stefancs avatar image

Connecting a charger to Load (Out) side of a BatteryProtect


I have two questions:

  1. is a BatteryProtect able to handle reverse current when connected as shown below in case of the BP220? Reverse current would be up to 30A while charging, and Vout >Vin.
  2. In case 1. is possible, let's now now assume the BP220 disconnects due to cell under voltage ( and thus BMS Load Disconnect output being set free floating). When connecting the charger (14.2 or 14.4V) thereafter, what would the voltage drop between Vout and Vin be? Can the BP220 handle 30A reverse current in that case, up to the point when it connects again after 30sec?

Note: the design intent of the layout below is to protect the SmartLithium battery from standby current drawn by the charger in case of cell under voltage. With the original design from the manual, this could only be achieved by a main switch (or pulling the fuse) between battery+ and BP. I cannot use a Cyrix-Li-Charge because my charger applies a IU1oU2oU3 charging characteristic with U3 being below the Cyrix-Li-Charge 13.7V engagement level.

Image: BatteryProtect Manual figure 5, modified (second BP out connected to 1st BP out instead of 1st BP in as in original diagram).



BMSBattery Protect
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

12 Answers
mvader (Victron Energy) avatar image
mvader (Victron Energy) answered ·

Hi, here the official answer:

The BatteryProtect is not at all designed in a way that the current can flow both ways.

Doing so might result in fire. Do not do it.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Thank you for clarifying this.

Not to dispute the official answer, but what about a battery connected to an inverter/charger. When inverting, the current is flowing from the battery to the inverter and in charging, the current is flowing form the charger to the battery. If there is a battery protect in the circuit, isn't the current flowing both ways?

Hi @don1cobb, indeed its flowing both ways. Its why a BatteryProtect should never be placed there in the circuit.

don1cobb avatar image don1cobb mvader (Victron Energy) ♦♦ ·

Then how should I use the battery protect in this situation? It is currently a backup system and there is only batteries and the inverter/charger. I want to protect against over discharge of the battery. I currently am using the remote cutoff controlled by a BMV-712 to protect against charging below freezing, high voltage and high temperature.

You use the signals to control the inverter.

If you need more you can use:

-a breaker with a trip unit (only off, manual on)

-a contactor (but need a precharge device also then)

don1cobb avatar image don1cobb mvader (Victron Energy) ♦♦ ·

I've looked through all the material I could find from Victron and there is no mention of this issue. I would suggest that you put a warning in the manual, etc. I would not have bought it if I had know this. But better to loose a sale than be liable for damage to a house, etc.

Agreed. I also bought two (and destroyed one) before finding this out.

Hi both, thanks for this; I’ll make sure we improve the documentation.

I’m sorry to hear about the failed units.

I will be returning mine after reading this discussion - thanks to all. It seemed like the perfect solution to protect an RV battery from being discharged while unattended. Since most RV's in the USA have been sold with inverter/chargers installed for two or three decades, Victron needs to alert customers to this danger in all their advertising and web content.

wkirby avatar image
wkirby answered ·

If the Battery Protect internal circuit design is how I think it is then no, it would not work in reverse.

Edit: I misunderstood the question, ignore my above comment.

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

So ,what is proper set up for two BP-100 UNITS ,one for low and one for high voltage?

stefancs avatar image
stefancs answered ·

@WKirby: thanks for your input.

I could think of a circuit design where it might work though. As we can only second-guess the internal design, I think only Victron could provide a definitive answer. Otherwise I just might try it.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

stefancs avatar image
stefancs answered ·

I now tested it and it worked.

With Pin 2.1 connected to +12V current will flow in either direction, with voltage drop <0.1V. When disconnecting 12V input to Pin 2.1, Battery to Load is interrupted. Load to Battery is not, but now has a voltage drop of 0.6V. As expected for n-ch power MOSFET.

The remaining question is whether it will damage the BP over time / at higher currents. I tested at 12.4V and 1A.

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Hi @stefan.cs: FYI I just repeated the same tests, with an 18W load - a couple of high-wattage 12V car light bulbs and a 1 Ohm power resistor all in series. (aside: the bulbs alone tripped the BatteryProtect's short-circuit protection due to their high inrush current when cold; the resistor was just enough to reliably run :-).

I also confirmed that current flows in reverse (i.e. with power supply connected to Out, and the load connected to In), regardless of the state of the remote input.

I'm willing to bet what we're seeing is that the BatteryProtect uses a power MOSFET with its intrinsic or body diode conducting when reverse biased. As I understand it, these diodes are generally rated to about the same as the MOSFET or even higher, so it should be safe.

It does beg the question why Victron don't use the topology you suggested in the OP?

myshopsolaire avatar image
myshopsolaire answered ·

Victron Energy advise this type of cabling when the charger can't be pilot by remote by the BMS so it will not damage the BP. And it's more efficient to use a BP than a Cyrix Li Charge so your system is good.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

stefancs avatar image
stefancs answered ·

@MyShopSolaire: thank you for your feedback. Unfortunately I now have two contradicting statements, yours and an offline one by the VE distributor I bought the BP from who basically says doing this would be at my own risk. Not sure you noted the change I applied to the drawing, i.e. connecting "out" to "out" between the 2 BPs. Do you have knowledge of anybody using that setup?

@BenL: I assume you are right, also the BP data sheet says it uses MOSFETs. However I would strongly recommend not to apply nominal load to the BP in reverse direction while the BP is "off". Assuming a drop of 0.7V through the diode, at 200A it would generate around 150W of heat. I cannot imagine a unit the size of a BP220 handling this. Compare for example the size / weight / heat sinks of the Argo Diode products, which are Schottky.

In the meantime I tested the BP-65 at about 25% nominal load / 15A. With the BP "on", voltage drop in either direction is in the range of a few mV. With the BP "off", voltage in reverse dropped by 0.7V, as expected. Makes for around 10W. For < 1 minute this may be ok, the BP did not get very warm.

Anyway I will not move ahead with this layout for now. Without an official spec limit by VE on reverse current I take it to be zero. The risk of the BP later failing in "on" state due to prior damage from reverse current defeats the idea of battery protection.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

myshopsolaire avatar image
myshopsolaire answered ·

@ stefan.cs Yes, we have already advised a similar use for the BP as load or load disconnection. It's a little different, it was for a lithium system, we have to use BP the Multiplus Compact does not accept the lithium assistant.

This installation dates from May 2018 and we had no bad feedback and it has been approved by one of our sales manager of our country.

And we have also sold many battery protect for similar uses of lithium instead of the Cyrix Li Load or Li Charge.

schema-m-bourraly.jpg (145.9 KiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

@ MyShopSolaire , In your posted diagram "" schema-m-bourraly.jpg "" I don't understand why the positive from Multiplus goes to BP220 and BP100? Can you explain? I think there is no HV protection in your diagram if Multiplus gives a too high voltage. I would say diagram will work if Multiplus and MPPT goes only to BP100. Also the wire connected to BMS, load and charge disconnect need to be change from position!

I see in many diagrams that Victron advise to switch off the Charges with a Ve.bus or ""charge disconnect"" control signal.

I like yours with the BP for disconnect chargers I think only a control signal is NOT safe and much better to have always a ""citix li charge"" relay or BP that switch off the chargers. This because if there is a HV event, there must be something wrong in one off the chargers. If there is something wrong in one off the chargers it is most likely that a Ve.bus or ""charge disconnect"" will not work also. What is your opinion about this?

I like your drawing. How do you make this nice drawing, where can I download this program?

Don’t do this; the BatteryProtect is not designed to be used in a solution where the current can flow both ways

benm avatar image benm mvader (Victron Energy) ♦♦ ·

@mvader Can you show in the drawing above from "M Bouraly" what is the correct way to connect the Multi.

It goes directly on the battery terminals. The VE.Bus BMS will shut the Multi down, when required, by a digital signal over the VE.Bus communication.

Paul B avatar image
Paul B answered ·

As these BP pass current both ways. ie in and out then all should be fine. it really does not matter which is in or out based on your diagram at all

the reason they have a in and out is ONLY to supply power from the IN (the in normally comes from the battery) to pin one via a resistor so that you can turn the unit on and off manually or to remotely override the off or low state of pin 2 by joining pin 1 and 2, plus the internal programming features.

otherwise there would be no need for a in and out at all.

It can also be turned on by battery voltage on both the in and out sides so its fully bi directional.

well thats my thoughts.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Hi @Paul B: you're correct that the BP will pass current both ways. However one of those ways is uncontrolled :-)

i.e. current from In to Out will be switched off and on depending on the state of the remote input; current from Out to In is permanently 'on', with a voltage drop from the internal diode.

So presuming you installed the BP for the purpose of controlling the load, you do need to pay attention to the connections. The In and Out connections are not interchangeable.

Thanks Ben

I was not aware of this.

Also then the circuit diagram in this thread has a issue as the BP220 has the multi and the MPPT units connecting to the out connection of the BP220 and thus current will flow to the lithium battery and wont ever be turned off if a overcharge of a cell overcharge occurs or a high voltage shutdown is required.

hmmm am I correct here or am I missing something.

this is not good it sort of makes the BP unit not usable in a BMS system that needs to control the incoming current from the chargers,

I expect the schematic in the OP, which is a small change from the BP manual's figure 5, will work fine and should eliminate reverse leakage from the battery to charger at the cost a bit of charging power loss in the BP220 - probably no big deal from a modest mains charger.

Mind you, I suspect a simple power diode would do the same thing, more simply?

The other diagram labelled "M Bouraly" I have no idea - I haven't traced it out properly.

> it sort of makes the BP unit not usable in a BMS system that needs to control the incoming current from the chargers,

That's correct, a single BP alone is not usable for managing chargers as well as loads. You need to use a second units for the charger. Ref. figure 5 of the BP manual.

Sorry I was meaning the M Bouraly one

@BenL: the reason I cannot use a power diode is that the 3rd party charger needs power (17mA) at night to keep its microprocessor alive at night, so it will not start a new charge cycle at sunrise unless needed. Otherwise you are absolutely right of course.

@Paul B: I agree with you regarding the M Bouraly diagram - unless I am missing something, too.

stefancs avatar image
stefancs answered ·

Coming back to my initial question, I now assume the following would work. The Agrodiode Isolator puts Schottky diodes in parallel to the BP's MOSFET reverse diodes. It will take the reverse current in case any of the BPs disconnects.

In principle, it should look like this:

I would achieve the same by connecting a BMS 12/200 battery minus and connecting charger and loads to LB, leaving AB disconnected. The reason I am trying to find a different solution is lack of adequate installation space for one large unit as compared to several smaller ones..

bms2.jpg (124.5 KiB)
bms3.jpg (47.6 KiB)
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

uncle-bob avatar image
uncle-bob answered ·


Ok, I get that the BP series is not *spec'd* to do this or that. I've sat through very long lectures from corporate lawyers about how you never ever "extend" the specifications on a product in conversations with a customer. Staying employed is a good thing. Liability lawsuits are to be avoided. I get all that.

So what happens if you *do* put a charger on the load side? The answer is that the world does not immediately come to an end.

Let's say you hook up a BP-220 on the bench and feed up to 5A through it. In the "correct" direction, the voltage drop is about 3/4 what it is in the "reverse" direction. Indeed the power dissipated into the heatsink *is* higher when run this way. It's not the "gigantically higher" sort of thing you would get from a 0.7V diode drop suddenly coming into the picture. You still are in the tens of mV sort of range.

So indeed, the ratings "backwards" most certainly are not as high as the ratings "forwards". Unless they re-write the spec sheet to reflect this, there's an issue. It most certainly can overheat at a "lower than expected" current level. That's not quite the same thing as "does not work".

Does the "derating" required go up as current increases? That's a good bet. Is the net result a 50% derating or is it higher than that? All the test above suggests is that you are ok at 2.2% of the 220A rating.

Why even bother with this insanity ( I freely admit it's a bit crazy) ? The proposed "use two in series" fix shown with the solar charger is fine if that's what you have. If you are running a Multi hybrid inverter / charger .... not so much. Thus the ongoing investigation .....errrr ... insanity ....


2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

graham avatar image
graham answered ·

I wonder if the answer to this is a Cyrix Li load and a Cyrix Li charge wired in basic parallel then if the load control goes low the Multiplus can still charge via the Li Charge.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

offhighwayvan avatar image
offhighwayvan answered ·


I have some input into this discussion that is very important. I installed the BP100 in a system and connected the charging source to the load side as the original poster was asking about.

It resulted in a fire in the BP100 unit and nearly destroyed the vehicle. The situation was that the batteries (not Lithium) were in a low state 70% charge or so and the charge source (Xantrex Freedom inverter charger) went into bulk charge mode when connected to an AC source. It was putting in about 60A in the Bulk charge mode. I was elsewhere in the shop and after some period of time (maybe 20 minutes) I noticed smoke coming from the electronics area of the van. I disconnected the AC source and batteries as well as used an extinguisher to put out the burning plastic of the case of the BP100.

Clearly reverse current of 60A through the BP100 generated enough heat to damage the unit and start a fire.

I'm planning on installing a new BP100 unit and hooking all my charging sources (inverter/charger, solar and vehicle alternator) direct to the battery this time. At this point the BP100 a glorified battery cutoff switch because only my smaller fused loads will be disconnected from the battery if the battery reaches the cutoff voltage. One of the main reasons for the disconnect feature was to protect if the customer left the inverter on by mistake and parked the vehicle for a few days.

Anyhow, be careful with current flowing through the BP100 in reverse.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

That's what I'd call a definitive answer! Thanks for sharing.

I'm guessing that the BP100 was off / non-conducting when this occurred?

(@60A and an estimated diode drop voltage of ~0.6-0.7V, that's about 40W power dissipation; seems about right to start smouldering after a few tens of minutes for the plastic encased BP100)

Ben, it was actually on, conducting. I had it powering the interior lights.