diy solar

diy solar

Why not to use Daly BMS with MPPT controllers


So there is a major problem with these type of BMS.

Any thoughts?

My take:

Overcurrent triggered short circuit protection.

The pass through of 90V from the panels/failed MPPT blew the fets shorting the BMS.

Unregulated charge from the panels allowed the battery to go over-voltage and poof the cells.

My thoughts?

What MPPT?

What's a reasonable expectation for a BMS's rated peak voltage? 250V? What about high voltage string inverters that may feed 450-600V? Should the BMS protect against that?

He mentions an Electrodacus, but would it protect against that too? Sure... because you're forced to use low voltage on the most primitive solar charge technology around. No more MPPT and you're roped into 60-72 cell panels in parallel only.
 
Would be nice to know what MPPT controller did that.

He mentions an Electrodacus, but would it protect against that too? Sure... because you're forced to use low voltage on the most primitive solar charge technology around. No more MPPT and you're roped into 60-72 cell panels in parallel only.
I've read about those, but at first glance it seemed like a really shitty solution compared to just using a decent Victron or similar.
 
Yes, it would be nice to know the type of MPPT he used.

In addition, these DALY types of BMSs should have a fail safe system that prevents these runaway failures or a design that inherently cannot have this type of failure.

Is there anyone who can be contacted regarding to bugs of the DALY bms's? e.g. I would like to set the SOC after I fumbled with the system. I do know that there is an option in the android app, but it is not working.
 
Sounds like a higher quality SCC is in order? Too bad the BMS didn't do its job though, I guess thats exactly what its made for.
 
He mentions an Electrodacus, but would it protect against that too? Sure... because you're forced to use low voltage on the most primitive solar charge technology around. No more MPPT and you're roped into 60-72 cell panels in parallel only.
Electrodacus does not have to be used with their PWM controllers. I will be using it with a Victron MPPT which the Electrodacus can cntrol.
 
Electrodacus does not have to be used with their PWM controllers. I will be using it with a Victron MPPT which the Electrodacus can cntrol.

Electrodacus doesn't have PWM controllers. They simply short the panels to battery until absorption is reached, and then they open circuit or engage diversion load.

Pretty much the most primitive solar charging tech there is.

Use it to control a Victron, and you need to concern yourself with the exact same thing (though MPPT failure chances are diminished). Should the MPPT fail in the same way, the Electrodacus will be subjected to voltage much higher than 24V. Are you confident that it can take it? I'm not, and since it can't even open a circuit like a FET BMS, it will allow the array to overcharge the cells in exactly the same way.
 
Use it to control a Victron, and you need to concern yourself with the exact same thing (though MPPT failure chances are diminished). Should the MPPT fail in the same way, the Electrodacus will be subjected to voltage much higher than 24V. Are you confident that it can take it? I'm not, and since it can't even open a circuit like a FET BMS, it will allow the array to overcharge the cells in exactly the same way.
Electrodacus SBMS0 cannot be exposed to a dangerous voltage. Electrodacus has no charge or load voltage travelling through it.
 
Electrodacus SBMS0 cannot be exposed to a dangerous voltage. Electrodacus has no charge or load voltage travelling through it.

1625640592689.png

I get the impression you don't understand the issue or the Electrodacus.

The DALY failed because it went open circuit and stopped the flow of current. Once the flow of current stopped, the BMS saw full panel voltage, and it shorted the BMS allowing current to pass.

As you can see the Electrodacus connects to all cells. If there is a short in the MPPT, the two circled terminals will experience a momentary spike as full PV voltage is applied. It will rapidly drop to battery voltage as it did for the DALY. The question becomes, can the Electrodacus survive momentary exposure to >120VDC when it's designed for 30V-ish?

Additionally, the Electrodacus is powerless to stop the shorted array from over-charging the batteries.

An Electrodacus controlling a MPPT controller offers no better battery protection than the DALY in this failure mode. The only question is will the Electrodacus survive the same failure mode.
 
I do not expect that any FET based switch (or BMS) is able to handle this. Most 24V FET switches use FETS rated in the order of 40-50VDC. Even a 48V switch, which probably uses 70-80V FETS, would maybe not survive this.
A good MPPT should not put 90VDC on the output when it is disconnected but if it fails for some reason, the FET switch will also be destroyed. A BMS with mechanical relays connected to it may have worked. However even many mechanical relays which are rated 12/24VDC cannot open safely with 90VDC if the gap is too small between the contacts. It depends on the current which is flowing and if the current keeps flowing in that case. For example, 1A 90VDC may work but 30A 90VDC not.
 
I set my Morningstar controllers to have a max output regardless of setting of 3.60 x 16 or 57.6 while not perfect I think the way the architecture of this controller is designed it shuts off any thing above this. mind you this in not bulk, absorb or whatnot.. it seems to be an upper absolute. now if that works after being subjected to damage is another question.
 
I'm a retired EE/CS with my own custom 120VDC, off grid power system; I'm new here, and I'm considering LiFePo4 and a new custom 39 cell BMS/balancer for my 120VDC battery bank.

I think this is a most valuable failure report and and failure analysis video posted by BlueFlower. Bravo. I appreciate it and have learned from it!

The switching MOSFET in any PWM or MPPT will typically fail as a dead short. Most are a buck converter design topology (step down only) since that is the simplest and highest efficiency. When the switching MOSFET in a buck converter fails, it causes full on, full PV voltage sent to the BMS. I'd bet few commercial BMS systems are designed to cope with this typical failure of the PV charge regulator. They should, and it's important, which is why Blueflower's report is important. He's illustrating a serious flaw in the illusion of BMS protection. If a BMS with high voltage protection isn't available, then you'd have to add a voltage limit protection on the PV regulator output; and this is not a standard product. The external voltage limiter must disconnect or short the PV regulator output before voltage rises above the capability of the BMS, fast enough to protect the BMS. Overvoltage will typically fail-short the MOSFETs in the BMS disconnect switch, so now you have directly connected your PV arrays to your batteries.

Something to consider seriously for systems with typical high voltage PV down regulated by a bucking (topology) PV regulator. A PV regulator fail can destroy the BMS and all of your batteries. Software assigned voltage limits on the PV regulator offer no protection at all for this failure. A horrible, costly failure. I'd sure not want to be using a cheap, bucking PV regulator over half it's rated capacity. You shouldn't risk a major (to you) battery investment without addressing this PV regulator failure mode.

My first thought for a solution is a fuse and overvoltage detecting zener diode SCR crowbar circuit after the PV regulator to protect a BMS. The SCR would clamp down the overvoltage while waiting for the fuse to blow. The crowbar has no affect on the circuit in normal operation and won't fail as a full on as a solid state (MOSFET) switch/relay will. This is not a very difficult circuit, and if someone has an urgent need for it, I'll try to work it up. There may be some industrial DC voltage limiters/crowbar protection circuits available

Thanks again to BlueFlower.
BruceM
 
You can find many, many examples via google. An SCR based circuit would be used for a DC crowbar. The SCR is very old type of power transistor, used in the days when bipolar was all we had. It's a gate controlled diode. Once triggered with current at the gate, it stays turned on until current through it goes to zero. There are some devices prebuilt which may be suitable. You have to get the specifics for your application, what DC voltage trip point, how much current to blow the fuse.
 
There are other options to implement a crowbar (fuse blowing) type over voltage protection.

In practice you'd set the PV regulator current limit below the capability of the PV arrays and the fuse value just above that, so that the crowbar device(s) can then blow the fuse if the PV regulator fails to limit the output voltage. The crowbar device(s) should trigger at a voltage well below the limit of the BMS input voltage, or somewhat above your normal full current charging voltage.
 
I use two of these (I'm monitoring loads and charging separately) for a completely independent system as a fail-safe in case BMS or SCC or whatever fails to do it's job for whatever reason to protect the battery bank, especially when I'm dinking around with different options: http://bed-electronics.com/product/sbs6-64al/... going to verify what would happen if it was exposed to 90V... specs state for their overvoltage protection
"64V TVS Diode, safe ONLY IF circuit breaker is used"
 
I'd look up the capacity of the TVSS diode. Many commercial products are a farce, horribly designed. EE's are just like any other professionals; 10 percent are knowledgeable in their field and reliably competent. I like TVSS devices, and use them often for circuit protection where fast reaction is needed.

Your overvoltage protection device(s) MUST be able to blow the fuse or breaker, and that can take time. More time that most people realize.
So it takes some detailed study time to insure that you get it right, and when the time comes, your BMS and battery bank will really be saved.
Puting in a single TVSS diode in a 200A SS relay, and thinking that it will trip a high amp current breaker may not work. Breakers are very slow to trip unless the fault current is many times the rated. If the TVSS fails before the breaker trips, oops.
 
Do you really think it is very time critical?
To blow a fuse within 1 second is ok I think. If it blows at all.
Otherwise a crowbar circuit that triggers an explosive fuse will do the trick within a few microseconds....

But, if the current is not enough to blow the fuse in the first place, then that fuse was wrongly designed in the battery system. Perhaps an explosive fuse is not really an bad idea. Does anyone knows if they are being sold somewhere?
 
Last edited:
Back
Top