If Sig doesn't have a zero offset tweak facility, you could 'hack' it by threading a thin wire through the  Sig CT (i assume it uses a CT on the meter tail), and use a low voltage transformer and resistor to pass 10W worth of a.c. through the CT (42mA?) to make it 'think' you were importing another 10w.

  mab1 said  

If Sig doesn't have a zero offset tweak facility, you could 'hack' it by threading a thin wire through the  Sig CT (i assume it uses a CT on the meter tail), and use a low voltage transformer and resistor to pass 10W worth of a.c. through the CT (42mA?) to make it 'think' you were importing another 10w.

Thanks Mab, that is so devious - I love it. I have used that technique with DC sensors (using DC) with no tweak adjustments for similar offset reasons.  My system is 3 phase so it gets complicated quickly but I know there are small 415V transformers available usually for delta connection but if I use them for star they will have naf all idling current and still easily supply what I need for such a tweak.

I'm embarrassed to say I have no idea about the installation and have never removed any covers.  There was a time where I often pulled covers etc off something to see what made it tick before I even used it, I was probably a little worried about Sig warranty issues too. I know they have remotely looked at my system, what if they notice my fix and "helpfully" tweak it back to zero remotely for me lol.

I am contemplating doing this at least for a short-term fix. I will guess the CT's are in the gateway box as the system needs to automatically isolate the grid from the house if the grid goes down and that gateway box is full of contactors - so where's my screwdriver....
Edited 2026-05-03 09:09 by wiseguy

Hi Mike,
I read your comments with interest.
My installation is only a small grid-tie but I have noticed the same difference with metering between the power authority (Aurora) and my Enphase solar control system.
I access the Enphase metering via HTTP which requires a free "token" from Enphase. The token only lasts 12 months so it is an annual task to renew it.
I have similar field of data to play with.
The reactive power swings + and -.
Unfortunately, the active power spends too long on "import". I need to remove a few more trees on the western side.

I have 12 micro inverters which are only interrogated every few minutes. Not a problem if you are only after accumulated energy but not good for instantaneous power.

I have my own metering CT on the heat-pump where I calculate power every 500mS.
For the EV charger, I use a Lanx power meter with 2000 pulses per kWh. This works well for the steady load of the charger.

I have thought about putting my own CT on the incoming mains and use the same setup as the heat-pump but accumulating power in both directions.
The authorities would not be impressed with any mods on their side of things.
For now I make do with the inconsistencies.

I only use the information for power scheduling and trying to decide if I should increase the size of my system or change the tariff.

Jim

Hi Jim, thanks for the post and comments. I think I misunderstood your post, I re-read it a few times so I will ask a question for clarity. I did not see the word "battery" used - I was expecting the mention of an Enphase battery, you aren't using one?

It sounds like what your Enphase system is reporting and what you provider is reporting don't match and I'm guessing not in your favour. That is why you want to separately monitor your energy import/export for an independent report?

The power network people do not like you doing anything to their wiring but I believe the current measuring CT's used are (split) that can be clipped to a conductor without making any wiring changes. For voltage information you can make your measuring device circuit wired with a 3 pin plug into a handy socket and then with an isolation transformer like a 2mA to 2 mA part you have phase information volts and amps calculate as desired.

My understanding is that you are allowed to build something that connects to the mains with a 3 pin plug but nothing hard wired to the house wiring. You are not allowed to sell such a device without appropriate safety tests and approvals etc, but you can use it for your own purposes and the Darwin principle applies - please take care

I assume the shading issue was mentioned because of the deficiency of solar without any battery storage to carry you over the short fall dips, prime the chinsaw.....
Edited 2026-05-03 23:39 by wiseguy

I received my reply from SIG (on a Sunday afternoon !) and I have to say I am totally unimpressed but not surprised, here it is;

Hi Michael,

Thank you for your detailed response.

At this time, there are no settings available that allow the system to consistently import or export a fixed amount of power to the grid regardless of operating conditions.

Grid import and export are primarily influenced by solar generation and the site’s load consumption. Additionally, the maximum grid export active power is determined by your DNSP (SAPN), as they regulate the system’s discharge to the grid.

We have also performed an advanced wiring diagnosis on your system, and no abnormalities were found.

To help optimize your system’s performance, we recommend updating the firmware to the latest version.

Please let us know if you need any further assistance.

What a load of patronising obfuscating gobbeldygook ! The sentence starting "Grid import and export.... has zero to do with my problem, what a waste of words and implying the networks control what I export (near zero...).

I have deleted the rest of my original post but suffice to say I am not done with SIG yet.

All their emails end with "If you need further assistance, please let us know"

Maybe they meant to write "You won't get any further (or past) assistance" so let us know...

Below is an excerpt from my recent reply;

Quote "This is expected behavior and not a measurement fault"

When the utility reports that we have imported 28.50kWH in a month but for that month the SIG system reports a total of 1.67kWH that is a measurement fault.

So Mab's solution is starting to look betterer and betterer (to quote Poida).
Edited 2026-05-04 09:29 by wiseguy

Hi Mike,
I should have been clearer.
I only have solar and no battery. At the time, I felt that I would use much of the surplus energy to charge the car. This has work out well for 4 years but now I have an EV with a much larger battery, it is time to re-calculate things. Currently Aurora give 9c/kWh feed-in but I expect that will drop again soon.
The boss has allowed me to put some panels on a north facing roof and hide the batteries on a back wall out of sight.

It is 30 years since I worked for the power authority but I still know how far I I allowed to go with my fiddling. The local inspector was an apprentice when I retired and he is the only employee remaining from my days.

My main interest is data acquisition and control.
The local provider gives me fancy charts of usage but 24 hours behind the times so no use for management. They don't provide the data in a usable form so I can't play with spreadsheets.

The Enphase people do provide the raw data monthly and I can use the api they publish to get real-time data.
It is when I compare the Enphase and Aurora data when I get the monthly power bill that I can see a difference. It is not enough to cause concern, just enough to get me interested.

Re the trees, I have removed all that can be done with a chainsaw but there are others that are too close to infrastructure. They will need big machinery and an equally big bill. 50 metres high leaning towards the shed...

Jim

Hi Jim, is your EV capable of supplying power to the home and grid?
If so  there would be no need for a home battery.
Our car can't do V2H or V2G, so we will just stick with our fully off grid system
Maybe one day but at the moment our car doesn't have to have internet connection of back to factory communication so we are happy with that.
Good luck with becoming more self sufficient
Pete

  Godoh said  

Hi Jim, is your EV capable of supplying power to the home and grid?

It has one 10amp 240V outlet in the luggage area and I can purchase an external one to plug into the charge socket.

I can pension off one of my generators.

I think I will only be using it for power outages.

Jim

  wiseguy said  

When the utility reports that we have imported 28.50kWH in a month but for that month the SIG system reports a total of 1.67kWH that is a measurement fault.

This is similar to the standard energy metering in other consumer grade inverters.  You can ask them to verify but it's pretty clear the inverter does not have revenue grade metering.  Some inverter makers (e.g. Enphase) do offer optional revenue grade metering at a premium.  The lack of revenue grade metering is often a key cause of discrepancy from the utility reports.

Good point, but even allowing for the percentage of accuracy between the two, 1.67kWH ver 25.6kWH seem way beyond that, something insidious appears to be going on here, or am I just reading to much into it?

  KeepIS said  

Good point, but even allowing for the percentage of accuracy between the two, 1.67kWH ver 25.6kWH seem way beyond that, something insidious appears to be going on here, or am I just reading to much into it?

This is edited - previous maths was crap! Should have used a calculator.

Have to agree, I believe if I designed an energy metering device I could get it to 10 times more accuracy even before the first calibration takes place. I have been throwing the figures in my head. As we are drawing ~ 1kWH (according to the utility meter) over the 16 hours we are not exporting, that is ~ average of 62W continuously.
Reminder to self dont try to do power and energy calculations in head!

Note that if the system is hunting around zero and spends equal time exporting and importing at ~50% duty cycle the 62W becomes 124W spikes of export/import

I could buy that as it wobbles back and forth trying to balance to zero. I am thinking if I use the MAB idea and start with a forced export of ~ around 60W might be required do negate their 60 odd average watts of import.

I did get an email from SIG last night saying my issues are being promoted from the help desk to the technical area, who knows maybe we can make more progress without resorting to a lot of work to solve someone elses problem that seems to have become my own...
Edited 2026-05-05 19:54 by wiseguy

  TassyJim said  

Hi Mike,
My main interest is data acquisition and control.
The local provider gives me fancy charts of usage but 24 hours behind the times so no use for management. They don't provide the data in a usable form so I can't play with spreadsheets.
Jim

Is it worth considering something like a Fronius smart meter or similar device.
Not sure which device may be best to implement or what protocol would be involved but for you guys should be a piece of cake   KeepIS will probably chime in to suggest a suitable device....

Yes an electrician would need to wire it in but once installed it would be be a great tool for accurate monitoring if you are able to make it work.

I have a Fronius 3 phase one if you can use it, I also have an EvalSTPM33/34 development board if its any good to you. It is the earlier model with A D9 connector not the later USB type. You could maybe utilise it with a clamp on CT and change a part or two if required to match the clamp on to original CT which does not clamp on, put it in a box attach a cord and 3pin plug and voila!. Ok maybe optos required for txd and rxd, I cant remember, played with it years ago.
Edited 2026-05-06 01:08 by wiseguy

  KeepIS said  

Good point, but even allowing for the percentage of accuracy between the two, 1.67kWH ver 25.6kWH seem way beyond that, something insidious appears to be going on here, or am I just reading to much into it?

There definitely could be other contributing factors. However, the reported energy difference is similar what I have seen reported for other consumer grade inverters, where the main fix is usually switching to revenue grade metering that significantly reduced the difference.  So, I assume a combination of CT, ADC, calibration, sampling and algorithm differences are the main culprit.  I've seen commercial inverters that can't even provide accurate true RMS AC power measurement especially at low power levels and/or low power factors.  I believe such capability is a pre-requisite for accurate energy metering.

It would be interesting see if the Sig inverter can provide accurate true RMS AC power measurement at low power levels.

  analog8484 said  

I've seen commercial inverters that can't even provide accurate true RMS AC power measurement especially at low power levels and/or low power factors.  I believe such capability is a pre-requisite for accurate energy metering.

I agree and I have posted about the inability of these AC power units being unable to calculate PF correctly and AC power at low power levels.

These thoughts and info from wiseguy are similar to that I was visualising.

  wiseguy said  

Note that if the system is hunting around zero and spends equal time exporting and importing at ~50% duty cycle the 62W becomes 124W spikes of export/import

I could buy that as it wobbles back and forth trying to balance to zero. I am thinking if I use the MAB idea and start with a forced export of ~ around 60W might be required do negate their 60 odd average watts of import.

 

Is it just a coincidence that metering and export/import errors and design inevitably favour the energy supplier?

And what does this say about the slant in calculations for power usage and feed in power with our older solar grid ties and insidious introduction of smart meters over the past 18 years?

I am really, really, close to telling the corporations where to shove their grid  
.
Edited 2026-05-06 08:59 by KeepIS

Today I contacted the SA Energy Ombudsman to find out what might be able to be done to stop the consumer rip off. His answer was to shop around for a better energy deal.
He continued that this issue does not fall under his jurisdiction, that as the energy markets are now unregulated they have no control. I then contacted our local MP who promised it will be raised with the government minister for Energy as they also conceded that there was definitely room for improvement.

With regard to accurate energy and power factor measurement, back in 2005 when I created the Power-Mate energy meters I initially used a Cirrus CS5460A which gave brilliant results for power measurement. It could resolve to 10mW of power ie with a 240K resistor as a load the power showed ~ 240mW +/- 10mW if we had a load of 2.4kW it could still resolve 1 watt difference. Power factor was much more difficult and could only be achieved with maths.

Then Cirrus logic bought out an improved CS5463 which did a lot more and calculated apparent, reactive, harmonic, PF and various other measurements for free - well ok for ~$4. From memory, reactive & PF needed a few watts of load for best accuracy.

The reason for bringing this up is that there is absolutely no excuse for the likes of SIG and others to not provide detailed and accurate information to the users.  For the sake of a few low cost parts and some code it is all measured and presented in 24 Bit registers, you just have to read them out and format their units for display, this is not hard!

Hi Mike long no see but when you brought up that first toroid to me it got me hooked line and sinker to where YES I did go the the MadBooard design.

Now on my farcebook local group page it was asked if I could help with a SIG system  problem where I found it was a basket case of a problems with export and import problems. The support was that bad I asked the family is it better just use the panels to go off the grid.

I did suggest a Ali inverter and they are now 2 years with the grid turned off on a rural property.

So it does have me wondering with that inverter you designed and how good it is why did you go for that SIG system at all when the stuff you designed just worked so well and still does for many folk and will for many years.

Regards Bryan

Response from SIG a few minutes ago - a small hope is starting in my mind.

Thank you for your thoughtful feedback and for sharing your suggestion regarding a self-consumption with grid isolation mode. We appreciate the time and effort you've put into analyzing your system's performance and proposing a potential solution.

We have already escalated your case to our HQ team for further review, and we will also forward your suggestion for a grid isolation feature to our product development team for consideration. While we cannot guarantee immediate changes, your feedback is valuable and helps us improve our products.

If you have any further information or questions, please feel free to reach out. We will keep you updated as soon as we receive feedback from HQ.

Stay tuned......

Bryan, developing my inverter took a big slab of what's left of my life. My wife said you can keep tinkering in your workshop but I intend to go travelling, "I will send you postcards".

I actually want to travel too, I don't want to be a slave to our system and having to turn switches and push buttons - and that's when I'm here. I was after a set and forget solution and something that will keep our fridges running for 5-7 days if the grid goes out whilst we are away for a month or three.

The decision did not come easy but the complications of interfacing my single phase inverter onto our 3 Phase house can and has been done, but manually. Automating it was going to be messy and complicated and a can of worms, so I made a rational choice. The only gotcha was that we are importing energy, which on that principle alone, given that I was told we would be self-sufficient, I find very annoying. I want a solution that I don't have to implement, if possible. It's their mess hopefully they will clean it up.
Edited 2026-05-06 22:47 by wiseguy