Why not to use Daly BMS with MPPT controllers

[Davismltc]

if you reconnect the PV in bright sunlight, you will fry the BMS

I'm new to solar stuff. Is this common that the BMS would fry upon reconnecting the PV in bright sunlight, or is it the SCC that would fry? I'm assuming the issue is overvoltage?

Edit: Never mind... I see in Miles' scenario that the SCC's FETs are already shorted closed, which would then connect the PV panel high voltage back to the BMS in bright sunlight. (e.g. Not the accidental trip that codfish referred to)

I think an Arduino could detect an SCC failure, disconnect a HV panel relay/contactor and trigger an alarm, instead of reconnecting the PV panel and frying the BMS. Adding software control to the failsafe allows a variety of possibilities.

Also, in searching about the SCC frying (which also appears to be the original problem of this thread), I found this:

Trying to destory my SCC's: I can't do it!

So I took every EPEver/Victron/PWM/Renogy solar charge controller, and tried to fry it by connecting it only to solar panels, and not to a battery... guess which one survived? EVERY SINGLE ONE! I couldn't destroy any of them, and I left them connected to a high voltage array (80 volts) for 4+...

diysolarforum.com

Anything can fail... Are we flogging a dead horse here?

 

[Miles B]

No, the horse is definitely not dead, as the original video demonstrates. The point is to make your system tolerant of any single fault. Currently, some systems are not tolerant of a SCC failure, and could cause fire in this case.

 

[Davismltc]

A more simple approach...no Arduino...no crowbar:

Connect a normally open continuous duty relay in series between one leg of the PV and the SCC .
Connect one leg of the relay coil to the + battery terminal.
Connect the other leg of the relay coil to the BMS P- terminal.

As long as the BMS has the P- terminal connected to the battery negative, the PV panel is connected to the SCC.
When the BMS disconnects from the battery for whatever reason, the PV panel is disconnected from the SCC.
The relay coil only draws power from the battery when the BMS and PV panels are connected.

If the PV panel doesn't have an isolated ground, an alarm/alert could be connected to the normally closed terminal of the relay.
(In this case, the negative terminal of the PV panel is connected to the relay's N.O. terminal and the SCC negative to the PV is connected to the relay's common terminal) The alarm/alert activates whenever the BMS disconnects.

Carry a spare relay...in the event of continuous duty relay failure ;-)

 

[Just John]

No, the horse is definitely not dead, as the original video demonstrates. The point is to make your system tolerant of any single fault. Currently, some systems are not tolerant of a SCC failure, and could cause fire in this case.

I'd say 99% of systems wouldn't survive the problem if using a FET based BMS.
Contactor based BMS likely would.
If you monitored your pack voltage, you have plenty of time to figure out something is wrong.

Most of them also wouldn't survive a hurricane, or a nuclear reactor meltdown either.

 

[mrzed001]

I'd say 99% of systems wouldn't survive the problem if using a FET based BMS.
Contactor based BMS likely would.
If you monitored your pack voltage, you have plenty of time to figure out something is wrong.

Most of them also wouldn't survive a hurricane, or a nuclear reactor meltdown either.

And a lot of us use FET based BMS like Daly, Heltec, ...

Connect a normally open continuous duty relay in series between one leg of the PV and the SCC .
Connect one leg of the relay coil to the + battery terminal.
Connect the other leg of the relay coil to the BMS P- terminal.

So copy the BMS relay state to the PV side.
Not a bad idea but will not work with a FET based BMS.
That can allow discharge while deny charge.

As long as the BMS has the P- terminal connected to the battery negative, the PV panel is connected to the SCC.
When the BMS disconnects from the battery for whatever reason, the PV panel is disconnected from the SCC.
The relay coil only draws power from the battery when the BMS and PV panels are connected.

In this last part I am not sure. It does not care if PV is connected or not. If BMS relay open then PV relay open. Or do I miss something ?

More of a visual type ... post a connection schema pic if I am wrong and not seeing something please

 

[Davismltc]

My take....
SCC fails shorted (this initial failure may have been avoided with the N.O. relay in place, if the cause was a battery disconnect)
SCC is shorted and battery continues to charge until BMS disconnects.
When BMS disconnects, relay opens, causing the PV panel to disconnect from the SCC (the relay disconnect may take 10ms or so to fully disconnect the PV panel from the SCC).
Whether or not the BMS fails shorted depends on the ability of its FETs to withstand the PV voltage during the 10ms or so relay disconnect.

IF the relay fully disconnects before the BMS FETs fail shorted, the PV panel disconnects and the battery and BMS are saved.
IF NOT, the FETs short, the PV panel reconnects, the BMS is toast, and the battery is overcharged.

Not a 100% solution, but it's simple, inexpensive, and adds an extra layer of protection while waiting for the "better" BMS to come to market.
Does add a small parasitic draw to the battery to keep the continuous duty relay energized 24-7

Other issues?

I'm new to solar components...not sure what you mean mrzed001:
"Not a bad idea but will not work with a FET based BMS.
That can allow discharge while deny charge."
-and-
"...It does not care if PV is connected or not."

 

[mrzed001]

I'm new to solar components...not sure what you mean mrzed001:
"Not a bad idea but will not work with a FET based BMS.
That can allow discharge while deny charge."

FET based BMS can allow charge while deny discharge and replaced.
So battery full or high V event, BMS disables charge ... but still enables discharge so it causes the relay to remain ... so PV still connected to SCC.
The PV relay state copies the BMS "relay" ... or in a FAT based BMS the "discharge relay" state.

 

[Davismltc]

Okay, I get it. Thanks mrzed001. I thought it disconnected altogether, but I see that's not the case. Still learning...LoL

 

[curiouscarbon]

Okay, I get it. Thanks mrzed001. I thought it disconnected altogether, but I see that's not the case. Still learning...LoL

I too was confused by the way FET BMS behave.

FET BMS can disconnect both one way lanes and be (effectively) 100% disconnected for charge and discharge.

But also can allow only charge.

But also can allow only discharge.

Normally a wire is connected or not, how could it be physically connected and allow charge but not discharge?

Kind of like replacing a dirt road with two one way lanes. My brain still has a hard time accepting understanding of that cool feature.

hope this helps, i’m still learning too and been at it for over a year haha

 

[mrzed001]

Okay, I get it. Thanks mrzed001. I thought it disconnected altogether, but I see that's not the case. Still learning...LoL

It is an interesting idea
The PV side relay (like 500Vdc 20A) is much cheaper than the battery side relay (500Vdc 200A ... because it would need to break even if all 500V PV is going through).
But still needs something to control it ...

 

[Hedges]

I too was confused by the way FET BMS behave.

FET BMS can disconnect both one way lanes and be (effectively) 100% disconnected for charge and discharge.

But also can allow only charge.

But also can allow only discharge.

Normally a wire is connected or not, how could it be physically connected and allow charge but not discharge?

There is something called an "Ideal Diode".
Normal diodes have a voltage drop 0.4V to 1.5V when conducting.
A FET with gate voltage controlled depending on direction of current flow can conduct with about 1 milliohm resistance (no fixed voltage drop like a diode) and block current flow in opposite direction.

So BMS FET can be controlled like that, to function as a diode.
I suppose a relay could be controlled the same.

Either way (or even with a real diode), there is some "shoot-through" current in wrong direction before it turns off.

 

[Davismltc]

If at first you don't succeed... LoL

Okay, back to codfish's idea...with a little modification:

Normally closed (continuous duty) relay in series connection between SCC and the PV panel.
Trigger the normally closed relay from a voltage comparator (e.g. via a transistor with enough current capacity to operate the relay).
One input is + battery to a precision voltage reference to comparator
Other input is + battery to a resistor voltage divider network to comparator (divider output voltage is set for max safe battery voltage)
Ground is negative battery
When battery voltage via divider network exceeds max setting, comparator triggers, energizes the relay, and opens the PV panel to SCC circuit.
Divider network set to trigger above normal charge voltage (e.g. trigger at 3.6v per cell)
When voltage drops below threshold, PV panel reconnects to SCC

Modest cost, saves the battery, not the BMS.
Protects from overvoltage BMS error
Low current draw
Relay could trigger alert/alarm

 

[Davismltc]

There is something called an "Ideal Diode".
Normal diodes have a voltage drop 0.4V to 1.5V when conducting.
A FET with gate voltage controlled depending on direction of current flow can conduct with about 1 milliohm resistance (no fixed voltage drop like a diode) and block current flow in opposite direction.

So BMS FET can be controlled like that, to function as a diode.
I suppose a relay could be controlled the same.

Either way (or even with a real diode), there is some "shoot-through" current in wrong direction before it turns off.

How would the "Ideal Diode" be utilized in, or as a "fail safe" circuit? Is it to replace the relay, or ?
Could you elaborate a little further?

 

[Hedges]

I was referring to how a FET BMS can disconnect charging, but connect discharging. Rather than just opening at over-voltage (or under temperature), it can open when it sees current flowing in the "charge" direction.

If the fail-safe circuit works by crowbarring PV, backfeed from battery through failed SCC could be an issue. I that case a diode between PV (including crowbar) and SCC could be used be prevent backfeed. A diode always loses power due to its voltage drop. An ideal diode there should eliminate the drop.

It seems the protection schemes for over-voltage of BMS (or battery, in my AGM application) would be either a series switch that opens PV (redundant to failed switching transistor in SCC) or one that crowbars PV (and may experience backfeed of battery through SCC.)

Here's a possible solution from Midnight.

I think this is a 100A, 150V (per pole?) polarized breaker, coupled to two breakers which can be activated by a remote signal. One is "12V shunt trip", other detects ground-fault current through 1A breaker. If either activated, it would disconnect PV array.

But to protect BMS FETs, would have to be fast enough that voltage doesn't rise too high - maybe determine trip speed and add sufficient capacitance? For my AGM over-charge application, should work as-is.

Product - MidNite Solar

MidNite Solar Product Page.

www.midnitesolar.com

http://www.midnitesolar.com/pdfs/MNDC-GFP100-Dual_manual_.pdf

 

[Davismltc]

Thank you Hedges!

One problem I see with my last variation on codfish's idea is in the case of the shorted SCC, the battery will be charged at full PV panel voltage.

To fix that I'll propose a mod to the circuit design. I hope you guys don't mind a newbie's ramblings....

A normally open continuous duty relay is placed across the PV panel terminals, so when energized, it shorts the output of the PV panel.
A 15A fuse is placed on one leg of the input to the SCC, below the relay (to handle normal PV current)
The relay is triggered by the output of a comparator anytime battery voltage exceeds 3.6v per cell

When battery voltage exceeds 3.6v per cell, relay energizes and shorts out PV panel output.
If the SCC has shorted as well, the battery current will blow the 15A fuse and disable all further PV output to the battery.

Being a newbie, I'm not familiar with the shorted SCC characteristics. Would battery current blow the fuse on the PV panel side of the SCC if its +/- input was shorted by the relay?

 

[Hedges]

Battery side of SCC would have a capacitor bank. Shorting that would blow at least a fast-blow fuse.

Shorted SCC is our speculation, but I think there is an FET connecting input PV (and capacitor) through an inductor to battery.
Shorting PV input should draw battery current (if battery still connected) through inductor (which should have linear current ramp with constant voltage across it), through the failed FET.

 

[Davismltc]

Sounds like the 15A input fuse would blow "IF" the battery was still connected when the shorting relay energized.

I'm a newbie, so please bear with me...
If the BMS was doing its job, the battery shouldn't be at 3.6v per cell. (or whatever trip voltage selected above normal charge voltage termination)
In order to get to 3.6v per cell, wouldn't the battery have to be connected to the PV?

Same advantages as before:
Modest cost, saves the battery, not the BMS.
Protects from overvoltage BMS error
Low current draw
Relay could trigger alert/alarm

Issues/comments regarding this latest "fail safe" attempt?

 

[SpongeboB Sinewave]

I mentioned this on his YT channel but something like this on either the battery or PV side could help on a system where you might be worried about over-voltage of the battery and its low V handling BMS.

MidNite Solar DC Breaker with Remote Trip (MNEDC125RT)

The 125A, 125VDC MNEDC125RT breaker from MidNite Solar features Remote Trip that allows it to work with MidNite's MNBDM and the Rapid Shutdown System.

realgoods.com

Doesn't have to be THIS particular breaker of course but a remote trip switch or breaker of some kind that can handle the over-voltage and current.

A little bit of DIY needed but not too bad.

The charge controller would have to have two (2) faults. First the top FET of the buck converter would have to short (typical FET failure mode) and also the internal relay contacts would have to weld together.

boB

PS... I just now noticed there are MANY MANY responses to this so a disconnecting switch may already have been mentioned.

 

[Just John]

I mentioned this on his YT channel but something like this on either the battery or PV side could help on a system where you might be worried about over-voltage of the battery and its low V handling BMS.

MidNite Solar DC Breaker with Remote Trip (MNEDC125RT)

The 125A, 125VDC MNEDC125RT breaker from MidNite Solar features Remote Trip that allows it to work with MidNite's MNBDM and the Rapid Shutdown System.

realgoods.com

Doesn't have to be THIS particular breaker of course but a remote trip switch or breaker of some kind that can handle the over-voltage and current.

A little bit of DIY needed but not too bad.

The charge controller would have to have two (2) faults. First the top FET of the buck converter would have to short (typical FET failure mode) and also the internal relay contacts would have to weld together.

boB

PS... I just now noticed there are MANY MANY responses to this so a disconnecting switch may already have been mentioned.

Electrodacus uses a similar "last ditch" arrangement. Trips a breaker to cut off all input current.

The solution exists, and is cheap.

 

[Davismltc]

Does anyone see problems or issues with the proposed "fail safe" solution below to protect the LiFePO4 battery from overcharge?

A normally open continuous duty relay is placed across the PV panel terminals, so when energized, it shorts the output of the PV panel
A 15A fuse is placed on one leg of the input to the SCC, below the relay (to handle normal 10A PV current)
The relay is triggered by the output of a voltage comparator anytime battery voltage exceeds 3.6v per cell (e.g. 14.4v for 4 cell battery)
(The relay and comparator are powered directly from the +/- battery terminals)

"IF" battery voltage exceeds 3.6v per cell, relay energizes and momentarily shorts out PV panel output until voltage drops below 3.6v
"IF" the SCC was shorted, the battery current blows the 15A fuse and disables all further PV output to the battery

In normal operation, only comparator draws power from battery (milliamps)
Alarm/alert could be triggered along with the relay (LED / buzzer)

======================================================

(FYI - My truck camper build: 4s 200A LiFePO4, Daly 4s 100A BMS, 2x 350w PV, 2x EPever 30A Tracer AN SCC, 3kw 12vdc/120vac inverter)