.

So I have learned a lot even though it has also caused some frustration as I am not a member of the Sense target audience. For some, just having a tool to help increase awareness of electrical consumption, hopefully reduce electrical consumption is enough. And I am appreciative of the potential it has in those arenas.

But, Sense is not designed for totally off-grid, remote microgrid systems. It is virtually impossible to find anyone in the Sense community who is familiar with the normal power quality of an off-grid solar system. So, it is almost impossible for me to find help to understand what is going on.

Is the frequency of the Sense monitor set to operate within a narrow range around 60 hz, grid frequency?

This off-grid solar electric system utilizes frequency shifting for the communication pathway between the Tesla AC inverter and the SMA DC inverter. It was raising the frequency to 65 hz to shut down the flow when the Tesla batteries were full and house loads low. I got in touch with Tesla and they lowered the Frequency shift to a max of 62.5 hz which makes my back-up power supply and other devices happier. But still, like right now, the frequency is riding at about 62 hz as the sun is shining bright and my batteries are full.

Power quality is VERY important in the Off-grid world, including as a warning flag for possible problems with pumps, wiring, grounding, etc.

Unfortunately, I just came to the conclusion that Sense is not interested in the minority of its customers who place such value on the power quality info. I am always an outlier on the graphs and have come to accept that the mainstream world, the bell on the graph, is what most companies have to cater to with their business outreach.

But, it is sad that the value and potential of the power quality lab efforts doesn’t seem to be appreciated by the company. Perhaps that is why it has been sitting in the ‘testing lab’ for so long.

I’d like to see it come back on-line just as it has been. I’d prefer to see it move forward as you describe, including the in-house notification when the system goes down. And I’d prefer for the company to work to improve its communication process, especially within the "sense support’ team.

But, I fear I am watching Sense cut the strings on the power quality lab entirely. It will be a loss for me for sure and possibly many others.

But, as @kevin1 points out, the voltage data and frequency is still there and perhaps the only option to move forward will be for us to pursue the pi / home assistant project, with Sense integration.

[kevin1]

Hi @Kathy,

I’m not sure about the 2000W mystery - things get complicated with Sense on battery systems since Sense doesn’t see all the current flows the Tesla controller sees (3CT pairs vs 2 CT pairs), at least from what I have seen.

As far as the reason for the voltage divergence on my grid on Nov 22, I’m guessing it might have been my neighbor responsible. I don’t see any potential causes in my Power Meter for that time period (but I do see the cause for the 3:30A dip in both voltages).

Thank’s for also solving the mystery of why your unit has Ethernet.

[Kathy]

I’m not familiar with Tesla EV controller and CT’s, or how they compare With what may be inside the powerwall 2 batteries.

I’m as sure as I can be that the 2000 w mystery had something to do with the sense and/ or the ethernet connection.

The Tesla has always reflected the same data as the sense, until the other day.

And as I noted when the Sense was Re-booted the 2000 w disappeared; the Tesla and sense have continue to show the same data.

I didn’t realize that this was a pro monitor since it was a donation to our R & D project. It is nice to have the ethernet option but I am unable
To access the Sense Pro app as it is for installers only.

[Kathy]

So what was the drop at 3 am, and what funded on between 330’am and when it dropped again around 8 or 9 am (can’t see graph at the moment.)

[JuliaAtSense]

We would like to follow up on the issues you are experiencing with the Power Quality feature in Labs. First, we have identified the issue and are updating the service and implementing some updates to add more stability going forward. We have also streamlined our internal processes to better support Labs, and in the future, you can contact support with any issues you experience in Labs and we will respond in a more timely way.

[obscuredtrip]

@JuliaAtSense @CarolAtSense @aisha Can you PLEASE explain your streamlining process? Sense employee Anthony stated at 9:29am only a few users have reported this issue and support needs more info to properly document this issue in the internal ticket to engineering that just opened TODAY.
POLL: Is your Labs power quality graph current? PLEASE PLEASE reach out to your higherups and ask them to reach out to to us directly and publicly because it is IMPOSSIBLE to believe ANYTHING @JuliaAtSense or anyone in support says anymore.

[wood.jj]

I am currently seeing the same no update issue since 11-9-2022.

[Kathy]

Thank you for the up-date Julia.

I hope the customer input proves beneficial to Sense on their continuous improvement journey.

My power quality graph is still frozen at 157 AM on November 10th.

Kind Regards,

Kathy

[killercarver]

Stuck on 10NOV here too. Trying to find out where/how to put in a ticket…

[obscuredtrip]

To formally submit a ticket:
From the web app, on the top right scroll over your email address and click settings. On the left go to feedback, then select Support as the category.

Sense has intentionally made contacting support from the mobile app a process. Go to settings, then help, then support & Faq’s, then scroll 3/4 of the way down and tap ‘open the assistant’, that brings up a form where you can submit a ticket.
You know Sense is becoming sense LESS when event the ‘Contact Us’ link in the help section doesn’t actually provide a means to ‘contact’ Sense it just pulls up the annoying help bot.

You can also e-mail them (be sure to send from the same e-mail your account is registered to) at support@sense.com but that does not guarantee a ticket will be created.

To view the status of your tickets you can go to help.sense.com and click on ‘My help portal’.

[JuliaAtSense]

I apologize for the miscommunication here @obscuredtrip. We know now that this is a widespread issue. We understand that the Power Quality feature is important to our users, and rest assured, I have escalated this issue internally within our team.

[Kathy]

Current time: 1015 EST 12/13/22

I just noticed that my power quality lab is now displaying graphs back to 12/28/22.

However, I still have large sections devoid of data, other sections with erratic patterns.

The monitor is connected via a cat5e ethernet cable.

The internet upload speed tests are good.

We have not lost power or experienced any appliance/device problems.

I will resume my observations of the variables that might be contributing to the gaps and the erratic data.

Thanks to the Sense team for working on the power quality lab feature. I appreciate it.

Thank you @JuliaAtSense , @Emory and everyone who has provided input.

[kevin1]

Hi @Kathy,

I looked at the Power Quality in the app and the web app this morning and it appears that Sense has generated new complete dashboards for me for Nov 28th through Dec 10th, though there are still indicators that there are underlying code fixes needed (the date range is clearly wrong - Dec 10-Dec 10 and the “last {{0}} days” artifact remains. But @JuliaAtSense has highlighted that Sense is working on fixing things. Here’s my view.

I don’t have any dropouts in any of my dashboards, so I suspect yours might stem from data dropouts on your end. I have a hard time believing that your internet service caused the big drops in data you are seeing, or else you would see big issues on all your Sense data. I’m just speculating here, but I think that the main Sense app and view generation, have been “hardened” to deal with short data dropouts, but that Labs has not (since it is still more of a beta-test subsystem). So maybe even short chunk of missed/unknown data hoses up the graph generation until there’s another complete chunk.

I probably have a little more empathy for Sense than most because my wife runs a SaS prescriptive analytics company. I have heard her stories about having to rewrite cloud-based subsystems over and over again for different goals. Sometimes the steps below are combined, though one often doesn’t know the bottleneck for the next stage until they get there,

• First it is written quick and dirty to show the concept can work - prototyping for one customer
• Next it is re-worked for extensibility and better structure so they can add new features
• Re-written again for performance and concurrency to enable overnight or hourly updates
• Re-written again for cost (reduced connect times to AWS servers and more interactive stuff pushed to the app and browser)
• Re-architected for the next level of scale - dozens of users to thousands of users and beyond.

[Kathy]

hi @kevin1,

Yeah, thought I would share my news in case others had not noticed that there has been a step in the right direction. And, yes, it appears there are still some bugs to work out, re dates, etc.

You have such a nice power quality graph, relatively speaking. I don’t know what I would do with such normality

Yes, I thought the same with regards to the internet/wifi/ data drops. I am thankful to have a monitor with the ethernet port so we could rule out the wifi connection. And, glad the lab is back up so we can see if the ethernet cable made any difference … which it didn’t. But it’s always nice to rule out variables when dealing with complex and highly dynamic systems.

I wondered about the “hardened” or “smoothed over” components that may be functioning on the primary Sense tools but perhaps not on the lab … as of yet.

I am still wondering about frequency ranges and now the low wattage parameters … if that makes any sense.

As you recall, the Tesla inverters communicate with the SMA inverters with frequency shifting for solar curtailment, etc. The challenge with the possibility that it may be frequency screwing up my power quality graphs, is that the frequency only increases for solar curtailment when the batteries are mostly full and it’s still sunny outside. There doesn’t appear to be any gaps or freakouts that directly correlate to those time periods.

In addition, I consume very little electricity at times. Also no direct correlation with what is happening with the power quality graphs but I was wondering if there are any low wattage black-outs… just not enough of a draw for sense to pick up the voltage.

It’s sweet that your wife runs a SaS prescriptive analytics company. I had to look it up to get an idea of what it might involve. Pretty amazing.

My ex was a math and computer science prof so I’ve been around programming since the dinosaurs walked the Earth. But, it’s a whole new ball game now and totally over my head.

I don’t even try to understand what goes into all of it, the programming, the analytics, the machine learning. But I think I am fascinated by it all, and a bit concerned.

The Tesla, SMAs and all the ‘smart’ stuff feels out of reach to me, out of control. Now I have a new heat pump, multi-zone mini-split and the company and the installers can’t even keep up with the instruction manuals, etc. I had to go digging for info about how to get at the rotary fan to clean it.

So I acknowledge that the engineering and programming that goes into a Sense monitor is over my head, complicated and a work in process, especially with the lab features.

I think I am also more empathetic due to the complexities of communication, at-distance team process and the general decline in functionality across all of our societal systems.

I just figured I’d provide updates from this field test site in the event anyone else was interested in the experiments.

I am a little surprised that Sense has not sent a replacement monitor to see if that would make a difference. They say they have tested it and it is okay but the support folks have told me all kinds of things that didn’t necessarily pan out as truth.

Or, as usual, it is another example of my old school hand’s on way of thinking. Send me a different monitor, I’ll plug it in and we’ll see what happens. If nothing changes, indicating the original monitor is good, I can send the new one back. Just a thought.

Thank you for the ongoing inputs and the enjoyable conversations.

Best wishes to you and your wife; wish her luck for me.

[kevin1]

@Kathy,
You point to a second possible problem with your data (thanks for reminding me) - The RMS voltage displayed is computed using a base frequency, which in the US is a range around 60Hz. Are your seeing your Power Meter drop out in the same regions the Power Quality graphs drop out ?

If no, then I suspect that the dashboard rendering for Power Quality is less robust than the Power Meter. It either coughs up on missing data or on voltage frequency out of range.

I’ll continue to make the case for you to try Home Assistant with the Sense Integration or write your only little home-brewed Python to sample the Sense Signals data if you want to sort things out and get better data on what’s going on. More on writing your own home-brewed code below.

Integration for Home Automation

@bayesianlogic.1, You’re right. If you can know what your are using, and when, you can do effective power management. Have you tried using the python API that Michael also referenced ? If you want realtime access to power consumption and production it works just fine: GitHub - scottbonline/sense: Sense Energy Monitor API. If you want to dig through historic data, you can do a little scraping. Every bar graph on the web app is easily navigable and scrapeable vis selenium or rselenium, Our …

[kevin1]

@Kathy +
I forgot the niceties - I was in a hurry this AM. And I realized that I didn’t read your message to the degree of detail required. I missed this:

@Kathy

he challenge with the possibility that it may be frequency screwing up my power quality graphs, is that the frequency only increases for solar curtailment when the batteries are mostly full and it’s still sunny outside. There doesn’t appear to be any gaps or freakouts that directly correlate to those time periods.

Appreciate your best wishes - same to you as we get ready for the holidays and a New Year. Your programming history made me chuckle:

@Kathy

My ex was a math and computer science prof so I’ve been around programming since the dinosaurs walked the Earth. But, it’s a whole new ball game now and totally over my head.

Let me know if I can help you get a second set of eyes on your Sense monitor data via either path (Home Assistant or Python). But I’m only useful for really simple Python

[obscuredtrip]

@kevin1

though there are still indicators that there are underlying code fixes needed (the date range is clearly wrong - Dec 10-Dec 10 and the “last {{0}} days” artifact remains.

Not sure who @JuliaAtSense talked to. As of yesterday Sense Support was unaware of this, they are letting Engineering know.
Hopefully this can be attached to the original internal ticket, otherwise a separate internal ticket would have to be created (after x amount of other users open their own tickets) which may take months (or longer) to cross the engineers desk and fix as it (a display bug) doesn’t meet Sense’s internal policy for priority triage.

[JuliaAtSense]

Thank you @kevin1 and @obscuredtrip, I alerted our team to this separate issue and they are looking into it.

[phil.garver]

@JuliaAtSense,

Thanks for the update. Appreciate the continued work. Mine has also finally updated. This is a great feature and look for more like this in the future!