diy solar

diy solar

25kWh daily output, but I only used to need 10kWh per day.

Yup
If it's not a bidirectional meter, it only counts up.
No matter which way the current is flowing.
 
Hi jfpetesn, I'm not sure how to install a shunt in between the battery cable and the battery - would I just cut one of the existing cables in half and put new battery lugs on them, so I can screw them onto the shunt terminals? (And obviously use a decent crimp tool.)
I think you're right, I would at least be able to see what is actually going into the batteries, there is obviously something going wrong with the data that is coming from the inverters and/or the batteries to my PC, and I'm now thinking that I can't really trust any of it!
Re the attached screenshot - this is from the start of November. Four LEDs were lit on two US5000 batteries (I only had two then), yet the 'Battery W' shows 12.29kWh capacity in the batteries, which is impossible, because they only hold 9.6kWh when full. In the first screenshot in my previous post, the 'Battery W' field shows 8.38kWh in three batteries, at 20% SOC, which is also impossible.
If I could get Solar Assistant to display proper battery information, I would use that, but that shows something like -18,000W coming out of the batteries, when they are actually charging on a sunny day, and only rarely shows a green bar for the battery section - so I can't trust that either.
Tomorrow I will definitely get my laptop connected to the inverters with 1m or shorter leads and see what I get by doing that.
 

Attachments

  • 3-11-2022 Only four LEDs lit at end of day despite over 12kWh into batteries.PNG
    3-11-2022 Only four LEDs lit at end of day despite over 12kWh into batteries.PNG
    125 KB · Views: 3
Hi jfpetesn, I'm not sure how to install a shunt in between the battery cable and the battery - would I just cut one of the existing cables in half and put new battery lugs on them, so I can screw them onto the shunt terminals? (And obviously use a decent crimp tool.)
I think you're right, I would at least be able to see what is actually going into the batteries, there is obviously something going wrong with the data that is coming from the inverters and/or the batteries to my PC, and I'm now thinking that I can't really trust any of it!
Re the attached screenshot - this is from the start of November. Four LEDs were lit on two US5000 batteries (I only had two then), yet the 'Battery W' shows 12.29kWh capacity in the batteries, which is impossible, because they only hold 9.6kWh when full. In the first screenshot in my previous post, the 'Battery W' field shows 8.38kWh in three batteries, at 20% SOC, which is also impossible.
If I could get Solar Assistant to display proper battery information, I would use that, but that shows something like -18,000W coming out of the batteries, when they are actually charging on a sunny day, and only rarely shows a green bar for the battery section - so I can't trust that either.
Tomorrow I will definitely get my laptop connected to the inverters with 1m or shorter leads and see what I get by doing that.
My Solar Assistant gets its battery information from a Victron Smart Shunt. All other information comes from the AIO inverters.
 
get my laptop out to the inverter box tomorrow and test it with 1m long cables.
Yup.
While I know nothing about the equipment you are using, I have good mechanical sensibilities… which when I read,
Man, that would suck green toe jam. I didn’t know this could happen.
…my mind tells me you should explore if there is a ‘zero export’ option in the inverter/charger setup. Is it even supposed to be able to gridtie/sell??!
Because what you’re reporting makes little sense (in my little knowledge base) unless it was a weakened lead acid battery bank.

First thing I’d do is play with the short ethernet cable and eliminate that unlikely problem.
Second, you mentioned “flat” ethernet cables; if the short cable doesn’t eliminate the ghost batteries/data I’d just plain stop using the multi-thingie software and move on to another monitoring solution. OR can you plug in independently and view saved data from each pylontek? And prove what’s what? See if the data starts making sense?

Third, I might be inclined to check into a bunch of gadge-wizzits like these or this one that I bought and monitor what’s going actually going on, coming and going. This is where separate components are ‘easier’ to diagnose weirdness because you can monitor charging Wh separately from inverter Wh

I'm not sure how to install a shunt in between the battery cable and the battery
Bolt the shunt to the battery, cable to the shunt- no cutting
 
Hi 12VoltInstalls, my Pylontech battery terminals are designed to take Pylontech press on cable connectors, so I can't bolt anything onto them.
I was thinking a week ago about buying one of those battery monitors, much cheaper than the Victron Smart Shunt, and probably good enough for what I need, i.e. just temporary testing.
My inverters are MPP Solar PIP5048MKX models, they are 'off grid' and cannot export to the grid, I asked MPP Solar a lot of questions before I bought them. (They take electricity FROM the grid, but won't export to it.)
I will post more results tomorrow when I've plugged my laptop into the inverters and the batteries.
timselectric, I may buy a Victron Smart Shunt for my Solar Assistant software if I can't get anything else to work, thanks for the tip.
 
When the PV is charging the batteries during the day, I have often checked our electricity meter to make sure the dial isn't turning, and it isn't, so I'm pretty sure no electricity is going back into the grid.
If it's an old glass meter with the big spinning metal disc. It's not bidirectional, and only counts up.
So definitely wouldn't want to be paying for any excess production.
If it's not spinning, that's the best it can get.
But your inverters are off grid inverters. So, you have nothing to worry about.
 
designed to take Pylontech press on cable connectors, so I can't bolt anything onto them
Then use the other end :)
good enough for what I need, i.e. just temporary testing
and maybe more! I have used a bunch of different kinds of these inexpensive pre-value-added subassemblies over the years and I’ve been shocked how dependable they are.
 
At about 11pm last night, the 'Battery W' value was 8.3kWh, but the grid was still powering everything, and was still charging the batteries at 2A, I don't know how to stop the inverters from doing that. But some time after midnight I checked the 'Battery W' value again, and it was around 0.3kWh! So it had presumably reset to zero at midnight, and was now showing how much charge the grid had put into the batteries since then.
I have attached the 'Totals' page from MultiSIBControl, about six weeks of data - look at the 'Battery Chg' and the 'Battery Dchg' fields. In every single case, the discharge amount is a fraction of the charge amount. I installed the third battery on the 15th November, so prior to that I only had two 4.8kWh batteries, so you can still see that there was loads more going into the batteries (or somewhere) than they could actually hold.
I have added the Battery Charge and Discharge graph, which makes it much more obvious that something is very wrong, the discharge bars are so small every day.
I have no idea what the 'Pylon Chg' and 'Pylon Dchg' values mean, since they are so different from the 'Battery Chg' and 'Battery Dchg' values, but then, if they were the same values, there would be no reason to have a separate field for them.
On the main display of MultiSIBControl, the 'Battery W' value is climbing up, currently at 3.94kWh, and I am getting 2014W from my 12kW array, on a very misty morning. I am now going to finally get the laptop out to the inverter box and test MultiSIBControl on that. I will report back later. Thank you all for your help, I do appreciate it.
 

Attachments

  • 29-11-2022 11am Battery W climbing.PNG
    29-11-2022 11am Battery W climbing.PNG
    178.3 KB · Views: 3
  • 29-11-2022 Totals page.PNG
    29-11-2022 Totals page.PNG
    178.4 KB · Views: 3
  • 29-11-2022 Battery charge discharge graph.PNG
    29-11-2022 Battery charge discharge graph.PNG
    141.7 KB · Views: 2
Last edited:
some time after midnight I checked the 'Battery W' value again, and it was around 0.3kWh! So it had presumably reset to zero at midnight, and was now showing how much charge the grid had put into the batteries since then.

Well I think that multislab software and/or aio inverter setup is whacked out. I’d be offloading it on marketplace already, or negotiating a refund. I’d be totally out of patience.
 
I tried my laptop out, next to the inverters and batteries - NO battery errors at all, so the battery errors that are showing up in MultiSIBControl are most likely due to the flat ethernet cable (15m) I have been using - I expect it is picking up interference from other electrical cables in the outside box I have the inverters in. I have ordered a shielded, round CAT6 cable from Ebay, that should fix that problem.
MultiSIBControl v1.7 is out and I have installed that. The MultiSIBControl.net website definitely says that the value arrowed in the screenshot attached is the TOTAL amount of kWh that has gone into the battery that day. See second attached image below, from their website.
 

Attachments

  • Battery charge amount.png
    Battery charge amount.png
    8.1 KB · Views: 6
  • s21m.jpg
    s21m.jpg
    235.8 KB · Views: 6
This morning my new MultiPro clamp meter arrived from CPC (very fast delivery), and when the inverters said they were giving 14A to the batteries, the new clamp meter only showed from 1A to 1.4A in the battery cables. I then looked at MultiSIBControl, and noticed that the 'Pylon W' box was showing 92W going into the batteries, whereas the 'Battery W' box showed something like 1,400W. So I believe that the 'Pylon W' box shows the actual amount of electricity that is going into the batteries, whereas the 'Battery W' box shows what the INVERTER is claiming it is sending to the batteries.
Yesterday I turned off the three batteries for a few minutes, because all the lights had gone out on them for a minute, then two LEDs on each battery came on for a while, then they went out again. While the batteries were switched off (at the battery switches), I tested the cable current with my old clamp meter, and that showed zero, but the inverters were both claiming they were sending 20A.
So, only a tenth of the watts produced by the PV and available for the batteries after the PV supplies the load to my house, is actually going into the batteries! No wonder they aren't working well.
 
About half an hour after my previous post, the inverters were saying they were sending 23A to the batteries (PV was well over 2kW), and my new clamp meter showed 13A in the battery cables. So a vast improvement on earlier. Now that I know that the 'Pylon W' readings in MultiSIBControl are what the batteries are actually receiving from the inverters, I am going to go back over my two months of records in MultiSIBControl and look at what the 'Pylon W' values have been doing each day. Prior to this I didn't know what they were and ignored them, because they were so different to the 'Battery W' readings and the inverters' alleged output to the batteries.
I have done more research today with my new (trustworthy) clamp meter. I changed the bulk charge and float charge voltages to only 51V, as Brianonmull said above that Victron had suggested this, and I thought, it can't hurt the batteries. I set up the inverters so that they were charging the batteries, supposedly at 40A (Setting 1: Usb, Setting 2: 40A (max charging current), Setting 11: 40A (max utility charging current), Setting 16: Sbl UCb (allow utility to charge battery), and the inverter was claiming that 50A was going into the battery - my new clamp meter showed 44.5A, so pretty close. Why the inverter was sending more than 40A, I do not know - PV input was about 1.5kW at the time, though presumably that shouldn't take Seetting 2 over 40A.
Anyway, the batteries were now charging at 44.5A from each inverter, so just over 4kW going into them in total. The Pylon W 'total' amount started to increase, so all of this agrees with each other; the inverter's claimed charge current, the clamp meter's readings, and the Pylon W total in MultiSIBControl increasing. Also, this tells me that the temperature of the batteries at the time I did this (which was about 12C) isn't have much, if any, effect on the charging. This also shows me that the batteries DO accept a charge without problems, if it is high enough. I didn't run this experiment for very long, maybe half an hour, but tomorrow I am going to fully charge the batteries from the grid, (via the inverters), probably at 20A, making a note of how long it takes to fully charge them, and what the SOC was when I started. The batteries themselves are actually accepting charge, but only on their terms... so 90% of the PV that I get each day, which isn't used directly on the load of my house, is lost. Who knows where it goes!
But the Pylon W 'total' amount is the current total of kWh in the Pylontech batteries - it goes down when the battery is discharging. It is not a daily total of kWh that have been put into the battery.
 
51v on a 15s battery is fine, if it can float long enough to reach 100%. It would take several hours, probably.
Measure the amps going into the batteries. As it decreases, the batteries are getting closer to full.
When they reach 100%, there will be near zero amps flowing in.
 
timselectric, thanks for that info, I will use my clamp meter tomorrow during the 'long charge' to monitor things on the battery cables, to confirm they match the amps that MultiSIBControl is showing me.
Re the attached screenshots: '30-11-2022 Pylon W 1290W Battery W 2413W 13A on clamp meter.png' shows that MultiSIBControl's 'Pylon A' reading (top right hand corner) agrees with my clamp meter - 13A from each inverter equals 26A total.
'30-11-2022 Pylon W matching Battery W.png' shows approximately the same watts for 'Battery W' (which I believe is what the inverter claims it's sending to the batteries) and 'Pylon W'. But eight minutes later, I had '30-11-2022 Pylon W back down to 96W.png', only a feeble 96W actually going into the Pylontech batteries, I have no idea why it dropped so much.
Then I had '30-11-2022 Pylon W now -149W even though Load W is much less than PV W.png' - the inverters were taking energy from the batteries, even though the PV was 1393W and the house load was only 404W. They should have been charging the batteries, not discharging them!
'30-11-2022 Charging from grid working fine, Pylon W 4282W, Setting 11 is 40A.png' shows what happened when I was charging the batteries using the grid, should have been 40A max, in Setting 2, but I was getting 44.5A on my clamp meter (or thereabouts), which agrees with the 85.93A in the 'Pylon A' box in the top right hand corner of MultiSIBControl. There is quite a big loss of energy here, 4,571W from the grid plus 1,146W from the PV is 5,897W, but 4,282W 'Pylon W' plus 569W 'Load W' is 4,851W, so a loss of 1kW.
'30-11-2022 Changed from Battery mode to Grid, Pylon W kept taking -417W regardless.png' - the inverters are discharging 417W from the batteries, even though the Grid Watts plus the PV Watts exceed the Load by almost 200W.

But the most important screenshot of all, is '30-11-2022 Daily PV vs Pylon charge.png'. Going back to 16th Oct when I first started using MultiSIBControl, and 28th Oct I got an RS232 to USB adaptor connected to the first of my (then) two US5000 batteries, you can see what I hadn't noticed at the time (because the batteries were working better back then), that a fair amount of the PV generated was going into the batteries, maybe 70-80% for a week. The empty part is where I was using Solar Assistant instead of MultiSIBControl. Then things started going downhill, as the graph shows.
 

Attachments

  • 30-11-2022 Daily PV vs Pylon charge.PNG
    30-11-2022 Daily PV vs Pylon charge.PNG
    164.4 KB · Views: 5
  • 30-11-2022 Changed from Battery mode to Grid, Pylon W kept taking -417W regardless.PNG
    30-11-2022 Changed from Battery mode to Grid, Pylon W kept taking -417W regardless.PNG
    170.8 KB · Views: 5
  • 30-11-2022 Charging from grid working fine, Pylon W 4282W, Setting 11 is 40A.PNG
    30-11-2022 Charging from grid working fine, Pylon W 4282W, Setting 11 is 40A.PNG
    161.4 KB · Views: 4
  • 30-11-2022 Pylon W now -149W even though Load W is much less than PV W.PNG
    30-11-2022 Pylon W now -149W even though Load W is much less than PV W.PNG
    168.7 KB · Views: 4
  • 30-11-2022 Pylon W back down to 96W.PNG
    30-11-2022 Pylon W back down to 96W.PNG
    172.9 KB · Views: 4
  • 30-11-2022 Pylon W matching Battery W.PNG
    30-11-2022 Pylon W matching Battery W.PNG
    157.4 KB · Views: 4
  • 30-11-2022 Pylon W 1290W Battery W 2413W 13A on clamp meter.PNG
    30-11-2022 Pylon W 1290W Battery W 2413W 13A on clamp meter.PNG
    157 KB · Views: 5
I think there is a wiring problem. The MPPT controller is not going to take 2500W from the solar panels, unless it has a load somewhere to get rid of it. As the PV amount rises, the amount going into the Pylon W rises, but it's always 1000-1500W lower than what it should be.
I am presuming that the MPPT controller can't take more from the solar panels than it has a load to give to. I have felt the air coming from the fans on the inverters, it is mildly warm (bearing in mind that my inverter box is completely open when I open the door and lid, and it's 6C outside), so I don't think there is 1.5kW being turned into heat, especially as the holes at the bottom of the inverters are quite small.
If the BMS in the batteries was refusing to take, say, 1,500W from the inverters, and only allowing 500W, then surely the PV amount should be lower, and the total Load W and Pylon W, should be close to what the PV W is (maybe 80-90% of it, allowing for losses in the inverters.) But there is a consistent 1,500W going missing, all the time.
So yesterday when I was using the grid to try charge the batteries at 40A, the batteries did charge, but weren't receiving the full 50A that the inverters claimed to be sending, my clamp meter was showing 44.5A. (The inverters must have been using 10A from PV as well, to make 50A.)
So I think I will have to get the electrician who fitted everything back out to have a look at it, and I'm praying that he has miswired something and that it's an easy fix.
re the screenshots below, please bear in mind that 'Battery W' is what the inverter claims it is sending to the batteries, 'Pylon W' and 'Pylon A' are what is actually going into the batteries, I've verified it many times with my clamp meter.
 

Attachments

  • 2-12-2022H 800W heater turned off, still only 1078W going into Pylon W.PNG
    2-12-2022H 800W heater turned off, still only 1078W going into Pylon W.PNG
    164.3 KB · Views: 2
  • 2-12-2022G 800W heater on again, 1078W going into Pylon W.PNG
    2-12-2022G 800W heater on again, 1078W going into Pylon W.PNG
    164.7 KB · Views: 2
  • 2-12-2022F 2934W PV causes Pylon W to take 1088W, still 1500W going missing.PNG
    2-12-2022F 2934W PV causes Pylon W to take 1088W, still 1500W going missing.PNG
    170.9 KB · Views: 2
  • 2-12-2022E PV W now back up to 2512W, 1500W going missing now.PNG
    2-12-2022E PV W now back up to 2512W, 1500W going missing now.PNG
    168.8 KB · Views: 2
  • 2-12-2022D 800W heater turned OFF, notice PV W has dropped, it climbs back up to 2500W in ten ...PNG
    2-12-2022D 800W heater turned OFF, notice PV W has dropped, it climbs back up to 2500W in ten ...PNG
    167.4 KB · Views: 2
  • 2-12-2022C 800W electric heater on, note Pylon W of 646W.PNG
    2-12-2022C 800W electric heater on, note Pylon W of 646W.PNG
    168.8 KB · Views: 1
  • 2-12-2022B 1kW loss when testing with 2kW fan heater.PNG
    2-12-2022B 1kW loss when testing with 2kW fan heater.PNG
    163.3 KB · Views: 1
  • 2-12-2022A 1.6kW going missing.PNG
    2-12-2022A 1.6kW going missing.PNG
    176.7 KB · Views: 1
  • 2-12-2022I 800W heater turned on, now taking 155W FROM Pylon W despite 2274W PV.PNG
    2-12-2022I 800W heater turned on, now taking 155W FROM Pylon W despite 2274W PV.PNG
    152.4 KB · Views: 2
Last edited:
I think there is a wiring problem.
I tend to agree. I'm not especially familiar with Pylontech, but similar products I have looked into have a bunch of opportunities for poor internal connections. Add in the field terminations and you really have a lot of opportunity for hot connections.

As a general rule, all connections should really be torqued at installation, and then a follow-up re-torquing after a week or two of cycling when you deal with things that go between 0-100% load multiple times in a day. Doing it again every year or so used to be recommended practice, but now most people just do thermal IR scans instead and only re-torque if there is evidence of a problem.
 
12VoltInstalls, while I tend to agree with you, how do you explain the PV W figure being consistently so much higher than the house load plus the actual battery charging load, which I measure frequently with my new clamp meter (which I trust)? Why would the MPPT controller take 2.5kW from the solar panels if there was only 1kW of load to send it to? If you look at the screenshots in my last post, you can see how I test the PV watts by turning on an 800W electric heater, and then I went to check that the electricity meter wasn't turning at all (it wasn't), so the power was coming from the PV. And I've done other tests previously, where the PV was 4kW and I turned on a 3kW immersion heater, and my house's background load was 500W, and the electricity meter didn't turn and the PV stayed at 4kW, so I trust the readings on MultiSIBControl - plus, I did many tests with a single solar panel and an Elejoy solar panel multimeter, to see how many watts came out of the panel at different angles, different times of the day and different weather conditions, so I know how much to expect from 20 of the same panels, and it all tallies with what I get now.
Now, at 3pm, I have
Load W 371W
PV W 1,112W
Pylon W -279W (that's minus)
So it's as if the inverter is feeding a load of 1kW somewhere - it's taking from the batteries now, even though the PV watts are still much higher than my house load of 372W.
 
Back
Top