Vassen from Sandton

The thread effectively focussed on discrediting Pylontech because of changes to warranties, not being a 1C product, things like that while boosting Hubble. Trying to make it seem like Pylontech will not honour warranties because of a number of fine prints in their documentation.

I don’t really understand these things. Surely the cells in Pylontech and Hubble can’t be that different - One BMS allowing a battery to work harder compared to another one isn’t necessarily a “plus”.

Aaah ok, I had a gander at the thread on the other side.

Out of the three listed initially, the Hubble is the one I’d rate third. That is not to say it is a bad battery, but I have a lot of experience with Pylontech, and also a bit of very positive experience with BSL. Honestly, from a technical standpoint, BSL has one of the best BMSes I’ve dealt with, the only thing I don’t have data on (which for Pylontech we do, thanks to the Ausie battery tests) is how they will hold up.

There are people who sell a Hubble with a Victron inverter, and then they program good voltages on the inverter because the battery cannot communicate with the rest of the stack. Which is fine on a technical level (with some caveats)… but many manufacturers won’t honour a warranty unless you have that comms cable…

I remember years ago, when I was a young Linux-convert who took every opportunity to bash Microsoft Windows, that someone told me I was going about it wrong. You don’t pick the OS first. You pick the best application first. Then you pick the OS that the application works best on. The OS is secondary. The application (which does the job) is primary.

I think some of that spills over to this debate. You don’t pick the battery in isolation. It needs to fit the rest of the stack.

Besides, that 1C argument is rubbish. Load shedding is 2 hours at least. Nobody specs a battery in a solar system for 1C.

Edit: Where 1C comes into the debate is when you are comparing batteries. You cannot directly compare two batteries if one is tested at 1C and the other is tested at C/10 (which at least one manufacturer actually does!). And you cannot directly compare the expected cycle life if one tests to 100% DoD at 1C and the other tests to 80% DoD at C/2. That is true. But once you KNOW the parameters… you can usually make a pretty good estimation as to where they will land. A battery that does 2000 cycles at 100% DoD and 1C discharge rates… will usually do over 7000 if you treat it properly, which is where EVERYONE else specs theirs too…


I agree that it cannot be a one size fits all. Each person and situation will call for a different implementation.

C rating was used as an argument to favor Hubble but when you do a comparison based on the 1C, which the data sheet says gives 3000 cycles, the new argument becomes but that’s at 1C and at 0.5C, it’s a lot more than 10000 cycles. When I ask where does 10k come from, I get given a spec sheet for a totally different battery.

Personally I have enough batteries to not care about C ratings and the pylons work well in my situation. From a price perspective, for me pylons worked out better.

When doing a comparison of the warranty, from reading both the warranty documents both appear to have strengths and weaknesses. Conveniently people only speak about pylontech weaknesses and not the Hubble. The biggest part for me was that Hubble warranty on the BMS is only 2 years. They are a 1 year old company that already have some deprecated products so my concern was what happens after 2 years should the current product be deprecated.

I never stated that it was a bad battery. It’s just one that needs to prove itself in a market and the proof in my opinion can only come with time as more people use it and they have the opportunity to resolve issues.

Maybe we can start a separate thread on battery comparisons focusing on pylontech, bsl, freedom won, Hubble and other currently popular brands and look at the data sheets / warranty clauses. Price is very subjective as they change pretty rapidly. I bought my last 2 batteries at 16100 retail 3 weeks ago and now they are just over a R1000 more thanks to loadshedding.

Well, there is a bit of difference. Sure they all use Lithium, but the cathode material varies.

Take a look at this link at battery university that gives a good explanation.

Isn’t both LiFePO4? I understand that the cells can be manufactured better or worse, but don’t Hubble and Pylontech just assemble the batteries from cells purchased in bulk (i.e. not manufacture their own cells) and slap their BMS on it?

Hubble are using NMC cells and not LiFePO4 as far as I understand.

No. Pylons are LiFePO4 which they actually manufacture themselves from raw materials. I watched a video once about their manufacturing process. They also design their own BMS. So it seems like they manufacture the battery end to end. Not sure if it’s still true but it was a year ago.

Hubble, from what I’ve read are using BYD LiNMC Nickel Manganese Cobalt cells and they then slap a unknown BMS on.

1 Like

Nope. They do not. While the chemistry is the same, the cells in a pylontech battery are very different. The differences do go deeper than just the BMS.

If that is true… the entire comparison falls flat. You can’t compare those two.

Yep, I have one of their batteries here. They are using a 13S configuration of 117Ah NMC cells for their 48V, 5kWh battery.

1 Like

Well, they do use chips purpose-made for at least the balancer, and the maker of the chip probably has a reference design on which they based theirs… but that is if you want to nitpick. The design is their own, they do the firmware…

Did you guys know you can update Pylontech firmware over the CAN-bus now? I mean the tools aren’t quite there yet, but it’s in the newer protocol versions as well as in the newer firmware versions. BYD’s Premium LV can update the firmware using an app on your phone. I don’t know any others (other than Victron’s SmartLithiums of course) that has that ability.

What I’ve learned so far: … ok ok ok, stompies I picked up along the way …
Use A-Grade cells, cells that are of similar internal resistance installed in the same bank.
Balance them properly and thoroughly upfront.
Don’t run the bank at min/max charge/discharge. Let the cells “idle” between say min of 30% SOC and max of 90% SOC - the users “see” a min of 20% and max of 100% - so that the cells can last their cycles … and less “support”.
Close the box and don’t touch it again … or the warranty is affected.

Ah well yeah in that case, comparing them becomes MUCH more difficult…

Okay okay. I didn’t mean they manufacturer every single component themselves. :joy:It’s definitely not a diy equivalent of getting some cells and the bms and packaging it together and claiming to have developed a battery.

Ive seen posts of people doing it. It was more for compatibility issues though between US3000c not working with B versions. I didn’t have any issue with mine. With these things I go with the “if it ain’t broken don’t fix it”.

1 Like

Tell us more please… My 3000b’s need an update for the GX to get all the info from the BMS. :wink:

Vassen So glad to see you are here … Me Also here …

Hi Ryan

I actually joined after I saw that you got kicked out and joined here.

You should feel lucky though. At least you got a mod to tell you what happened. I was just wiped out. No warning, no notification. Nothing. Just a login failed when I visited the site. :joy:

Hopefully we can have civilized discussions here and continue to learn and share what we learn.

There is a vast knowledge and experience needed for a “Proper” installation. You can install some cables and a box, but you can also burn your house down or heaven forbid cause someone to be shocked with AC/DC. My house is 230v but my Dad is 380V. That cant be good shock.

There is a way to transfer the firmware blob to the battery by splitting it up in chunks and sending it one by one. The battery sends back a frame to ack the one you sent. When you are done, you ask the battery to verify the checksum of the blob it received. Then if all that checks out, you ask the battery to transfer the firmware to the “lower module”. So essentially it’s a two-part setup, with one CPU doing the comms bit and another doing the more sensitive stuff, and you use one to upgrade the other. I don’t know the full details yet. I just know it is possible. I don’t know another battery so far (other than the newer BYD stuff, and of course the SmartLithiums) that have this. OK, the Discover Energy AES battery has a USB cable and software, that’s about as close as it gets.

1 Like

Well, if you just need to simply update the firmware, I’ve seen a thread on the other side under the battery section “pylontech 3000c and 3000” where the guys have successfully updated the firmware using battery view. Firmware is also provided together with battery view. Apparently pylontech has provided these to resolve some issues of using mixed banks.

1 Like

That’s the traditional way. It uses the RS232 port on the device, which in this day and age (can you believe it, this was the defacto standard of the day) actually requires someone to go out and buy extra hardware. But the CAN-bus is there already, so people already have “the hardware”. Hence my excitement about this development.