A Tale of Two (Relay) BMS’s

[Vincegod]

I joined this forum a year ago and shortly afterwards purchased from a fellow forum member the following items:

• 16 x 280ah Docan batteries
• Victron 48v inverter
• JBD 100a BMS

During the summer last year, I had been charging the batteries via 3x 655W solar panels on my garage. All the solar power collected was either used directly or dumped into the batteries for use later in the day. This has worked out well but as winter arrived and short sunless days the setup became unused, and savings became nil.

Around that time British Gas had been hounding me to have a Smart Meter installed, so I signed up for the smart meter thinking I can subscribe to a cheap rate night tariff. 14 days later I had the smart meter installed and I immediately “switched” to Octopus who have a much better choice of tariffs (another 7 days). I then had Octopus move me to the Economy 7 tariff which was another few weeks. Final after nearly two months with little sun or battery power I was able to charge the batteries overnight on cheap rate electricity from the grid. However, this did not cover all my usage and some days I had to utilise the more expensive day time rate but still I halved my electricity bill.

Previously I had directed a couple of friends to Docan as the batteries they supplied are quality and they also have a good name here. So, I placed an order with Docan for the same items one friend had recently purchased from them: 16x CATL Lifepo4, busbars & nuts, 2x JBD 200A Smart BMS with BT & RS485 (one for existing batteries), trade assurance, so what can go wrong!

Delivery was earlier this month and a few weeks before I expected which was good. On opening the wonderfully packaged batteries I could see they all were flat and good quality but what where these BMS’s they have a Relay, really?

I’ve not seen BMS’s with relays for some time and expected a MOSFET BMS as all others are and the same as what my friend had received. I then queried Docan why I had received these who said I didn’t ask, and these are cheaper! Did I really need to ask, who would think they are providing old technology and cheaper what $2 cheaper, really? This went back and forth getting nowhere and clearly Docan would not have them back or refund me. Best they could do was give me $32 each off if I ordered again, so would cost me another $260 to get what I had original thought I would get. I mentioned about going through the trade assurance route and Docan said they would do nothing. I filed a dispute and trade assurance gave me back $150 so half of the originally estimated cost.

As I only had these bricks of a BMS, I built the new pack on the bench with one. The BMS seemed OK and I was able to bring this new battery pack online in parallel with the first pack which has a JBD Smart BMS 100A (MOSFET). I ran both for a day or two and decided to rebuild the first pack with the 2nd brick of a BMS but it was very tight in the Powerwall cabinet. On bringing both packs back online with the Relay brick of BMS’s it was clear they don’t like each other as I could only get one working at a time. I did manage to get both online at the same time by powering everything off but after a charge only one would start discharging. At this point I immediately ordered a JK BMS and the Relay BMS’s are going.

What have I learned:

• Make sure you get photos, model numbers and full details from a Chinese supplier as they could give you whatever rubbish they have in stock.
• Don’t trust Chinese suppliers to give a dam once they have your money and I was glad I paid for trade assurance.
• Relay BMS’s may be great if you want to power your golf cart but for a Powerwall they clunk, are clunky and only any good if you have one pack.
• I was glad I made it easy to replace BMS’s without having to replace all the balancing wires.

Vince

 

[Vincegod]

What I expected and what my friend received a few weeks before I ordered




This is the brick of a BMS a received (actually two of them)

 

[SparkyJJO]

Contactors aren't necessarily "old tech" - they are necessary for higher current than FETs can do. That said, they do seem to need a minimum current to work right which is probably the hangup you're having.

I got two of them (300A ones, overkill really) but I'm suspecting I might have problems like that too - I did not know of the issue until after I got them. Guess we'll see.

 

[Ampster]

I actually have more confidence in a contactor based BMS, especially at high Amperage.

 

[Vincegod]

Balance lead harness making it easy to replace BMS




The Brick which has now been replaced by my original JBD Smart BMS 100A (with MOSFET’s)

 

[Vincegod]

Contactors aren't necessarily "old tech" - they are necessary for higher current than FETs can do. That said, they do seem to need a minimum current to work right which is probably the hangup you're having.

I got two of them (300A ones, overkill really) but I'm suspecting I might have problems like that too - I did not know of the issue until after I got them. Guess we'll see.

There is a 10W resistor in parallel to the contacts which it uses to detect current flow before switching on the contactor (depending on charge/discharge switch). In a parallel battery setup current flows from the other packs first and it takes a significant draw before it switches on the contactor. I believe we a situation where both packs present different resistances depending on the state of the contactor which is causing the issues.

 

[Quattrohead]

That resistor is the "soft start" for when you connect an inverter for the first time.
I like the JBD contactor BMS, I have 2 different ones and they both work well.
However they rarely run together, one seems more lazy than the other at charging or supplying load even though all settings are the same.
But ultimately, between these 2 and 4 Felicity batteries (that also have good BMS) they figure it out eventually.

 

[Vincegod]

That resistor is the "soft start" for when you connect an inverter for the first time.
I like the JBD contactor BMS, I have 2 different ones and they both work well.
However they rarely run together, one seems more lazy than the other at charging or supplying load even though all settings are the same.
But ultimately, between these 2 and 4 Felicity batteries (that also have good BMS) they figure it out eventually.

The resistor has a side benefit of limiting the inrush current but is primarily used to detect the current flow and direction for switching on the contactor. This is great for single pack operation and very large current flows but does not work well in parallel pack and low current applications.
I see Andy did a video of the Qucc relay BMS and could not understand why the relay was chattering (rapid on/off) when it was charging. Take the situation when your solar is just producing the same energy as your load and you have set the BMS to charge only or are at minimum SOC. As the solar produces more energy than the load the relay/contactor will turn on and when the energy drops the relay will turn off. this can happen many times per minute, how long is that relay/contactor going to last?

 

[Quattrohead]

The resistor has a side benefit of limiting the inrush current but is primarily used to detect the current flow and direction for switching on the contactor.

Nope, that is what the big fat gold bar thingy is for that is connected between the contactor and output cable. It has a notch lazer cut in to calibrate it.

Qucc relay BMS and could not understand why the relay was chattering (rapid on/off) when it was charging

That is a wrong setting in the FET output, it needs to be set to 10s or more.

 

[Vincegod]

The shunt is used to measure accurately the current, however the 10W 10ohm resistor allows current to flow when the relay it disengaged.

If you examine the photo attached you can see that the negative battery goes through the shunt and then through the relay contacts. There is a black wire on the input side of the relays contact which comes from the 10 resistor with the other side of the resistor connected to the output side of the relays contacts.

This arrangement increases the batteries resistance by 10ohms when the BMS is off which is why when in parallel it will not turn on until there is significant current draw. This is not a good design for parallel batteries in a ESS setup.

 

[Vincegod]

I actually have more confidence in a contactor based BMS, especially at high Amperage.

Solid state switching should be more reliable as there is no possibility of contact arcing, much quieter, faster and the MTBF (mean time between failures) is vastly greater. Unless you are switching 400 Amps then MOSFET would be a safe bet.

 

[SparkyJJO]

FETs Can fail shorted though. Really not that much different than a welded contact.

 

[Ampster]

Unless you are switching 400 Amps then MOSFET would be a safe bet.

It is just the opposite. All EVs use contactors because of the risk of MOSFETs failing closed in high Amperage situations as noted by @SparkyJJO

 

[BentleyJ]

I swapped out the cheap golf cart contactor with a DuraCool magnetic latching contactor. The pulse control board does use some power all the time but its 7mA vs. over 100mA. There are also conventional contactors with economizers that have fairly low holding currents but it's surprising how fast a small parasitic load adds up over 24hrs, weekly and monthly.

 

[PyroRider]

Solid state switching should be more reliable as there is no possibility of contact arcing, much quieter, faster and the MTBF (mean time between failures) is vastly greater. Unless you are switching 400 Amps then MOSFET would be a safe bet.

Andy from OffGridGarage had a 200A JK BMS running at a bit over 200A and then open under that load and it immediatly blew at least one of the mosfets.
For my battery pack, I will go with the JK BMS Mosfets AND a high current relay