Buffer battery or inductor in Lifepo4 system to protect equipment if BMS shuts down

Hi,

So i accidentally pressed the emergency stop (connected to Lynx BMS-NG REM terminal) whilst the Skylla IP65 was charging at 68Amps and all loads were running. Inevitably therewas nowhere for excess energy to go and I damaged some non victron loads.

Only person to blame here is me. Anyway, my understanding is that this is a reason why a buffer battery is often advised or could this be an inductor?

Can anybody with experience in this suggest a solution to prevent/reduce damage to loads in similar events?

Also, should i choose, a buffer battery (what type, and where in the system is it installed) or an inductor?

Your thoughts much appreciated.

Thanks,

Alex

NOT an inductor…
The best method for over voltage protection would be either a varistor - if you can get one of the correct voltage, OR a FET based voltage clipper, designed for the voltage. There are some devices designed to protect alternators form inadvertent disconnection, or one can be home brewed if you have the skills.

In this situation the Lynx NG BMS should disconnect all charge sources through the ATC connection. If the Skylla does not have a remote terminal and is not controlled by VE Bus VE CAN or VE Direct then you should have a Smart Battery Protect on the output of the charger with the Battery Protect controlled by ATC. You could also look into the alternator mode setting for the relay as this can be set to shutdown 2 seconds before the contactor opens and run the ATC wire through this. I am not sure if details of this are in the Lynx BMS manual, it is certainly in documentation for the Wakespeed WS500 alternator controller.

Hi, thank you for the reply.

Yes, the Skylla is connected to the BMS by the ATC terminal. In my panic I did not check if it had been told to stop charging. I should turn off the loads and test this.

The Skylla is also connected by ve.can but that connection seems to be only for monitoring, and it doesnt participate in DVCC or those things.

I wonder how fast the Skylla would stop charing before unwanted voltage spikes pass through the system or if a battery protect would be any faster?

Alex

Hi, Thanks for your reply.

Ok, not an inductor. Will explore the other options you have suggested.

Am I doing something wrong or is there a reason why this sort of feature is not a standard item in 12v systems? I have never seen such included in any reference diagrams from Victorn or anyone else.

Alex

Hi,

So the simplest thing seems to be to connect either a Balmar Alternator Protection device APM-12 or a Sterling Alternator Protection Device across the outputs of the Skylla IP65 Charger

Do you concur?

Alex

Edit, I did a search for Dump Load, now that I know about that term, and this link seems to best sum up the situation, although for me the charging source is not an alternator but an AC/DC charger but same rules should apply.

The alternator protection device would be a good start.

The regulation problem:

  1. Switch mode supplies: The output voltage surge depends on how much energy is transferred in the time it takes for the internal regulator loop to notice that the output is over voltage. If the SMPS is operating at full capacity, then this can be considerable.
  2. Alternators - similar to the SMPS, but the energy is stored in the magnetic filed of the alternator and can cause the output to spike to ~80V (depending on model) if the load is suddenly disconnected when the machine is at high rpm and high current.

An emergency stop button should NOT cause disconnection of batteries from charging sources, this covers the above problems from destroying equipment. 2) above, for the alternator is the best reason why a lithium battery should not be used as the primary energy storage for the alternator.
Charging sources do need a proper emergency shut down procedure, which usually entails removing / disconnecting the energy source from the converter BEFORE any converter load is disconnected.

As I mentioned above, look at the option to use the programmable relay in alternator mode to shut down the charger a few seconds before the contactor opens by running the ATC wire through the relay on the way to the charger. A battery protect option will be quick acting and should stop the over voltage from getting to the loads but the charger will still spike.

Hi,

Thanks again for the feedback. I’m not sure I follow you. The current set up is as below using the 4th method on the far right:

My understanding of the REM connection is that it will IMMEDIATELY open the contactors. I cannot find in the documentation anywhere if it will also send a command to the ATC ATD contacts as well.

If the Lynx BMS NG opens the contacts imediately then I dont see where/how to introduce the delay to give time to the chargers to back off.

The intention of this system is for use ONLY in emergency and not as part of the usual shutdown process of the system, thats why the EPO Emergency stop button is behind a cabinet door, but crew are trained what to do if they need for some reason to disconnect the batteries. This, I am told, is a requirement in ABYC and ISO standards.

In an emergency it could be argued that the downsides of a Dump Load/Spike event is the lesser of the evils, I am just trying to reduce damage if it occurs, accidentally or on purpose.

In my case the AC/DC charger was puting out 1kw and I had only 250w of load, so the rest was going in to the batteries, and so when I accidentally bumped the EPO button, the contactors opened and the charger was left understandably hanging. For the purposes of language clarification, no alternator is involved here, but I see that Alternator type functionality could be adapted to my use case.

How does that work with the Lynx BMS NG, Distributor and Power IN design by Victron? The manuals clearly state that the Batteries stay one one side of the BMS and all Charging and Loads on the other side

I have Lynx Class T Power in one onside, then the Lynx BMS NG and then the Lynx Distributor with loads and sources.

The advise re the ATC is nothing to do with the remote switch on the BMS.

You have said you have the BMS ATC connected to the charger but even with the ATC you still get overvoltage. This is what happens with alternators. To avoid this, the programmable relay on the BMS can be configured into alternator mode. See sections 3.4.7 & 4.3 for the relay. When the programmable relay is in alternator mode, the programmable relay opens 2 seconds before the main contactor. The ATC signal to the alternator is run through the programmable relay and the ATC relay in series. I am suggesting that you use this same method for the Skylla remote. You could get the same problem with your set up if the battery stopped charging due to high cell voltage or high or low temperature. Using the alternator mode may give enough time for the voltage to dissipate before the contactor is opened.

In the drawing below are the connections for a Lynx BMS with a WS500 alternator. A positive supply is taken from the BMS aux power through a 0.5A fuse labelled WS500. This connects to the ATC relay. The ATC relay then connects to the common terminal of the programmable relay. The NO terminal of the programmable relay (grey wire) is the ATC signal to the alternator but should be able to be used like a normal ATC signal to a Skylla charger.

I know this has been marked as solved with an extra solution next to that. But I would like to propose another way to solve this with a device that’s common in industrial automation.

I’m currently building my home system. To make the E-STOP less invasive on the switching components I use a E-STOP relay from Schneider Electric with 4 direct NO contacts and 4 delayed NO contacts. The delay has a setting range from 0 to 900 seconds. With the direct contact I switch off the converters (MultiPlus, MPPT, Fronius, Orion, chargers etc.) this reduces the current through the BMS to virtually nil. After 3 seconds the BMS opens and simultaneously the PV is disconnected from the controllers. Only thing alive will be a 24V lead acid with the Cerbo and some control components.

This relay also simplifies a controlled restart. The reset switch should best be in the vicinity of the battery, so that the battery can be observed when restarting from dark.

I think this device can also be used aboard to switch off alternators and such.

Ideally the PV should be removed from the MPPT’s before the BMS is opened.
However, with the DC bus also having the inverters present, and also having been turned ‘off’ by remote control, this should not pose a problem.

Thanks for your reply. I know the timing of this could be a bit more favorable, but this is in an emergency, and I think this is the best way to do this. And when switching things back on, I have to go to the attic to manually switch on the PV strings. That is only possible when the E-STOP is reset and the BMS is back online.

I designed a varistor based protector that would be suitable for arresting/redirecting a surge. Refer to the article in Silicon Chip magazine, Solar Panel Protector - March 2026 - Silicon Chip Online

For this application, the device would be placed across the DC cables, (instead of protecting the solar assets), as close as possible to the source of the high voltage, (the charge controller). Select a varistor that is several volts above the maximum battery voltage under all normal operating conditions. An in line fuse must be installed for this installation, since the battery has a very low impedance and could easily release the smoke from the device or the connecting cables. The fuse will NOT activate in time to protect the downstream devices but it will disconnect in time to protect the cabling should a fault occur. The varistor WILL activate in time to absorb the surge.

Hi everyone,

Thank you all so much for a lot of valuable information in this thread, I will take a look at it all.

Many thanks again, but keep the discussion going and see if there are any other smart ideas.

Alex

Hi All,

Could you comment on this drawing? Specifically the location of the SPD and the smart battery protect. The distances between the chargers and the Lynx distributor are not far.~1-1.5m cable length one way. Concept is to aggregate all the chargers to a single point then take it back to the main Lynx system where the BMS and batteries are.

Do you think thats reasonable or do I need to put an SPD on output of each charger?

Advice much appreicated.

Alex

Alex,

A surge protection device has the ability to protect a circuit against a fast rising voltage, (induced current), from numerous sources. The devices, (IDEALLY), need to be customised to provide the maximum protection to the intended circuit, as an example, some commercially provided SPD devices are triggered at 1000V. If your circuit is sensitive to voltages over 150V, then this device is not going to offer much protection. It will help when a voltage surges above 1000 volt but will remain inactivated for a surge of 800V.

I would recommend that you consider the highest voltage that will be expected on the line, (I didn’t see a voltage on your circuit), if it was a 48V nominal battery connected to the Lynxs distributor, then I would be looking for an SPD that would activate at no more than 75V.

With regard to the protection offered and the location, you have to consider WHAT you are protecting against and WHAT is to be protected. In your circuit you have an AC/DC charger, (this looks like the main source of power by the amperage rating of the fuse). If you are concerned about spikes on this source then the SPD should be located on the AC side of the charger, preferably as close to the connection point of the AC cable. This would allow all the cable from the connection point to assist in the reduction of the surge on the AC line, think of the cable as a resistor/inductor will help reduce the surge.

If the surge was from the DC/DC chargers, ie an inductive surge if the battery was disconnected, (as per the question I responded to), then the SPD should be located as close to the DC/DC output as possible.

If the surge is from the MPPT unit, (assuming solar panels), then the SPD should be located on the solar panel side and be about half way between the panels and the MPPT, (or close to the MPPT to keep it out of the weather). The reason for the half way point is because a nearby lightning strike will induce a current and hence voltage into the cable, it works like an aerial, the mid point will absorb spikes better, protecting both ends equally.

If the surge is, (like the article I responded to), is caused by the battery being removed and inductive spikes resulting, then you would place near to any inductive sources, (as per DC/DC charger above).

The important part to remember is that the SPD should be fabricated with consideration of your voltage limits. If installed in a solar string, then the activation voltage should be set close to Voc +10%. If installed on a battery line, then Vbatt(max), +10%. For DC applications like the battery, you would only need a varistor connected from the positive to the negative line, for an AC line or the solar panels, you would use the full 3 varistor, (see article I referenced earlier), including the paths to Earth. The cabling for the SPD should ideally be at least 4mm2. Some commercial units that I’ve seen use very lightweight wire, it would carry very little fault current and would also have a high resistance, two features that would not help. If connecting between battery terminals or other low resistance/high current sources, then install a fuse in line to protect against faults in the SPD.

Ideally you would want the SPD activation voltage to be below the capacity of the attached electrical goods but this cannot always be achieved. For example, you could be using an MPPT with a maximum string voltage input of 150V, if your solar panels have a Voc of 100V, then you could select an SPD with 125V activation. This would be ideal. But of course, the existing string voltage could also have a Voc of 140V, so the SPD activation level have to be above the ideal 150V input level of the MPPT charger. Even in this condition, the SPD will provide a very high level of protection to the device from a spike because the MPPT has capacitors on the input which take time to charge up and exceed the rated 150V, (they might be good for 200V before failing), the SPD activates before the voltage inside the MPPT has time to rise and hence stops the bulk of the surge.

So where to place and what to protect are important but you need to consider the above before selection and installation.

BUT… If you are talking about preventing a spike from a direct hit by lightning, then sorry to say there is nothing that can be done cheaply to help there. An SPD is limited because it can only provide short term protection. Eg nano seconds. If someone connects a 230VAC line to the battery, the SPD will activate but it will soon fail too. They are only good for short duration spikes.

Hope this helps.

Ian

Hi Ian,

Thanks for the detailed reply. I have read and am in the process of digesting your article on making the Solar Panel protector. I am also considering the Balmar Alternator protect or the Sterling Power Alternator protect devices in the same position.

My drawing was in the context of the original post rather a long way up, and in this case it is to protect DC loads from damage by so called “Dump Loads” caused by the battery BMS opening the contactors when the charging sources are all pumping away at full capacity.

This happened to me because I accidentally pressed the emergency stop button while chargers were putting out, causing the spike.

This is a 12v system, and my idea in the drawing is to remove the ATD contact from the AC/DC charger and move it to a Battery Protect, and also add a device like your Solar Panel Protect device, following your advice on the selection and sizing.

I understand that Lightning is a different story all together. I am just seeking to handle the “Dump Loads” from the chargers.

Best,

Alex