48v battery voltages help please

[rounds]

I know this has been covered, but I still am seeing so much conflicting information and could use some help. I'm still new to ebikes, just got my first one, and am learning. It's a 48v. I have seen multiple graphs that supposedly correlate battery voltage to percent of usable battery left. However, they are conflicting and seem inaccurate. Here are two that I have seen and have been referencing:

However, neither of these charts makes perfect sense (first chart shows 1 volt correlating to 5% sometimes and 7.5% other times nonlinearly; second chart shows 100% at 53 volts which I know is wrong).

Where I need help is here. I went on a long ride the other day and the battery died much earlier than I expected, and I had to peddle an 80 lbs fat tire 15 miles home without assist (not a good time). When I got home and put my FLUKE DMM on the battery it read 43.5 volts. At that voltage the BMC was cutting the battery, my display was showing "error 26," and that was that. Based on the first chart I should have still had 30% left but based on the second chart I was all the way down to 12% left. If I am at 12% I don't mind that the BMC cut the battery because I don't really want to ever go below 20%, but if I had 30% left than I think there is something amiss?

Could someone please help me with this? Is my battery defective, or is 43.5 volts a point at which the BMC should cut the battery? Am I missing something?

Thanks all and kind regards,
craigr

 

[JRA]

It is BMS to begin with Battery Management System. The BMS from different manufacturers will cut out at different voltage levels and apparently yours does @ 43.5. So instead of trying to make sense of it via a graph/chart just pay attention to the gauge and don't be 15 miles from home next time you run out? BTW as you might have noticed the power drops off incrementally as the voltage drops also so try and keep and hills and headwinds to a minimum at low voltage.

 

[rounds]

It is BMS to begin with Battery Management System. The BMS from different manufacturers will cut out at different voltage levels and apparently yours does @ 43.5. So instead of trying to make sense of it via a graph/chart just pay attention to the gauge and don't be 15 miles from home next time you run out? BTW as you might have noticed the power drops off incrementally as the voltage drops also so try and keep and hills and headwinds to a minimum at low voltage.

Thanks for your help. Like I said, I'm still learning.

My question really is this, is my BMS cutting the battery too soon? 43.5 volts seems too soon to cut off the battery based on the first chart. Like, I don't want to have 30% capacity left when my BMS decides it's time to shut off the battery. I want to be able to use another 10%, or if I choose ride until the battery really is discharged if I need to.

That's why I am wondering if my BMS/battery is defective. 30% in the tank is too much left on the table. But if there was only 13% left, than I am OK with that and won't hound the manufacturer.

Thanks again,
craigr

 

[harryS]

Craig, there are two low voltage shutoffs in your ebike. The first is in your ebike controller, typically set at 40-42 volts (for a 48V bike). Look on the controller label to find it. The second is in the battery BMS, but this one operates at the cell level, with the minimum being 3.0V/cell, or 39V for a 48V battery which uses 13 series groups.

There is also voltage sag going on. My typical 10Ah battery will sag .5 -.7volt from its no load value, when pushing about 4A, and maybe drop 3V at 20A. If I've got that battery down to 43V, it will shut off if I bang the throttle and ask for 20A.

In short, the above charts can't be applied to all batteries. The real performance depends on the quality of your internal cells. I've got one 36V battery where I paid an extra $100 for premium Samsung GA cells, and yesterday, I ran it all the way down to 30V without any shutdown. I also have its companion, with generic chinese cells that quits at 35V.

I don't think your battery performance is out of line. How many Ah is it?

 

[rounds]

Craig, there are two low voltage shutoffs in your ebike. The first is in your ebike controller, typically set at 40-42 volts (for a 48V bike). Look on the controller label to find it. The second is in the battery BMS, but this one operates at the cell level, with the minimum being 3.0V/cell, or 39V for a 48V battery which uses 13 series groups.

There is also voltage sag going on. My typical 10Ah battery will sag .5 -.7volt from its no load value, when pushing about 4A, and maybe drop 3V at 20A. If I've got that battery down to 43V, it will shut off if I bang the throttle and ask for 20A.

In short, the above charts can't be applied to all batteries. The real performance depends on the quality of your internal cells. I've got one 36V battery where I paid an extra $100 for premium Samsung GA cells, and yesterday, I ran it all the way down to 30V without any shutdown. I also have its companion, with generic chinese cells that quits at 35V.

I don't think your battery performance is out of line. How many Ah is it?

OK so the controller has low voltage protect at 40v. The battery is rated 672Wh so I guess ~14Ah. It's OEM and has a sticker "LG Cells Inside" for what that's worth. Its a hub 750W Bafang motor. Turboant N1.

The controller is rated for 22 amps but was set to 20 amps on delivery. I raised it to 22 amps as well as the top speed. What you are saying makes sense. Today on a long ride I was more carful with speed. I also upgraded the freewheel from 14-28 to 11-28 giving me two gears higher (11 and 13). I did 15 miles today and used less than 50% of the usable battery voltage while doing 15-20MPh whenever possible, and using PA5 at 25MPh mostly just when passing. I also used lower PA and peddled harder.

When the battery died the other day I was on PA5 with only 14T so couldn't help much at top speed. There was also a VERY strong headwind on the way down. On the way home I wonder if I had used just PA1-2 and had helped more, if the battery would have kept going longer? Like I say, I'm just learning.

 

[rich c]

There is a condition known as voltage sag. Riding at full speed on max current going to the motor, the voltage will sag down to the shut off point in the BMS. Let the eBike sit a bit, and you will see some recovery. I assume that is what you think is 30% left in the pack. It could be 30% sitting there, but try to take off at full speed again and watch it drop dramatically. If you lower the PAS to an eco mode and ride at 12mph, you'll get closer to a full drain of the pack. Which by the way is hard on the battery pack. A great number of owners never let the pack go below 20% for max battery life. If you are riding a low end bike, there isn't a great amount of precision in the control system. Try measuring the battery pack with a multi-meter and see what you get. The display may not be that accurate either.

 

[tomjasz]

If you lower the PAS to an eco mode and ride at 12mph, you'll get closer to a full drain of the pack. Which by the way is hard on the battery pack.

What? What is a full drain? LVC?

 

[6zfshdb]

What is the fully charged voltage of your battery? Are you sure it's really a 48V? Some manufacturers, like Pedego for instance, rate their batteries at 48V when in reality, they are actually 52V. A 43.5V cutoff is around 10% for a 52V battery but closer to 30% for a 48V.

As others have said, the voltage / percent rating of a lithium battery is not linear and varies with cell type and quality. If you ride out to the 50% point, you will likely not have enough juice to get back. On average, my batteries have between 35 and 40% power remaining when my display reads 50%. To be on the safe side, you need to make your own chart through experimentation.

 

[rounds]

There is a condition known as voltage sag. Riding at full speed on max current going to the motor, the voltage will sag down to the shut off point in the BMS. Let the eBike sit a bit, and you will see some recovery. I assume that is what you think is 30% left in the pack. It could be 30% sitting there, but try to take off at full speed again and watch it drop dramatically. If you lower the PAS to an eco mode and ride at 12mph, you'll get closer to a full drain of the pack. Which by the way is hard on the battery pack. A great number of owners never let the pack go below 20% for max battery life. If you are riding a low end bike, there isn't a great amount of precision in the control system. Try measuring the battery pack with a multi-meter and see what you get. The display may not be that accurate either.

Thanks for your reply Rich.

I am not using the battery gage on the display to make my determinations. I have a degree in mechanical engineering and am owner of an engineering firm that does all electronic related engineering. As I said in my first post I am checking the battery with a FLUKE DMM. The FLUKE is what is telling me that I have 43.5 volts left. That is when it cuts off and my display shows "error 26." If I kept riding for a while 'that day' with the bike turned off and then rebooted the display I would get some small peddle assist for a little while, but not long at all. The fact that the battery has 43.5v in it and I was not getting much assist after letting it rest disturbs me and makes me think I may have a bad cell(s) or connection inside the battery. I would open the battery but it has multiple warranty stickers.

If you mean sag in that the moment I start to pedal and that cuts the battery than I would say maybe, it's possible. However, if you mean sag as in letting the battery cool for a while, than I don't think so.

For what it's worth, this is where the battery gauge on the bike shipped from the factory; 41.0v, 43.0v 45.7v, 47.0v, 49.0v. I tried adjusting these to make it so my battery would read empty at 20% discharge (theoretically) and increased all of the values. With the stock settings, I would have had two bars remaining when the battery cut out.

I know riding below 20% is hard on the battery and it's not my plan to do so unless it's absolutely necessary. However, based on my readings and depending on what chart I reference than I was still at or above 30%. When I left the house, I had the battery charged to 85%. I built this charger to help me target my charging voltage better and never go near 100%:

$10 DIY Smart Charger Alternative / Ver. 2.0

I've been experimenting with an easier way to achieve a 70% - 90% charge or a storage charge for my battery other than just timing it and I came across an inexpensive solar charge controller that I thought I might be able to adapt. In the end I am able to achieve my goal, but it didn't end up...

electricbikereview.com

I know I said I'm still learning, but I have done my research and do have a firm grip on electronic and mechanical principles. So far I am not convinced that this battery pack is good. Or perhaps the BMC is poorly designed. But again, I got very light assist after letting the battery rest and power cycling the display which would indicate to me that the batter was indeed used up. It seems much too early based on voltage.

Thanks again,
craigr

 

[rounds]

What? What is a full drain? LVC?

Tom I've read a lot of your posts and would appreciate your opinion of my situation. You seem to know a lot about batteries.

Thank you!
craigr

 

[rounds]

What is the fully charged voltage of your battery? Are you sure it's really a 48V? Some manufacturers, like Pedego for instance, rate their batteries at 48V when in reality, they are actually 52V. A 43.5V cutoff is around 10% for a 52V battery but closer to 30% for a 48V.

As others have said, the voltage / percent rating of a lithium battery is not linear and varies with cell type and quality. If you ride out to the 50% point, you will likely not have enough juice to get back. On average, my batteries have between 35 and 40% power remaining when my display reads 50%. To be on the safe side, you need to make your own chart through experimentation.

Neat idea, but alas I think not. The battery charger that shipped with the bike outputs 54.5v. Fully charged if I let it go to 100% the battery will read 54.5v. I did fully charge the battery the other day because I have read that some less advance BMC's don't maintain balance well and occasional 100% charges are required. So to cover this I charged to 100% then road soon after to get the charge back below 90%. I have not been able to discharge the battery as much as the day it cut out again as my schedule has not allowed it. Yesterday when I took the bike to the doctor I knew I needed ~13 miles round trip, so I had to charge to 85% again to insure I'd make it home. Today after that ride my battery is reading 46.1v on my FLUKE DMM which is very close to the same as what it read right when I got home.

Thanks for the idea though. That would have been cool. I'd rather have a 52v battery ;-)

craigr

 

[Gionnirocket]

If your pack is taking a full charge on one attempt to reach 54.5v then you can ass_ume that the cells are in good shape, balanced and of equal resistance. So in my mind, this leaves the BMS to be the culprit here. 43.5v seems a bit high as it's typical around +/- 1v of 40v. But if you were loading it hard, a 3v sag is understandable.
Another consideration is temperature... If you were riding the battery hard down to 43v it may have gotten warm thus lowering the internal resistance and letting the remaining current flow more easily and creating greater sag as well.
What all that the others have said about voltage sag and % not being linear is great information.
Me personally I think those charts are crap. In general (with a little rounding) a 48v batteries range is 34.5 - 54.5v giving you a 20v spread with each 1v equaling 5%. But since % is not linear it's just as informative as the charts
Staying above 39.5v or 25% battery actual capacity is what's needed to protect the battery. The only way to know exactly what is going on is to open the battery and test the individual cell groups at shutdown... But since that isn't advisable at this point, I'd do what the others have suggested and just keep note of conditions when it shuts down. It could very well be that you have an over protective BMS

 

[tomjasz]

I'd rather have a 52v battery

Greatly overrated. Whoopdee F-ing doo 4 volts.

 

[rounds]

If your pack is taking a full charge on one attempt to reach 54.5v then you can ass_ume that the cells are in good shape, balanced and of equal resistance. So in my mind, this leaves the BMS to be the culprit here. 43.5v seems a bit high as it's typical around +/- 1v of 40v. But if you were loading it hard, a 3v sag is understandable.
Another consideration is temperature... If you were riding the battery hard down to 43v it may have gotten warm thus lowering the internal resistance and letting the remaining current flow more easily and creating greater sag as well.
What all that the others have said about voltage sag and % not being linear is great information.
Me personally I think those charts are crap. In general (with a little rounding) a 48v batteries range is 34.5 - 54.5v giving you a 20v spread with each 1v equaling 5%. But since % is not linear it's just as informative as the charts
Staying above 39.5v or 25% battery actual capacity is what's needed to protect the battery. The only way to know exactly what is going on is to open the battery and test the individual cell groups at shutdown... But since that isn't advisable at this point, I'd do what the others have suggested and just keep note of conditions when it shuts down. It could very well be that you have an over protective BMS

Thanks! That all makes a lot of sense. Sign, perhaps I'll open the battery and see if I can modify the BMS... but probably not. Based on what you are saying I already ran the battery below "safe" levels so hopefully I can avoid that again.

 

[6zfshdb]

Thanks! That all makes a lot of sense. Sign, perhaps I'll open the battery and see if I can modify the BMS... but probably not. Based on what you are saying I already ran the battery below "safe" levels so hopefully I can avoid that again.

I don't see why a battery maker would design a BMS with a cutout voltage set below a "safe" level. Either it's defective or they are just trying to sell more batteries.

 

[Gionnirocket]

Thanks! That all makes a lot of sense. Sign, perhaps I'll open the battery and see if I can modify the BMS... but probably not. Based on what you are saying I already ran the battery below "safe" levels so hopefully I can avoid that again.

From what we know... the BMS shut down perhaps prematurely which would be well above safe levels.

 

[rounds]

So, I suspect that I have identified the problem. If anyone is interested, TurboAnt asked for a video showing the problem. It's here, but in a nutshell, my display shows error 26 and I pull the battery and measure the voltage with my DMM. http://www.cir-engineering.com/bin/bike/battery_problem.mov

While I had the battery out later that day I pressed "the button" on the battery that indicates charge level with LED's; green, yellow, and red. The LED showed yellow, so the battery considered itself as being usable still. This led me to believe that the BMS still 'thinks' that the battery is good to go at 43.5 volts and is not causing the shutdown.

I then remembered that I had pulled the controller and photographed the label on it for my records. Note that on it, low voltage protection is shown as 40 volts.

Others with my bike who live in hilly areas have had issue with this controller overheating because it's also only rated for 11A continuous. Luckily for me there are no big hills in Chicago, so I have not had that problem, but this controller is a POS.

I'd like to replace this controller with as little splicing as possible to keep it close to OEM. This is what the termination looks like, and it does have headlight, taillight, and brake light.

If I must slice, splice, and solder I will. Can anyone point me in the direction of what controller may be compatible with my system? I cannot find a datasheet for this controller anywhere.

On that note, I suspect that the controller's LVP is hard wired and not programmable. I'm planning to open it today to see if it's potted and if I can modify or even disable the LVP circuit in the controller. Maybe I will get lucky and even find a couple of trimmers that are adjustable. I am assuming some here would object to this idea, but I would like to hear the cons please. The BMS in the battery is good and I would think knows better than the controller what is safe for the battery?

Craig, there are two low voltage shutoffs in your ebike. The first is in your ebike controller, typically set at 40-42 volts (for a 48V bike). Look on the controller label to find it. The second is in the battery BMS, but this one operates at the cell level, with the minimum being 3.0V/cell, or 39V for a 48V battery which uses 13 series groups.

There is also voltage sag going on. My typical 10Ah battery will sag .5 -.7volt from its no load value, when pushing about 4A, and maybe drop 3V at 20A. If I've got that battery down to 43V, it will shut off if I bang the throttle and ask for 20A.

Thank you Harry, for this early comment regarding the controller.

If your pack is taking a full charge on one attempt to reach 54.5v then you can ass_ume that the cells are in good shape, balanced and of equal resistance. So in my mind, this leaves the BMS to be the culprit here. 43.5v seems a bit high as it's typical around +/- 1v of 40v. But if you were loading it hard, a 3v sag is understandable.
Another consideration is temperature... If you were riding the battery hard down to 43v it may have gotten warm thus lowering the internal resistance and letting the remaining current flow more easily and creating greater sag as well.
What all that the others have said about voltage sag and % not being linear is great information.
Me personally I think those charts are crap. In general (with a little rounding) a 48v batteries range is 34.5 - 54.5v giving you a 20v spread with each 1v equaling 5%. But since % is not linear it's just as informative as the charts
Staying above 39.5v or 25% battery actual capacity is what's needed to protect the battery. The only way to know exactly what is going on is to open the battery and test the individual cell groups at shutdown... But since that isn't advisable at this point, I'd do what the others have suggested and just keep note of conditions when it shuts down. It could very well be that you have an over protective BMS

I don't see why a battery maker would design a BMS with a cutout voltage set below a "safe" level. Either it's defective or they are just trying to sell more batteries.

From what we know... the BMS shut down perhaps prematurely which would be well above safe levels.

Sorry. I made this confusing when I said I had already gone below a safe voltage. I actually misread Gionnirocket's voltage post because my battery's voltage was so far off. I was so tired when I read the post that when I saw 39.5v as a good minimum voltage value to keep the battery at or above 25% I actually was thinking my 43.5v was lower... I looked at the 9 and 3, but missed the 10ths place .

Thanks for all the help everyone. It's tough to be new at something.

Kind regards,
craigr

 

[rounds]

Also, this is the motor from the QR code on it, "RM G060.D 6.5 1F6V5291242."

 

[Pine_marten]

This thread is a good argument for a high capacity battery. I got a 48v 20ah battery ( was the largest offered with the motor kit ). I have only ever charged it to 80% and it seems to have great reserves. If I ever replace the battery i will look for a 30ah triangular unit.

 

[indianajo]

Having a handlebar mounted DVM that reads continuously can help you see if voltage sag at heavy loads is causing your problem. The reading that you take with a portable DVM after you remove the battery after the controller shut off is a no load reading, and not particularly helpful. The DVM built into the throttle did not discharge the battery unduly powered up 100% of time. Unfortunately rain destroyed that throttle and I haven't found another with a DVM. I found climbing big hills with 80 lb groceries with battery under 43 v the controller would red light & shut down(40 v) temporarily, but recover if I used less wattage on the next hill. That was when I was using a 26 A controller & a 1300 W motor.
I've had two batteries that voltage sagged to 7 & 11 v any time more than 50 or 100 watts was drawn. I believe those two batteries had bad welds inside. I got my money back for one, but had to devise a load test with resistors and a DVM to prove the other was bad, since it would snap back to good voltage as soon as the load was removed. I didn't get money back for the second battery. My 3rd 48 v (13s liion)( battery is 4 years old with about 250 charge cycles, and will run reliably down to 40.3 v with 10 amp controller (which I did last week). It sounds as if your battery is built of cheaper cells that sag in voltage under load worse than what I've got. Samsung, blah blah cells are thrown around in advertisements for batteries, but nobody ever cuts them open to see what you actually bought. Samsung makes a low current flashlight cell, too. I paid $300 each for the first two 48 v 17 ah trash batteries, and $630 for the good 17.5 AH battery. Unfortunately that vendor is no longer selling 48 v batteries, only 52.
I've been into all my controllers except the potted ASI, and there is not much you can control or change. All bike controllers I've had came with insta-leak trash rail caps, 470 uf @ 63 v usually. I changed those to panasonic or nichicon industrial caps on my first controller, which stopped the big spark the second time the controller was plugged in in 5 minutes. I also added heat sink compound between the spreader bar on the mosfets & the decorative (but non-functional) aluminum case. That controller developed a "fault 07" at 4 years and I have been unable to repair it. Unfortunately the sellers of controllers on the US ebay are illiterate idiots. Ebay never debits my bank account without authorization. If you wish to give your debit card number to the hackers in *****, TJ reports there are some people there that know what they are selling.
I most especially do not recommend the ASI controller, that has pins so close together that they melted off in the rain into the harness of the motor & destroyed both 7/11/21. 45 A is powerful, but that controller is suitable only for Death Valley. I couldn't replace the leaky caps in the ASI, it was potted. Lots of people recommend Grin Tech in Vancouver, who are knowledgeable, but they were only selling potted controllers last month too. If they have actual no-leak industrial rail caps I would be very surprised.
I'm currently using a 48 v 500 W controller I bought with only rectangular connectors. I bought a set of replacment rectangular connectors on ebay in 2 pin, 3 pin, 4 pin 6 pin and 9 pin male & female, with 100 each pins. The pins can be soldered and hand crimped with 4 applications of a slip joint plier. I don't like those molded juli connectors, they cannot be modified because rubber is injected in the housing after the pins are soldered or crimped. Butt splices in wiring IMHO are not reliable, especially at 30 amps. I've had good luck in the rain with the rectangular connectors on hall effect sense, throttle, ignition sw, My motor wires use .157" bullet connectors, that can be bought from Dorman at the local auto supply. My recent 500 w controller has a lot of weird connectors like "alarm power" "3speed" "self learn" "door switch" "reverse" "hi brake" "low brake" that are useless. Some others connectors don't match any chart on ebay and are labeled only in ****ese. "door switch" I found after weeks of experimenting has to be powered for the controller to function. "alarm power' was always hot at full battery voltage, and I burned the pins off trying to measure it, so I cut it off, put a new 2 pin block on, and drew the power for the "door switch" from that through 100 ohms. I have a toggle switch under the seat to turn the controller on or off. I think the green male bullet connector is for the handlebar DVM, but the ground for that has to be derived from the throttle, or from the "alarm power". Reverse didn't work & 3 speed did nothing when the 2 powered wires were alternately connected to the ground. I had to reverse rotation direction by swapping wire colors on both phase wires and hall effect wires. Try that with a juli connector! The rectangular connectors you can pop the pins out with a tiny screwdriver, bend the tangs back out, reverse a couple of colors and reinsert them to lock.