Connecting PV panels to resistive underfloor heating

I’d like to use PV panels connected to a resistive underfloor heating. The system should be fairly small (one room) built from 4 panels 300Wp (mpp voltage 32.6V). During the night the underfloor heating should/could be powered from mains supply, based on a setpoint temperature. The expectation is that for many nights it might not need to be turn on at all due to a heat accumulation in the floor mass, but the option to do so, would be great.

I’d like to use a hackable system that allows me to learn and later maybe expand to other areas (adding batteries, more power etc). What can I use from the available designs? I am looking at the MPPT 2420 HPX and I could fit into the input rage by series-parallel connecting the PV panels for around 65V at which point I’d need to be able to take in maybe 20A tops. I have some freedom in building and connecting the heating elements but overall power should be adequate to the panels - 1.2kW - which seems a bit stretch.

What would I have to change to be able to base my system on the designs available here?

Hi Zbynek,

do you know the open source “DMPPT” by Electrodacus? That’s designed for exactly for your described purpose:

1.2 kW is a bit too much for the MPPT 2420 HPX. It could work with two of them in parallel, though.

I knew about it but never got to exploring it properly - every time I tried I was turned off by the webpage.

You mention it is open source - but where is the source? The webpage is very very strange - containing mostly pictures that link to pdf user manuals and most of them seem to contain the same content (tens of pages). I am having a very hard time getting any useful information from there. I can guess from the names of the products that stuff with SBMS in the name is likely to be Solar Battery Management System (so it is a battery charger?) and DMPPT is the maximum power point tracker that switches the resistive loads. But it beats me what DSSR20 is - it is mentioned only twice and one of the mentions is paypal button.

So, is that stuff hackable? Is it even open source?

You are right, it’s quite difficult to understand and the documentation is not ideal. There used to be a Google+ (or how was Google’s facebook equivalent called?) forum some time ago, but that disappeared when google shut down G+.

For most the designs, you can find the links to the sources (HW and FW) on the last page of the manual.

And there are some Youtube videos that explain e.g. the DMPPT idea.

OMG, open source project without github (or equivalent) repository?

I get the idea - you build a variable resistance load by connecting fixed ohmic loads in parallel. Then add a MCU and bunch of mosfets to control it, to maximize the power sinked into the load. That is the idea, isn’t it? I see that when batteries are added they are treated just as another kind of load.

The idea itself seem cool but I am missing any kind of developer facing documentation or anything supporting hacking it in any way :frowning:. Have you got to actually looking at how it works internally? Is the SBMS different in any way from your MPPT 2420?

The SBMS is quite different from the Libre Solar MPPT. Most importantly, the SBMS is not an MPPT charge controller. It’s more like a switch that disconnects the solar panels after a certain battery voltage is reached. As far as I know, it doesn’t even use PWM charging, so there is no constant voltage (CV) charging phase.

The software is quite messy, so I didn’t have a closer look. But I kind of like the hardware ideas.

I decided to separate BMS from charge controller to have a second measure of protection. If in the SBMS the MOSFETs towards the solar panel fail shorted, the battery would be overcharged without any possibility to prevent it. If there is a charge controller + a BMS, we don’t have a single point of failure and the entire system becomes more safe.