External LiFePo4 battery?

[Philip W]

Hi, everyone ! I have a KingSong 14C for 3 years. The battery is getting worse and worse. Right now I am thinking about adding an external battery to switch over. That is , once the internal battery is exhausted, I will flip a switch and the external battery will take over.

The ones inside are both 67.2v Lithium Ion, but I am really not happy with the degradation and potential hazard of burning. So I am looking at LiFePo4 batteries, which are supposed to be better. No memory effect, no fire hazard, can recharge over 2000 times...

So I found one seems to be good as a replacement for my 60v batteries:

https://www.aliexpress.com/item/33029347856.html?spm=2114.12010612.8148356.2.27827841YMgDw4

The size is not an issue, since I will put it on top of the EUC. The voltage is similar. I will make sure it will be charged only up to 67v, instead of 73v. Now the only thing I am not sure is that the battery has a maximum discharge of 150AMP ! This is way over the regular EUC's motor. My question is, since the battery's discharging current is much higher, will it damage the control board or the motor? Or the board has circuit to prevent excessive power from the battery? Thank you very much for any of your insight !!

Edited August 17, 2020 by Philip W
change topic, add tags

[Chriull]

55 minutes ago, Philip W said:

My question is, since the battery's discharging current is much higher, will it damage the control board or the motor? Or the board has circuit to prevent excessive power from the battery? Thank you very much for any of your insight !!

Kingsong has fuses and current limiting in firmware.

But i have no idea how "effective" the current limiting is and/or how much the design relies on the weaker batteries.

Maybe you have to take some spare fuses with you, or maybe it blows the mosfets/motor wires. The motor itself should sustain it.

1 hour ago, Philip W said:

I will make sure it will be charged only up to 67v, instead of 73v.

Seems to be a 20s battery pack.

If you only charge it to/with 67.2V this could lead to low capacity usage? LiFePo4 have a very different voltage discharge/charge characteristic - be sure to educate yourself so this works out and not leaves you with a huge pack without usable energy...

[Philip W]

2 hours ago, Chriull said:

Kingsong has fuses and current limiting in firmware.

But i have no idea how "effective" the current limiting is and/or how much the design relies on the weaker batteries.

Thanks for the info ! Yeah, good to know that Kingsong has some kind of current limiting calculation in the firmware. The thing I am very concerned is that,  if I switch it to the much bigger battery, will the wheel still behave the same, or the wheel will suddenly be 5-times more powerful than before (30Amp vs 150Amp), even with the same voltage?

2 hours ago, Chriull said:

Seems to be a 20s battery pack.

Yup, it's a 20s pack with 3.2v nominal voltage. The beauty of this setup is that, though the max charging is different (73v vs 67v), the cut-off voltage is the same (50v). So even thought I lose 5v of maximum power, the lower end is the same. It should behave the same like regular Lithium-Ions, right?

Also I am curious about what will happen when I charge it to 73v. Will the EUC handle the 5v difference? I have an old broken EUC that I can experiment on. Maybe the board can take the 5v over-voltage?

 

[Chriull]

6 hours ago, Philip W said:

will the wheel still behave the same, or the wheel will suddenly be 5-times more powerful than before (30Amp vs 150Amp), even with the same voltage?

One has about the same power limit. The only difference is the (presumably lower) internal resistance of this battery pack.

For any speed one drives the motor generates a voltage U_back_emv=kv * rpm.

With that a maximum motor current (~torque) I = (U_battery_no_load - U_back_emv)/(R_coil+R_internal_battery) can flow.

So there will be some difference, but far away of factor 5. Btw: this 30A are the max continous current specifications. Peak currents should be much higher?

Just looked at the specs of the battery again it has 60 A continous and 150A peak. That should be high enough, so that the BMS does not cut off in high current situations.

But there is no (i did not see at least) detailed specification of the used bms. Just mentioned that it protects from under discharge - which is not used with any (not "ancient") euc bms. By the voltage sag in burden situations the bms could cut of by this and one flies forward... :(

7 hours ago, Philip W said:

Also I am curious about what will happen when I charge it to 73v. Will the EUC handle the 5v difference? I have an old broken EUC that I can experiment on. Maybe the board can take the 5v over-voltage?

The board should take most probably the overvoltage. Maybe the dc/dc converters could not run "optimal" but very very imho should stand it. You could look at the datsheet of the used ic. And check the used switching mosfets.

However the board measures the battery back voltage and will issue an overvoltage alarm - beeping, speech message, tiltback and probably cut off.

Ks14?d? had a history of some batches/motherboard revision measuring this battery voltage to high and issuing this overvoltage alarm with just fully charged batteries. Afair the wheels just started to buckle /stutter forward backward once turned on. Should be many of these reports to be found somewhere here.

[Philip W]

Wow! This is a lot of info to absolve ! Thank you so much for taking the time and effort to give me the answers. You rock !

[Philip W]

Just an update. Thank you @Chriull again for your great advice! It helped me in the decision of modding my KS-14C's battery config. See below for more details.

Last year I opened the KS-14C and found out that the 2 batteries are actually connected in a serial way: battery one's discharging port connected to the battery two's charging port, and battery two's discharging port connected to the main board. Therefore, when I was charging the EUC, I was actually charging battery one first, then the battery one was charging battery two. And the EUC was actually powered by battery two only. Battery one is more like an external battery.

I had been wondering about this for a long time, thinking there must be a reason for this kind of setup. Maybe when two batteries are connected in parallel, it will be overpowering for the wheel? Thinking their design must have some merit. So I hadn't change it for a year. But the battery charge is getting worse: now the first battery can only holds 63v after charging, and the second battery, being charging by the first, holds only 60v of charge. So, after a "full" charge, I still got 60v in Kingsong's App, and the range measurement vs voltage was all messed up.

After reading Chriull's opinion, I realized that the torque/current will increase, but it will not just double. So I just need to stay in the setting of "moderate riding", the board should be able to take it. So finally I opened up the wheel again, and soldered all the wiring, making those two batteries parallel instead of "serial". After I finish it and turn it back on, now the Kingsong app is reporting 63v of charge, way better than before. The riding experience is similar. I don't feel much difference. ( or maybe I didn't push it hard, I am just a casual rider anyway.)

I suggest all Kingsong 14 riders check your battery config as well. Connecting 2 batteries in serial, in my opinion, is not a good way of EUC design.

[Chriull]

42 minutes ago, Philip W said:

Last year I opened the KS-14C and found out that the 2 batteries are actually connected in a serial way:

Yes - the "classic" (?historical?) way. I had always called this "daisy chaining" as serial could eadily lead to grave misunderstanding.

Quote

battery one's discharging port connected to the battery two's charging port, and battery two's discharging port connected to the main board. Therefore, when I was charging the EUC, I was actually charging battery one first, then the battery one was charging battery two.

Only if the voltage of the second pack is higher as the voltage of the first pack including the voltage drop at the protection circuitry of the second pack. Once this is "overcome" they both charge simultanously, just the charge protection circuitry has to take the whole charge current burden.

Quote

And the EUC was actually powered by battery two only. Battery one is more like an external battery.

No. Both battery packs power the euc together. Just if the second pack has a higher voltage as the first it will deliver the power alone. This happens by regenerative braking - the second pack gets charged alone and will power the wheel alone until the voltage drops again.

This could or maybe was circumvented by removing the reverse polarity protection diode from the seconds pack BMS. But i never looked at this in detail - if or when this was done. Afair chris from @1RadWerkstatt told me in these times that he delivered ks wheels this way.

Quote

I had been wondering about this for a long time, thinking there must be a reason for this kind of setup. Maybe when two batteries are connected in parallel, it will be overpowering for the wheel? Thinking their design must have some merit. So I hadn't change it for a year. But the battery charge is getting worse: now the first battery can only holds 63v after charging, and the second battery, being charging by the first, holds only 60v of charge. So, after a "full" charge, I still got 60v in Kingsong's App, and the range measurement vs voltage was all messed up.

Pff... Was really time to get rid of these packs! Sounds like some dead cells!

Afair these packs did not have a synchronisation wire? If so pack two was permanently disconnected from pack one by his charge protection circuitry!

Quote

After reading Chriull's opinion, I realized that the torque/current will increase, but it will not just double. So I just need to stay in the setting of "moderate riding", the board should be able to take it. So finally I opened up the wheel again, and soldered all the wiring, making those two batteries parallel instead of "serial". After I finish it and turn it back on, now the Kingsong app is reporting 63v of charge, way better than before. The riding experience is similar. I don't feel much difference. ( or maybe I didn't push it hard, I am just a casual rider anyway.)

Putting them in parallel needs a synchronization wire! So pack two with most presumable dead cells still cuts off his overcharge protection, but gets still charged via his discharge wires via pack one! So that's inherently dangerous!

Quote

I suggest all Kingsong 14 riders check your battery config as well. Connecting 2 batteries in serial, in my opinion, is not a good way of EUC design.

No! Please do not do this!

As, most presumably in your case pack 2 needs to be replaced (or repaired)! No way out of this!

Edited August 20, 2020 by Chriull

[Philip W]

22 hours ago, Chriull said:

Putting them in parallel needs a synchronization wire! So pack two with most presumable dead cells still cuts off his overcharge protection, but gets still charged via his discharge wires via pack one! So that's inherently dangerous!

The sync wire is something I don't understand. When both packs has the same BMS board, why not just let both charge at the same time? Maybe pack one will disconnect after fully charge, but it shouldn't affect the pack two keep being charged, right? Why use a wire to disconnect at the same time?

22 hours ago, Chriull said:

This could or maybe was circumvented by removing the reverse polarity protection diode from the seconds pack BMS.

I measure the charge port of the pack two. It has a voltage of 6v, same as the pack one. So I think both packs are the same: they both have BMS circuit board to protect them. The downside is that pack one will have a 3v drop from 67v, and pack two has another 3v drop from 63v which pack one provides.

Yeah, I know both battery packs have dropped voltage. Isn't it expected for a 3-year-old wheel? I still want to squeeze some extra life from those packs, then switch to LiFePo4 packs.

22 hours ago, Chriull said:

Afair these packs did not have a synchronisation wire? If so pack two was permanently disconnected from pack one by his charge protection circuitry!

Now I read your comment again. I think I start to understand your point. Well, is there a way to find out if the pack two has over-charge protection in a safe way? I am pretty sure pack two has BMS board, otherwise the charge port should show a voltage of 63v, instead of 6v. So when pack two has a BMS board, it should have over-charge protection as well, right?

Also thanks for the warning ! Now I will not ride it or charge it until I get these issues sorted out.

Edited August 21, 2020 by Philip W

[Chriull]

@Philip W with some chance your primary pack is ok, but your second pack is just dead. And dangerous.

Try to charge them indivudually - under supervision in non flamable environent! Check battery pack voltage at full charge, after disconnecting the charger, after some hours and next day.

[Philip W]

OK, I will do this test very carefully once I am back home from Connecticut. Thanks for the tips ! It has been quite an interesting discussion!

[Philip W]

OK, now it's the update. I am happy to tell all of you that both of my battery packs are working properly. They both cut off at 63v.

So this Sunday I will take the wheel for a spin. Also, I just finished making my piano-bench seat for EUC, version 3.0. New cushion is much narrower, not as comfy as before, but much nimbler. I will post some pictures for it this weekend.

[Chriull]

5 hours ago, Philip W said:

OK, now it's the update. I am happy to tell all of you that both of my battery packs are working properly. They both cut off at 63v

63V?! Then they are both dead?

You checked your charger - does it deliver 67(+)V?

[Philip W]

On 8/29/2020 at 5:52 AM, Chriull said:

63V?! Then they are both dead?

You checked your charger - does it deliver 67(+)V?

Yup, the charger is at 67v, but I think the BMS cutoff at 63v for some reason. Maybe after a few years they all behave like that? Cause my old EUC also did the same 63v thing.

However, yesterday I did ride the wheel for about 15km in a park testing my new seat. Seems the total range should be around 20km, instead of the supposed 60km. Well, I am still OK with the 20km. Most of my rides are within 10km anyway. And I am planning to get LiFePo batteries to replace them in the future.

[Chriull]

1 hour ago, Philip W said:

Yup, the charger is at 67v, but I think the BMS cutoff at 63v for some reason. Maybe after a few years they all behave like that?

Sounds like one dead cell (4.2V missing), so the BMS has to cut off.

1 hour ago, Philip W said:

Cause my old EUC also did the same 63v thing.

Same thing - the wrakest cell seem to die first...

1 hour ago, Philip W said:

However, yesterday I did ride the wheel for about 15km in a park testing my new seat. Seems the total range should be around 20km, instead of the supposed 60km. Well, I am still OK with the 20km. Most of my rides are within 10km anyway.

Although the 18650 seem to be quite stable and take aome misuse, riding and (?especially?) charging still bears the risk of tgermal runaway (battery fire). This risk should imho rise with every further use...