docnjoj
1 GW
U really need to charge 12 hours to get balancing to occur! Also your starting voltage is 2-3 volts low. Let em charge overnite at least once.
otherdoc
otherdoc
Is Jimmywu66 & EastUnitedMusic the same person?
- Thread starter Sacman
- Start date
Fatal flaw in Chinese LiFePo4 battery/BMS design.
There is a fatal flaw in the battery design. I mean - battery/BMS system.
Battery itself and BMS itself are fine.
How I found it.
Finally, after 36 days of waiting, I got my 24v 20Ah LiFePO4 battery from jimmywu66 and started to test it.
I charged it (it took about 30 minutes for charger to turn red LED green, which means that battery was shipped fully charged) and installed it into my 500W scooter. I found that BMS cuts off even with a moderate load.
I was surprised, because the cut off current suppose to be 65A and my WattsUp showed different max. current values every time, but all between 35-50A.
I didn't understand why BMS cuts off and I started to look at the board trying to figure out how it works.
It has an over-current protection (four shunt wires for current sense). Signal goes into 16-pin chip, which shuts 4 MOS-FETs off in case of any load problem.
The heart of BMS is a 16-pin micro, which controls MOS-FET switches in the load and charger circuits.
BMS has 8 balancing circuits, which also act like over-charge and under-voltage protection. Each circuit has 2 opto-couplers (one for under-voltage protection to cut the load, one for over- voltage protection to cut the charger). The outputs of each type (under- and over-) are connected in series and the combined signals go into micro.View attachment pic1.jpg
For the test I used 2Ohm 50W adjustable resistor. I connected this load manually for a short period of time of 1-2 seconds. The load resistance I measured by measuring voltage drop across it with constant 1A current from the bench power supply. The load voltage was measured by o-scope and the load current was calculated.
I found that the BMS cuts off at about 30A instead of 65A claimed.
At the same time battery voltage was not to low - 22.0v.
I shorted current shunt resistors - the same cut off result. Uhm????
Then I decided to measure individual cell voltage on the BMS board @ 50A load:
Measuring 1st circuit voltage:
View attachment pic2.jpg
Measuring 2nd (and so on) cell voltage:
View attachment pic3.jpg
Surprise!
Result was astonishing!
Test results for the battery as-is @ 50A load (voltage is measured @ BMS board:
Iload = 50A
Cell # 1 2 3 4 5 6 7 8
Voltage 1.80 2.90 2.86 2.92 2.88 2.89 2.90 3.02
BMS cuts off.
Firstly I thought my battery's 1st cell is BAD L
Then I looked at the long wire with connector going to from BMS to the battery - hmmm???
I measured the voltage drop on this wire @ 50A load - almost 1volt!!!!
Now it's getting clear.
The first voltage monitoring circuit (if you go from negative terminal) has no (-) sense wire. It is using (-) battery wire instead. Other circuits have both (-) and (+) sense wires:
View attachment 6
As a result, when the battery is loaded, the voltage drop across the unnecessary long negative battery wire (including connector, put for no reason) is getting subtracted from the first cell voltage. In my case at 50A load the first cell voltage "dropped" to 1.80v and BMS cut off. The real cell voltage I measured after I installed sense wire was about 2.90v.
View attachment pic5.jpg
Connector - is one thing, contributing into the voltage drop. Another thing is the wire itself.
It looks like 10AWG - the same external diameter, but if you look at the metal, it is, most likely, 14/16AWG. Chinese wire on the left, 10AWG - on the right.
View attachment pic6.JPG
So, I had no choice but to replace battery wires.
Battery rework.
It was a tough decision. I didn't want to cut-open my battery till the end, when I realized there is no other way. And it actually wasn't bad and difficult.
Caution! Always use safety glasses! No rings and jewelry!
I used 10AWG red wire for (+) terminal, but two 12AWG short black wires in parallel for the (-) terminal, since two 12AWG wires have less resistance than one 10AWG wire and more flexible. I also installed (-) sense wire for the possible future usage.
Original (-) wire removed:
View attachment pic7.jpg
Sense wire is soldered:
View attachment 2
Two negative 12AWG wires and sense wire:
View attachment 1
The same was done to the (+) wire.
Final look:
View attachment pic10.jpg
Test results after (-) and (+) battery wires reworked (measured @ BMS board):
Iload = 50A
Cell # 1 2
Voltage 2.74 2.88
BMS over-current protection.
After all the reworks I decided to try the battery again with my 500W scooter.
Guess what? BMS cuts off again!
I decided to check if it's under-voltage or over-current protection.
I set up my load to 0.36Ohm and loaded the battery - BMS cut off.
I shorted current sense resistors with a piece of thick wire - no cut off.
With 0.36Ohm load the current was ~60A.
I put a long piece of thinner wire across the current shunt and loaded battery again - BMS cut off.
I shortened the wire step by step until BMS stopped to cut off and tried it with my scoter - fine.
NOTE: due to post limitations I can not attach any more pictures, so I'll continue at the next post.
Thanks!
There is a fatal flaw in the battery design. I mean - battery/BMS system.
Battery itself and BMS itself are fine.
How I found it.
Finally, after 36 days of waiting, I got my 24v 20Ah LiFePO4 battery from jimmywu66 and started to test it.
I charged it (it took about 30 minutes for charger to turn red LED green, which means that battery was shipped fully charged) and installed it into my 500W scooter. I found that BMS cuts off even with a moderate load.
I was surprised, because the cut off current suppose to be 65A and my WattsUp showed different max. current values every time, but all between 35-50A.
I didn't understand why BMS cuts off and I started to look at the board trying to figure out how it works.
It has an over-current protection (four shunt wires for current sense). Signal goes into 16-pin chip, which shuts 4 MOS-FETs off in case of any load problem.
The heart of BMS is a 16-pin micro, which controls MOS-FET switches in the load and charger circuits.
BMS has 8 balancing circuits, which also act like over-charge and under-voltage protection. Each circuit has 2 opto-couplers (one for under-voltage protection to cut the load, one for over- voltage protection to cut the charger). The outputs of each type (under- and over-) are connected in series and the combined signals go into micro.View attachment pic1.jpg
For the test I used 2Ohm 50W adjustable resistor. I connected this load manually for a short period of time of 1-2 seconds. The load resistance I measured by measuring voltage drop across it with constant 1A current from the bench power supply. The load voltage was measured by o-scope and the load current was calculated.
I found that the BMS cuts off at about 30A instead of 65A claimed.
At the same time battery voltage was not to low - 22.0v.
I shorted current shunt resistors - the same cut off result. Uhm????
Then I decided to measure individual cell voltage on the BMS board @ 50A load:
Measuring 1st circuit voltage:
View attachment pic2.jpg
Measuring 2nd (and so on) cell voltage:
View attachment pic3.jpg
Surprise!
Result was astonishing!
Test results for the battery as-is @ 50A load (voltage is measured @ BMS board:
Iload = 50A
Cell # 1 2 3 4 5 6 7 8
Voltage 1.80 2.90 2.86 2.92 2.88 2.89 2.90 3.02
BMS cuts off.
Firstly I thought my battery's 1st cell is BAD L
Then I looked at the long wire with connector going to from BMS to the battery - hmmm???
I measured the voltage drop on this wire @ 50A load - almost 1volt!!!!
Now it's getting clear.
The first voltage monitoring circuit (if you go from negative terminal) has no (-) sense wire. It is using (-) battery wire instead. Other circuits have both (-) and (+) sense wires:
View attachment 6
As a result, when the battery is loaded, the voltage drop across the unnecessary long negative battery wire (including connector, put for no reason) is getting subtracted from the first cell voltage. In my case at 50A load the first cell voltage "dropped" to 1.80v and BMS cut off. The real cell voltage I measured after I installed sense wire was about 2.90v.
View attachment pic5.jpg
Connector - is one thing, contributing into the voltage drop. Another thing is the wire itself.
It looks like 10AWG - the same external diameter, but if you look at the metal, it is, most likely, 14/16AWG. Chinese wire on the left, 10AWG - on the right.
View attachment pic6.JPG
So, I had no choice but to replace battery wires.
Battery rework.
It was a tough decision. I didn't want to cut-open my battery till the end, when I realized there is no other way. And it actually wasn't bad and difficult.
Caution! Always use safety glasses! No rings and jewelry!
I used 10AWG red wire for (+) terminal, but two 12AWG short black wires in parallel for the (-) terminal, since two 12AWG wires have less resistance than one 10AWG wire and more flexible. I also installed (-) sense wire for the possible future usage.
Original (-) wire removed:
View attachment pic7.jpg
Sense wire is soldered:
View attachment 2
Two negative 12AWG wires and sense wire:
View attachment 1
The same was done to the (+) wire.
Final look:
View attachment pic10.jpg
Test results after (-) and (+) battery wires reworked (measured @ BMS board):
Iload = 50A
Cell # 1 2
Voltage 2.74 2.88
BMS over-current protection.
After all the reworks I decided to try the battery again with my 500W scooter.
Guess what? BMS cuts off again!
I decided to check if it's under-voltage or over-current protection.
I set up my load to 0.36Ohm and loaded the battery - BMS cut off.
I shorted current sense resistors with a piece of thick wire - no cut off.
With 0.36Ohm load the current was ~60A.
I put a long piece of thinner wire across the current shunt and loaded battery again - BMS cut off.
I shortened the wire step by step until BMS stopped to cut off and tried it with my scoter - fine.
NOTE: due to post limitations I can not attach any more pictures, so I'll continue at the next post.
Thanks!
Continue the previous post.
View attachment pic11.jpg
I rode my scooter on a hilly road - no cuts off. WattsUp results:
- the maximum peak current - 60A
- the minimum voltage - 20.4v
- the max power ~1200W.
Wow!
Now I love it!
Final measurements @ 60A load
Iload = 60A
Cell # 1
Voltage @ BMS 2.65
Voltage @ sens wire 2.76
It is 11mV voltage drop along the battery (-) wire @ 60A,which I consider acceptable.
Balancer
I connected my bench power supply, limited to 20mA, across the cell monitoring circuit and Fluke to measure voltage. I slowly increased the voltage to find when balancing circuit opens. The balancing resistor is 33Ohm and in theory, the balancing current is 3.8v/33Ohm = 115mA, so I concluded that when the voltage stabilizes, I am reaching 20mA balancing current - the circuit breakdown point.
Balancing voltage - 3.80v
I thought that small 6-pin chip is kind of comparator to control opto-couplers when cell voltage is getting over- or under- the specified limits.
I applied voltage to the circuit and slowly changed it while measuring the opto-coupler output resistance with Fluke.
I expected circuit to have sharp thresholds when opto-coupler turns on/off. I found that the opto-coupler output resistance is changing gradually and it's really impossible to say what to consider ON or OFF state. It needs further investigation.
"Results" are: over-voltage opto-coupler output resistance reaches 10kOhm @ 3.77v.
under-voltage opto-coupler output resistance reaches 10okOhm @ 1.96v.
To be continued…
Conclusion:
The (-) wire from the battery to BMS board resistance is the most critical thing. The less - the better. The more resistance - the more voltage drop, which subtracted from the 1st cell voltage, which will result in premature BMS cut off, before you've used battery's full capacity.
As a deduction - the lower current - the more battery capacity you can use before BMS cuts off.
Thanks!
Niky
View attachment pic11.jpg
I rode my scooter on a hilly road - no cuts off. WattsUp results:
- the maximum peak current - 60A
- the minimum voltage - 20.4v
- the max power ~1200W.
Wow!
Now I love it!
Final measurements @ 60A load
Iload = 60A
Cell # 1
Voltage @ BMS 2.65
Voltage @ sens wire 2.76
It is 11mV voltage drop along the battery (-) wire @ 60A,which I consider acceptable.
Balancer
I connected my bench power supply, limited to 20mA, across the cell monitoring circuit and Fluke to measure voltage. I slowly increased the voltage to find when balancing circuit opens. The balancing resistor is 33Ohm and in theory, the balancing current is 3.8v/33Ohm = 115mA, so I concluded that when the voltage stabilizes, I am reaching 20mA balancing current - the circuit breakdown point.
Balancing voltage - 3.80v
I thought that small 6-pin chip is kind of comparator to control opto-couplers when cell voltage is getting over- or under- the specified limits.
I applied voltage to the circuit and slowly changed it while measuring the opto-coupler output resistance with Fluke.
I expected circuit to have sharp thresholds when opto-coupler turns on/off. I found that the opto-coupler output resistance is changing gradually and it's really impossible to say what to consider ON or OFF state. It needs further investigation.
"Results" are: over-voltage opto-coupler output resistance reaches 10kOhm @ 3.77v.
under-voltage opto-coupler output resistance reaches 10okOhm @ 1.96v.
To be continued…
Conclusion:
The (-) wire from the battery to BMS board resistance is the most critical thing. The less - the better. The more resistance - the more voltage drop, which subtracted from the 1st cell voltage, which will result in premature BMS cut off, before you've used battery's full capacity.
As a deduction - the lower current - the more battery capacity you can use before BMS cuts off.
Thanks!
Niky
docnjoj
1 GW
Hey Niky1961! Thats a really good piece of engineering! Jimmywu or whomever should send U some extra batts as a reward! Those cells look "better' than the usual bags of lithium.
otherDoc
otherDoc
docnjoj
1 GW
Did U connect the new sense wire? or did u let the 2 (12 ga) wires take its place with better connector? Ping uses sense wires on all cells.
otherDoc
otherDoc
I just read Niky1961's post. I actually emailed Niky1961 through ebay last night because I read his ebay feedback to JimmyWu and he mentioned some possible problem with the bms. I felt I may have a bms problem too.
I just need to know if I should be happy with the approx 6.5 miles (with very little pedaling) out of my 36 volt 10ah JimmyWu battery. Or, if I need to start taking things apart too. I'd rather not have to mess around with anything! From dogman's previous post, I should expect no more then 8 miles so I'm thinking that how my battery is now, is about as good as it is going to get, but, may get a little better with some more charge cycles. I just don't know what exactly I should expect. If I could squeak out a couple more miles by shortening the negative wire it would be worth it. I already know that my starting voltage of 41.5 is low. I'm not ready to make any modifications yet (and I shouldn't have to) so I will just keep reading on for now.
I just need to know if I should be happy with the approx 6.5 miles (with very little pedaling) out of my 36 volt 10ah JimmyWu battery. Or, if I need to start taking things apart too. I'd rather not have to mess around with anything! From dogman's previous post, I should expect no more then 8 miles so I'm thinking that how my battery is now, is about as good as it is going to get, but, may get a little better with some more charge cycles. I just don't know what exactly I should expect. If I could squeak out a couple more miles by shortening the negative wire it would be worth it. I already know that my starting voltage of 41.5 is low. I'm not ready to make any modifications yet (and I shouldn't have to) so I will just keep reading on for now.
Ypedal
100 TW
Hey Niky1961 , that was a most excellent post !!!! ( I'd like to move it to the EV basics section actually )
Considering you found this much so far, have you considered opening up the whole thing for a full inspection ?
Considering you found this much so far, have you considered opening up the whole thing for a full inspection ?
To docnjoj:
No, I didn't connect the sense wire. To be able to do that, you need to separate (cut on BMS) over-current and voltage monitoring grounds.
It is one ground trace, connected to multiple components, so you need to cut it at some place and, probably, put a jumper at another. I was to lazy to figure it out, but I would do it, if...
Since the voltage drop on 2 12AWG wires is only 11mV, I decided not to do it for now.
Thanks,
Niky
No, I didn't connect the sense wire. To be able to do that, you need to separate (cut on BMS) over-current and voltage monitoring grounds.
It is one ground trace, connected to multiple components, so you need to cut it at some place and, probably, put a jumper at another. I was to lazy to figure it out, but I would do it, if...
Since the voltage drop on 2 12AWG wires is only 11mV, I decided not to do it for now.
Thanks,
Niky
TylerDurden
100 GW
R2R2 said:
360Whr / 6 = 60Whr per mile... surely you are more efficient; or are you WOT alla time?
I would expect at least 12mi... with aero-bars and low speed >15mi.
dogman dan
1 PW
Guessing that since Niki has 8 cells in his 24v battery, I would think the 36 does have 12. If so, then there is something fishy about the final charge voltage. Is the charger putting out only 42v? Maybe all you need is a 44.5 ping charger.
In any case with a WE brushed at full throttle, range will suck with any battery. And a ten ah will allways be too small for that sise motor to get the best possible range.
I'd be real worried about the c rate I was pulling out of a 24v battery that is doing 50 amp discharge. Expect to see a lot less than 1000 cycles. But do let us know over the next year just how quick the battery dies, since real world tests to destruction are good to have. That is why I put the thread on the death of my WE brushed motor. I hope my loss of $250 bucks can save another from the same fate.
In any case with a WE brushed at full throttle, range will suck with any battery. And a ten ah will allways be too small for that sise motor to get the best possible range.
I'd be real worried about the c rate I was pulling out of a 24v battery that is doing 50 amp discharge. Expect to see a lot less than 1000 cycles. But do let us know over the next year just how quick the battery dies, since real world tests to destruction are good to have. That is why I put the thread on the death of my WE brushed motor. I hope my loss of $250 bucks can save another from the same fate.
To Ypedal:
Thank you for the nice comment!
If you think it would be better to move it to basic EV section – no problem. I am new to this forum and I don’t know how to do it. I can open a new thread and copy it there. Is it the right way?
No, I haven’t considered opening the whole battery. I’m not sure if you can see it from my pictures, but it’s glued and has soft plastic planes on each side.
I’ve seen enough to say that welding quality is good, but they are using a little bit thin metal straps for 60A peak current, which may increase internal battery resistance.
Actually, there are some other tests I am planning to do in a future:
1. Internal resistance bench test (Full charge/ half charge/ almost discharged)
2. Battery Wh real ride test – when the BMS cuts off
I’ll post the results when I get them.
Thanks,
Niky
Thank you for the nice comment!
If you think it would be better to move it to basic EV section – no problem. I am new to this forum and I don’t know how to do it. I can open a new thread and copy it there. Is it the right way?
No, I haven’t considered opening the whole battery. I’m not sure if you can see it from my pictures, but it’s glued and has soft plastic planes on each side.
I’ve seen enough to say that welding quality is good, but they are using a little bit thin metal straps for 60A peak current, which may increase internal battery resistance.
Actually, there are some other tests I am planning to do in a future:
1. Internal resistance bench test (Full charge/ half charge/ almost discharged)
2. Battery Wh real ride test – when the BMS cuts off
I’ll post the results when I get them.
Thanks,
Niky
docnjoj
1 GW
I'm wondering if the builders of that battery didnt purposely keep the wire resistance high as a safety measure to "overprotect" their cells. Is that possible, Niky??????otherDoc
I'm definitely going to open up my battery again this week! Thanks for the info.
I had a another thought - I don't think charger we got from JimmyWu is for LIFEPO4. I think its an SLA charger (see charger pic posted by Sacman). Although I'm not 100%, the charger look same as this - http://cgi.ebay.com/Battery-Charger-Electric-Scooter-Bike-36V-2-5A_W0QQitemZ300200312172QQcmdZViewItem?_trksid=p3286.m20.l1116
The charger I got from Ping had a label that says Lithium Ion charger. I just got this charger... will be interesting what my final voltage will be.
I had a another thought - I don't think charger we got from JimmyWu is for LIFEPO4. I think its an SLA charger (see charger pic posted by Sacman). Although I'm not 100%, the charger look same as this - http://cgi.ebay.com/Battery-Charger-Electric-Scooter-Bike-36V-2-5A_W0QQitemZ300200312172QQcmdZViewItem?_trksid=p3286.m20.l1116
The charger I got from Ping had a label that says Lithium Ion charger. I just got this charger... will be interesting what my final voltage will be.
To docnjoj.
I don’t think so, because excessive connector/wire resistance contributes only in incorrect 1st cell voltage reading at load and doesn’t affect over-current protection circuitry.
Over-current circuit measures voltage across the shunt resistor on BMS board.
I e-mailed Lau aka jimmywu66 my research results. He thanked me and told me that he doesn’t have technical knowledge, but will show it to his technician.
My guess would be – they don’t have any engineers involved and just assemble cells, using somebody’s else BMS and charger.
Thanks,
Niky
I don’t think so, because excessive connector/wire resistance contributes only in incorrect 1st cell voltage reading at load and doesn’t affect over-current protection circuitry.
Over-current circuit measures voltage across the shunt resistor on BMS board.
I e-mailed Lau aka jimmywu66 my research results. He thanked me and told me that he doesn’t have technical knowledge, but will show it to his technician.
My guess would be – they don’t have any engineers involved and just assemble cells, using somebody’s else BMS and charger.
Thanks,
Niky
To R2R2.
You wrote:
From my first of three rides, here are some specs:
0.0 miles 40.7 volts (wasn't fully charged, full charge is around 41.8 volts I think, maybe 40.
2.3 miles 39.1 volts
4.5 miles 38.9 volts
5.4 miles 38.4 volts
The volts readings were taken by getting off the bike and putting a volt meter up to the battery.
After reaching the 38.4 volts, the battery would keep cutting off.
The battery chemistry stays the cell nominal voltage is 3.3v.
36v battery has 12 cells 12*3.3=39.6v.
Of course, freshly charged battery has higher voltage. My 24v battery has 27.4v, which drops to 26.4v (3.3v per cell) after short period of time under load.
As I read, LiFPO4 charger suppose to have 3 stages:
1. Constant Current charge till cell voltage of 3.70v (for 12 cell battery – 44.4v)
2. Constant voltage charge (3.70v per cell) till current drops below 0.05C(?)
3. Floating voltage stage (3.45v per cell). In this stage your charger light is green.
My 24v battery charger from jimmywu66 gives me the exact right voltages of 29.6v in stage 2 and 27.6v in stage 3.
About balancing circuit:
Since it starts working (shunting overcharged cell with 33Ohm resistor) only when cell voltage is over 3.75v (spec from lifepo4.hk), it will never heat up if your battery is not severely unbalanced or your charger is broken and provides more than 3.7v per cell.
To me it also does not make any sense to keep battery over long period of time with charger in stage 3 (3.45v per cell) with hope to better balance your cells.
Charge is already completed!
What I mean, if your charger light turned GREEN, its output voltage dropped down to floating stage level and if for some reason your cells are still unbalanced, keeping the battery connected will not balance it.
Example:
24v battery, 8 cells.
7 charged cells with 3.7v = 25.9v
1 under-charged cell with 2.5v
Total = 28.4v
But your charger provides only 27.6v. There will be no current.
To leamcorp:
Charger from jimmywu66 I got doesn’t have any LiFePO4 words, but switches green at right voltage and provides right floating voltage.
But,
I didn’t check if it actually switches to float charging state when stage 2 Constant Voltage current drops below certain level or right away after 29.6v level is reached.
It might be a problem with balancing, because stage 2 is designed to balance your cells until they all have equal voltage and no more current is going to the battery.
Thanks,
Niky
You wrote:
From my first of three rides, here are some specs:
0.0 miles 40.7 volts (wasn't fully charged, full charge is around 41.8 volts I think, maybe 40.
2.3 miles 39.1 volts
4.5 miles 38.9 volts
5.4 miles 38.4 volts
The volts readings were taken by getting off the bike and putting a volt meter up to the battery.
After reaching the 38.4 volts, the battery would keep cutting off.
The battery chemistry stays the cell nominal voltage is 3.3v.
36v battery has 12 cells 12*3.3=39.6v.
Of course, freshly charged battery has higher voltage. My 24v battery has 27.4v, which drops to 26.4v (3.3v per cell) after short period of time under load.
As I read, LiFPO4 charger suppose to have 3 stages:
1. Constant Current charge till cell voltage of 3.70v (for 12 cell battery – 44.4v)
2. Constant voltage charge (3.70v per cell) till current drops below 0.05C(?)
3. Floating voltage stage (3.45v per cell). In this stage your charger light is green.
My 24v battery charger from jimmywu66 gives me the exact right voltages of 29.6v in stage 2 and 27.6v in stage 3.
About balancing circuit:
Since it starts working (shunting overcharged cell with 33Ohm resistor) only when cell voltage is over 3.75v (spec from lifepo4.hk), it will never heat up if your battery is not severely unbalanced or your charger is broken and provides more than 3.7v per cell.
To me it also does not make any sense to keep battery over long period of time with charger in stage 3 (3.45v per cell) with hope to better balance your cells.
Charge is already completed!
What I mean, if your charger light turned GREEN, its output voltage dropped down to floating stage level and if for some reason your cells are still unbalanced, keeping the battery connected will not balance it.
Example:
24v battery, 8 cells.
7 charged cells with 3.7v = 25.9v
1 under-charged cell with 2.5v
Total = 28.4v
But your charger provides only 27.6v. There will be no current.
To leamcorp:
Charger from jimmywu66 I got doesn’t have any LiFePO4 words, but switches green at right voltage and provides right floating voltage.
But,
I didn’t check if it actually switches to float charging state when stage 2 Constant Voltage current drops below certain level or right away after 29.6v level is reached.
It might be a problem with balancing, because stage 2 is designed to balance your cells until they all have equal voltage and no more current is going to the battery.
Thanks,
Niky
I ignored this thread for like 6 pages thinking it was just another ebay-trash-fest :wink: , but there is some seriously good information in here. This thread needs to be re-named something more relevant and moved to sticky or into EV Basics or something..02
I got 12 miles out of the 48v 10a battery from jimmywu66. I only pedal to get the bike going and when there is its going up a hill with some wind resistance. I probably avg about 17-19 miles/hr. I didn't use full throttle much, only only when going incline (small hills). I only went full throttle when there were no cars around and thats very rare. I live in Memphis so hills aren't that bad. How do yall know how steep these inclines are anyway? Well, bacdk to my ride, I had to pedal 2 miles home and it sucked hard core. I hope with more charging the distance may increase to 15miles.
More on the JimmyWu66 specials. First of all, I left my new 36v 10ah special on it's charger for 18 hours in hopes of balancing out the cells. It seemed to work. Previously, 7 hour charges were not balancing out the cells despite the charger indicating that the battery was through charging. Previously (yesterday), when taking the battery off the charger after charging for 7 hours, the voltage started off at 41.6 volts and dropped down to 40.4 volts after 15 minutes. An 18 hour charge seemed to remedy this. After 18 hours the initial voltage was 41.5 volts and 5 hours later it was 41.3 volts. So that seemed good.
I didn't have time to go out on a ride this afternoon, but, I wanted to drain the battery and recharge it to see if this would help brake the battery in more. My technique of draining the battery may not have been the smartest because it caused the BMS to melt some wires. What I did was put the bike upside down on its seat and left the throttle on full power for an hour and a half. I realized that this was not getting much of anywhere, so, I put the brakes on somewhat to help put a load on the battery. This went on for another half hour and I started to notice a burning smell. I immediately turned the bike off and looked inside the batter case only to find the following (see pictures):
I may have deserved this.
I didn't have time to go out on a ride this afternoon, but, I wanted to drain the battery and recharge it to see if this would help brake the battery in more. My technique of draining the battery may not have been the smartest because it caused the BMS to melt some wires. What I did was put the bike upside down on its seat and left the throttle on full power for an hour and a half. I realized that this was not getting much of anywhere, so, I put the brakes on somewhat to help put a load on the battery. This went on for another half hour and I started to notice a burning smell. I immediately turned the bike off and looked inside the batter case only to find the following (see pictures):
I may have deserved this.
Opps. Continued from above. I clicked the send button by accident. Anyways, I may have deserved this from this cruel torture test I performed. At the same time, I don't think something like this should have happened. If the BMS was really good, it should have prevented this I would think. What do you all think? Did I deserve this? I have email Mr. Wu to see what he has to say. As it is now, it seems like I could fix this up myself somehow. Maybe electrical tape the burnt wires. They are melted but no wires are really bare except for one small area.
Also, I had to cut through the original protective rubber casing around the BMS to look inside. It seems that this protective casing is needed. Any ideas for a new protective casing? Below is a picture of how I found the protective casing (it had come apart a little from all the heat).
When I aborted my testing my front WE 600 Watt hub was very hot. This was a pretty cruel test I must conclude but I don't think this should have happened. Maybe I am wrong. The battery is on the charger now and seems to be charging. What does anyone think? Electric tape the burnt wires? Get new wires and re solder myself? Meanwhile, I'll be waiting to her from the WU. I am at 10:22 PST (Pacific Standard Time) and from my past experiences with the WU he usually starts responding around 11:00 PST which would be something like 1:22 pm Hong Kong time (Sunday).
The tells of the WU continue.....
Also, I had to cut through the original protective rubber casing around the BMS to look inside. It seems that this protective casing is needed. Any ideas for a new protective casing? Below is a picture of how I found the protective casing (it had come apart a little from all the heat).
When I aborted my testing my front WE 600 Watt hub was very hot. This was a pretty cruel test I must conclude but I don't think this should have happened. Maybe I am wrong. The battery is on the charger now and seems to be charging. What does anyone think? Electric tape the burnt wires? Get new wires and re solder myself? Meanwhile, I'll be waiting to her from the WU. I am at 10:22 PST (Pacific Standard Time) and from my past experiences with the WU he usually starts responding around 11:00 PST which would be something like 1:22 pm Hong Kong time (Sunday).
The tells of the WU continue.....
dogman dan
1 PW
Ouch! I have wondered why a small ventilated box can't be provided for the bms, much like the ones our chargers come with. But hey, it is duct tape prices. I put mine on some little wood blocks to let it get some air on both sides so a warm wire could cool a bit. In my climate I figured some air around the electronics was a good idea. This idea may not help your problem, but who knows? it can't possibly hurt.
When my WE hub gets really hot, the resistance goes up, and the amps it draws goes up. Maybe running the motor without wind to cool it caused it to pull more than normal amps at the very end of the test. Still, idealy your controller would cut out before damage to the power source. I've had lots of controller cut offs without my bms melting.
When my WE hub gets really hot, the resistance goes up, and the amps it draws goes up. Maybe running the motor without wind to cool it caused it to pull more than normal amps at the very end of the test. Still, idealy your controller would cut out before damage to the power source. I've had lots of controller cut offs without my bms melting.
Try something like this - they come in all sizes (not just from RS) - http://www.radioshack.com/product/index.jsp?productId=2062285&cp=&sr=1&kw=box&origkw=box&parentPage=search
You could mount the BMS using standoffs and plastic boxes can be drilled for ventilation or installing small fan.
You could mount the BMS using standoffs and plastic boxes can be drilled for ventilation or installing small fan.
To R2R2:
Looks like one of BMS load switching MOS-FET transistors was burnt. Is it right?
If it’s only one transistor out of three, it might mean that two others are not working (turning on) and all the load current is going through the first one.
Check for damaged traces and cracked solder joints.
Another thought – your BMS has the same looking 4 current sense wires in parallel. If they have the same resistance as mine, the over-current threshold could be the same as I have initially for 20Ah battery - ~50A, which is too high for 10Ah. My BMS also has 4 transistors in parallel, not three like yours.
In my understanding, BMS should not get hot at all. If MOS-FETs are correct, they ON resistance should be in milliohms and power dissipation should be very low.
I ran my scooter yesterday uphill, draining about 50Amps for 5 minutes straight and my BMS has no signs of over-heating and its heat shrink still not shrunk and can be taken off easily.
I would suggest you to measure the voltage drop on MOS-FETs under load with known current to calculate ON resistance.
If it’s possible to read MOS-FET type (on my BMS it’s sanded out), you can get a Datasheet and compare.
Thanks,
Niky
Looks like one of BMS load switching MOS-FET transistors was burnt. Is it right?
If it’s only one transistor out of three, it might mean that two others are not working (turning on) and all the load current is going through the first one.
Check for damaged traces and cracked solder joints.
Another thought – your BMS has the same looking 4 current sense wires in parallel. If they have the same resistance as mine, the over-current threshold could be the same as I have initially for 20Ah battery - ~50A, which is too high for 10Ah. My BMS also has 4 transistors in parallel, not three like yours.
In my understanding, BMS should not get hot at all. If MOS-FETs are correct, they ON resistance should be in milliohms and power dissipation should be very low.
I ran my scooter yesterday uphill, draining about 50Amps for 5 minutes straight and my BMS has no signs of over-heating and its heat shrink still not shrunk and can be taken off easily.
I would suggest you to measure the voltage drop on MOS-FETs under load with known current to calculate ON resistance.
If it’s possible to read MOS-FET type (on my BMS it’s sanded out), you can get a Datasheet and compare.
Thanks,
Niky
dogman dan
1 PW
Indeed the heat I was trying to dissipate was that which I saw happening when the pack was in balancing mode on the charger. But the wire going out of the bms can get warm when running, so I liked the idea of not having that stuffed between the battery and the bms. Just in case.
