Ebike Circuit Protection, Fuses and Circuit Breakers

Electric Bicycles have batteries capable of delivering a lot of current in a short time. Good design calls for a fuse or circuit breaker within a few inches of the battery to protect wiring, batteries and in some cases body parts from becoming damaged. Some battery chemistries can "vent to flame" when shorted, so this can be a fairly serious hazard.

It is important that fuses or circuit breakers be properly selected for the application. Both voltage and current are important. Using a protection device with too low a voltage rating can cause the device to fail and not provide the desired safety.

Many common fuses are not up to either the voltage rating or the current rating demanded in electric bicycles. This topic will be a place for folks to discuss this and share the products they have found that are useful for this application.

The fuse should be close to the batteries to provide maximum protection. It doesn't do much good if the protection is downstream of the short. Ideally it should be either in the middle of the battery chain or immediately at one end.

Here is one that I found that I plan to use in my system:



http://www.littelfuse.com/data/en/Data_Sheets/1575701_Strip_fuses.pdf

Rated at 80 volts and various currents from 35 to 500 amps these units are designed for electric vehicles. They cost about $6 each at Mouser. A mounting plate can be fabricated from a piece of plastic and some bolts. Or bolt wires to it and slide inside a piece of PVC or ABS pipe to insulate and mechanically protect it.

My design is still evolving, but at this point I plan to put a fuse right inside the battery box. I will build it into the wiring that the batteries plug into. I'll keep the Turnigy Lipo connectors as-delivered from HK, and then make a series/parallel harness that brings them out to one connector, or perhaps two - one smaller for precharge and the big guy for the main current.

Originally I was planning to use MaxiFuses, and those are okay at lower voltages. Since I have fallen for a higher turn motor and I need to move to higher voltages the Maxi's are no longer suitable for my setup. I have noticed that fuseholders often fail in high current usage, so make sure you select a high quality holder.

Circuit Breakers are another choice. After I started this thread Solar systems became very mainstream and the availability of inexpensive solar circuit breakers with DC ratings increased. I used one on the CroBorg build, and it has worked well, though a 63 amp unit did trip at 80 amps climbing a steep hill. Usually the time at full current is too low, but one day I tripped it. So I moved to a 100 amp unit. At this point I think the breaker is a better choice unless the fuse is rated high enough to only pop on a short overload.

So, what solutions are people using in their bikes?

Links to older threads would also be good to tie the information together:

http://endless-sphere.com/forums/viewtopic.php?f=3&t=22749

I have used thermal circuit breakers, which are resettable, but most are only rated for 50Vdc (250Vac) and 30A. You can get these from Farnell - element14, RS components, etc.
Eg. tyco W58 series 30A, Farnell p/n 143-6215 costs $7.92 AUD +gst.

Updating my old thread.

There have been a lot of failures with cheap fuseholders, even if the fuses have adequate DC voltage and current ratings, the holders often don't make adequately low loss connections, resulting in heating and failure at what should be safe operating conditions.

The best ebike circuit protection recently has been found to be Solar type AC/DC rated circuit breakers. These have ratings in the range we need, and are fairly inexpensive due to the growing solar power market.

Any part number recommendations for 48V/52V and 50A ?

The common values near there are 50 and 63 amps, I would go with 63 to have a little headroom. I used a 63 amp unit on the Borg, with the Sabvoton set for 80A. There is a time and temperature element, so a 63 amp breaker will hold 80A for a minute or two. It wasn't enough for a big hill, so I went to a 100A unit.

Here's one that says 50A but shows a 63A unit in the photo:

https://www.amazon.com/MidNite-Solar-Breaker-150VDC-MNEPV50/dp/B00BSYSTN2

And here's the 63A version that I have used:

https://www.amazon.com/MidNite-Solar-Photovoltaic-Circuit-Breaker/dp/B004EQRAQS

The 50A should hold at 50A for awhile, but if you operate at 50A for long periods of time then the 63A unit might be a better choice.

There are also some generic units that might be okay, check the ratings to see if they are accepted by the appropriate NRTL (Nationally Recognized Testing Laboratory) for your country (for those outside the US). Major products will have this rating (example: UL), others may not, or they may claim it and not have it. The model number of the device should be listed on UL's website if they are accepted, so you don't have to take the vendor's word for it.

Thanks for the info.
I have a simple Chinese circuit breaker waiting to be mounted but it seems wiser to order such a solar circuit breaker.

I suspect most of these breakers are made in China, the real question is does it have the appropriate DC ratings, and is it built with good quality and testing.

Note also that DC breakers usually have a polarity, they often use a magnet to force the disconnect arc into a quenching structure, and the current flow direction is important to make this work properly.

It looks like we need one of the guys in the business to step up and do some tests of the various fuses and breakers.

For myself the generic auto fuse plus a couple of magnets works for me but...........my setup is "only" a BBS 02 and 20 amps/48 volts. There are a lot of things to consider when clearing a fault on a DC circuit....rate of rise of current, prospective final current, the inductance on the circuit, open circuit voltage etc. Certainly some of the higher kW installations will need a much more serious look at their requirements before selecting what protection is required.
As I understand it my battery BMS has an electronic O/C trip built into it which should do a better job than anything else....on the proviso that it doesn't fail as well !
I have spent some 40 years designing and commissioning in the heavy electrical business where DC breakers rated to carry 10kA and clear 100kA are not uncommon and a pair of ear plugs and ear muffs are a good idea when one of these breakers clears.

The cells in my battery aren't the best but at ~100 milliohms internal impedance per cell and four in parallel with a nominal 48 volts of battery, prospective short circuit current of some 48 / 0.025 = ~ 2kA would be the sort of current that could be expected if the BMS doesn't work or is not fitted. With a very low inductance of only a single loop of wire ( call it 10 uH to keep it simple ) then the circuit rate of rise of current would be ~5 amps per micro second (very fast!). Ultimately its the wire size that is going to limit the current that flows (too thick a wire might not be the best thing !).

DC circuit breakers come in many flavours from the very simple thermal tripped breakers to the ones with magnetic trips, blow out coils, blow out magnets, arc shutes etc etc. The very simple thermal ones are probably not the best idea since they have inbuilt resistance ( the heater to drive the bimetallic trip) and they don't normally have more than a pair of contacts that open under a spring load. (which can often weld!) Units that have magnetic trips are better in that the more overcurrent the faster they work.

The voltage rating of a fuse or breaker is a complex issue. But one that largely revolves around the length of arc generated by the action of the fuse blowing or the contacts of the breaker opening. So in the simplest view the longer the gap between the fuse terminals or the spacing between the open contacts of the breaker the better the voltage rating.
Using a magnet or a blow out coil to force the arc into a longer path improves things dramatically.

Good domestic house AC breakers normally rely on a combination of the voltage passing through zero and a blow out coil/arc shute and will often carry a lower
DC voltage rating as well. They normally come in several fault current ratings and if one chooses a 5 kA rated one ( and that has a magnetic trip ) with a 50~80 volt DC rating then it will be a fairly good bet that it will work in this case.

From an industry perspective a correctly selected fuse usually offers better protection in these sorts of circuits, but are more of a pain to replace and fault find the circuit than breakers. So the choice is open.

Regards bob

Testing would be excellent, we have empirical testing showing a lot of ebike fuse systems failed due to too much heating in the fuseholder contacts. Not to say they cannot work, but the commonly available inexpensive moderate size fuseholders aren't up to the ebike duty cycle. They work fine in audio systems with high peak to average power, but ebikes have a different duty cycle characteristic, causing the holders to melt and overheat on heavy hill climbing or high speed running.

The solar AC/DC circuit breakers are a much closer fit than a household breaker, and have been successfully employed by many. The ones I have looked at are equipped with both thermal and magnetic trips, contain DC arc handling features, and the spec sheets are available.

If you are volunteering to do some testing, we'd all love to see your results. The testing I've done is operational, not bolted faults, so I don't have much direct data aside from the one slow thermal overcurrent trip, which worked fine, and reset fine as well. The Borg, which has such a breaker, has an 18S4P Turnigy Multistar 66V 32AH pack which is 75V hot off the charger, no BMS, and 7 gauge wiring with hydraulic crimped interconnects and XT90 connections that are fairly short and direct, so it is capable of considerable operational and fault current. Due to the capacitor bank in the controller I don't think the circuit breaker will ever see the inductance in the motor or much inductance at all beyond the wiring itself, so the potential inductive energy will probably not be a major factor in the post-trip arc, and we won't have the solar characteristic of sharply rising panel voltage when the load is cut off.

The feature I'm most concerned about (assuming they've properly handled the fault current interruption, which is well specified and tested) is the initial charging current for the motor controller caps, as these are the type of high current peaks that could weld contacts, and I'm not sure that solar systems have such sizable capacitors to charge at turn-on, plus their energy source (panels) is current limited unlike our battery source. However some solar systems have battery banks, so perhaps these breakers are also used for that). I don't recall seeing a spec covering the allowable peak very short-term current on the breakers, but that would be a good number to know. I personally try not to put a lot of cycles on the breaker, and if I did I would put a manual precharge current limited bypass on it to preserve the contacts. However there are folks who have essentially tested this by using their breakers as on-off switches, and I have not seen the failure reports that you do see for most regular switches used in this service. At the higher voltages, they don't last long before their contacts weld closed.

Yes, well, no fine (a south Africanism, I spent 40 years there). And I'm not volunteering anything yet, my home project list is still deep, only exceeded by the house bucket list.

I have been a little bit careful about looking at breakers used for PV since they are generally operate on low fault currents but potentially high open circuit voltages. The common house breaker (at least the ones spec'd for 220/380 volt operation) are a lot better than one would think.

Due to the capacitor bank in the controller I don't think the circuit breaker will ever see the inductance in the motor or much inductance at all beyond the wiring itself, so the potential inductive energy will probably not be a major factor in the post-trip arc, and we won't have the solar characteristic of sharply rising panel voltage when the load is cut off.

Agreed the major issue is getting rid of or limiting the energy in the fault and that will be basically the prospective fault current. I get worried when folks talk about putting larger and larger current rated fuses into circuit as a "good" thing. What we want is to get rid of the current fast before things get out of hand. The comment about fuse holders melting is important..... we want fuse holders that will stay cool but fuses that will blow quickly.....old , old, story of fuse discrimination. I once did work on a small DC rated (5 amp) breaker working at 110 volts DC and was amazed at how much faster the breaker cleared as the fault current increased to the kA range. The larger auto fuse holders may work well being rated for 50 /63 amps for those who really need the amps.

Breaker contacts do weld ! and its one of the things that sort out the Indian and Chinese breakers from almost every one else. Proper mechanical design that move the contacts as the connection is being made plus other goodies like precious metal tips to the contacts etc. sort the ones that work from the others. You only get what you paid for , if you are lucky ! I once worked for a major breaker and contactor manufacturer. They made the mistake of moving a production line to china. A couple of years later there were fakes on the market sold in identical housings but with poor bits inside. So even branded parts can be suspect.

If I get motivated enough it will be to charge a "fat" bank of caps and turn on a big thyristor ( 1~2kA rated) and try to make some fuses explode. Looking at volts and amps is ok for me but the fast video would be an issue.

Bob

Any part number recommendations for 48V/52V and 50A ?

Click to expand...

I've had very good luck with breakers from this company. Giant selection.



https://www.bluesea.com/products/category/14/Circuit_Breakers



Amazon carries some of thier products too.



https://www.amazon.com/s/?ie=UTF8&keywords=blue+sea+breaker&tag=googhydr-20&index=aps&hvadid=34206353027&hvpos=1t1&hvnetw=g&hvrand=371673150674127064&hvpone=&hvptwo=&hvqmt=e&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9030938&hvtargid=kwd-6182837693&ref=pd_sl_21t2zxgllj_e

Alan B said:

The common values near there are 50 and 63 amps, I would go with 63 to have a little headroom. I used a 63 amp unit on the Borg, with the Sabvoton set for 80A. There is a time and temperature element, so a 63 amp breaker will hold 80A for a minute or two. It wasn't enough for a big hill, so I went to a 100A unit.

Here's one that says 50A but shows a 63A unit in the photo:

https://www.amazon.com/MidNite-Solar-Breaker-150VDC-MNEPV50/dp/B00BSYSTN2

And here's the 63A version that I have used:

https://www.amazon.com/MidNite-Solar-Photovoltaic-Circuit-Breaker/dp/B004EQRAQS

The 50A should hold at 50A for awhile, but if you operate at 50A for long periods of time then the 63A unit might be a better choice.

There are also some generic units that might be okay, check the ratings to see if they are accepted by the appropriate NRTL (Nationally Recognized Testing Laboratory) for your country (for those outside the US). Major products will have this rating (example: UL), others may not, or they may claim it and not have it. The model number of the device should be listed on UL's website if they are accepted, so you don't have to take the vendor's word for it.

Click to expand...

Those breakers are amazing and do work very well, even for dozens of events without degrading breaker performance, but it's not without caveats. Vulnerable to corrosion and dirt/sand impacting function, and 10x-100x slower acting than a fuse designed around protecting silicon parts. Both are neat and have there place, I put a DC solar breaker right off the pack in my wifes NXY frame bike (a 125A DC solar breaker version) and use it as the pack disconnect for service and it's been great because it's kept clean and dry inside the frame and nothing too delicate is being protected with it.

Just a side-note. ES member teklektik built a 2WD Yuba Mundo, and he is a professional electrical engineer. He conducted lots of research, and he chose the "Blue sea" breaker listed above, used for many years with satisfactory results.

https://www.electricbike.com/teklektiks-dual-motor-yuba-mundo/

Yes the blue sea seems a nice breaker and should work fine. A choice from an old employer would be from Schneider, they do 60 volts per pole. But note that you can get ganged poles from them so two poles wired in series = 120 volt DC rating. (and of course 3 would get you 180 volts !!) Note these are what I meant by a domestic AC breaker.
View attachment Pages from Acti9_Communication_System_Catalog.pdf

Regards Bob

Blue sea doesn't actually make breakers or fuses, they just put a sticker with there name on parts other folks made, but typically pick quality stuff to put there name on.

Again though, if your application is OK with short-circuit events taking 10-100x longer, then the breakers are awesome, if you want speed it's still a fuse.

Circuit Breakers are renewable, just turn back on. Fuses are non-renewable, must be added to the landfill.

Neither are really fast enough to protect the electronics, either is fast enough to keep from melting all the wires and connectors.

Excellent example of which there are many in the ES library. Thanks for the details.

Weird. I made this post after Teklektik's below, and somehow it slipped in above it. Strange. Anyway, it refers to the post below...

Alan B said:

Neither are really fast enough to protect the electronics, either is fast enough to keep from melting all the wires and connectors.

Click to expand...

+1

liveforphysics said:

...[breaker] and 10x-100x slower acting than a fuse designed around protecting silicon parts...

Again though, if your application is OK with short-circuit events taking 10-100x longer, then the breakers are awesome, if you want speed it's still a fuse.

Click to expand...

This is certainly true if using an expensive high-speed silicon protection fuses. But for ebike use and more modestly priced conventional fuses for basic wiring protection it becomes a specsmanship issue where specific parts choice plays a significant role in speed. Spec sheets certainly show that quality conventional speed fuses can beat a magneto-thermal breaker - although there is measurable overlap in response times. Neither will protect electronics but either can protect wiring from short circuit events. Between ubiquitous cheap fuses and the timing overlap with quality breakers, the speed picture is not as clear in real world use as we might like to believe.

FWIW, I had 50A Maxifuses in my packs to protect against mishap even while the packs were disconnected and a 50A Blue Seas breaker on the frame to protect the harness and serve as an ON/OFF switch. Marine and solar breakers are (typically) intended specifically for switch service - unlike many breakers - and so offer a nice compact disconnect for those who wish it. I fully expected the fuses to do the heavy lifting protection-wise and the breaker was backup protection and primarily a convenient on/off switch.

Not the case at all. I had a number of at least 175A faults (according to CA before it shut down) that consistently tripped the breaker. The fuses never blew or showed any visible sign of distress.

I eventually had the MaxiFuses melt the holders at the friction connections and switched to the bolted connections of miniANL fuses. The higher available ratings of ANLs also gave the opportunity to bump the rating to a more comfortable 160% of expected max. No melting fuseholder problems since but also no short circuit events to compare the miniANL response times with the breaker.

• I generally object to anecdotal evidence as 'proof' of anything, but this seems to illustrate at least a single situation where two serial MaxiFuses and a Blue Seas breaker of the same rating repeatedly showed the breaker to be unequivocally faster in short circuit events. The rating was chosen close to the operating current so there was not a lot of current margin to overload with an associated increase in trip time because of overspeced ratings. The wisdom of using MaxiFuses vs other choices, the quality of the fuses, or the particular subcontractor that produced the breaker for Blue Seas can all be called into question - I'm just reporting some experience.

So - just a Tale of Two Fuses for your enjoyment - make of it what you will...

About magneto-thermal breakers...

Thank you to everyone who has contributed, from links to known-good product recommendations through education on the theory of DC circuit interruption technologies.

Couple years back I inadvertently shorted a 6s3p battery of A123 26650 cells by dropping a piece of balance lead wire across the contacts. Experience was over in a moment – a section of the shorting wire vaporized with a flash-bang leaving only smell of ozone and my shout of “Woah!!!” Reconstructing events afterward I realized that even that small battery had a theoretical impulse discharge of 19V * 360A = 6.8kW. Convinced me of the power inherent in bike-level batteries, and that it was probably a smart idea to install a fuse / breaker in bike power systems. We really do ‘ride the lightning’ even in mid-power bikes. But I couldn’t find much solid specific advice, so I never got around to implementing fuse / breaker on any of the bikes I’ve built.

I do hope to correct that!

What about this automatic breaker 63A and ~4$ 125V DC rated.

If the spec's are true then it looks very good. Permanent magnet blow out style breakers are the "right" stuff for DC battery systems. The only Caveat is........don't get the polarities wrong ! if you do it will likely melt after a couple of bangs. I tend to suggest ones without permanent magnet magnetic blow outs to the general population....someone will always get the polarity wrong and blame everyone else.

Bob

For a few bucks more, I like this series from Midnite Solar.
More compact, easy panel mount, and stud terminals on the back instead of the top/bottom clamping connections.

Midnite Solar MNEDC 5A thru 100A

What about 10x38 1000V DC rated solar fuses? Largest ive seen are 25A but ~1$/piece.
Never seen anyone added servo control for circuit breaker? Why? because high current latching contactor/relay is damn expensive.
Im gonna do it.

teklektik said:

For a few bucks more, I like this series from Midnite Solar.
More compact, easy panel mount, and stud terminals on the back instead of the top/bottom clamping connections.

Midnite Solar MNEDC 5A thru 100A

Click to expand...

Nice find! We used to use those in chassis, many moons ago. Good form factor.

Oh yes !, for spec junkies like me these look very good. but its the difference between $20 and $4 for the Chinese stuff. As always you get what you pay for if you are lucky!

Regards Bob