DIY Electric Car Forums banner

1 - 19 of 19 Posts

·
Registered
Joined
·
1,858 Posts
Discussion Starter #1
I've been following the heated BMS debate here and thought it would be beneficial to gather data on those of you who have driven with no BMS on your batteries.

I post this because a lot of people are in the situation I'm in. Ready to make the switch but not sure which path to take. Hopefully you guys can steer us with your experiences. If you would take a minute and give us some data it would be greatly appreciated!


  • How many miles, Km have you driven without a BMS?
  • What sort of problems have you experienced?
  • What instrumentation do you feel is most beneficial/necessary to safely do this?
  • Do you still drive/recommend no BMS or have you changed your mind?
I've ordered my pack and have a month or so until they arrive. I still have not decided whether I really need a BMS or not. I am convinced I don't need shunting ability during the cell's early life but not sure if it would be beneficial when the cells have a few hundred cycles on them and are aging a bit.

If you drive with no BMS we really need to know if you have alternative measuring equipment installed or just drive very conservatively or generally what it is that makes your decision to go without a BMS a viable option. If it wasn't a good decision please help us and tell us why. Thanks in advance! :)
 

·
Registered
Joined
·
903 Posts
In a few days I will have had my TS-LFP100AHA battery pack in my Gizmo for a year. I now have 5500 miles on them. For the first 2700 miles I top balanced my 2p18s pack to 4.00vpc with a Black Sheep Technology cell level BMS system. I didn't like charging that high so I had my Zivan charger properly programmed for charging curve #612 and 19 cells so a cut off voltage of 69.3V. I then installed two more cell pairs to make my pack a 2p20s pack. I also adjusted the voltage calibration pot in the charger to an ending charge of 69.7V so now I charge to 3.485vpc, well below the shunting voltage of the BMS modules. Before my first drive with the full pack I top balanced everything with the 4.00vpc BMS setup and a bench top lab grade power supply.

I then took weekly measurements of the end of charge voltage of each cell in the pack. The Zivan nicely ramps the current down to 0A to keep the pack voltage typically between 69.6 and 69.7V and is very consistent. During the last stage of charging, the CV stage, it keeps the voltage constant and will start charging again if the voltage drops a little but it only puts in a few mA of current. It is during this time, when no current is going into the pack, that I take my voltage measurements. After a couple of weeks I saw that the cells didn't drift much, if at all, so I've gone to approximately monthly measurements. My voltage measurements have been out the the thousandths of a volt. The biggest difference between the top and bottom cells has been 0.081V and the last reading the difference is 0.071V. The high and low cells in the pack have been the same for the past 4 months. The previous high and low cells are still very close to the current high and low cells. I'll see as the battery pack temperature goes back up if the high and low cells change back to the original two.

About 600 miles after installing the 2p20s pack I installed a CycleAnalyst to track Ah in and out of the pack. This is a great device because it gets all its power from across the entire pack, relying on nothing else, so it doesn't put any imbalanced drain on the batteries. Between Ah and voltage sag I have a relatively good idea how my pack is doing.

However, when my pack temperature is somewhere below 10 degrees C a greater than 1.5C battery current will cause the LVT of the BMS boards to trigger even though the pack is fully charged. The pack voltage has generally been around 58.6V when it triggers which matches with the conservative 2.93vpc LVT point. Under heavier loads the voltage drops even farther. The problem I see is that with this particular simple system I have no way to know if all the BMS boards are triggering or if only one is. Furthermore, the BMS boards can only measure terminal voltage and not the true voltage of the cell so they will give a false signal when under load. This makes it quite easy to begin ignoring the warning which may make it easier to kill one or more cells at some point in the future. Right now I have plans to build a Lee Heart type Batt-Bridge circuit to compare the voltages of each half of the pack. The probability of two cells, one in each half, going out of range simultaneously is low. I plan on having an analog meter with a zero center and whose needle will swing one way or the other depending on how far out of balance the half-packs are. This will at least show me if a cell is going bad and what side of the pack it is in. I can then go look for it manually. This is much better than my BMS which merely says one or more cells are below 2.93V. If the half-packs' voltages are the same I don't care if the pack voltage reading is 56V if I have not used much energy out of the pack. It merely means both halves are working equally and I don't get lulled into ignoring my current BMS warning. If a cell is going low on one side then I will know I need to take care of it before charging my pack unattended to avoid a potential fire. A Batt-Bridge type circuit also has the added benefit that there are fewer points of failure since there are only three fused wires going to the battery pack rather than a long daisy chain of one or more wires from cell to cell or a whole bunch of spaghetti going to a central system. Furthermore, it would eliminate the potential uneven drain on the cells due to variations in one BMS module to the next and it would eliminate the possibility of the BMS, if it has shunting capabilities, draining the cell dead as has happened to some people. It would not take much to install a monitoring circuit which would turn off a relay to the charger if a cell went high at the end of charge.

I have been constantly trying to find info about how this specific flavor of lithium ion cells behave. It is extremely important to note that I'm only talking about LiFePO4 cells and not any other chemistry. Unlike other battery chemistry such as lead acid, NiMH, NiCd, etc. these cells use a Li ion transfer and don't have a shuttle reaction. Therefore there is no mechanism for self discharge. Only discharge external to the cell, assuming no mechanical failure in the cell. This is consistent with the fact that when I had flooded lead acid batteries in my Gizmo it went 4 miles/kWh measured from the wall. With the same charger I now go a little under 6 miles/kWh measured from the wall. Furthermore, I'm not seeing any drifting of the cells from each other. Naturally time will tell but I don't expect to see this behavior change. Being a curious Physicist I will be checking periodically, however.

At this point I don't feel that the added risks associated with a cell level BMS of any sort is worth it. A good charger which has a consistent turn off voltage, a Batt-Bridge type of circuit, an Ah counter, ammeter, and a volt meter are all that is necessary to have a safe LiFePO4 battery pack in an EV. If your pack has been top-balanced I would bottom balance it to minimize the possibility of reversing a cell in the event that you push the range of the vehicle. Bottom balancing also gives the added benefit that as the capacity of the pack diminishes the normal 80% DOD creeps toward 100% DOD and cell reversal is more likely if only Ah counting is done. Naturally, if the driver is observant, it will be quite obvious from voltage sag and the Batt-Bridge circuit that the pack is nearly depleted before reversing a cell.


So in summary...
How many miles, Km have you driven without a BMS?
5500 miles so far, the last 2800 without using the BMS top balance function.

What sort of problems have you experienced?
BMS LVT when batteries are cold with only a 1.5C current. Also, I don't know how many or which cells are sending the LVT signal.

What instrumentation do you feel is most beneficial/necessary to safely do this?
Besides a charger that has a consistent end of charge voltage cutoff it is necessary to have at a minimum a pack volt meter, Ah counter and Ammeter. I think a half-pack voltage comparison circuit is also in order to catch an errant cell so it can be dealt with. It might be beneficial to have a limp mode triggered after a certain number of Ah have been used for those who are inclined to "drive until empty."

Do you still drive/recommend no BMS or have you changed your mind?
My mind has changed from the BMS side of things to the string level, or no BMS side of things. Since I have top balanced my pack and thus pushed all the differences in cell capacity to the bottom end of the SOC curve I will be bottom balancing my pack when the weather warms up. I will still monitor cell voltages at end of charge because I'm a physicist and want to see if my current position remains valid. I'm quite sure it will but it will be helpful to validate my position and help others using LiFePO4 batteries go a safer more cost effective route.

HTH.
 

·
Registered
Joined
·
2,494 Posts
In a few days I will have had my TS-LFP100AHA battery pack in my Gizmo for a year. I now have 5500 miles on them. For the first 2700 miles I top balanced my 2p18s pack to 4.00vpc with a Black Sheep Technology cell level BMS system. I didn't like charging that high so I had my Zivan charger properly programmed for charging curve #612 and 19 cells so a cut off voltage of 69.3V. I then installed two more cell pairs to make my pack a 2p20s pack. I also adjusted the voltage calibration pot in the charger to an ending charge of 69.7V so now I charge to 3.485vpc, well below the shunting voltage of the BMS modules. Before my first drive with the full pack I top balanced everything with the 4.00vpc BMS setup and a bench top lab grade power supply.

I then took weekly measurements of the end of charge voltage of each cell in the pack. The Zivan nicely ramps the current down to 0A to keep the pack voltage typically between 69.6 and 69.7V and is very consistent. During the last stage of charging, the CV stage, it keeps the voltage constant and will start charging again if the voltage drops a little but it only puts in a few mA of current. It is during this time, when no current is going into the pack, that I take my voltage measurements. After a couple of weeks I saw that the cells didn't drift much, if at all, so I've gone to approximately monthly measurements. My voltage measurements have been out the the thousandths of a volt. The biggest difference between the top and bottom cells has been 0.081V and the last reading the difference is 0.071V. The high and low cells in the pack have been the same for the past 4 months. The previous high and low cells are still very close to the current high and low cells. I'll see as the battery pack temperature goes back up if the high and low cells change back to the original two.

About 600 miles after installing the 2p20s pack I installed a CycleAnalyst to track Ah in and out of the pack. This is a great device because it gets all its power from across the entire pack, relying on nothing else, so it doesn't put any imbalanced drain on the batteries. Between Ah and voltage sag I have a relatively good idea how my pack is doing.

However, when my pack temperature is somewhere below 10 degrees C a greater than 1.5C battery current will cause the LVT of the BMS boards to trigger even though the pack is fully charged. The pack voltage has generally been around 58.6V when it triggers which matches with the conservative 2.93vpc LVT point. Under heavier loads the voltage drops even farther. The problem I see is that with this particular simple system I have no way to know if all the BMS boards are triggering or if only one is. Furthermore, the BMS boards can only measure terminal voltage and not the true voltage of the cell so they will give a false signal when under load. This makes it quite easy to begin ignoring the warning which may make it easier to kill one or more cells at some point in the future. Right now I have plans to build a Lee Heart type Batt-Bridge circuit to compare the voltages of each half of the pack. The probability of two cells, one in each half, going out of range simultaneously is low. I plan on having an analog meter with a zero center and whose needle will swing one way or the other depending on how far out of balance the half-packs are. This will at least show me if a cell is going bad and what side of the pack it is in. I can then go look for it manually. This is much better than my BMS which merely says one or more cells are below 2.93V. If the half-packs' voltages are the same I don't care if the pack voltage reading is 56V if I have not used much energy out of the pack. It merely means both halves are working equally and I don't get lulled into ignoring my current BMS warning. If a cell is going low on one side then I will know I need to take care of it before charging my pack unattended to avoid a potential fire. A Batt-Bridge type circuit also has the added benefit that there are fewer points of failure since there are only three fused wires going to the battery pack rather than a long daisy chain of one or more wires from cell to cell or a whole bunch of spaghetti going to a central system. Furthermore, it would eliminate the potential uneven drain on the cells due to variations in one BMS module to the next and it would eliminate the possibility of the BMS, if it has shunting capabilities, draining the cell dead as has happened to some people. It would not take much to install a monitoring circuit which would turn off a relay to the charger if a cell went high at the end of charge.

I have been constantly trying to find info about how this specific flavor of lithium ion cells behave. It is extremely important to note that I'm only talking about LiFePO4 cells and not any other chemistry. Unlike other battery chemistry such as lead acid, NiMH, NiCd, etc. these cells use a Li ion transfer and don't have a shuttle reaction. Therefore there is no mechanism for self discharge. Only discharge external to the cell, assuming no mechanical failure in the cell. This is consistent with the fact that when I had flooded lead acid batteries in my Gizmo it went 4 miles/kWh measured from the wall. With the same charger I now go a little under 6 miles/kWh measured from the wall. Furthermore, I'm not seeing any drifting of the cells from each other. Naturally time will tell but I don't expect to see this behavior change. Being a curious Physicist I will be checking periodically, however.

At this point I don't feel that the added risks associated with a cell level BMS of any sort is worth it. A good charger which has a consistent turn off voltage, a Batt-Bridge type of circuit, an Ah counter, ammeter, and a volt meter are all that is necessary to have a safe LiFePO4 battery pack in an EV. If your pack has been top-balanced I would bottom balance it to minimize the possibility of reversing a cell in the event that you push the range of the vehicle. Bottom balancing also gives the added benefit that as the capacity of the pack diminishes the normal 80% DOD creeps toward 100% DOD and cell reversal is more likely if only Ah counting is done. Naturally, if the driver is observant, it will be quite obvious from voltage sag and the Batt-Bridge circuit that the pack is nearly depleted before reversing a cell.


So in summary...

5500 miles so far, the last 2800 without using the BMS top balance function.


BMS LVT when batteries are cold with only a 1.5C current. Also, I don't know how many or which cells are sending the LVT signal.


Besides a charger that has a consistent end of charge voltage cutoff it is necessary to have at a minimum a pack volt meter, Ah counter and Ammeter. I think a half-pack voltage comparison circuit is also in order to catch an errant cell so it can be dealt with. It might be beneficial to have a limp mode triggered after a certain number of Ah have been used for those who are inclined to "drive until empty."


My mind has changed from the BMS side of things to the string level, or no BMS side of things. Since I have top balanced my pack and thus pushed all the differences in cell capacity to the bottom end of the SOC curve I will be bottom balancing my pack when the weather warms up. I will still monitor cell voltages at end of charge because I'm a physicist and want to see if my current position remains valid. I'm quite sure it will but it will be helpful to validate my position and help others using LiFePO4 batteries go a safer more cost effective route.

HTH.
Nice write up... thanks for the candid report. This is good information for others to help them understand based on your experiences.
 

·
Registered
Joined
·
194 Posts
Car: Blackhawk 100ah cells
12000 miles on pack without a bms. Only problem is in winter when unheated cells are about 1ah apart from the heated cells ever 15 cycles of 40ah.

Added mini bms now put 6000 miles additional on the pack (about 18000 total.) No issues just let the amp hour meter in the car count down to zero on the weekend 1 time a week and cells have stayed in line. All other days I cut it off at .3ah left to charge. That last .3 ah takes about 30 minutes.

Scooter TS 60 ah cells 2.5 years no bms about 4500 miles no issues. top ballanced them 1 time at about the 1.5 year mark but not really needed. No idea where they were when I bought the scooter.


I drive 18000 miles a year with a 54 mile round trip commute daily mon- friday.

No real issues to speak of but I monitored my cells like a crazy man. Now with the bms I dont car. I get in and drive and plug in and wonder away to do other things.

I check the bolts 1x a month and check the communication wiring/connectors/placement of the bms boards 1x a month as well. That takes about 30 minutes.
 

·
Registered
Joined
·
1,858 Posts
Discussion Starter #6
Thanks David! That's very good data. It's people like you who give of their time to write a detailed post that is so very helpful to anyone interested. I also checked out your blog site which was very informative. I too am going to clean my terminals and use Noalox. I use it now on my lead pack.Here's a link for others to view.
 

·
Registered
Joined
·
8,106 Posts
  • How many miles, Km have you driven without a BMS?
  • What sort of problems have you experienced?
  • What instrumentation do you feel is most beneficial/necessary to safely do this?
  • Do you still drive/recommend no BMS or have you changed your mind?
About 5K miles.
None.
Amp meter and Ah counter, I use the EV Display.
I can't recommend one way or another, it does take some work to run without a CMS ;) but I know some people who use them but, because of previous failures, still do manual checks even more often than I do.
If I were convinced of a system that could do no harm and were cheap enough I might be tempted to get one.
 

·
Registered
Joined
·
440 Posts
I've been following the heated BMS debate here and thought it would be beneficial to gather data on those of you who have driven with no BMS on your batteries.

I post this because a lot of people are in the situation I'm in. Ready to make the switch but not sure which path to take. Hopefully you guys can steer us with your experiences. If you would take a minute and give us some data it would be greatly appreciated!


  • How many miles, Km have you driven without a BMS?
  • What sort of problems have you experienced?
  • What instrumentation do you feel is most beneficial/necessary to safely do this?
  • Do you still drive/recommend no BMS or have you changed your mind?
I've ordered my pack and have a month or so until they arrive. I still have not decided whether I really need a BMS or not. I am convinced I don't need shunting ability during the cell's early life but not sure if it would be beneficial when the cells have a few hundred cycles on them and are aging a bit.

If you drive with no BMS we really need to know if you have alternative measuring equipment installed or just drive very conservatively or generally what it is that makes your decision to go without a BMS a viable option. If it wasn't a good decision please help us and tell us why. Thanks in advance! :)

I think we need to define BMS or not into 3 stages:

!: Auto BMS= circuit board mounted on the individual cell. (several brands and technology out there).

2: Manual BMS= keep a close eye on cell voltage and monitor cells continually.

3: No BMS= I just drive and maybe check Voltages every 6 months with a VOM

francis
 

·
Registered
Joined
·
4,293 Posts
I check the bolts 1x a month and check the communication wiring/connectors/placement of the bms boards 1x a month as well. That takes about 30 minutes.
have you had terminal bolts loosen?
do you use loctite?
what kind of lock washers?
 

·
Registered
Joined
·
194 Posts
have you had terminal bolts loosen?
do you use loctite?
what kind of lock washers?
I always have to put a bit more onto them every month from 120inch pounds here and there. Id say worst losened to 110 inchpounds. I use lockwashers I got from home depot nothing special. No loctite. So of the 52 cells perhaps 3 or 4 need a bit more to get them to 120 again. It doesnt seem to be the same 3 or 4 every time.
 
G

·
Guest
Joined
·
0 Posts
OK Guys, Lets keep this on track rather than hijack the thread.

Help! Mileage Driven, Problems & no BMS for Lifepo4
 

·
Registered
Joined
·
542 Posts
Only problem is in winter when unheated cells are about 1ah apart from the heated cells ever 15 cycles of 40ah.
Hi Dex
I think you said you had to bleed charge from the unheated cells.
When you mentioned this on another thread, I asked if the heater uses only the heated batteries, or the whole pack. This would then be a parasitic load problem rather than a battery problem.
If you answered, it was lost in the noise of the BMS debate, sorry.
Gerhard
 

·
Registered
Joined
·
194 Posts
Hi Dex
I think you said you had to bleed charge from the unheated cells.
When you mentioned this on another thread, I asked if the heater uses only the heated batteries, or the whole pack. This would then be a parasitic load problem rather than a battery problem.
If you answered, it was lost in the noise of the BMS debate, sorry.
Gerhard
yes i missed that question. The battery heat works off of 220Vac. So it only works when the ev is plugged in. Which is 21 hours a day in the winter. I plugin at work and at home. The heater doesnt work off the batteries.
 

·
Registered
Joined
·
8,106 Posts
That makes me wonder about something. The colder cells have higher internal resistance, which means for the same current the voltage is driven higher, but of course they aren't really getting more charge so they really aren't at a higher SOC than warmer cells. If the CMS turns on and burns off charge to keep the voltage in line with the others, or if you bleed off charge manually, the colder cells actually end up at a lower SOC than warmer cells. Unless each CMS has temperature compensation.
 

·
Registered
Joined
·
903 Posts
That makes me wonder about something. The colder cells have higher internal resistance, which means for the same current the voltage is driven higher, but of course they aren't really getting more charge so they really aren't at a higher SOC than warmer cells. If the CMS turns on and burns off charge to keep the voltage in line with the others, or if you bleed off charge manually, the colder cells actually end up at a lower SOC than warmer cells. Unless each CMS has temperature compensation.
I was also wondering what would happen if all the batteries were taken to the same temperature and then checked for balance. Would they all turn out the same? Drive for the week or month then find someone with a heated garage and park there over night then check the status of the cells.

I have only one battery box but I could see the cells around the edge being at a different SOC than the ones in the center due to temperature differences. My Zivan still charges to the same voltage regardless of temperature and since it tapers down to zero amps if necessary the batteries aren't pushed higher than with warmer temperatures.
 

·
Registered
Joined
·
1,858 Posts
Discussion Starter #17 (Edited)
Two things. MikeP posted about a customer who has a pack with 10000 miles on it. Drove first 1000 with a BMS when a cell failed at which time he removed it because he was told that's what killed the cell. He's driven the last 9000 BMS free but no details what he had for instrumentation.

I've looked into the new Cycle Analyst which I think will fit my needs. Gizmo uses one. I emailed them with a question of the wh counting which I read stopped at 10kw. Justin told me the high voltage unit which is rated for 200V is good for 100kw, actually 99,999wh.

They have a new large display suitable for mounting in a project box that displays volts, speed, kw used, amps and Ah among other things. It also has a low voltage cutout of some sort built in. And it's inexpensive, very much so I feel for all the things it does.

I mentioned the split pack monitoring with LED readout idea as a future integration possibility to them and Justin replied that a BMS is a must. I suppose it would be necessary if you're top balancing, trying to charge the pack to the max and occasionally discharge it close to 100% DOD which appears to be the mainstream thinking.
 

·
Registered
Joined
·
2,318 Posts
The colder cells have higher internal resistance, which means for the same current the voltage is driven higher, but of course they aren't really getting more charge so they really aren't at a higher SOC than warmer cells. If the CMS turns on and burns off charge to keep the voltage in line with the others, or if you bleed off charge manually, the colder cells actually end up at a lower SOC than warmer cells. Unless each CMS has temperature compensation.
Yes, but if you are top balancing the cms won't start shunting until a cell is at or above the knee of the curve, since the change in cell voltage with temperature is small, so the difference in SOC will be negligible. For example, the minibms starts shunting at 3.5V, so unless the temperature difference is really large, I think even the colder cells should be close to the knee when shunting starts.
 
1 - 19 of 19 Posts
Top