Building Electric Moped. Battery Questions

[pedbuild]

Hello,

I want to build an electric moped/ebike to get it fully registered using the Enduro frame.

http://imgur.com/a/ljLWAGv

Since this will be used within the city. It only needs to reach a maximum of 20MPH most of the time and 30MPH in certain areas. I realise that I will need a hub motor of about 750W for the rear wheel.

I need quite a bit of range since I will be using this all day. I was thinking around 50-60AH at 48V.

Would this simply be the case about upgrading the controller for higher wattage such as 2000W+

Since the controller for the 750W only says it supports 30A.

Is there any issue using a 2000W+ controller on a 750W motor since I really need the big AH battery?


Or would I need to make 2 smaller batteries and use a switch and use them 1 by 1?



The issue is I don't want to use a say 2-3000W motor because that can simply drain the battery too fast and I will not even be going anywhere near its full speed.

If you have any questions, do let me know.



The battery I thinking of making is going to be using these 26650 Cells at 5000Mah each

https://www.aliexpress.com/item/3279335 ... hweb201603_

The BMS will be rated for 80A using a Daly Bluetooth BMS.

 

[vfr400]

You're getting amps and amp-hours mixed up. You can have as many amp-hours as you want with whatever system you want. It only affects how far you can go, not the power.

30mph takes requires 1000w of output power or 1300w of input power, which at 48v is about 30 amps. If you use 30 amps, your 60Ah battery would last 2 hours, but you wouldn't be using it all the time.

Choose a motor with the right speed characteristic. For a 48v bike with a 26" wheel, you'd need 400 - 450 rpm for 30 mph, which is a Kv of around 8 rpm/v.

It would be much cheaper and easier to buy a ready-made bike or scooter.

 

[vfr400]

I haven't checked out those cells yet, but I think you should do some proper research on them. There's no way in the world that you can get a 5000mAh 50A battery for less than £2, otherwise every ebike in the world would be using them instead of none of them.

See what they say on the Candlelight forum, camping forums or battery forums, where they test such batteries and give you the actual results.

 

[Nealh]

The cell are crap for EV use, sags heavily at 15a to 3.2a in 15 mins use.
Limit them to 3a per cell in parallel per group and the cell achieves a bit over 90 mins run time.

At 30a the cell is pure crap sag is instant to 3.5v , it hits 3.2v in about 2.5 - 3 mins.
One would need 15 cells in parallel for an average 2a discharge per cell for usage time of 2h.20m.
Test of LiitoKala INR26650-50A 5000mAh (Cyan) (lygte-info.dk)

Mooch hasn't even bothered to test I can see why.

 

[vfr400]

I bought 10 to test, as I need a 7S 5Ah battery of a tiny low power fold-up bike I have. They were £28, then it added £7 shipping, then £7 tax, so they ebded up at £42.

Hot LiitoKala lii-50A 26650 5000mah lithium battery 3.7V 5000mAh 26650-50A rechargeable battery suitable for flashligh - AliExpress 44

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

I don't mind risking £42 for the sake of science, and I need to get rid of some money because I have too much. Soon, we'll know if they're any good.

 

[pedbuild]

I haven't checked out those cells yet, but I think you should do some proper research on them. There's no way in the world that you can get a 5000mAh 50A battery for less than £2, otherwise every ebike in the world would be using them instead of none of them.

See what they say on the Candlelight forum, camping forums or battery forums, where they test such batteries and give you the actual results.

Hi,

Thanks for the above replies.

So, yes I do believe I was getting mixed up between A and AH. Also, these batteries, you can check on YouTube too, they do actually have 5000mAh of capacity (Actually more, most are around 5500) and I've tested that with a simple drain/charge test on them. I used these to build a 36V 20Ah pack in the past so thought they would be alright for this build.

Since I see both you and @Nealh wouldn't really recommend them. Would you suggest this:

LG INR21700-M50T 4850mAh - 7.3A - 21700 & 20700 - Li-ion - Rechargeable batteries | NKON

.

eu.nkon.nl

It's an LG 21700 with 7.3A Discharge and 4850mAh capacity.

Can you tell me if my calculations are correct then since I did some more research on A and AH.


Target: 48V 60Ah Battery

13S 12P

Therefore: 48V at 60Ah at 87.6A (7.3 x 12)

Giving me 2880Wh

Therefore I need a BMS rated for at least 90A+ and not 60. I would also need a controller that supported 90A too then.

What wattage of motor would you guys suggest for me then? Since I only have 2.9kWh. I do only want to go 20MPH and preserve as much battery as possible for work in the city and going to uni. I will be using this in the UK London and since traffic is always there and many turns and not so many straight roads I doubt I will have any chance of a long throttle ride without hitting a red light to stop.

Please do let me know if I'm correct or not on the calculations above.

@Nealh Are those LG batteries with the link I posted any good in your opinion. I see that then these 26650 cells may not be ideal for me if I'm understanding your post correctly. If these LG cells aren't good too what would you suggest for me to get them or what did you use to build your kits?

Thanks for the help.

 

[pedbuild]

The cell are crap for EV use, sags heavily at 15a to 3.2a in 15 mins use.
Limit them to 3a per cell in parallel per group and the cell achieves a bit over 90 mins run time.

At 30a the cell is pure crap sag is instant to 3.5v , it hits 3.2v in about 2.5 - 3 mins.
One would need 15 cells in parallel for an average 2a discharge per cell for usage time of 2h.20m.
Test of LiitoKala INR26650-50A 5000mAh (Cyan) (lygte-info.dk)

Mooch hasn't even bothered to test I can see why.

Hi. I replied to you in the above post. Thanks

 

[Nealh]

The Samsung 50E looked good on the cycle aging tests carried out on ES over a year ago.

 

[vfr400]

What I said is that that you wouldn't get a 50A 5000mAh battery for less than £2. These cells are roughly equivalent in size, price, weight and performance to a pair of other branded 2500mAh cells, so noy a lot of advantage except for the convenience in welding up a pack with fewer welds and strips. I'll reserve full judgement on them until I've tested them. I can imagine that there are a lot of fake ones around already. At only £4.20, It wouldn't surprise me if mine are fakes. I've seen them much more expensive than that.

 

[Nealh]

If you used the liitokala's before what was the battery specs ?
I only saying they are crap by looking at the graphs I see from unbiased testing. I can't believe in the 50a rating.

There aren't any 21700's that can match the 26650 for current and capacity yet, most of the 5000mah cells are sub 10a.

 

[pedbuild]

The Samsung 50E looked good on the cycle aging tests carried out on ES over a year ago.

Found the cells:

Samsung INR21700-50E 4900mAh - 9.8A - 21700 & 20700 - Li-ion - Rechargeable batteries | NKON

.

eu.nkon.nl

Hi,

So since those are 9.8A each and I will have 12P I would need a controller effectively for 120A+ then? The only ones out there I can see are the Sabvoton and that's really it. Most other kits only support around 30-40A

 

[pedbuild]

What I said is that that you wouldn't get a 50A 5000mAh battery for less than £2. These cells are roughly equivalent in size, price, weight and performance to a pair of other branded 2500mAh cells, so noy a lot of advantage except for the convenience in welding up a pack with fewer welds and strips. I'll reserve full judgement on them until I've tested them. I can imagine that there are a lot of fake ones around already. At only £4.20, It wouldn't surprise me if mine are fakes. I've seen them much more expensive than that.

Thanks will look forward to those. I did find basically the 50E suggested for just over £6 so will opt for those due to the brand, performance etc but since I'm doing this 6 months down will have plenty of time to research. Also, was I correct on the prior calculations just want to ensure I'm on the correct path and not being over/under powered. Thanks

 

[pedbuild]

If you used the liitokala's before what was the battery specs ?
I only saying they are crap by looking at the graphs I see from unbiased testing. I can't believe in the 50a rating.

There aren't any 21700's that can match the 26650 for current and capacity yet, most of the 5000mah cells are sub 10a.

The battery I built was 36V @20AH 10S 4P using those 26650 Liitokala Cells. I believe the 50A is just peak, but sustained wouldn't even match that at all. So yeah I see your point. Not sure if that was ever tested to be true but even then it wouldn't be for a long sustained time.

 

[Nealh]

You have set your eyes on 48v one could go 52v if a suitable one can be got.
At 48v a 30 or 40a controller should be ample, the bikes finished weight might need to be known as it will be a factor.

 

[Nealh]

Found the cells:

Samsung INR21700-50E 4900mAh - 9.8A - 21700 & 20700 - Li-ion - Rechargeable batteries | NKON

.

eu.nkon.nl

Hi,

So since those are 9.8A each and I will have 12P I would need a controller effectively for 120A+ then? The only ones out there I can see are the Sabvoton and that's really it. Most other kits only support around 30-40A

You need as many cells in parallel as you can afford and fit in the space the battery is to go in. with 12P although in theory it can can provide 120a your never going to need it and should just look at a max 40a controller and a BMS that can cope with 50a continuous for some head room. The reason to have more cells in P is so that they aren't put under stress and are only drawing 3 - 3.5a per each cell.

Tbh you are best posing your question on ES for the guys there to give advice as that is there niche, they tend to use bigger motors like 3 - 5 kw but you will get good advice.

 

[Nealh]

My most powerful bike is the one in the avatar, the Swizzbee.
I see up to 32a in low gear and ascending from the brushed 270w motor, controller is 24v and can take about 80a. Though typically of flatish terrain current drawn varies from 6 - 16a dependant on speed and gearing.
I can get around 57miles range but at a more sedate pace of sub 17mph. I have two batteries in Parallel.
#1 is a Panny PF 2900mah 10a cell main frame battery of 8p7s so 80a and 23.2ah capacity for 556wh.
#2 wired in parallel is a LG HE2 2500mah 20a cell rack pack of 7s7p so 140a and 17.5ah capacity for 420wh.
Together as one a 220a pack of 40.7ah or 976wh for about 18wh per mile which is quite expensive usage for me.

The PF pack can be seen in the link below. I used copper Buss for the series current flow , the X X parallel connections don't need to be great as little flows in parallel.
The battery was constructed as two separate banks of 4p 7s but wired to BMS as one. The Y lead discharge wiring though not shown were of equal length to draw equal current from each bank.
A rare piece of history - Swizzbee | Pedelecs - Electric Bike Community


Though both batteries have high current capabilities I have only got max 45 - 60 a BMS in each one as the controller can only the draw it's max rating so you don't need a BMS to match the battery max current in P but a BMS a bit higher then the controller rating.
Speed and current is the big killer for battery wh usage, once you start drawing 20a continuous or more the range will go down drastically even with a 60ah battery.

My other bikes are typically all 20a ish rating and I can get a reasonable 8-12 wh per mile as an average of 14mph by drawing 2.4 - 3 amps average.

 

[pedbuild]

You need as many cells in parallel as you can afford and fit in the space the battery is to go in. with 12P although in theory it can can provide 120a your never going to need it and should just look at a max 40a controller and a BMS that can cope with 50a continuous for some head room. The reason to have more cells in P is so that they aren't put under stress and are only drawing 3 - 3.5a per each cell.

Tbh you are best posing your question on ES for the guys there to give advice as that is there niche, they tend to use bigger motors like 3 - 5 kw but you will get good advice.

Oh right. I have been getting a bit confused then I see. So even using a 20, 30A controller would be fine too then as an example? As for the BMS in regards to ever wanting to upgrade the motor in future. If I buy a lower-rated BMS say 40A then I want to upgrade my motor to 3kW or 4kW in the future I will face problems and need to change the BMS. So to eliminate that, since the pack I am going to build will be wrapped with several layers of heat-shrink and protected would it be best to use say an 80A or 90A or 100A BMS on the battery with a 30A controller. The controllers are easily changeable, plug and play but the BMS that's a whole new job.

So basically.

90A BMS with a 30A controller so the motor can only draw a maximum of 30A. Then the 90A gives me more wiggle room in the future. Let me know your thoughts on this and if this has any downsides to it as I think this may be best. Thanks again for the help.

 

[pedbuild]

My msot powerful bike is the one in the avatar, the Swizzbee.
I see up to 32a in low gear and ascending from the brushed 270w motor, controller is 24v and can take about 80a. Though typically of flatish terrain current drawn varies from 6 - 11a dependant on speed and gearing.
I can get around 57miles range but at a more sedate pace of sub 17mph. I have two batteries in Parallel.
#1 is a Panny PF 2900mah 10a cell main frame battery of 8p7s so 80a and 23.2ah capacity for 556wh.
#2 wired in parallel is a LG HE2 2500mah 20a cell rack pack of 7s7p so 140a and 17.5ah capacity for 420wh.
Together as one a 220a pack off 40.7ah or 976wh for about 18wh per mile which is quite expensive usage for me.

The PF pack can be seen in the link below. I used copper Buss for the series current flow , the X X parallel connections don't need to be great as little flows in parallel.
The battery was constructed as two separate banks of 4p 7s but wired to BMS as one. The Y lead discharge wiring though not shown were of equal length to draw equal current from each bank.
A rare piece of history - Swizzbee | Pedelecs - Electric Bike Community


Though both batteries have high current capabilities I have only got max 45 - 60 a BMS in each one as the controller can only the draw it's max rating so you don't need a BMS to match the battery max current in P but a BMS a bit higher then the controller rating.
Speed and current is the big killer for battery wh usage, once you start drawing 20a continuous or more the range will go down drastically even with a 60ah battery.

My other bikes are typically all 20a ish rating and I can get a reasonable 8-12 wh per mile as an average of 14mph by drawing 2.4 - 3 amps average.

Yeah, I see your points and that's a nicely built battery. Since I am going for potentially being future proof with the battery best bet too would be to use Copper like I can see you have used in your battery build. More expensive but for future proof ness then can beat pure nickel.

 

[Nealh]

You can't weld copper very well unless it is thin and in a nickel sandwich, laying nickel either above or below as well allows the spot weld to penetrate if you up the joules to penetrate but if the copper receives the power directly from a diy spot welder it doesn't take very well. The guys on ES experimented as there is a very good thread on the various mini diy spot welders to buy and their usage.
Battery wise the favourite cell used by quite a few is still the Samsung 29e because of it's price and performance, performance wise the guys over there use 20 odd cells or so in parallel.

 

[Nealh]

BMS wise there is no harm in going with a higher rated one those cost is more, the main ting is there isn't a minimum current draw. One might even consider a smart BMS so one can monitor each P cell as and when you wish. The speedict Neptunes might be the best there is for this.