dtu as "surplus" charger of a LFP battery ?

[typxxi]

This idea has come to my mind when I helped a family member who has rented her flat where she has some limitations how to use solar power.

The main goal is to avoid any additional DC cabels going through the wall / window and running a LFP battery (inside).
She wants it clean and tidy which means short cables from the 4 x 400 Wp panels to the 2x HM-800 inverter and from there into the plug of the balcony.

• 400 Wp (4 panels but 3 of them in different directions / angles like South 90° and West 90° and kind of roof style West 8°)
• 2x HM-800
• Shelly pro 3em
• Huawei Charger or anything like that
• 3,6 kWh LFP 16S 51V battery with JK BMS

#1 the hoymiles is generating and feeding the energy into the local grid
#2 could the dtu now being used to charge the LFP battery with the Huawei LFP charger based on current surplus (production -consumption) ?

It's just about a clean installation without all the cable mess that usually would come along with this kind of installation.
Of cause AC detour to charge a battery is obviously not the most efficient way.

But if you can not drill a hole and have 3 different panel directions then you will also end up with a 24 V battery system (the voltage from 1 panel is about 40V so not enough to charge a 48V battery), which then would require an even bigger cable diameter than the 48V system.

Inverter and panel mess should stay outside, charger and battery inside.

Would that kind of surplus charging be possible ?

[spcqike]

I don’t think so as you would still need a third inverter. You can’t connect the battery and PV Module to the inverter at the same time. You would either need to switch between on or another or have a third inverter.

If you are going for a third inverter, why not simply use a Victron multiplus? This can than AC charge the battery and feed power back to the grid.

[Manos1966]

why not simply use a Victron multiplus?

Hi @spcqike , how does the Victron multiplus handle Nulleinspeisung? Is it included or is extra HW needed?

I just noticed, verkauf-bochum offers for 479EUR the MultiPlus-II 48/3000/35-32
Compared to the cost of Huawei + Hoymiles, it is an interesting alternative, especially when I see how expensive the Huawei became (compared to the days we found it for 80-90EUR).

[Solarteur]

So far I am operating 4 PV modules on 2 HM-600.
A Shelly EM3 is mounted in the meter cabinet. This part should stay as it is.

A few days ago I put the following system into operation as an extension:

• 1 Pylontech US2000C,
• 1 Huawei R4850G2,
• 1 additional HM-800.
  Everything is set up as described in the OpenDTU-OnBattery-WiKi. Either the inverter or the charger on the 230V AC side is switched on via a changeover relay. This relay is switched via a Shelly 1 PM depending on the position of the sun: charge when it is bright, feed in at night.

During the day, the battery is charged with "solar surplus" and in the evening/night the household requirements up to a maximum of 600W are fed in from the battery --> inverter until the battery has an SoC of 20%.

I am amazed at what is possible with the great “OpenDTU-onBattery” project. Thanks to the developers!

[typxxi]

Thanks a lot so far

Meanwhile I have drawn a sketch to make it clearer.
What about this scheme which I have modified from the 4th example shown in the github reference?

The only issue right now might be the missing support for HMS-2000-4T or is that only caused by the missing CMT2300A module ?
Is that a question of time / developement or a technical limitations we can not overcome?

Right now I should be able to solve it at least with a second opendtu I guess.

The other choice would be to move the whole stuff outside in a cabinet (battery with a victron mppt 100-30) to run it as a 24V battery system with 5 panel in parallel (to solve the issues of 5 panels in 4 directions angles) and a Hoymiles HM 800 or 1200 as the main inverter.

[Manos1966]

The only issue right now might be the missing support for HMS-2000-4T or is that only caused by the missing CMT2300A module ?

I think you have not understood the Wiki explanations well enough.
It is currently not possible to use at the same time:

• the HW necessary to control the Huawei R4820
• the HW necessary to control the HMT/HMS Series Inverters

There is a lot of explanations why at the Wiki but, "it is as it is" right now.

Can it be fixed? Yes of course. Can you program? (I guess not, otherwise you would not be asking).

So, the VERY EASY solution is:

• leave the HMS-2000-T4 aside for other projects
• put two HM800 in the roof to handle four panels,

Problem solved! 👍

[Manos1966]

to solve the issues of 5 panels in 4 directions angles

regarding this, you need to read more (you are not right).
Although it is NOT RELATED to OpenDTU-OnBattery, there is some information in the Wiki to help you start with your search:
https://github.com/helgeerbe/OpenDTU-OnBattery/wiki/Victron-MPPT-solar-charger#eastwest-orientation-of-panels-with-only-one-victron-mppt-charger-

PS. I like the picture. It is a nice example of situations many people living at rented places experience!

[typxxi]

I think you have not understood the Wiki explanations well enough. It is currently not possible to use at the same time:

• the HW necessary to control the Huawei R4820
• the HW necessary to control the HMT/HMS Series Inverters

I have read the introduction and tried to understand the key points, I have also read the Huawei psu part too, and both mentioned the same:

The example had mentioned the limitation as above the graphics as
"(using the Huawei AC charger in combination with the CMT2300A radio board for HMT/HMS Series Inverters is not supported at the moment"

and in the Huawei document it is quite similiar

"Using the Huawei AC charger in combination with the CMT2300A radio board (for HMT- HMS- Inverters) is not supported at the moment"

Therefore I was aware of the issue and had placed an 2x HM 800 als alternative in the picture and asked about it.

• leave the HMS-2000-T4 aside for other projects
• put two HM800 in the roof to handle four panels,
  Problem solved! 👍

Not really cause the HMS-2000-T4 inverter and 6 panels had been ordered and delivered 3 weeks ago. It was sold as a kit before she had ever thought about a battery and how to implement that. I am only familiar with the big 15 kWh rack batteries and our diy 25 kWp array which I had built in the previous years. Therefore she had asked me after she had oreder it, but was also aware about the wiring and cable mess in the test phase back then in 2019 & 2020.

But why do I need to control the HMS 2000 output really cause even the PV Sol Premium simulation has never shown a peak above 1000 W due to the poor orientation (4 panels are facing West 290°, 2 of them in vertical angle and the only 1 facing south is vertically mounted and suffering from shadows from 10 am to 3 pm, when the top 2 panels will get close to 400 Wp which are also facing west 290°).

If I simply ignore the HMS 2000 4T as that is simply generating then I am able to read and control

#1 SmartMeter EM3 Pro
#2 the Huawei charger via CANBUS and
#3 the HM 800 for the generation during night or in case of peak demand

Our HM 800 is also a 2 mppt inverter and regarding the panels and directions I have used the HMS-2000-T before in a similiar situation for a fence and a garage string out of old panels. And there it had worked quite well with the opendtu as a reader and a bridge for the HA integration, before I moved ahead to Growatt interter.

I do not know what should be wrong with the HMS-2000-4T
and yes, I had read that before and the issue that misreadings might occure but I do not care about zero greed limit.

Afaik I can use the dtu to work with Shelly Pro 3EM, the 2nd HM inverter AND with the psu to charge the battery.
Therefore I have removed the "communication line" from the panel inverter to the dtu on battery in this diagram.

But i will also do a 2nd planing for a 24 V outdoor (battery & inverter) system over the weekend as an alternative

Thanks !

[Manos1966]

Oh, I see what you mean.... and you are right!
Why did you not say this earlier? 😜 (just joking)
In Automatic mode the Huawei power is controlled by the Power Meter and information provided in the web-interface. If excess power is present the PSU is turned on and maximum Huawei power equals the Power Meter value.

This should work! (I need to add some explanations in Wiki) 👍

@schlimmchen , @spcqike @anybody_with_a_Huawei how does this explanation sound?

• Using the Huawei AC charger in combination with the CMT2300A radio board is not supported at the moment. In other words:
  • You can not connect a HMT/HMS Inverter to the OpenDTU-OnBattery, the Huawei rectifier is attached to. Subsequently, you can not use a HMT/HMS Inverter to control the "zero feed-in" (Nulleinspeisung)
  • You can however use a HMT/HMS Inverter for the solar power generation (which you will need to monitor using a separate OpenDTU)
  • For more information: https://github.com/helgeerbe/OpenDTU-OnBattery/wiki/Device-Profiles#other-information

[typxxi]

im my opinion the power generation HM 800 inverter would act

1. as a kind of "booster" during daytime if more is needed than the solar panel are delivering via the HMS 2000 where the Shelly and openDTU will determine the amount of boost needed from the HM-800 now and
2. during night time the openDTU will run the same way where the HM-800 is the only source available like shown below

And then I can add another HMS openDTU with that CMT2003A chip to monitor the HMS solar power generation.

The last remaining point is mentioned in the Huawei document but I was not able to solve:
make sure you get an ESP with at least 2 SPI which I could not really find or do not now how to look for that.
I know about the RAM versions and the more the better, I know about the D and U letter , but not sure what explains if a board has 2 or even more SPI.
And of cause I read about the importance for 38 pin board to not run out of pins.

So far so good - this concept looks good, if I will get all the parts and pieces missing like the battery and the exchange of 2 panels.
But meanwhile we had looked deeper into a way to get the DC cables in or how to handle with 5 panels x 2 cables = 10 DC cables each about 4mm² thick.

But this will be part of another comment cause that would start from scratch, but looks even a bit better due to less investment and complications involved.

[typxxi]

Same story as above but this time as the usual DC concept with higher efficiency and less investment.
This one should work out of the box with Victron mppt, JK BMS, 24V LFP battery and only 1x HM-800 / HM-1200 / HM 1500.

We found a way (hopefully) to get the DC cables through the wall into a cabinet, but the size / space for the cable is very limited so I had to do the math and calculate and optimize a lot. First I needed to find out the peak production of each panel at any given time during the year (peak Watt). That is required to determine which panel could go with which other panel togehter through 1 cable to stay within the usual 2 - 3% voltage drop. The goal ist to get rid off the 5x 2 cables from the panel to best 2x cables only, but this did not work out.

The issue here is that these 5 panels will never produce 5x 10,2A = 51 A cause during peak 12 am the 105° = WestNorthWest facing vertical panels will get close to 0 sun except reflections. Therefore the 51A are gone and considering the 2nd vertical panel is facing WSW and the 3rd West , there was no way to every achieve 50A or 40 A. It looks like peak will be around 32 A - but for 3 days only and just 12 hours a year. Therefore a 150V 35 A mppt should be enough. I had to calculate the cable diameter based on peak load which will be quite even after the simulations I hade made, roughly 15,5 A to 16,5 A for each of the 2 strings with 4 mm²

We are merging 3 (weaker orientated) panels into 1 "string" and the other 2 into a second "string" as shown below.

And then we have to push just 4 cables through this tiny spot where we were already able to push a small but hard wire through.
2x 4mm² for each string would be OK, where 2x 6 mm² insteat 4x 4 mm ² would not be enough.

The mppt is now a bigger one and also capable to deal with 48V, even though we have to stay with a 24V battery due to the 5 panels in parallel providing just 42 Volts.
Or a step up booster would be required and then you loose at the booster what you can gain by the 48V efficiency.

[Manos1966]

@typxxi I have no idea what simulations you are doing but, I have the impression you are looking the panels-story from a wrong angle.

First of all however: This is not the best forum to learn about Panel orientation and handling. All your questions (which I had myself also, some months ago) are better answered at the Victron Forum.

even though we have to stay with a 24V battery due to the 5 panels in parallel providing just 42 Volts.

You are working on the wrong start hypothesis....

And in order to help you with the start of your further search (at the Victron forums):

• What will happen if you connect all panles, irrespective of Orientation, in series?
• do not forget to include the temperature coefficiency of Voc to your calculations (for safety)
• and of course the very good tool Victron has developed to help you ensure your selected configuration is within safety limits: https://www.victronenergy.com/mppt-calculator

[Manos1966]

@typxxi Hmmm, I see....
There are some things we need to set straight here.

1. The nice thing in this forum is the polite form used by all participants. Everybody expresses his opinion based on his knowledge and experience AND nobody insults other members.
   I do not remember anyone giving you a "free card" for the later so, please refrain from posting derogatory remarks about other members in this forum. If you like insulting people, find other forums to do that.

2. "says the bloody amateur without any clue"
   Ofcourse I am an amateur. This is actually a forum of amateurs who try to share knowledge and help each other.
   If you are looking for "professionals", maybe you should get in contact with a professional company which sells Victron systems and inverters.

3. "you would have not started the wrong discussion about Voc in a system of 42 V and a mppt that can live with 150V."
   I wrote:
   -- What will happen if you connect all panels, irrespective of Orientation, in series?
   -- do not forget to include the temperature coefficiency of Voc to your calculations (for safety)
   -- and of course the very good tool Victron
   You need to consider all points TOGETHER, not select one to use as argumentation about my knowledge on Victron MPPTs

4. "510 kWh less production based on your stupid advices ! You are the master for sure."
   This is a complex one... :)
   -- If you knew what you are doing, you would not be posting on the forum. Is this a "mental mast*** excersize" to prove to the rest of us how well you know what you are talking about or, are you asking for our opinion?
   -- In risk of repeating myself, "I have the impression you are looking the panels-story from a wrong angle". But, since you are sure of yourself, I respect your superior knowledge. Your additional 510 kWh (at 42V) is the "game changer" you are a master. 🥇

5. "And I am not aware of what I am talking ?"
   We agree to disagree!

6. "But how do you wanna charge a 48V battery when you only have a 42V mppt voltage available if you want maximum Solar Power generation and not such stupid move to a serial = string system ?"
   -- Hmm.. In risk of repeating myself, "I have the impression you are looking the panels-story from a wrong angle"
   -- Wait, there was a guy who told you "This is not the best forum to learn about Panel orientation and handling. All your questions are better answered at the Victron Forum." Oh, hold on, that was me! 😆

7. "but I had never talked about set of strings due to all the different azimuth and angles."
   This is why I am repeating "I have the impression you are looking the panels-story from a wrong angle".
   This is a good example of what I mean by saying "we use the polite form of expression in this forum".
   (In a DIFFERENT FORUM, one could have used the expression "you have no fuc#### idea what you are talking about", but as I said, this is NOT the way we express ourselves in this forum and we respect everybody's opinion, no matter how bad it is ).

8. "I guess you should stop trying to help people with panels and solar power beyond what you really have understood."
   Who was that guy who posted "This is not the best forum to learn about Panel orientation and handling. All your questions are better answered at the Victron Forum." Damn, I am repeating myself, again 😆
   And please, no "guessing", it is not your strong point, it is even worse than your "knowing"

9. "people who have learned well can learn about how to design such a system down to the cable diameter and make sure that things will work out."
   This is exactly what scares me, reading your posts!

10. "we have to stay with a 24V battery due to the 5 panels in parallel providing just 42 Volts. Or a step up booster would be required and then you loose at the booster what you can gain by the 48V efficiency."
    I am tempted to say "You are absolutely right, you know everything better, good luck, live long and prosper!" 👍
    I will however, (with my amateur "hear say" knowledge 🙄 ), recommend you have a look at https://github.com/helgeerbe/OpenDTU-OnBattery/wiki/24V-Battery-issues-with-Hoymiles

[swingstate]

24V doesn't make much sense respectively beyond 500W the inverter will become highly inefficient - I had to learn that the hard way with the first version of my setup. Get a 48V in place (or live with sub 500W unless you want to was energy).

[github-actions[bot]]

This discussion has been automatically locked since there has not been any recent activity after it was closed. Please open a new discussion for related concerns.