Page 99 of 99

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 12 Nov 2019, 23:07
by wolfandy
Thanks for all your feedback, Coulomb. I really appreciate it

But just for clarification: Are my original following statements not incorrect?
  • In the morning, any PV would supplement my consumption (battery still empty)
  • Once PV exceeds my consumption, it would charge the battery
Would this not mean mixing sources in the morning? As my Axpert cannot do this, would it then not be rather this?
  • In the morning, while PV is less than consumption, PV charges the battery
  • Once PV exceeds consumption, PV covers consumption and any excess is used for charging the battery
Sorry for being a bit pedantic, but I am still trying to figure out whether PV makes financial sense for me (and if yes, what size). And this would mean that until my PV production reaches my consumption, I will continue to use the grid - which means that I cannot really claim this portion of the PV production as a saving (the battery would get fully charged throughout the day anyways)

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 00:30
by Mantas
weber wrote:
Tue, 12 Nov 2019, 15:23
Hi @Mantas. If you want to have any chance of paralleling the new machine with the old ones, you must not buy a 64 V machine. That means you must not buy a PIP-5048MS. You need an Axpert MKS 5K that does not have the 64 V option, or a rebadged equivalent such as the Mecer SOL-I-AX-5P (not the SOL-I-AX-5P64 or the SOL-I-AX-5KP).
Hi Weber.

Thanks for that. I would not have made that mistake, but better safe than sorry, I suppose 😉

I have my replacement RCT 5000 mks installed and this one is authentic in all Axperts :lol: All the right stickers in the right places, right colour screws, correct serial number assembly, correct serial number in WatchPower; so I think I’m good to go.

I have a new question however. The new guy is on firmware U1 74.31

Do I still flash PB1_72.70b to both the 5KW and 4KW in order to parallel them?

The new inverter also shows U2 04.12. Can I leave that as is? My PV array only connect to this inverter. For some reason I get no PV readings on the other inverter, but that is a problem for another day.

Thanks once again for all your assistance, I truly appreciate it!

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 00:32
by Mantas
wolfandy wrote:
Tue, 12 Nov 2019, 17:25
Hi all

I am a new member from South Africa :)
Hello fellow South African ;)

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 01:05
by Mantas
wolfandy wrote:
Tue, 12 Nov 2019, 23:07

Would this not mean mixing sources in the morning? As my Axpert cannot do this...
wolfandy, what do you mean yours can't mix sources? I had the exact same model (it recently went faulty after nearly 5 years) and it quite happily switched between sources. In the morning it would start charging the batteries from about 6am with the bit of solar I got and fed my appliances from council power. From about 9am I would have enough solar to charge the batteries and supply what I needed to run computers and lights. If I switched on the kettle, for instance, it would quite seamlessly switch over to council power, for that bit of time.

I do use mine in Solar priority output mode though and not in SBU. I need my batteries for backup with our loadshedding woes.

As far as charging the batteries is concerned, I have set mine to charge from solar first and utility at max 10A. I still save enough to make a half decent dent in my utility bill. I could probably save alot more if I ran in SBU but like I said, I need my batteries full at night in case of a power failure.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 06:48
by coulomb
wolfandy wrote:
Tue, 12 Nov 2019, 23:07
But just for clarification: Are my original following statements not incorrect?
  • In the morning, any PV would supplement my consumption (battery still empty)
  • Once PV exceeds my consumption, it would charge the battery
Would this not mean mixing sources in the morning? As my Axpert cannot do this, would it then not be rather this?
  • In the morning, while PV is less than consumption, PV charges the battery
  • Once PV exceeds consumption, PV covers consumption and any excess is used for charging the battery
You didn't mention powering the loads from AC-in. I was thinking of the situation (as I have here) where it's extremely rare to switch to AC-in at all, certainly not every night. In those cases, the inverter-charger is blending battery power and PV power to supply the loads, and if/when the PV power is sufficient, all load power comes from PV, and the battery is charged from any excess.

So let's split into two cases:
1) Inverter is in "line mode". AC-in supplies all the load, no PV supports the loads at all. All PV power charges the battery. If enabled, the AC-in can charge the battery at the same time as the PV. It doesn't matter whether PV power exceeds the load or not; nothing special happens at that point. The inverter-charger will switch out of line mode only when the battery voltage exceeds the value in setting 13, the "back to battery" setting. Ideally, this state is to be avoided if possible.

2) Inverter is in "battery mode". The battery and/or PV supplies the load; AC-in is not used at all except to synchronise the inverter output for possible quick and minimal-glitch switching to AC-in if required. In this mode, if the PV power exceeds that of the load (ignoring losses for simplicity), then the excess charges the battery. If the PV power is less than the load, the battery supplies the difference, and discharges a little. The difference can't be taken from AC-in. The inverter-charger stays in this mode until the battery voltage drops below the value in setting 12, the "back to utility" setting. Ideally, this is the mode you'd prefer to have your inverter-charger in most of the time. But of course, to do that, you need sufficient battery storage and solar panels, appropriate to your loads.
And this would mean that until my PV production reaches my consumption, I will continue to use the grid -
As noted above, you will continue to use the grid until the battery voltage reaches the value in setting 13. If you want, you can make this fairly low, so that you don't spend much time using grid power.
which means that I cannot really claim this portion of the PV production as a saving (the battery would get fully charged throughout the day anyways)
I think you should look at it as a saving: the energy you put into the battery will save you from having to use grid power at a later time. There are losses of course, so the best way to use solar power is to be in battery mode, organise your loads to coincide (as much as is practical) with solar availability, so that the loads are powered directly from the panels. That way, you don't have to put the energy into the battery and take it out again with losses, and you don't cycle the battery unnecessarily.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 07:30
by weber
Mantas wrote:
Wed, 13 Nov 2019, 00:30
I have a new question however. The new guy is on firmware U1 74.31

Do I still flash PB1_72.70b to both the 5KW and 4KW in order to parallel them?

The new inverter also shows U2 04.12. Can I leave that as is? My PV array only connect to this inverter. For some reason I get no PV readings on the other inverter, but that is a problem for another day.

Thanks once again for all your assistance, I truly appreciate it!
Thanks for the kind words.

You can actually do a little better than PB1_72.70b. You can flash them both to LC1_72.70c, as you can read here:
https://powerforum.co.za/topic/2554-axp ... ment-42385
But note that the bugs HannesE thought he saw, in the way things were displayed on the LCD, are actually features once you understand them, as Coulomb explains 2 posts later.

It's possible that it might also work if you flash them both to LC1_73.00e. But this has not been tried.

I note that once you start reflashing, there is no way to restore your existing 74.31. But 74.40 is available.

Yes you can and should leave U2 04.12 as it is. That's good SCC firmware.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Wed, 13 Nov 2019, 16:18
by coulomb
derBastler wrote:
Sat, 02 Nov 2019, 21:56
We have VW E-Golf Batteries, VW Passat GTE, VW E-Up, BMW i3 and all other manufacturer batteries. A very typical batterie nominal voltage is 44,4V. Upper limit 50,2V, Lower Limit 38,4V

It would be great to change the PIP-5048MS to this values to bring a hugh mass of batteries into second use applications.

Do you think this is possible?
Weber and I have been considering this possibility. One thing that bothered us is how such a low battery voltage will generate enough bus voltage to generate a 230 V sine wave. In fact, we began to wonder how these inverter-chargers work down to as low as 40 V (the lowest possible value for setting 29, the Low DC cutoff voltage). We know from experience that the bus voltage is usually very close to 8.0 times the battery voltage, presumably due to the high frequency transformer turns ratio of 1:8. 40 V times 8 is only 320 V, which is not enough to generate the peaks of a 230 V sine wave (a bare minimum of 230 x √2 = 325 V is required).

Weber conjectured further that the 64 V models (which can charge the battery up to a maximum of 64.0 V) must have a lower turns ratio in their transformers, because 64 V times 8 is 512 V, and the inverter bus capacitors are rated for 500 V. Indeed, he found that his 64 V model usually had a bus voltage of very close to 7 times the battery voltage. Yet these models still allow the battery voltage to go as low as 40 V. With the battery voltage at 40.0 V, the bus voltage would only be 280 V... how does that produce a 230 V sine wave?

There is a circuit between the DC-DC converter and the inverter proper, called the buck converter in some service manuals. Its main purpose seems to be to buck a high bus voltage down to an appropriate voltage for utility charging the battery. In other words, it's a buck converter when power flows from the inverter output (connected to AC-in when utility charging) towards the battery. We wondered if that converter could act as a boost converter when power flows from the battery to the load (i.e. in battery mode). There seemed to be clues in the firmware that supported this idea. However, when we looked at the block diagram in this post, it sure didn't look like the buck converter was bidirectional.

To sort out what really happens with a low battery voltage, Weber kindly reconfigured his battery from 16S of LFP (nominally 51.2 V) to 13S (nominally 41.6 V).His 64 V model inverter claimed it was generating 230 Vac on the LC Display. With the battery voltage at 44.4 V, the bus voltage (as revealed by the QPIGS command) was 311 V. (311 / 44.4 = 7.005). How was it generating 230 V? So we set up Weber's DSO to look at the generated waveform. This is what we found:

Clipped output 42.9.png
Clipped output 42.9.png (4.52 KiB) Viewed 270 times

We measured the peak to peak voltage at 576 V, or 288 V peak to zero. The bus voltage would have been about 42.9 x 7.0 = 300 V, so there is an overhead of some 12 V.

Using the default output voltage of 230 V, to avoid clipping you'd need about 230 × √2 + 12 = 337 V. With a 58.4 V model (1:8 transformer), that means the lowest battery voltage to avoid clipping would be 337 / 8 = 42.1 V. That would be 3.51 Vpc for a 12S NMC or NCA battery, which seems to be about 12% SOC, per the table below.




With a 64 V model (1:7 transformer) the situation is much worse, even when you tell the inverter to only output 220 V, as shown in the following table. Using an output voltage of 220 V instead of the default 230 V helps, but only a little.

                      Onset of clipping
Model    Out V  | Bus V     Bat V     Cell V (12S)    SOC (12S NMC)
----------------+---------------------------------------------------
64V       230   | 337      48.1 V    4.02 V           90%
64V       220   | 323      46.2 V    3.85 V           75% 
58.4V     230   | 337      42.1 V    3.51 V           12%
----------------+---------------------------------------------------

In summary: Unless you don't care about the output clipping, using a battery whose nominal voltage is less than 48 V isn't going to work well with a 64 V inverter model.

[ Edit: many minor rewordings; thanks Weber. ]

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Thu, 14 Nov 2019, 09:25
by weber
I'd state the conclusion a bit differently from Coulomb. I'd say the 64 V models are an abomination. They claim to work with nominally 48 V lead-acid batteries, and they claim to be pure sine wave. And yet they can't even produce a pure sine wave at 48 V, if they have any kind of load on them.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Thu, 14 Nov 2019, 23:31
by wolfandy
Thanks for your responses, Coulomb and Mantas

Apologies for the delayed response, but I am in Ghana at the moment for work...
wolfandy, what do you mean yours can't mix sources? I had the exact same model (it recently went faulty after nearly 5 years) and it quite happily switched between sources. In the morning it would start charging the batteries from about 6am with the bit of solar I got and fed my appliances from council power. From about 9am I would have enough solar to charge the batteries and supply what I needed to run computers and lights. If I switched on the kettle, for instance, it would quite seamlessly switch over to council power, for that bit of time.

I do use mine in Solar priority output mode though and not in SBU. I need my batteries for backup with our loadshedding woes.

As far as charging the batteries is concerned, I have set mine to charge from solar first and utility at max 10A. I still save enough to make a half decent dent in my utility bill. I could probably save alot more if I ran in SBU but like I said, I need my batteries full at night in case of a power failure.
This is exactly what I have in mind :) With the exception that I am planning 2 different settings depending on the Eskom situation:
  1. If Eskom is in loadshedding times, then as per your above to ensure that the battery are always full
  2. If the loadshedding situation is relaxed (as it is at least in my area at the moment), then SBU to maximise savings
I am running a Pi with ICC software connected to my Axpert, so switching between these settings should be easy.
You didn't mention powering the loads from AC-in. I was thinking of the situation (as I have here) where it's extremely rare to switch to AC-in at all, certainly not every night. In those cases, the inverter-charger is blending battery power and PV power to supply the loads, and if/when the PV power is sufficient, all load power comes from PV, and the battery is charged from any excess.
I just do not think that my 3.5kWh Pylontech will get me through the night. And I am a bit hesitant at the moment to add more battery capacity as I am not sure that it makes financial sense. My existing Pylontech I bought to cover me when we have loadshedding - and this investment is absolutely worth it as it ensures quality of life. But another Pylontech would give me an extra 2.8kWh of capacity (at 80% DoD), which translates into R5.60 savings per night (at currently about R2/kWh from the grid). At a local price of R19.500 for the Pylontech, this means a payback period of around 9.5 years. And I just do not think that this makes sense at the moment. If the local power price goes up further (which it surely will), additional capacity will become more attractive. And as I understand that with the Pylontechs you do not have to worry about mixing batteries of different ages, I think adding capacity at a later stage makes more sense for me.
Inverter is in "battery mode". The battery and/or PV supplies the load; AC-in is not used at all except to synchronise the inverter output for possible quick and minimal-glitch switching to AC-in if required. In this mode, if the PV power exceeds that of the load (ignoring losses for simplicity), then the excess charges the battery. If the PV power is less than the load, the battery supplies the difference, and discharges a little. The difference can't be taken from AC-in. The inverter-charger stays in this mode until the battery voltage drops below the value in setting 12, the "back to utility" setting. Ideally, this is the mode you'd prefer to have your inverter-charger in most of the time. But of course, to do that, you need sufficient battery storage and solar panels, appropriate to your loads.
As per above, this is also the mode that I would prefer - except that I will not have sufficient battery capacity to get me through the night. So I would eventually switch to grid every night

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Sat, 16 Nov 2019, 05:42
by derBastler
coulomb wrote:
Wed, 13 Nov 2019, 16:18
was bidirectional.

To sort out what really happens with a low battery voltage, Weber kindly reconfigured his battery from 16S of LFP (nominally 51.2 V) to 13S (nominally 41.6 V).His 64 V model inverter claimed it was generating 230 Vac on the LC Display. With the battery voltage at 44.4 V, the bus voltage (as revealed by the QPIGS command) was 311 V. (311 / 44.4 = 7.005). How was it generating 230 V? So we set up Weber's DSO to look at the generated waveform. This is what we found:

[ Edit: many minor rewordings; thanks Weber. ]
Thank you for conducting this experiment. At the moment we are working on solutions to this problem. I have 3 of the 64V Version inverter with 450V-MPPT.

At the same time I built a 14S battery pack to see if the inverters work properly with NMC at all. Fortunately, the E-Golf 7 has two types of battery modules built in, 4S3P and 2S3P. You can use them to build 14S3P batteries (Nominal voltage 51.8V).
vw-hochvoltbatteriesystem-e-golf-7f744e24.jpg
vw-hochvoltbatteriesystem-e-golf-7f744e24.jpg (55 KiB) Viewed 151 times
I directly put that to the inverter. DONT do that for 24h use without a fuse! I saw a lot of burning batteries in my career.
IMG_20191115_222117.jpg
IMG_20191115_222117.jpg (3.72 MiB) Viewed 151 times

In summary, I can say that this setup works well. I still have to test it in three phases but it looks good till yet. If you set the Bulk charging voltage=Float charging voltage, it seems deactivated. Additional i deactivated the "equalization". For NMC and NCA you simply use a CC-CV charging.

We continue to work...

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Sat, 16 Nov 2019, 15:34
by paulvk
Now that I have found some software to capture the animated screen
here is it working
Note got rid of dry cell battery now have animated one and got rid of light globe house makes more sense
https://youtu.be/irECmmy1-ds

This will be my base screen for normal monitoring

Regards Paul

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 19 Nov 2019, 00:48
by wolfandy
Hi all

I have now decided that I will go PV and would like to ask for a bit of advice on sizing the PV system for my Axpert. I have received advice from a local installer (who has done a number of installations in my area and seems knowledgeable) but would like to sanity check it.

My Axpert is the 5kVA / 4kW model with max 3kW PV and 145V max SCC. I would like to run a number of larger consumers off PV during the day (AC, pool pump, etc), so I would like to get as close to the 3kW PV as possible.

Here is what my local installer has told me:
  • He recommends JA Solar mono PERC 5BB modules
  • Put 3 panels in series (roughly 40V each) to get as close to 145V SCC as possible - then put multiple strings of these in parallel
  • He says that I cannot exceed 3kW installed capacity as it could otherwise destroy the DC board in my Axpert - and that I need to take into consideration the power tolerance of the PV panels
  • According to him, I need to calculate with a +5% power tolerance (due to conditions in South Africa). However, the specs on the JA panels reads +5W - which is a significantly difference
  • He recommends 9 x 315W panels (2,835W), which with his +5% tolerance calculation would get me to 2,977W
  • However, if I follow the JA spec of +5W tolerance, I would only get to 2,880W
How 'strict' is that 3kW PV capacity limit? Would 9 x 330W panels, which incl. +5W tolerance would total 3,015W be a viable alternative? Or would the risk of damaging my Axpert be too high?

Thanks :D

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 19 Nov 2019, 06:15
by coulomb
wolfandy wrote:
Today, 00:48
Put 3 panels in series (roughly 40V each) to get as close to 145V SCC as possible - then put multiple strings of these in parallel
I don't agree with the idea of getting as close as possible to 145 V; the solar charge controller starts derating (using less power than rated) after 130 V, and will use zero power at 145 V. The multiple strings are needed for these models.
He says that I cannot exceed 3kW installed capacity as it could otherwise destroy the DC board in my Axpert - and that I need to take into consideration the power tolerance of the PV panels
It's definitely not a strict limit like that. However, how far you can go past the nominal power is not known. For a few other models where the manufacturer actually supplies a maximum PV power rating that isn't the same as the nominal power, it's about 12.5% greater than nominal. For example, 4000 W nominal and 4500 W maximum power. So I'd say that in general 10% more than nominal should be fine. I have this model myself, and one of them has been running 3200 W of panels for years without problems. That's 6.7% over nominal.
How 'strict' is that 3kW PV capacity limit?
As stated above, not very. But you should not go too high, as then overshoots could become a problem.
Would 9 x 330W panels, which incl. +5W tolerance would total 3,015W be a viable alternative? Or would the risk of damaging my Axpert be too high?
It certainly won't blow up your solar charge controller. However, the more important question is how many cells are in the 330 W panels. The 310 W panels you mentioned (PERC 5BB) are 60 cell, with a Voc of 40.3. This is perfect for the 145 V max solar charge controllers: even in winter, the maximum voltage will still be just under 130 V. But 72 cell panels have a Voc more like 46 V or more, which means the maximum voltage at 0°C is over 148 V, and with that you really do risk blowing up your solar charge controller. Even if it doesn't blow up, you'll find the controller is stressing out too much, and will be reducing its output to prevent damage. So check the specifications for the number of cells.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 19 Nov 2019, 08:51
by weber
I agree with Coulomb. 145 V is an absolute maximum limit for array open-circuit voltage. You need to know the coldest dawn panel temperature you can expect at midwinter in your location and use it to adjust the panel's datasheet Voc (which is at 25 °C) using the datasheet temperature coefficient of Voc. 3S of 60-cell panels is usually OK. 3S of 72-cell panels is usually not.

There isn't really any limit, in terms of safety, to the nominal power of the PV array. The SCC will not draw any more current than it can handle. It's just a question of diminishing value for money because the output will be clipped in the middle hours of every non-cloudy day. In Australia, up to a 33% oversizing of the array is permitted under the MRET subsidy scheme.

Coulomb, can you please spell out what you mean by overshoot. What quantity overshoots what threshold, and why is this bad?

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 19 Nov 2019, 10:09
by coulomb
weber wrote:
Today, 08:51
Coulomb, can you please spell out what you mean by overshoot. What quantity overshoots what threshold, and why is this bad?
Solar charge current is the quantity that overshoots. It's bad because it could damage the battery (unlikely because it's short duration), or cause the BMS to nuisance-trip. I'm having the nuisance trip issue right now; even setting the maximum charge current to zero, it overshoots by over 300 W). The overshoot of solar current causes the battery voltage to overshoot the absorb voltage setting.

This is exacerbated today because I'm running a load that switches from 1000 W to zero several times a minute (the load is a cheap induction cooker; every setting below 1000 W is "slow PWMed").

When I say "nuisance trip", I mean it's a nuisance to hear the contactors clunking repeatedly, and if the solar charge controller didn't overshoot, it wouldn't happen. However, these trips are highly preferred over a cell getting overcharged and therefore ruined. I'll try a little manual balancing shortly; it's always cell 5 that goes over-voltage.

Solar charge current undershoots as well, but that never threatens the safety of the battery, and the tiny quantity of lost energy is replaced within a minute.

Re: PIP-4048MS and PIP-5048MS inverters

Posted: Tue, 19 Nov 2019, 15:23
by dekin
Hi coulomb
Using an online PV calculator with a maximum and minimum temperature recordered in my location as 33 and 22 degrees, connecting a 375W panels in 3S gives a maximum voc of 144 which is below the 145V for the Axpert King inverter. Is it advisable to still go this route?
I keeping reading from online forums that having a high voc panels gives you maximum harvest as against low voc panels. how true is this?