Backup down under

[wattmatters]

If you're paying for all that AC stuff, that must be eating into any gain you get from solar. Just do it yourself and don't tell anyone. It's your house!

My grid solar PV paid for itself in about 4 years. I'll let you guess when it was installed:



The newer addition of off-grid battery will take a bit longer to recover, more like 6-7 years. I'm OK with that. The off-grid system's primary function is for outage backup, not saving money (hence the title of this thread). But it's doing both quite well. This month so far it is returning close to $2/day and that's going to be more like the daily average from here on as all our energy tariffs have just gone up by quite a bit.

 

[wattmatters]

Love to see what you find out.
Could be in UPS it's trying have waveform output ready for switchover and it's deforming the grid one trying to sync it?

I think I set it back to APL mode because when the batteries reached low power I suffered a loss of power when it switched to grid, but in APL it's a smooth transfer.

I just changed it over. And the IoTaWatt started to capture voltage data again.

See the blue trace. It would normally get the occasional reading and flatline but since the change it's been able to read the voltage:



This is very promising!

I am not going to get my hopes up too high just yet because it can sometimes capture voltage for short periods before relapsing into the loss of signal.

I'll let it run and if it reads reliably for an hour or so then I think I can revert to using the direct voltage reference.

 

[wattmatters]

Could be in UPS it's trying have waveform output ready for switchover and it's deforming the grid one trying to sync it?

I will also take the oscilloscope out to capture the VT's output waveform with the inverter in APS mode when I get a chance. It definitely seems to be working now. This is the last three hours:



You can see how the IoTaWatt has had much difficulty getting a lock on the blue VT trace but once the change was made at ~11:05 AM it's been recording just like the other phases.

This has made my day.
Thanks!

 

[wattmatters]

Love to see what you find out.

OK, some voltage output waveforms for comparison:

Utility First UPS mode:

Utility First APL mode:

While the APL mode waveform is a little distorted, it's not nearly as distorted as the UPS mode waveform. The asymmetry between the first and second halves is also halved to ~ 200 µs (I zoomed right in to get as good a measurement on the zero crossing points as I could).

 

[nezjnr]

Nice

My grid solar PV paid for itself in about 4 years. I'll let you guess when it was installed:

View attachment 144167

The newer addition of off-grid battery will take a bit longer to recover, more like 6-7 years. I'm OK with that. The off-grid system's primary function is for outage backup, not saving money (hence the title of this thread). But it's doing both quite well. This month so far it is returning close to $2/day and that's going to be more like the daily average from here on as all our energy tariffs have just gone up by quite a bit.

View attachment 144166

Is this an actual value return from your battery? What's generating that for you?

Glad to hear APL mode worked out, has it allowed you to now use it as a voltage reference or did it bug it after a while?
Really interested to know what's going on in UPS mode but the manual doesn't explain anything.

 

[wattmatters]

Is this an actual value return from your battery?

Yes, that's average daily value of the grid imports replaced by energy discharged from the battery, accounting for some other factors as mentioned below. My expectation was to get somewhere in the range of a $500-600/year bill reduction.

What's generating that for you?

The energy source used to charge the battery is from a small (2.2 kW) off-grid PV array which was originally in place to keep my backup battery at float, ready for outages. It also used to power my pool pump but nothing else. It would produce ~75-85 kWh/month.

However this PV array had a LOT of unrealised capacity.

It was "use it or lose it", effectively free energy as the PV array was going to be there whether or not I used it to its full capacity, so now I am using it. With the addition of the LiFePO4 the off-grid PV array now supplies ~ 400-450 kWh/month. Of that ~ 230 kWh/month goes into the battery, with the balance being consumed directly which it can now do because the inverter enables that as well.

Occasionally I do a little bit of supplementary charging from the grid-tied PV array but not a lot, and given our net metering credit for export is a fraction of the import cost it's no big deal to ensure the battery is fully charged if the small off-grid PV array won't quite make it (I have automations managing that, only using the minimum grid-tied PV required). I have allowed for that in the calculations.

Glad to hear APL mode worked out, has it allowed you to now use it as a voltage reference or did it bug it after a while?

Yes, I have now changed all those circuits over to use the direct voltage reference VT and it has been reporting correctly all day no issue.

Really interested to know what's going on in UPS mode but the manual doesn't explain anything.

Me too. I shared my findings at a local forum with some gurus on this broad family of inverters. If they don't know, pretty much no one will.

Thanks again for the suggestion - it was "Today's Top Tip"!

 

[wattmatters]

By the way, I took a waveform sample when the inverter cut over to SBU mode earlier this evening:

Much nicer waveform.

The Iotawatt voltage measurement earlier this afternoon covering the period when the inverter switched over from Utility First to SBU mode:

You can see why I was keen to have the direct voltage reference working.

 

[Hedges]

By the way, I took a waveform sample when the inverter cut over to SBU mode earlier this evening:

Much nicer waveform.

The Iotawatt voltage measurement earlier this afternoon covering the period when the inverter switched over from Utility First to SBU mode:

You can see why I was keen to have the direct voltage reference working.

10.1 Vrms on scope? What is that?

IotaWatt shows 242V & 244V.

Waveform is a linear ramp of voltage rather than a sine wave.
Current, I've seen a number of difference shapes depending on load, but voltage I expect to see something closer to a sine wave. Yours has distortion.
Does your scope have any digital features, e.g. to show harmonics with FFT?
I've also transferred waveforms to the PC, where FFT and other analysis can be done with Matlab, Python, Excel, etc.
Teledyne LeCroy offers free download of "Maui Studio". I haven't operated it on PC yet, but on their Windows scopes, do all sorts of graphing and processing.

 

[wattmatters]

10.1 Vrms on scope? What is that?

Yes, that's the output from the isolation voltage transformer used to step down the mains grid voltage to a safe level for measurement with a scope.

The following is more PSA than specifically directed at you:

That's the safest way to measure a mains voltage signal with an oscilloscope. You of course need to know the step down factor if you are interested in knowing what the peak to peak and RMS voltage is on the high voltage side of the transformer.

In my case I wasn't interested in the actual voltage (I already know that from regular voltage monitoring) but rather I wanted to inspect what the AC waveform looked like and whether this was causing issues (it was).

Being a 230V country (and my grid voltages can be up to 260 V RMS) we can't just place regular oscilloscope test probes on the mains supply as the peak to peak transient voltages can easily exceeds the scope's voltage input limit (bad news for the scope) and the potential for ground loops can also release the magic smoke from the very probe you are holding, not to mention zap you with mains voltage.

If you really want to measure mains voltage directly with a scope (without use of a step down isolation transformer) it requires use of a special (read very expensive) differential probe designed for it plus of course a LOT of care to ensure there are no ground loop issues with the scope's ground connection.

If the supply is 110-120 V AC, then on the 10x setting most scopes probably can be used to read the AC waveform directly but the need for great care on ground loops still exists and is why use of an isolation step down transformer is much preferred.

 

[wattmatters]

Waveform is a linear ramp of voltage rather than a sine wave.

Which one?

All of them have some distortion from a pure sine. The grid signal isn't perfect either.

What I was looking for was two issues which might cause the IoTaWatt to have difficulty getting a lock on the voltage signal:
- Issues with zero crossing
- Issues with waveform asymmetry (in particular the periods between zero crossings)

Does your scope have any digital features, e.g. to show harmonics with FFT?

It's a cheaper tablet scope, it can show an FFT plot but unfortunately/annoyingly it provides no scale for the FFT plot. All you can see is whether harmonics are obvious.

It does not have the ability to capture thousands of samples for that type of detailed signal processing/analysis.

A couple of examples to show what I mean about the limited FFT info (the purple plot at the top):

 

[Hedges]

Being a 230V country (and my grid voltages can be up to 260 V RMS) we can't just place regular oscilloscope test probes on the mains supply as the peak to peak transient voltages can easily exceeds the scope's voltage input limit (bad news for the scope) and the potential for ground loops can also release the magic smoke from the very probe you are holding, not to mention zap you with mains voltage.

If you really want to measure mains voltage directly with a scope (without use of a step down isolation transformer) it requires use of a special (read very expensive) differential probe designed for it plus of course a LOT of care to ensure there are no ground loop issues with the scope's ground connection.

If the supply is 110-120 V AC, then on the 10x setting most scopes probably can be used to read the AC waveform directly but the need for great care on ground loops still exists and is why use of an isolation step down transformer is much preferred.

Which I have done. At first glance, 300V limit for many 10x probes seems OK for our 120V, because it is 170Vpeak. But power line transients can hit kV and might kill it. Shell of BNC is ground for most scopes (isolated differential input for some handheld.) Don't want to connect ground to neutral either.

I have one of those expensive HV differential probes at home and at work, so I prefer to use that.

I also have some 120/240 to 9/18V toroids at work, which I've also used for isolation and/or step-down. Useful to have better control of low voltage using Variac as well (testing saturation of chokes.)


I prefer scopes that let me adjust FFT span independent of time domain window. Only some of ours at work do. They do have deep memory, so I can get 20 seconds or so for better low-frequency FFT.

Try offline processing. If Maui Studio is available to you overseas, all its signal processing capabilities might be useful. You'd have the capture of your low-end scope, but signal processing of a high end one. Maybe it can do a better job with your limited capture window. On the LeCroy scope, I'll average many successive FFT, which gives better SNR despite noise possibly not being synchronous with the time-domain trigger.

 

[wattmatters]

Try offline processing.

Not sure it's possible.

It's a FNIRSI-1013D, a portable tablet style scope:
http://www.fnirsi.cn/productinfo/556152.html

No option for anything other than a basic waveform capture.

It's a little on the basic side but is all I need. Nice and portable, fairly easy to use because there are not too many fancy features.

 

[Hedges]

240kbit storage depth, USB.

If you can pull out a data file, you can post-process it on your PC.
storage of 1000 screenshots and 1000 waveforms.
Twice it mentions transferring screenshots. I haven't read that waveforms can also be transferred.

 

[wattmatters]

Twice it mentions transferring screenshots. I haven't read that waveforms can also be transferred.

I'll have another look but AFAIK the only files I can see on the unit are images.

 

[wattmatters]

If you can pull out a data file

Just checked and unfortunately the only files I can access are images in .bmp format. No data files.

 

[Hedges]

Some mentions of hacking the code including USB driver.
Maybe somebody already got it to transfer wave data to USB or SD card.

FNIRSI-1013D "100MHz" tablet oscilloscope - Page 45

FNIRSI-1013D "100MHz" tablet oscilloscope - Page 45

www.eevblog.com

Tablet Oscilloscope Claims 100 MHz, But Is It?

[LearnElectronics] grabbed a FNIRSI tablet oscilloscope from a vendor from China. The device has a seven-inch touchscreen and claims to be a two-channel 100 MHz scope. But is it? Watch the video be…

hackaday.com

Such a shame if it can't do that.

 

[wattmatters]

Some mentions of hacking the code including USB driver.

That's above my pay grade I'm afraid.

 

[nezjnr]

Interesting my inerters were in SBU mode but just bypassing to grid, for some reason the head unit has changed and my output voltage is now set to 220V.

When I switch all the grid feedins off that it went to battery mode, but the LED is blue instead of white.

 

[wattmatters]

Interesting my inerters were in SBU mode but just bypassing to grid, for some reason the head unit has changed and my output voltage is now set to 220V.

Weird.

I actually would like a 240 V AC output as it would reduce the voltage jump between grid and inverter output but I understand the design of these inverters restricts their ability to generate the higher peak to peak voltages needed, especially as battery voltage drops.

 

[nezjnr]

Weird.

I actually would like a 240 V AC output as it would reduce the voltage jump between grid and inverter output but I understand the design of these inverters restricts their ability to generate the higher peak to peak voltages needed, especially as battery voltage drops.

Hmm, I was just going through my settings and realised 13 - Setting voltage point back to battery when selecting "SBU" (SBU priority) in program 01 had also changed it's setting to FUL instrad of the 51V I had it set too, which was why it wouldn't operating in battery mode because it was set to only switch if at full charge.

I still can't manage to change the voltage back to 230, for me grid voltage is usually around 230V give or take for the majority of the time, 240V is only reached on sunny days when everyone's solar is raising grid voltage, you can see 5-10V swings when clouds pass over.

Love to know how and why my settings changed, especially since 230V is the default setting...