Any MPPT chargers with BMS connection?

[kuranaga]

Hi

what I realize building LFP packs is the lack of real comm between BMS and solar chargers. Usually BMS will have some sort of charge control but this either means internal mosfet or some relay output etc. Why after so many years companies like eg Victron do not make their chargers with a BMS input control so instead having to disconnect the charger by relay it will be just digitally driven by BMS signal (cell overvoltage). It could be so simple, but no we have to use all the shady solutions like (faulty) BMS internal relays, external relays and so on...

 

[DLTooley]

The charge controller and the BMS are redundant. The charge controller is primary.

A BMS can track your load and similar. I don’t know if there is any benefit integrating that with a controller. The Victron suite of products provides that functional integration, but that’s over my budget.

 

[toms]

There are plenty of BMS / MPPT systems that integrate.

The simplest one i have is the REC-BMS controlling a SMA AC coupled system.

There are dozens of others available.

I wouldn’t have it any other way for a large system.

 

[kuranaga]

anything for storage systems (not AC)?

 

[toms]

Why wouldn’t you use AC coupled with your storage system? Much simpler with large PV.

 

[Diysolar123]

they really have two very seperate and different jobs...
Your BMS has the sole job of protecting your "expensive" battery pack and watching the health of the individual cells, that is all it should care about; it is protection of the energy storage "black box".
Your charge controller is trying to maximize the amount of power you can get from your solar array and output that power within the bounds you specifiy (voltage and current) into this "black box".

when you start looking at highly integrated systems, well, thats the land of expensive commercial gear for the big boys...
As a DIYer I think you gain a lot of design flexibity by breaking down the components into lower cost functional modules that you then put together, and can easily swap out.

 

[toms]

I’ve seen too many cheap systems fail for me to bother - i just can’t afford to buy cheap gear.

Over the life of my system the difference in cost between the cheapest BMS that will do the job, and one that has CAN protocol communication with my inverter/charger works out to $4 / year.

Having the ability to taper initial and final charge current alone is likely to extend the life of my batteries long enough to cover that.

For a whole house off-grid power system there is no way i’d use a BMS that couldn’t control charge current.

 

[hour]

Yeah, you can diy it. Victron text protocol allows you to request certain data from the MPPT over wire. So any MPPT (even the old bluesolars, as long as they have a VE Direct port), in to a $3 microcontroller that also reads data over wire from the chinese smart BMS units and makes decisions.

You can listen to the victron mppt data as it constantly shouts out (24-7, just prints every parameter endlessly) or you can request specific data and ignore all the noise with their protocol. Ex: I can pass something like 83592395 and get back a value like 21000 for PV voltage aka 21volts.

In learning about that I found the command for enabling and disabling charge (and it worked, flawlessly). You just speak that series of numbers to it instead of the numbers above and ta-da on or off. So yes, reading the cell voltages from your BMS can immediately trigger your MPPT to disable charging, same as if you had pressed the toggle switch "Enable Charger" in the Victron app.

I vaguely remember doing this one night maybe 8 months ago, and was really pleased with it. Then I found in their docs a way to specify charge current and that's what I really needed. If I read from my BMS that a cell was getting wonky and my charge current is exceeding what my BMS can balance out, reduce charge current. Re-evaluate 15 seconds later or whatever. This would achieve the most holy of in-operation pack balances IMHO, same as what I see using my iCharger X6 to charge my 12v packs. Reduce current and let things smooth out.

Unfortunately I couldn't figure out how to pass specific values because of the weird encoding they use.. If it was as simple as "8359" + 03 for 3 amps + "395" for example, yea bingo. I never got it figured out and rage quit. But very TL;DR just spend a day and DIY it.

 

[kuranaga]

what about victrons with BT? there is no more a VE direct port.

 

[hour]

what about victrons with BT? there is no more a VE direct port.

Don't those have ve bus or whatever their other communication method is? I think people with the larger high end victron stuff probably have their color control gx thing, or use a $35 raspberry pi and load the software on that (google victron raspberry pi).

And if that's the case there's even more hacking done on pi's per Victron's open source section on their site (also google that). Shouldn't be hard to find something on github. And raspberry pis can also read chinese bms data. Or you can still handle bms reading through a cheap microcontroller that has wifi (esp8266, esp32) and serve data over local network to the pi. Queried every 5 seconds or something.

Sorry I never graduated from a pi + esp32-type board and always did my bms and ve comms on the board. heavy lifting web operations and data storage on the pi. Now down to just the esp32 type board (particle photon) because I got things dialed in and no longer needed a billion snapshots of the system every 24 hrs.

 

[kuranaga]

ok Ive checked, there is still the ve direct port (physical connector) but it is designated as jumper space. Is it still a full VE direct? Then I guess the jumper just shorts TX and RX so they can detect the jumper presence by reading back from the serial port, my guess...

 

[kuranaga]

ps the cable I did:

https://i.postimg.cc/yNXWL3NQ/IMG-20211109-091526-1.jpg

is just a max232 serial lvl converter, so the ve.direct is just a serial port

 

[hour]

ok Ive checked, there is still the ve direct port (physical connector) but it is designated as jumper space. Is it still a full VE direct? Then I guess the jumper just shorts TX and RX so they can detect the jumper presence by reading back from the serial port, my guess...

I have two 75/15's, a 100/20, and a 100/30.. i know I've seen (and must have removed) a jumper from one or all of them.

Looking at my amazon order history I'm 99% certain that you need a JST PH 2.0 cable to fit the ve direct port. Victron also sells an isolated ve direct cable that you can hack apart I guess. Mine has been fine for just over 2 years now 24-7-365.

If you're going that route and do have a ve direct port to use, I'd really suggest the particle photon. They're pricier than arduinos but 1) I can send you pics and code snippets to get going, and 2) you can use their cloud and set up functions and variables. It's invaluable for logging when you don't want to be sitting beside your battery with a computer. Can pull all your victron and bms and whatever real time data online and format it in a quick webpage, read it raw, whatever. And function wise, can manually turn on/off mppt, work on code and deploy over cloud and then check it out. I hate sitting 3 feet from an arduino for an entire day.

 

[kuranaga]

I have two 75/15's, a 100/20, and a 100/30.. i know I've seen (and must have removed) a jumper from one or all of them.

Looking at my amazon order history I'm 99% certain that you need a JST PH 2.0 cable to fit the ve direct port. Victron also sells an isolated ve direct cable that you can hack apart I guess. Mine has been fine for just over 2 years now 24-7-365.

yeah it is a ph2 cable but ready to go cables are not expensive. I would not build it again, not worth the hassle