Another Ebike conversion kit death

[Az.]

go on facebook or twitter and start calling every one a paki

That would be disrespectful.
Though I guess I might get easily imprisoned for sedition

 

[chris667]

Do you ban your clients from storing eBikes or thier batteries in the properties you rent ?

I have only had one so far. She has an Asgard shed, supplied and fitted by me, in what would be her parking space. Cost about a grand, plus a planning application.

I would not allow any bike (motorised or manual) to be stored in a communal area of any rented property. The one that caught fire at the start of this thread was basically a moped, I wouldn't let them store a moped in the hallway either.

 

[Plas man]

No doubt China man will print sticky label’s so the sellers can stick on old battery’s still in stock .
As for fire’s in the old days I’ve seen a couple of Tamworth manufactured Reliants‘s go up in flames .

 

[AndyBike]

you cant leave a bosch batt as the bms will drain the pack to nothing then brick the bms

Over a barrel

 

[StuartsProjects]

No doubt China man will print sticky label’s so the sellers can stick on old battery’s still in stock .

And to cut costs they will sell them on a sheet that also has the 250W motor stickers.

I wonder if you can get fireproof stickers ?

 

[soundwave]

Over a barrel

the frame batts can now be recelled but they need to be working and charging if the bms bricks its self then its a bin job

 

[clix]

I have bought several Yose power packs. A reasonably big brand who try and do things right, but when I have taken them apart, there are a lot of things I don't like from a safety point of view. Yose power are not alone in this. This has led me to dismantling most of my commercial packs, replacing the BMS and wiring, connectors, switches, fuses and then re-packaging.

Do you have any recommendation on what Bms to use? I bought a battery in hailong case from pswpower, that is supposed to have Samsung cells. I am not sure if I trust that Bms, and it probably lacks balancing. I would like to go with BT option, to individually check each block voltage and make sure that charger doesn't overcharge. Any recommendations? Also maybe tutorial where you explain what exactly you change/improve?
Thank you

 

[WheezyRider]

Do you have any recommendation on what Bms to use? I bought a battery in hailong case from pswpower, that is supposed to have Samsung cells. I am not sure if I trust that Bms, and it probably lacks balancing. I would like to go with BT option, to individually check each block voltage and make sure that charger doesn't overcharge. Any recommendations? Also maybe tutorial where you explain what exactly you change/improve?
Thank you

I'm a bit short of time this week, so I can't go into too much detail. But if you are wanting to modify your pack to improve it, here are a few pointers. @Nealh should be able to advise on a good choice of BMS. I have bought some, but I think he has tried a much wider range. I like to get something that is a waterproof sealed module, rather than an open board that can suffer water damage. Also it must supply a decent amount of amps above what you will need for your application. I go for basic, none of this blue tooth and programmable stuff. Great toys, but I prefer to keep it simple. Bluetooth running even at low current can kill your battery if it is not used for a long time. Plus, with fewer features, they are cheaper.

Set your charger to 41V to increase battery life. But, fit a way of interrogating individual cell bank voltages, so you can keep an eye of things going out of balance. I've never seen this with my packs, but it is a potential problem (I think it mostly occurs when poorly matching cells have been used to make the pack in the first place. A good manufacturer will choose matching cells and these can go hundreds of cycles without going out of balance).

Bear in mind, if you do put in a multiport for bank interrogation, if you accidentality short between pins you get full current going between the cell banks, which can be so much that you get a blinding flash and the melting of the port pins, or even the tips of multimeter leads! It could also melt the sense wires going to the BMS, so on my most recent build I put in miniature 100mA polyfuses on each of the sense leads, so if the worst happens, you don't do any damage.

There is the question of whether you go for a combined power out and charge port BMS, or a BMS with separate ports. Each has its advantages and disadvantages. I have tended to go for combined single port, because it keeps the battery pack itself as simple as possible. But you then need to think, how will I do charging? Do I want to be unplugging the output cable from the bike every time I want to charge? XT connectors are only designed for 1000 or so disconnects/reconnects. I have added a cable which has the power port connector and a separate charging port to get around this problem. Use a female connector on the battery output, to reduce the chance of shorts. For connectors, use something decent like XT60 or even XT90.

As for wiring, keep it as short as possible and use a decent gauge, depending on your application and how much current you need. For fuses, I've gone over to fuses that can be screwed down rather than the cylindrical ones you normally get which rely on spring fitting. These can corrode, or get worn due to vibration and make a poor contact. I did experiment with 30A trip circuit breakers, but I am not convinced they work as they should.

One of the biggest issues is what you use for an on off switch. Take the Yose power silverfish. It comes with a keylock main on/off switch. This is great, except that in no way was this designed to handle big currents, especially the surge you get when first sending power to a controller at switch on. I have found these switches stop working after a while as the contacts get burned away. Finding a decent high current DC switch is not easy. I think some manufacturers use a small switch to activate a MOSFET to turn the battery on. I am not so keen on this. For one thing, adding another MOSFET introduces another voltage drop and wastes power depending on the RDS on of the MOSFET, then, if the MOSFET fails in a short circuit mode, power is still connected.

I have been lucky in finding some army surplus heavy duty DC switches, but my supply has run out now. I am thinking for the next build I will use an XT90 connector in the power line from the pack. It will be a female socket and then the male connector will be shorted and embedded into a 3D printed fob, to make a kind of key. Push the two together and power is connected, pull it out and power is cut.

So as you can see from this brief overview, there is a lot to think about. I did not even discuss sealing the pack from moisture ingress, adding a temperature sensor, or how to package the battery...

 

[clix]

Huh, thank you for this lengthy reply.
Very good pointers, it will help me focus on most important things.
I will save it as a refference.

 

[WheezyRider]

I also need to mention, before you do any mods to your pack make sure it is discharged first. Even then, treat it as you would an unexploded bomb. Work on a clear, insulating surface. Keep all metal tools etc tidy and out of the way. Don't work on batteries if tired or distracted etc. Do the work somewhere where a fire could be contained should the worst happen, preferably not in the house. Any advice is for reference only and any work you do is at your own risk - if you are not sure about something, get further advice before proceeding!

 

[WheezyRider]

Something else I forgot to mention. If you change the BMS in your battery pack, you also need to look at your charger. I think it's not a bad idea to do this anyway, as I've seen a lot of chargers (even from well known suppliers) being sold that supply more than 4.2V per cell, which is the maximum recommended.

Sometimes, this is deliberate as a blocking diode may be added to the charging circuit in some BMS modules with a separate charging port, which drops the supply potential slightly (eg 0.5 to 0.7 V). This can be observed by seeing if you can measure a voltage on the charging port when no charger is connected. If you can't see the battery's voltage, there is a blocking diode. If you then used the same charger designed for such a battery with a single port BMS, it's likely you would end up overcharging. Not all dual port BMS modules have this blocking diode.

As a battery is only as good as it's charging system, I would recommend checking the output of your charger even if you are not going to mod the pack. This is not as easy as just connecting up a multimeter. The differences in voltage are very small (say 0.1V in a total voltage of 42.1V, which is not a good voltage for a 10s pack) and most multimeters only have an accuracy of +/-1% at best (so in this example the reading could be anywhere between 41.679V and 42.521V), so it's important to use a multimeter that has been calibrated properly and not assume you are safe because you see 41.7V on your multimeter. Most of the multimeters available these days are made in China and even with a quoted +/-1% accuracy in the 200V range - can you trust it? The inaccuracy could be a lot more than 1%.

Imagine if you don't mod your pack, but you "upgrade" the charger, as you need a faster turnaround between charges (delivery rider?) so you buy a 5A rapid charger from one of the online market places. If you can't check the output voltage accurately, you will be asking for trouble.

 

[AndyBike]

I wonder if you can get fireproof stickers ?

Given the thread reason, dont you think that a little crass ?.

I know the unscrupulous are more to blame here.

 

[guerney]

I wonder if you can get fireproof stickers ?

" the carbides of hafnium and tantalum can withstand the highest temperatures—close to 4,000 degrees Celsius—without melting. "

Inexpensive synthesis of high-melting-point carbide materials

Researchers from Skoltech and Tomsk Polytechnic University have used an unusual technique from the aerospace industry to synthesize hafnium-tantalum carbide, a hard-to-melt material for coating electrical and mechanical components that operate under extreme conditions. Inexpensive and effective...

phys.org

 

[Woosh]

I also need to mention, before you do any mods to your pack make sure it is discharged first. Even then, treat it as you would an unexploded bomb. Work on a clear, insulating surface. Keep all metal tools etc tidy and out of the way. Don't work on batteries if tired or distracted etc. Do the work somewhere where a fire could be contained should the worst happen, preferably not in the house. Any advice is for reference only and any work you do is at your own risk - if you are not sure about something, get further advice before proceeding!

I can see that you take a lot of precautions but I would like to add a word of caution for anyone who considers repairing or modding their battery and or chargers: product liability insurance will not honour their contract if you modify the product without being directed by your suppliers. You will be on your own. Batteries and chargers are sealed for safety reasons.

 

[Bonzo Banana]

This is for batteries manufactured between 4 to 6 years ago. How is that relevent here.

I think its a fair comment based on indicating that even high cost premium ebikes still fail despite making up a tiny percentage of ebike sales, here you see a model with a high level of failure.

Clearly that ebike in the news report has a direct drive hub motor and what appears to be a very large hub motor perhaps 2000W or more depending on if there is a controller to supply that. It's going to be easy to stress and damage all but the highest capacity battery packs but there is also a greater risk with higher capacity battery packs because there is soo much more combustible lithium material. True 250W motors that maybe peak at less than 400W and given a battery of say 400Wh should be much safer.

That is not the market though, we have both kit and so called legal pre-built 250W ebikes that get close to 1000W in consumption and heavily stress their batteries getting close to 100Nm of torque. Maybe battery capacity and torque should be limited to create a more stable long life ebike platform. Maybe limited to 40Nm. At the moment many ebike riders are obsessed with speed and torque and this leads to high discharge rates in lithium cells which over time can cause such issues.

I do think there is an issue with ebike kits mainly direct drive hub motors and mid-drive where people simply don't use the correct battery to suit such motors. It could be home-made, under capacity or lots of other issues. If a motor peaks at 30A then you should have a battery that easily sustains 30A without issue because as we know there is reduced capacity with age and the battery spec could be over-stated.

Perhaps we should build ebikes a bit stronger and simply use lead acid gel type batteries, much cheaper for the same capacity and no fires. It's funny how in India and China where ebikes commonly have lead acid batteries consequently they have much higher safety for these very cheap entry level ebikes. Typically these countries are not known for high safety standards but here in the UK and many other countries we are obsessed with having light ebikes and lead acid gel type batteries are just not acceptable on ebikes.

 

[WheezyRider]

I think its a fair comment based on indicating that even high cost premium ebikes still fail despite making up a tiny percentage of ebike sales, here you see a model with a high level of failure.

Clearly that ebike in the news report has a direct drive hub motor and what appears to be a very large hub motor perhaps 2000W or more depending on if there is a controller to supply that. It's going to be easy to stress and damage all but the highest capacity battery packs but there is also a greater risk with higher capacity battery packs because there is soo much more combustible lithium material. True 250W motors that maybe peak at less than 400W and given a battery of say 400Wh should be much safer.

That is not the market though, we have both kit and so called legal pre-built 250W ebikes that get close to 1000W in consumption and heavily stress their batteries getting close to 100Nm of torque. Maybe battery capacity and torque should be limited to create a more stable long life ebike platform. Maybe limited to 40Nm. At the moment many ebike riders are obsessed with speed and torque and this leads to high discharge rates in lithium cells which over time can cause such issues.

I do think there is an issue with ebike kits mainly direct drive hub motors and mid-drive where people simply don't use the correct battery to suit such motors. It could be home-made, under capacity or lots of other issues. If a motor peaks at 30A then you should have a battery that easily sustains 30A without issue because as we know there is reduced capacity with age and the battery spec could be over-stated.

Perhaps we should build ebikes a bit stronger and simply use lead acid gel type batteries, much cheaper for the same capacity and no fires. It's funny how in India and China where ebikes commonly have lead acid batteries consequently they have much higher safety for these very cheap entry level ebikes. Typically these countries are not known for high safety standards but here in the UK and many other countries we are obsessed with having light ebikes and lead acid gel type batteries are just not acceptable on ebikes.

I would not recommend going back to lead acid, they still can be dangerous. You've also got the extra weight and reduced range/cycle life - not factors likely to help increase Pedelec use. LiFePO4 is probably the way to go.

 

[WheezyRider]

I can see that you take a lot of precautions but I would like to add a word of caution for anyone who considers repairing or modding their battery and or chargers: product liability insurance will not honour their contract if you modify the product without being directed by your suppliers. You will be on your own. Batteries and chargers are sealed for safety reasons.

True, but I think these guys who want to go around on illegal 2000W motorbikes and pretend they are Pedelecs are not going to be thinking too much about warranty

Most people on here who want to open up batteries are usually those who have a pack that is not working properly, but is outside of warranty. It's important to keep people safe, but also not to encourage throwing stuff away that could be potentially fixed without difficulty.

 

[Bonzo Banana]

True, but I think these guys who want to go around on illegal 2000W motorbikes and pretend they are Pedelecs are not going to be thinking too much about warranty

Most people on here who want to open up batteries are usually those who have a pack that is not working properly, but is outside of warranty. It's important to keep people safe, but also not to encourage throwing stuff away that could be potentially fixed without difficulty.

I personally don't think lead acid gel type batteries are that dangerous. Lots of vehicles have them, motorbikes and mopeds for their electrical circuits and I can't remember hearing of any issues but admittedly haven't really looked.

Bosch make their batteries very disposable. Any anolomanies and they shut down and never work again. Some of this is safety related and some of it is because its proprietary and they want to force people to use their batteries and not enable cheaper third party batteries. However Bosch batteries still fail through the high discharge of their mid-drive motors. I've seen them taken apart and the cells have ignited within but the way they are sealed in glue denies the cells oxygen to create a real fire. So there is improved safety but the failure is still there. Some third party batteries for general ebike use also have very high quality construction with safety in mind. Ultimately we need the certification standards to be improved so no battery can be sold which doesn't meet higher safety standards. However a personal import from Aliexpress, ebay or even an Amazon marketplace seller where the seller is based in China bypasses such legislation. You are importing the product yourself as a private individual where as normal certification standards are regulated by applying them to the manufacturer or importer in this country or Europe.

For example if Poundland want to sell sunglasses even at a £1 they have to be fully certified and as Poundland import many of their products themselves they have to provide the right certification. However the same factory in China may make non-certified glasses that don't fully protect your eyes that you could import yourself and damage your own eyes using them.

Basically any improvement in legislation won't really apply to Aliexpress and many other sellers.

 

[Woosh]

Ultimately we need the certification standards to be improved so no battery can be sold which doesn't meet higher safety standards.

Beside manufacturers' specification sheets, there are two standard pieces of information on the cells: UN38.3 and MSDS Material Safety Data Sheets). All suppliers must have them as they are required for any shipment.
The issue arises mainly from poor quality control in manufacturing, mismatched cells and water ingress. No amount of paperwork can eradicate those risks completely. Compulsory product liability insurance is probably the most useful.

 

[StuartsProjects]

In one of those idle moments, I was wondering if a 10S1P battery would be a possibility for a real light eBike. So I put 36V 10S1P into alliexpress, and heck its amazing the bargains to be had.

You can get, for a mere £26.44, with free shipping, get an alleged 36V 30Ahr battery, what a bargain.

Even if it were 3Ahr, which seems more likley, why only £26 ?