DIY 48V Lipo Build

Hi all. Really hope i can pick the brains and get some Ideas for this.

Im almost halfway, and now my mind is playing tricks on me…

So my history with this build…
I got my hands on quite a few 20A LIPo Cells 3.7v.
Been testing them for the past 2 months, (not fun when you charge and discharge at 1a…) and found about 240 good cells.
Ive order 5A fast blow fused for individual Cell fusing.

My goal is:
to place them in 14S15P. then the max A draw per cell is 2.57A. The cells is rated for 0.2C (4A) (1A Burst)
My goal is to use them with my Victron MP2. 3000va

What is the best practice with 3.7v cells? - Ive read somewhere that its best to keep voltage between 3.9 and 3.5v per cell?
Should above be true, Do you get BMS’s that balance below 4V?
If not, do i use a Active Balancer?
What would be a good BMS to use to communicate with Victron.
Ive read that the BMS must be be able to handle Ripples?

Thanks for all the answers that will follow.

The less strain you put on the cells to longer they will last. So keeping in those limits should give you longer duration. But you can charge those cells to 4.1V.

On the LLT BMS that I own, you can set the voltage at which balancing should start. In fact you can set if you use LiPo or LiFePO4 cells (3.7V vs 3.2V per cell)

You are in luck (and at the correct location :slight_smile: ). The LLT currently has a driver for the Victron system that I created. It does all the comms. I’m working to add the Daly BMS as well. (Check my signature for details and there is a group buy to get some more of these BMS as well)

Thank you @Louisvdw for taking time to answer.
That sound like something i will use. Where can i get those BMS? What is the charge current?

But you can charge those cells to 4.1V. - Thanks. I will then push it up to 4.0v. There is not really alot going on above 4.0v when it comes to capacity. maby 3A.

Is there a place in SA that sell 5A fuses in bulk? I did order, but that will only be arriving end or March.

Sigh. You must have been talking to @TheTerribleTriplet :slight_smile:

The official term is DC ripple. It happens because in order to form a 50Hz output waveform, the current drawn from the battery will itself cycle from zero to some peak 100 times a second. Since V=IR and the battery has an internal resistance, that causes a ripple on the battery voltage.

The BMS cannot do a thing about it. That ripple depends on the internal resistance of the battery, which depends on the size of the cells. If you overdo it, then the ripple will show up, and the BMS can do nothing to prevent it.

What happens however is that sometimes something that the BMS does CAUSES a DC ripple. If the BMS decides to disconnect the battery (to protect the cells), this will very often result in a DC ripple. That is because with the battery disconnected, the DC-bus isn’t completely without storage. It still has a number of really large capacitors that act like a kind of battery, but of course the capacity is dramatically less than a lithium battery, and as a result the ripple created is significant.

So what is needed from the BMS is one that doesn’t freak out when nothing is wrong… because if it freaks out prematurely, it causes the SYMPTOM of a DC ripple. Which is actually not even the main problem. The main problem is that you cannot power your loads properly :slight_smile:

See the Group Buy BMS Group Buy - #6 by neliuszeeman

You get different models. I currently have a 100A model. The group by is for a 200A model. But there are 20A, 40A, 60A, etc. These BMS are very customisable. You can limit the current in the settings to any value (less than the model limit). So my 100A model is currently set to only change up to 50A and discharge up to 60A even though it can handle 100A.

1 Like

@plonkster , Thank you for the explanation. Has not been talking to @TheTerribleTriplet but did read his post. Thanks. I believe that this BMS will not freak out, and it enjoys the good life!!! :slight_smile:

@Louisvdw . Thanks. Will look onto the group buy. I need one, and the 100 or 200 sounds like it would work.

Thanks all for taking time to answers my questions.

1 Like

Plonk, you seem to forget that I sometimes do not colour within the lines and I ALWAYS get the other side of the coin presented.

Just like with a DC Ripple, your words, my experiences, what DC Ripple?
I got them in droves, like in plural, as in one after the other aka DC RippleS.

Once they start they will happen over and over until:

  1. I intervened, and fast at that, by dropping the SOC with substantial loads or
  2. switching all off,

I will experience one DC Ripple warning after the other until they got so bad the inverter switched off, resetting after 30sec, then another DC Ripple, switches off … resets … over and over … which was a cluster f…k not forgetting the stomach-dropping “helse groot skrik” each time that happened, one moment all is fine, the next moment everything is off …

Had to install, in haste, a changeover switch so that I could mitigate the damn RippleS, protecting the rest of the house all because of a damn cell and a BMS that could not get them balanced, not even after weeks.

Nothing is ever simple when things go off the beaten track with me … ek is al so damn gatvol dat NIKS eenvoudig is waar ek aan iets raak. :woman_facepalming:


Nothing is ever simple when things go off the beaten track with me … ek is al so damn gatvol dat NIKS eenvoudig is waar ek aan iets raak. :woman_facepalming:

Klink vir my soos n elektroniese ongeluks voël.

I know that feeling.
Hence nowadays, i do what i can. If after 3 tries or smoke or Sparks / blinding Sparks, I stop,
Google the daylights out of it. try again X times. X = amount of swearwords where is in the dictionary.
no smoke, great, works, celebrate!
not working, Google again, Ask a friend, Sleepless nights.
try again, Still not working.
Then do the, like my wifes calles it “wees a brommer vir n dag” and leave it in the “eendag as ek tyd het folder”

Next project!

Gets MUCH worse … :laughing:

Jis, that is funny!!!
Love it!!
My grandpa use to say, Mens is net een keer dom… dan leer hy!!!
And then the famous words of my teacher in school. If all fails everytime… Read the instructions!!!

Jy het al BAIE skoolgeld betaal!!!

1 Like

I think the simple truth that I try to get across is that DC ripple is often a symptom of something else that has already gone wrong. The ripple isn’t causing the problem. There is another problem that is causing the ripple.

Another way to think of it is this: A DC ripple alarm is essentially the same as a low battery alarm. Think about it… it means the DC voltage can be swung up and down too easily because there isn’t enough battery on tthe line to hold it steady.


1 Like

Slim Oupa.

My Rule: So you make a mistake …
1st time … we all learn.
2nd time making the same mistake, … some of us are just simply slow learners see, not all equal see.
3rd time the same mistake … my F_K “Marelize”… jy is dommer as ground ou maat.

Sometimes I feel no 3 coming on … :woman_facepalming:

YES, I have understood it as such since day one: The result, not the cause … unless the inverter was faulty.

AND Victron is experienced and well aware of such AND has safety margins built in, documentation on possible cause/s and potential fixes, to protect the inverters best they can.

They can now also add “Check individual cells in lithium bank for overvoltage/imbalances causing BMS to protect the bank resulting in a DC Ripple when BMS disconnects to protect cells from being damaged/set alight. Tends to happen at >95% SOC and <350w loads.” to the manual.

… for if cell imbalances are not fixed you can get multiple DC RippleS until the cells behave and the BMS does not disconnect anymore, BEFORE the inverter or 250/100 MPPT decides “ENOUGH, no-one loves me, I’m outta here!” … and leaves the premises. :wink:

Jiss, lets hope i wound get all of those DC Rippels

If a read correctly, @TheTerribleTriplet experiance it if cells are >95%, so im thinking of running my cells @3.9V, way from 4.2V.
That should, SHOULD, keep all happy.

I Just have to get the testing done. Just found out this over the weekend:

Almost halfway, and i found out, after just checking randomly, that some cells discharge at different rates. Same Cells, but for some , the discharge curves is different. All the Cells are 3.7V 0.2C 20A. Some are in blue wrapping, and other not.
But the Blue Cells have 2 different discharge curves…

  1. 4.0v and 3.65
  2. 3.9v - 3.5v

Raait kerels, laat ons weer begin…

My 2nd Life bank, I’ve done this to them: LiFePO4 Best Practices - #25 by TheTerribleTriplet

And “The Bulb V2”

The bulbs are 12v and I THINK brake lights, as both have 2 filaments and all 4 turns red.
Use it to drop 'n cell faster …

To recap, my cells, if one cells hits, on the old BMS, 3.6v, BMS disconnected.
New BMS sees 3.6v, beeps like mad, and on 3.65v will disconnect.
Have not seen that in weeks on the main 1st Life bank after I helped it with “The Bulb V1” (one bulb) and Louis software the taking over from there.

Having just checked on the JBD BMS I see Alibaba has a minimum order of 2 units.

When do you expect to go ahead with this group buy? It might be cheaper for me to pay for the postage from SA to Aus instead of buying two. Although I would be left with a spare. Anyway let me know when you want to make the purchase.

@neliuszeeman is interested and has a few things he wants to check to make the buy easier.

Ask for a sample unit, a couple of dollars more.

Me, I would get 2, I bought one, was so impressed, that I got a 2nd one … should have bought 2 from the start. :slight_smile:

Two sounds like the way to go then.

Reason I needed two, one was for a 48v system, then a 2nd system came about, being 24v.
These BMS’s can do 24-72v … should have bought two and not 2 x Samples.

When you order, get it with UART/Bluetooth and RS485 port added. That costed $5 per BMS.

And get the JBD UART Communication Box Intelligent Lithium Battery Protection Board USB 2.0 Interface,UART Tool too, for $5 extra.

Looks like this:

Or without the box: