Battery disconnect for nominal 48v systems?

Setting up a 48v nominal system. Two 48v batteries in parallel, the Jakiper server rack batteries. Each has a circuit breaker disconnect on the battery.

Most schematics also feature a main battery disconnect right at the positive busbar, using the red disconnects from Blue Sea. I would just add one of those in so that I can turn off the battery but they are rated at 48v max, and Blue Sea says they aren't for use with 48v nominal systems that can run up to 58-60v when fully charged.

So what are people doing for a battery disconnect for 48v nominal? Will's 48v workshop system seems to use a surface mount Blue Sea circuit breaker (like a 285 series), but those also seem rated for a max voltage of 48v.

Not something to worry about? Just use the disconnects on the server rack batteries (would need to ensure both are off!).

Take a look at Noark circuit breakers.

A human has to flip a master swtich, why not flip 2 switches to isolate the batteries?

I would double check with Blue Sea themselves because I thought someone has said their voltage rating is "nominal" rating so it is 48V nominal battery bank that it could handle.

jameshowison said:

Just use the disconnects on the server rack batteries (would need to ensure both are off!).

Click to expand...

That is what I would do.

rhino said:

I would double check with Blue Sea themselves because I thought someone has said their voltage rating is "nominal" rating so it is 48V nominal battery bank that it could handle.

Click to expand...

Victron, Blue Sea, Perko, and others all produce battery disconnect switches rated at "48v" which are commonly used on 48v-nominal systems.
I'd be interested to hear if any of these manufacturers actually forbid this practice, as it seems a bit misleading if they are not actually compatible with a "48 volt" battery.

Albright makes some of them too, rated 48v max, but then they also claim that it is "Suitable for voltages up to 48 Volts DC, commonly used on 12 Volts, 24 Volts and 48 Volts DC electric vehicle systems." according to their datasheet.

Clear as mud?

Thanks all. Some good options there. FWIW, I wrote to Blue Sea technical support with:

Disconnect for nominal 48v system (ie 58v max)

Are any of your battery disconnects rated for a nominal 48v system? ie voltages up to 58 or 60v?

I see 48v max on these listings.

Click to expand...

And they replied:

Hello James,

The only switch we have that is rated for 48V nominal systems is our 7700 or 7702 ML-Remote Battery Switches. They are rated up to 64V maximum, but they do require a 12V (7700) or 24V (7702) control system to be used remotely.

7700: https://www.bluesea.com/products/7700/ML-RBS_Remote_Battery_Switch_with_Manual_Control_-_12V_DC_500A

7702: https://www.bluesea.com/products/7702/ML-RBS_Remote_Battery_Switch_with_Manual_Control_-_24V_DC_500A

Please ask if you have any additional questions.

Best Regards,
Blue Sea Systems Tech Support

Click to expand...

Depends on how much you want to push the limits. I bought a switch from Aliexpress that is rated for 600A Continuous @ 50v max. It's the same as this one but half the cost if you wait for it from China. Amazon Battery Switch

I plan on never pushing more than 125A though it at 54v or so. For me, it was a reasonable assumption that it can handle that.
It's not entirely built yet, so I'll let you know if it goes up in flames.

I know Victron shows rotary switches in their 48V system schematics so they must exist. I have a switch from Marinco that also says simply "48V" so I am double checking with them if that is nominal or max.

It is rather pathetic that all of them are false advertising to an extent. not one of them are useful for lithium system that range to 58 volt for 48 volt systems. so basically our retail system is flooded with useless junk, ie an accident waiting to happen with no accountability what so ever. they are all really rotten manufacturers.

rhino said:

I know Victron shows rotary switches in their 48V system schematics so they must exist. I have a switch from Marinco that also says simply "48V" so I am double checking with them if that is nominal or max.

Click to expand...

Victron's "Wiring Unlimited" document seems to suggest that NH fuse disconnects are a good option for systems over 24v, at least for disconnects. Indeed, the ratings on a lot of those things are great; some of them up to 1500V DC and with a large (tens of thousands of AIC) interrupt rating.

The only problem is that I haven't found a decent domestic source for them yet in the USA.

In the USA, under 50 volts is low voltage so falls under a different set of rules, so a 48 volt system (which will have voltages over 50 volts) legally must be in conduit, batteries must be in a steel box with wiring to the inverters, charge controllers “must be “ in conduit where as a 24 volt system will not be subject to any high voltages anywhere, and the DC wiring does not have to be in conduit.

All qualified certification agencies will test to a much higher voltage than the “approved voltage” , I cant offically tell you this but I know for certain that most Bxxx Sxxx switches are tested to 72 volts but are certified to 48 volts.

Blue Seas switches are really rebranded Buss/Bussman rebranded, Buss does the certification and does push them over 72 volts

This question was posed to Robin Gudgel, president of midnight solar, about his breakers, and he responded with “don't lose any sleep over it, they will be fine”

Square D certifies their QO line to 48 volts DC , but tests to 72 volts, they used to be certified to 72 volts,

Please note, that is the QO line only, not the Homeline which are AC only.

The difference in AC and DC breakers is the “arc chute” which helps extinguish the DC arc, AC breakers do not have an arc chute as at zero crossing the arc will “self extinguish”

Tecnodave said:

Square D certifies their QO line to 48 volts DC , but tests to 72 volts, they used to be certified to 72 volts,

Please note, that is the QO line only, not the Homeline which are AC only.

Click to expand...

Interesting. On page 7-2 on this document (https://download.schneider-electric...CT1901_SEC-07.pdf&p_Doc_Ref=0100CT1901_SEC-07) it appears that the QO and QOB are DC rated at 48v and 5K AIC, but not the others like QOU, QOBH etc

I wouldn't worry at all about a 48V rated switch holding off 60V. I might worry about an AHJ refusing to sign off a permit based on that.
I would not want to open the switch under full load.
As people have pointed out, the only apparent issue is access to > 50VDC by the lay person.

Breakers and fuses, which open under overload, I would have greater concern. Some give varying AIC ratings for different voltages. Those with a single rating probably have a similar curve, but we don't know.

RobertGreen said:

Interesting. On page 7-2 on this document (https://download.schneider-electric...CT1901_SEC-07.pdf&p_Doc_Ref=0100CT1901_SEC-07) it appears that the QO and QOB are DC rated at 48v and 5K AIC, but not the others like QOU, QOBH etc

Click to expand...

Thanks for pointing that out, I missed that. I have used QOU for DC.....ouch, now only use CBI (Midnight Solar) for din rail, I like them much better. Now i'm going to have to do research on my Schneider (Germany) din rail breakers, that are used in some european electrical equiptment, they are totally different than anything that Square D has done.

RobertGreen said:

Interesting. On page 7-2 on this document (https://download.schneider-electric...CT1901_SEC-07.pdf&p_Doc_Ref=0100CT1901_SEC-07) it appears that the QO and QOB are DC rated at 48v and 5K AIC, but not the others like QOU, QOBH etc

Click to expand...

UL Listed to 48VDC for a single pole.
Another document says "Square D Certified" to 125VDC for 2 poles.

Tecnodave said:

Thanks for pointing that out, I missed that. I have used QOU for DC.....ouch, now only use CBI (Midnight Solar) for din rail, I like them much better. Now i'm going to have to do research on my Schneider (Germany) din rail breakers, that are used in some european electrical equiptment, they are totally different than anything that Square D has done.

Click to expand...

For AC, I've found QO270 had large variation in resistance from unit to unit, which caused imbalance in paralleled circuits (grid feed through relays of my inverters.)
The Schneider DIN rail 2-pole 63A breakers solved the problem.

RobertGreen said:

Interesting. On page 7-2 on this document (https://download.schneider-electric...CT1901_SEC-07.pdf&p_Doc_Ref=0100CT1901_SEC-07) it appears that the QO and QOB are DC rated at 48v and 5K AIC, but not the others like QOU, QOBH etc

Click to expand...

@RobertGreen

Go read that chart once again.....QOU is rated for DC, same specs as QO

Also Schneider’s Multi-9 din rail breaker is rated 60 volts DC in single pole and 120 volts DC tandem. These are what I find in German and Austrian electrical equipment in use here. These are listed in the Square D catalog but are manufactured in Germany.

Schneider Electric (Germany) bought Square D and incorporated them into the entire product line.

In my commercial work, I do stock only Square D, Schneider, and Siemens

Hedges said:

For AC, I've found QO270 had large variation in resistance from unit to unit, which caused imbalance in paralleled circuits (grid feed through relays of my inverters.)
The Schneider DIN rail 2-pole 63A breakers solved the problem.

Click to expand...

Are you speaking of QOU and Multi-9 breakers? I think the Multi-9 is a far superior design, I do see them in industry but they are timely to obtain. narrower than any other din rail , they are hard to substitute for. I confess to using Midnight Solar MNEAC breakers in a pinch where there were Multi-9 breakers in use.

I think it is QO270 and Multi-9 C60 63A


I already had a pair of the Schneider after visible blade disconnect as OCP for inputs of four inverters.
Outputs landed on two QO270 in the PV combiner box. I observed imbalance in inverter wattage display. I first fixed the imbalance by landing unfused on main lugs (use lugs with 2 holes & set screws each). Then added another pair of the Schneider in line with that.

Tecnodave said:

@RobertGreen

Go read that chart once again.....QOU is rated for DC, same specs as QO

Also Schneider’s Multi-9 din rail breaker is rated 60 volts DC in single pole and 120 volts DC tandem. These are what I find in German and Austrian electrical equipment in use here. These are listed in the Square D catalog but are manufactured in Germany.

Schneider Electric (Germany) bought Square D and incorporated them into the entire product line.

In my commercial work, I do stock only Square D, Schneider, and Siemens

Click to expand...

Good catch! I had noticed the "120/240 V~" marking on QO breakers before, and it hadn't occurred to me to dig deeper to find the DC rating.

Im surprised, Is that in your Sunny boy inverters. They use QO plug on buss breakers? I don't really touch grid tie solar, Living out on the edges it does not make any sense to me to grid tie. Most of my commercial electrical work is Agriculture or Industrial, with a smattering of remodels thrown in.

For very high reliability installations I use Siemens all copper buss farm panels, I prefer them over Square D, but I have to look for them, no locals stock them but nearby is Salinas, California, a “can do“ city

Around here the Strawberry farmers are king, we are the Strawberry Capitol of the world. Farmers that pop a water pump on a 4 day weekend will lose a million bucks in lost production before they see most electricians. Im available to select clients 24/7/365...that comes with a price. I do not need to mess with residential. I did a 2 am to noon 100 h.p. motor swapout and took home a months wages. And yes you can call 24/7/365 and get a loaner pump motor delivered within a few hours. Complete rebuild on a 1930’s vertical shaft 100 h.p. motor delivered to farm including loaner motor pickup and delivery...$2500, what a deal.....Shop will not even look at a motor smaller than 10-15 h.p Ag pump motor shop in Salinas....

Tecnodave said:

Im surprised, Is that in your Sunny boy inverters. They use QO plug on buss breakers? I don't really touch grid tie solar, Living out on the edges it does not make any sense to me to grid tie. Most of my commercial electrical work is Agriculture or Industrial, with a smattering of remodels thrown in.

Click to expand...

The Sunny Boy and Sunny Island don't have any AC OCP built in. Some Sunny Boy have a combiner with touch-safe fuse holders. Sunny Island has DC circuit breaker for battery (I don't know the exact rating.)

Various Sunny Boy require breaker/fuse of different ratings for grid connection.
Sunny Island requires breaker/fuse for both AC input and output. It can pass up to 56A through. Max 70A fuse on input (which gives 25% headroom). It says to use max 56A protection on the output (but that wouldn't work unless it was rated for 100% of breaker value; for thermal/magnetic we need 25% headroom.)

For paralleling feed-through from grid, it says to match wire length. I have about 40' on input and 20' on output (6 awg) and expected that to balance well, but the QO breaker dominated resistance and caused imbalance.

I've shipped out little motors for rebuild, like 2 HP pool pump. I got a 2-speed, but found 1800 RPM wasn't fast enough with DE filter at all plugged up. A pressure-side pool sweep came with the place, but kept blowing hoses. I got a suction-side, an to adjust the input bleed to where pump was cavitating to get it to work. Pool stayed clean of debris, but I burned up a couple pumps. One would run on low, but on high there seemed to be greater force, loading armature until it dragged (bearings going?) I now use a 3-phase with Hitachi VFD, set it where it pumps without making cavitation sounds. No automatic sweep, push the vacuum once a week. That's been working about 15 years of intermittent use.

I have dealt with a few Sunny boy installations and knew that the small ones do not have AC OCP but I have never seen a bank of 4 of them. Im surprised that the internal impedance of the QO breakers would affect the balancing of the units. Im still not quite clear as to how you replaced the QO with din rail. I have used bolt on adapters that allow a unfused feed to be bolted onto a QO bus, Is that what you are doing?

Not bolt-on adapters on the bus fingers, rather at the main lug ends. Then the DIN rails breakers in another box feeding it. This was for the 225A QO breaker panel on output of Sunny Island fed by the battery inverters.

(On input side, I already had a 100A visible blade switch which previously fed the main lugs when this was grid-tie only. For OCP on input when installing the battery inverters, I mounted a DIN rail inside the switch box and put two dual breakers there. With house now powered through Sunny Island, I also added an interlocked "generator switch" breaker panel so Sunny Island could be bypassed to put house directly on grid.)

I was surprised too. I expected 60' of 6 awg to dominate resistance.
You would expect excess resistance to result in heating and tripping at low current. Those QO breakers work fine individually although I haven't tested their trip threshold. I do have a 100A breaker that started tripping well below 70A, and wouldn't reset until some days later. I wonder if it slowly cooked and changed its characteristics, perhaps due to poor connection with wire (no burned insulation, though.) Home Depo wouldn't exchange, so I plan to contact Schneider to take advantage of their lifetime warranty.

Sunny Boy or Sunny Island?

Sunny Boy wouldn't involve multiple parallel paths from same source to same load. Each is independent, with its own PV array, feeding common AC grid. No problem there; each just drives its power into whatever AC voltage is present.

Sunny Island can have 1 to 4 in parallel for 120V system (220V for European model.) It can have 2 wired for 120/240V split-phase, or 4 which puts 2 in parallel on each phase. Each monitors current to/from grid which is not to exceed 56 Arms. When internal relays are connected to grid, load draws current through two in parallel, or AC coupled PV Sunny Boys backfeed current in parallel. Master SI reports total power to/from grid, and the other three Slave SI each report the power passing through them.

QO breaker panels are available "main lug" or "main breaker". I replaced the single large lug with the following dual lug.


I mounted a 6" x 6" x 2' wiring box above the breaker panel and put contactors in it for load-shed (house load disconnected at low battery so inverter can keep supplying AC to Sunny Boys while waiting for sun to come back up.) I added two dual 63A DIN rail breakers in the box so each SI has OCP.

I had thought SI could accept 56A from grid and add 50A (6kW/120V) for 106A load, but manual says 56A max output. Except surge of course.

Here's a big system, DC coupled with Sunny Island Charger rather than AC coupled with Sunny Boy. But it looks like multiple 3-phase clusters each with their own battery, which would then be connected together on AC bus with Multi-Cluster Box. No wires or relays in parallel.


This one is AC coupled to Sunny Boys.