diy solar

diy solar

Batrium BMS setup 8s3p

My application is a sailboat. For me it's essential to be able to communicate with my Victron equipment on board. If you go the route of multiple parallel packs with each a separate BMS, communication is an issue. I talked to REC BMS on this topic, and they said you need a BMS for each pack + on top of that another BMS that will act as master.

For my application, this is way I eventually choose Batrium with the decentralised approach. This way there is just 1 BMS talking on the Canbus to my chargers.



Steve, you're absolutely right. For my use case (again sailboat) I'm limited in space (existing battery box) and have only 2 real options:
  1. Custom build a battery (24 EVE 280Ah cells in 8s3p + Batrium monitoring). Batteries : 2000USD, BMS: 1500USD -> 3500USD for a total of 840Ah. Because the cells are not balanced, I can only use them between 20-90% or 590Ah usable energy. The total cost is 5,9 USD/Ah ( 3500USD/590Ah)
  2. Buy Battleborn batteries, but due to their dimension I can only fit 8. This gives me 400Ah (they are balanced so let's say 100% usage) for a total cost of 7200USD (900USD/piece). Cost is 18 USD/Ah (7200USD/400Ah).
The cost difference is huge and I can even be more conservative on the range in which I'll use the EVE cells. On top of that I have a battery system that communicates with my chargers (unlike the Battleborn dropins) and doesn't use mosfets to cut power (like all the popular chinese BMS systems), but a decent industrial contactor rated to 20kA. All the additional cell monitoring is just an added bonus.
I don't remember if you said you already bought your battery cells or not or still in the advanced planning stage. sorry been a really long day.
If you have not bought yet. Then you may want to consider spending a bit more for EV Grade cells or even step up to matched & batched from a manufacturer

Some companies like BB, for instance, use cylindrical LFP, they are smaller in AH capacity per cell but they parallel them up. These are more readily available in a higher grade level. the end result does cost more but your ending up with better. Alternately, step up to CALB, now be cautious there... there are different series of CALB, even blue-ees, their ratings & capacity for charge & discharge are different for different purposes, including EV use. Of course prices reflect such. It is a matter of compromises relative to budget capacity and ultimately what you "need". For me, in a boat, at sea, Murphy's Laws kick in, Davy Jones' locker starts sounding, I would NOT want to worry about my batteries ! No Feckin Way Mate ! LOL.

Sorry, but for many years while I was involved with defence, one of my tasks was to consider the situations we could encounter... The prime adage was "Crisis Management begins with Prevention & Planning" the first rules... and I know All about Murphy's Laws and how that blighter comes when "shit is real". as it were.
 
I would expect the cells to drift a little as you said, based on the difference of the IR and the error on the voltage measurements of the Batrium cell monitors. This is typical +/- 0.05V and max +/- 0.1V as per spec sheet. This accuracy seems poor for such an expensive system, although calibration is possible. Then again I'm not planning on running the system on the edge.

I updated the schematic with a pre-charge unit from REC:

View attachment 26055

Any more comments? I will be measuring the amps going through each battery pack and check wether there is an equal split. The cable lengths used to connect the 8S batteries will be equal.
Hi Jan: This conversation's been quiet for a bit but, if you're still around I wonder if you might answer a question for me. I've been working on my electrical design for several months now, and have drawn from this group heavily.... at some point I picked up, (I don't think it was from this conversation but, I can't find the one it was from), the need for a pre-charge unit for the contactor, and added one into my design. I see that you followed something of the same pattern. Now, I'm trying to remember the reason for this so I can explain it to someone who's working with me on tightening up my design. My memory was that it was there to essentially keep the contactor at a constant state of powered so that when the BMS tells it to shut off because the BMS has detected a problem, the contactor is ready to do so immediately where otherwise there might be a delay. Might you verify that that is correct, or augment/correct it from your knowledge, please? My system is similar but, smaller with just an 8S 24v pack and Watchmon 7 BMS.
 
Hi Lance,

please look here:



The pre-charge unit preloads the capacitors in your system avoiding a rush of current when the contactor switches on:

High input capacitance systems such as inverters, dc-dc converters, etc. can be exposed to large inrush currents during the initial power up procedure. If appropriate measures are not employed, these currents can overly stress or even damage the system components. The pre-charge unit eliminates high inrush currents by charging the input capacitor before the main contactor switches on, prolonging lifespan of the contactor and other components dramatically.
 
Hi Lance,

please look here:



The pre-charge unit preloads the capacitors in your system avoiding a rush of current when the contactor switches on:
Yes Jan, thank you..... the topic of REC pre-charge came up on a FB thread, and I ended up exactly where you've just shown me. Seems to me that while my comprehension of it's purpose was a bit off, it is still an important component to include.

Still, in re-reading that again, I remain puzzled and if you have the bandwidth to explain a bit more it will be greatly appreciated. They talk about the contactor switching on, and my understanding of it's purpose was that it was in a normally on position and prepared to turn off, shutting down all out flow from the batteries to the system when the BMS detected a problem with the batteries. Is this not correct?
 
Yes Jan, thank you..... the topic of REC pre-charge came up on a FB thread, and I ended up exactly where you've just shown me. Seems to me that while my comprehension of it's purpose was a bit off, it is still an important component to include.

Still, in re-reading that again, I remain puzzled and if you have the bandwidth to explain a bit more it will be greatly appreciated. They talk about the contactor switching on, and my understanding of it's purpose was that it was in a normally on position and prepared to turn off, shutting down all out flow from the batteries to the system when the BMS detected a problem with the batteries. Is this not correct?

The contactor should be normally-open so when not powered it is open.

When the BMS is first turned on and before it boots and figures out the system state that contactor should still be open.

Once the BMS establishes "all is in limits" it will close the contactor.

This is where the problems can occur.....

If you have a large capacitive load in the system it will draw a lot of power initially, and this will cause the contactor to arc as it's being closed. This arcing causes damage to the contacts in the contactor (think like arc welding). This leads to a sudden failure due to be welded "closed" or premature failure.

What the pre-charge does is bypass some power "around" the contactor, while open, to charge the capacitive loads. It usually uses a resistor to limit the current draw. This pre-charge remains closed for a set period of time; and then the main contactor is closed.

So that little REC box works as follows:

1) Signal into it to close the contactor is received
2) Instead of closing the contactor it closes an internal relay with a resistor in series across the contactors input/output. This charges the load capacitors at a controlled rate.
3) When the pre-charges timer reaches zero it closes the contactor and opens the pre-charge relay.
4) All power is now flowing via contactor.
5) If the delay is long enough capacitors are charged and contactor doesn't arc.

NOTE:

If you have a ESO switch you also need a pre-charge on this as well. Search on here how to do that with a resistor as well.
 
The contactor should be normally-open so when not powered it is open.

When the BMS is first turned on and before it boots and figures out the system state that contactor should still be open.

Once the BMS establishes "all is in limits" it will close the contactor.

This is where the problems can occur.....

If you have a large capacitive load in the system it will draw a lot of power initially, and this will cause the contactor to arc as it's being closed. This arcing causes damage to the contacts in the contactor (think like arc welding). This leads to a sudden failure due to be welded "closed" or premature failure.

What the pre-charge does is bypass some power "around" the contactor, while open, to charge the capacitive loads. It usually uses a resistor to limit the current draw. This pre-charge remains closed for a set period of time; and then the main contactor is closed.

So that little REC box works as follows:

1) Signal into it to close the contactor is received
2) Instead of closing the contactor it closes an internal relay with a resistor in series across the contactors input/output. This charges the load capacitors at a controlled rate.
3) When the pre-charges timer reaches zero it closes the contactor and opens the pre-charge relay.
4) All power is now flowing via contactor.
5) If the delay is long enough capacitors are charged and contactor doesn't arc.

NOTE:

If you have a ESO switch you also need a pre-charge on this as well. Search on here how to do that with a resistor as well.
That is so immensely appreciated Jan. I really had formed an almost opposite understanding of the contactor's role and the pre-charge unit. And, while I have it wired in correctly in my schematics, I was really flying blind and just going with what I'd seen another do. Thank you for taking that time to explain that!
 
Lance,

the purpose of the contactor is to have a switch in your system that can disconnect the battery no matter what. Most contactors require a constant voltage on their coil to remain closed. If you remove the voltage on the coil (by BMS, or manual switch, or ...) the contactor opens and the battery is disconnected.

The pre-charege unit is there for the opposite case. If you apply voltage to the contactor, it will close and connect the battery with all it's capacity to your system of users. In most electronics there are capacitors and they charge up when voltage is applied. The same is when you disconnect a capacitor, it will discharge slowly. This why if you call a helpdesk with an error on your TV, they say to disconnect the power and wait 10s before powering up again. The capacitor in the TV need time to discharge fully and cut all power in the TV. As long as they're discharging they will still power the electronics in your TV.

As to your question, when the BMS gives green light to the contactor to close, power will start flowing and the initial power draw will spike because of the capacitors in your system as they are all charging at the same time. To avoid this power surge, the REC pre-charge will trickle some current through before the contactor closes. This trickling will charge the capacitors before the contactor closes, avoiding the rush of current.
 
As an alternative to a contactor, you can also use an industrial bi-stable relay. They don't require constant current and do not empty the battery over time. But the pre-charge can also be used in this case; it avoids the spike in current when the relay closes.

This is a relay i'm using on a sailboat installation, where you typically don't want the power draw from the contactor:

 
As an alternative to a contactor, you can also use an industrial bi-stable relay. They don't require constant current and do not empty the battery over time. But the pre-charge can also be used in this case; it avoids the spike in current when the relay closes.
I'm really working to keep this as simple as I can, and the pre-charge seems like a very good add so, that is not something I'm trying to avoid. It seems to be the draw from the unit is fairly irrelevant in the scheme of things.
 
As an alternative to a contactor, you can also use an industrial bi-stable relay. They don't require constant current and do not empty the battery over time. But the pre-charge can also be used in this case; it avoids the spike in current when the relay closes.

This is a relay i'm using on a sailboat installation, where you typically don't want the power draw from the contactor:

It seems this remote battery switch requires an "off command", meaning a signal to open the contactor when the battery needs to be dis-connected. This means the BMS must provide the signal. The BMS I am looking at only have an "on" command (power provided when on).
 
Last edited:
The contactor should be normally-open so when not powered it is open.

When the BMS is first turned on and before it boots and figures out the system state that contactor should still be open.

Once the BMS establishes "all is in limits" it will close the contactor.

This is where the problems can occur.....

If you have a large capacitive load in the system it will draw a lot of power initially, and this will cause the contactor to arc as it's being closed. This arcing causes damage to the contacts in the contactor (think like arc welding). This leads to a sudden failure due to be welded "closed" or premature failure.

What the pre-charge does is bypass some power "around" the contactor, while open, to charge the capacitive loads. It usually uses a resistor to limit the current draw. This pre-charge remains closed for a set period of time; and then the main contactor is closed.

So that little REC box works as follows:

1) Signal into it to close the contactor is received
2) Instead of closing the contactor it closes an internal relay with a resistor in series across the contactors input/output. This charges the load capacitors at a controlled rate.
3) When the pre-charges timer reaches zero it closes the contactor and opens the pre-charge relay.
4) All power is now flowing via contactor.
5) If the delay is long enough capacitors are charged and contactor doesn't arc.

NOTE:

If you have a ESO switch you also need a pre-charge on this as well. Search on here how to do that with a resistor as well.
jwelter! I just realized, being brought back to this thread by a recent commenton it, that I'd read your "J" as being jan gils, and not jwelter..... doesn't matter a lot but: thank you for this very detailed and clear explanation! I do though wonder two things: 1. what are examples of a "large capacitive load" in the system? The only thing I'm seeing in mine is the inverter (24v/3000w.) So, assuming this would be the culprit, I'm then trying to see how the pre-charge unit is getting power to those capacitors constantly. Is it through the "system" positive/negative leads that I currently have diagramed to connect to the out side of the gigavac contactor (+) and the out side of the shunt (-)? This seems to make sense but, I thought I'd try to clarify if I still have your attention. The recent comment was regarding the switch Jan Gils was suggesting, and now I'm wondering about it again. Somehow it seems a cleaner, more modern, and less mechanical solution than the contactor.
 
It seems this remote battery switch requires an "off command", meaning a signal to open the contactor when the battery needs to be dis-connected. This means the BMS must provide the signal. The BMS I am looking at only have an "on" command (power provided when on).
I make the following comment from a position of very little knowledge of this stuff so, take that into account but, it seems to me that the BMS is not necessarily sending an "open" command or a "close" command but, instead is just sending a command to the relay/switch to do whatever it is that that relay/switch is designed/programmed to do.
 
jwelter! I just realized, being brought back to this thread by a recent commenton it, that I'd read your "J" as being jan gils, and not jwelter..... doesn't matter a lot but: thank you for this very detailed and clear explanation! I do though wonder two things: 1. what are examples of a "large capacitive load" in the system? The only thing I'm seeing in mine is the inverter (24v/3000w.) So, assuming this would be the culprit, I'm then trying to see how the pre-charge unit is getting power to those capacitors constantly. Is it through the "system" positive/negative leads that I currently have diagramed to connect to the out side of the gigavac contactor (+) and the out side of the shunt (-)? This seems to make sense but, I thought I'd try to clarify if I still have your attention. The recent comment was regarding the switch Jan Gils was suggesting, and now I'm wondering about it again. Somehow it seems a cleaner, more modern, and less mechanical solution than the contactor.

Hi,

Please see diagram below from REC manual.

What happens is when the BMS Input + signal is sensed by the BMS the pre-charge will draw power from the Battery+ contact and through a 66 ohm resistor to limit the current (14V/66 = approx 0.2A) to the System+ contact. This bypasses the contactor as you can see below. This happens for 4 seconds or the set-time if using the adjustable version. During this time the voltage allows any capacitors in the "system" to get charged. The contactor is open during this time.

Then when the timer expires it energized the contactor to allow the path to be closed. But because the capacitors already had this pre-charge it prevents the arc which damages the contactor.

When the BMS Input Signal goes low due to a problem detected by BMS or being shut off the pre-charge simply opens the contactor without delay as expected to disconnect the system from the battery.

Does this clear it up?



page5image4283949376
 
Does this clear it up?
Yes! Thank you... I had the basics of that pretty well understood already but, getting the processes filled in is immensely helpful. This is how I currently have my pre-charge unit diagrammed in too so, I've got that right. Pretty sure I got it from REC's site as well sometime back. The curiosity I have, and it really isn't relevant except to understand, is what capacitors are there in the "system" that want this pre-charge? The only one I can see, at least with any real draw, would be the inverter's capacitors, and those, to my understanding are pretty massive. Being a studio photographer, I have some sense of capacitors in our strobes' power packs and how they load and release energy; I'm assuming the inverter's capacitors work something in the same way except for with a slower, measured release. And also curious about the switch Jan pointed out.... can in be substituted for the contactor, and if so would it be better. My understanding is the pre-charge unit would still be in play.
 
I'm building a new battery bank with 28 EVE 280Ah cells, and I'm wondering wether the following setup is valid. At 0.5C the battery pack can continuously handle about 400A, but in reality I will only use 200A (0.25C). Everything is designed on 400amps though. The system is 24V.

View attachment 25977


The Batrium Watchmon 4 will monitor and actively balance all 28 cells. The BMS will also control both Victron chargers, but as a fail safe it also can close the Gigavac contactor (rated 400A).

Is this setup valid, or am I missing something?
  • Is it safe to use a common contactor for the parallel batteries?
  • Should I add a MEGA fuse to limit the current to 400A (or just use 1C as max current = 840A)? Or add fuses on each of the 3 batteries to avoid high currents between the packs in case a cell short circuits?
  • Do I need protection for inrush currents on the contactor and how to do this?
With this setup do you actually get voltage for each cell? Wont the parralleled cells just show the same values?

With this setup if one cell reaches over or under voltage you will lose all your power. I'm not sure if this is a concern or not but could you not link the packs with a seperate contactor that just pulls it offline allowing others to power the inverter?
 
That battery pack configuration will work well (watch your (+) & (-) wire lengths.
I would put a 200A Fuse on EACH PACK ! (I really came to love MRBF fuses from Bussman).
The Batrium system in this decentralized setup is perfect, you are getting what you paid for !

These 280AH Commodity cells WILL diverge and present increasing differentials as they cycle up & down, these will be most notable at the bottom & top of the SOC chart. Now, if you are being conservative and only using 80% total capacity (10% from bottom & from the top) the pack capacity is 224AH at 80% usage, you would most likely not see huge differentials. I will point out, however, that 1mv per AH is not unusual, so on the 280AH cells you can see 280mv differential between cells. Remember, these are NOT Matched, Batched & Binned according to their Voltage & IR values through cycles... Matching & Batching cells is a long costly process that can add up to $50 per cell cost for doing so.

Bottom & Top Balancing can level the cells and get them to a common full state prior to installation which will make maintaining their state easier for your hardware.

Active Balancing is being used on these larger cells as passive (dissipative) is quite slow and really pointless unless you start off without perfectly matched & batched cells. Active Balancers DO NOT FIX ANYTHING but come in handy when you have slightly mismatched cells that vary and this can compensate / correct for the low cells to hi cells. It is an ADD-ON Device that works in-line with the BMS but outside of the BMS controls. I am unsure how Batrium BMS will cope with an Active Balancer but it should not be a problem. With BMS' that have Passive Balancing, that has to be disabled when using an external Active Balancer otherwise things get terribly confused.

Reference info for you:

BU-803a: Cell Matching and Balancing – Battery University

Pre-Balancing Cells | Orion Li-Ion Battery Management System

Hope it helps, Good Luck
Steve

Purchased the Batrium Watchmon core w/ 16 Blockmons and it arrived last week. My batteries are not here yet, I’m doing a 4p16s so I would only need a 16s active balancing setup? One balancer per pack of 4 cells, being they’re 1120ah packs, is the QNBBM a good fit?
I did look up the QNBBM from Deli-Green today because I’ve seen people say good things about them. Just thinking that maybe I should have them here and on hand “if” needed. I will not put a full load on my system until maybe the first of 2022 as we are still under construction.
Thanks
 
Purchased the Batrium Watchmon core w/ 16 Blockmons and it arrived last week. My batteries are not here yet, I’m doing a 4p16s so I would only need a 16s active balancing setup? One balancer per pack of 4 cells, being they’re 1120ah packs, is the QNBBM a good fit?
I did look up the QNBBM from Deli-Green today because I’ve seen people say good things about them. Just thinking that maybe I should have them here and on hand “if” needed. I will not put a full load on my system until maybe the first of 2022 as we are still under construction.
Thanks

I have a similar Batrium setup but in strings. The Batrium has no issues keeping things balanced with the Blockmon's.

I'd not add an active balancer unless you have cells that require it.
 
I have a similar Batrium setup but in strings. The Batrium has no issues keeping things balanced with the Blockmon's.

I'd not add an active balancer unless you have cells that require it.
It’s good to know and I appreciate the information. Makes me feel better....I was thinking about ordering active balancer and just having them on hand just in case I were to need them at some point.
I’m using the Sol-Ark 12K, what inverter are you using?
 
It’s good to know and I appreciate the information. Makes me feel better....I was thinking about ordering active balancer and just having them on hand just in case I were to need them at some point.
I’m using the Sol-Ark 12K, what inverter are you using?
I run 2 Batriums - so I can see what's happening in my home powerwall packs and in my trailer. Balancing is misunderstood (in my opinion) and leads to long discussions about 'active balancers' and other topics. The key thing is that healthy cells (no self-discharge) that are well matched in capacity (within 2-5%) do not need much balancing. Its amazing - but its absolutely true.

For example, I have 84 packs @ 260ah each made of a variety of 18650 cells. Here's this morning's Batrium pic - .05v (50milli-volt) max difference:
1626965255070.png
The last touch-up balance I did was on March 22, 2021 (4 months ago) using auto-level to bring 20milli-volts back in to 0.04v max difference in the morning. The battery bank has charged/discharged 4,391kwh over those 4 months. It's 'drifted' up from 0.04v to 0.05v over that time. After another few months and it reaches 0.06v - I'll do a touch-up using Batrium auto-level balancing (can run at any voltage).

I could try for 0.02v or 0.01v max difference in the balancing... but it doesn't add anything to my operation and reaches a point of overkill. If they start at 0.04v max difference in the morning they may diverge up to 0.07v max thru the day but right back to 0.04v max difference by next morning. If forced it to 0.01v max difference - it would just diverge up to 0.04v during the day and then back to 0.01v.

1626966163256.png
After selecting "Auto Level" click the More button to the right of the drop-down to get this popup....
1626966204415.png



Here's the trailer Batrium - 0.02v max difference. No balance since last Feb and not likely to need any soon. Its 14s88p of all the same cells. As above - I didn't try to get any closer than 0.02v max difference when I originally balanced.
1626965947642.png
 

Attachments

  • 1626966192931.png
    1626966192931.png
    22.2 KB · Views: 2
My BMS handles up to 120A. I needed 200a so I install a 300a relay. Now my inverter/house is connected directly to the batter (via the 300a relay). The Relay trigger (-) is connected to the BMS. So the battery shutdown is managed by the BMS via the relay.
 
Back
Top