resistor size for bypass circuit ?

Chat about all things battery in here.
Need to know what type to use or size or capacity then again place your thoughts here
hybridsae
Posts: 6
Joined: Fri Jan 06, 2012 1:10 am

resistor size for bypass circuit ?

Postby hybridsae » Wed Feb 15, 2012 11:19 pm

Hello,
On charging, I'm trying to bypass 40Ah Lithium Polymer batteries with rated internal resistance of less than or equal to 2.5 milli ohms. What would be a good resistance value to bypass these batteries with in order to balance the cells upon charge??? Also, what should its wattage rating be? I'm guessing I would have to use very small size, high power resistors that can dissipate the heat well.

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: resistor size for bypass circuit ?

Postby retepsnikrep » Thu Feb 16, 2012 7:21 am

I assume this is with some sort of BMS controlling the resistors to bypass the current?
What is the maximum charged voltage for the cells?

R = V/I
W = V*I

So if you want a current of 400ma at 4V then R = 10

.4A x 4V = 1.6W of power will be dispiated in the resistor so it needs a minimum of 2W rating.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

hybridsae
Posts: 6
Joined: Fri Jan 06, 2012 1:10 am

Re: resistor size for bypass circuit ?

Postby hybridsae » Fri Feb 17, 2012 5:09 pm

Yeah we have a BMS controlling when the bypass resistor turns on. We are reading each individual battery voltage to a microcontroller and then we are going to send a signal to turn on an electromechanical relay (with a resistor) to bypass the battery cell depending on the voltage upon charge. Anyways, we have a bank of 32 cells so approximately 120V altogether, but each cell will only be at 2.5 - 3.8V with a resistance of 2.5m ohms. We will be charging the batteries at 40A of current and am unsure of how much current we should bypass in order to balance the cells upon charge. Are you suggesting R = 3.8V/40A so roughly 0.1 ohm? Should we try and bypass all the current? 2.5m ohms is basically just a wire. We should obviously use some resistance so we don't just completely short the battery right?

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: resistor size for bypass circuit ?

Postby GregsGarage » Fri Feb 17, 2012 11:44 pm

You could hold the charger current at a constant 40 amps for the duration of the charge and bypass the current around each cell, but that is about 150 watts of heat you need to safely dissipate from each cell. If all loads are on you would need to safely get rid of nearly 5kw of heat! :shock:

Firstly make sure you pre-balance all the cells first. Connect them in parallel and bring them all up to say 3.8 volts. You will then need to do much less balancing during charging. Also there is really no such thing as a perfectly balanced pack. If you get them all balanced at the top and discharge them you will find some get discharged before the rest, and vice-versa. You only need enough balance to make sure that you can use all of the capacity of the weakest cell.

Next, the 2.5m ohms internal cell resistance isn't a constant value. It changes depending on the state of charge. It goes up as the cell nears full charge.

Normally you charge the cells at CC until one of the cells reaches 3.8 (I charge to 3.65) then you switch to constant voltage, reducing current so the highest cell doesn't go over 3.8. At some point you need a load to bypass some current around the highest cells so the rest can catch up. I prefer to use a small amount of bypass current, maybe just a watt or 2 per cell. Sure it takes longer to balance, but the advantages are minimal fire risk and less chance of destroying a cell if a load accidentally sticks on. A 40 amp load on a 40ah cell can destroy it in an hour. A 0.4 amp load gives you 100 hours to notice the fault, :) and is more than adequate to keep the pack in balance.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

dbecker
Posts: 2
Joined: Sat Feb 18, 2012 7:48 am

Re: resistor size for bypass circuit ?

Postby dbecker » Sat Feb 18, 2012 4:52 pm

You might want to re-think your balancing design.

Using a bypass balancing resistor to bypass charge current requires a large resistor that can handle the heat, and risks a fault in the bypass circuit deeply discharging the cell.

A better approach is to expect that the BMS is actually intelligent. It shouldn't try to balance the pack during charge. Instead it should balance after the charge, based on what it has observed during operation and charge.

One strategy is to do a top balance when the pack is charged to near 100%. The BMS tracks when each cell moves off the flat part of the charge curve, observing the voltage slope over time dV/dt rather than relying solely on the absolute voltage. (This requires recording the voltages and making a retrospective decision.) As the charge cycle moves from CC to CV, it should note any cells that are still in the flat bulk charge region. As the cells hit the target full-charge voltage the BMS should note the time and Coulomb count for each cell. Only after the first cell hits the upper end of the target CV voltage, should the BMS start balancing. It should immediately remove the overcharge from all cells that are the upper end of the CV voltage range, limited by the coulomb count that it took to move from the target voltage to charge termination. It should then continue to remove a small proportion of the excess charge from any cells that are identified as being over 1% higher charge than average, based on the voltage rise knee.

Using this approach has many advantage over online bypassing.

The voltages can be monitored more accurately without the distortion of bypass current.

It minimises the time cells spend slightly overcharged. (Plating damage doesn't immediately happen from an overcharge, it happens over time from the continuing high voltage. A BMS that only bypasses only while charging leaves the top cells overcharged.)

Bypass current can be low, with smaller components and less likelihood of a erroneous full discharge. At the lower current, using an LED as a discharge element is feasible. This provides both a visual indication of balancing and puts a lower limit on the discharge voltage.

It gradually brings a pack into balance, which is likely to result in less cycle wear than attempting to match voltages each charge cycle.

hybridsae
Posts: 6
Joined: Fri Jan 06, 2012 1:10 am

Re: resistor size for bypass circuit ?

Postby hybridsae » Thu Mar 01, 2012 5:48 pm

So I have a lot more questions and concerns because the internet doesn't cover any of this stuff with any justice and obviously most people with designs keep them private. This forum is mint.

GregsGarage –
You are saying that I should completely take the batteries out of the pack and charge them in parallel? If you charged them in parallel, you would still keep them charging at a constant current? In parallel the currents could be different so would that not be difficult to control? Or would you charge them with 3.65V supply? What kind of current would you throw at them?
- I also don’t know how feasible this is since they already have series connected plates and also the monitoring boards will eventually be connected across the batteries probably. Do you have a BMS that runs them in parallel or series? Or do you take them out of the vehicle and balance them in parallel and then connect them and run it in the vehicle? I just tried to keep it simple and not sure what would work best. I just know we need a lot of voltage to run our motor on the other end so they need to discharge in series I believe.
Right now I have a pulse width modulating circuit to control the current and keep it constant (with a current sensor) in order to allow the batteries to charge fully from the generator (motor) and so I'll keep it at a constant current until they reach 3.65V but then I am very unsure of how much to decrease the current and how slowly in order to maintain the batteries at 3.65V-3.8V and not much more. I’m just unsure of how slowly or fast to decrease the current and keep the current going through the cells until one cell reaches complete full charge? As you decrease current, I should monitor the pack voltage and keep that constant with trying to decrease current? Will the batteries accept more (but lower) current while keeping its voltage around 3.8V? I also don’t know how high the resistance increases as it nears full charge to judge how much current will then flow through any bypass circuit I implement (to try and get the 0.4A or so). And 0.4A bypass would actually balance them a little?

Dbecker –
What you said confused me a bit. I will rephrase what you said and ask questions in between in order to see if I understand correctly. First off, is there a difference between target full charge voltage and target CV voltage? I know text isn’t the best way to communicate an idea but what I think you are suggesting is that you charge the batteries and just monitor voltage and keep track of voltage over time. Also, I have no idea how to keep track of coulomb count nor what that really even is. So you’re suggesting that my microcontroller will sample voltage from the cells during charge (how often?) and keep track of the voltage levels over the entire charge time? Then when one cell reaches target full charge voltage (say 3.65V??) the bms goes into constant voltage and lowers the current. Then what I think you are saying is that as the first cell goes from 3.65V (after constant current) to the target CV voltage which is maximum charge termination voltage (say 3.8V??) with NO bypassing the bms should then (and only then) balance the cells in this region? Should I then remove the excess charge from these cells and store it? or just dump it? or allow the higher voltage charged cells to transfer current and charge to the other lower charged cells (limited by the coulomb count to go from target voltage 3.65 to charge termination 3.8 as you said)? Or maybe when you say “remove” do you mean balance/bypass with a parallel LED or bypass circuit to remove its charge across itself? Then lastly, you suggest removing a small portion of charge from the cells that are 1% above the voltage rise knee voltage?

I’m just a bit confused dbecker on what the balancing technique or circuit is that you are suggesting and how you go about balancing or just simply charging them all up and then removing the excess voltage or charge on the top charged batteries. I know the risks of online bypassing but it was just a primary idea that was simple because my resources and knowledge of this material is very basic and primitive.

I’M REALLY SORRY FOR THE CONFUSION AND DUMB QUESTIONS AND MISUNDERSTANDING. I REALLY APPRECIATE YOU ALL HELPING ME HERE.

If any of you have skype and want to explain anything over skype or something audio or instant message just let me know if it would be easier than through forum text.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: resistor size for bypass circuit ?

Postby GregsGarage » Thu Mar 01, 2012 9:40 pm

hybridsae wrote:You are saying that I should completely take the batteries out of the pack and charge them in parallel?
No, I was only suggesting this as a one off before you install the pack in the vehicle. Paralleling the cells and then taking all of them to around 3.65 volts will make sure you start with them all at the same state of charge. Once a pack is balanced it should require very little in the way of balancing. If you've already installed the pack then I probably would just continue the build and see how they behave. With a bit of luck they are in balance and if not then you can decide what to do.

Today I was talking to Evan Tuer who has a Peugeot 106 running 40x 180ah cells. This pack has been in use for 3 years and he has come to the conclusion that he will balance it once a year whether it needs it of not. :shock: For daily use he doesn't bother with balancing the pack. The main thing with lithium cells is to make sure you monitor individual cell voltages and don't let them go too low or high.

I'll keep it at a constant current until they reach 3.65V but then I am very unsure of how much to decrease the current
Decrease the current as much as needed to maintain 3.65 volts.

I should monitor the pack voltage and keep that constant with trying to decrease current?
NO NO NO NO. Sorry for the shouting. You must monitor individule cell voltage and control your charger by the highest cell. Never let any cell go over it's max voltage which will happen if you only monitor pack voltage.

I also don’t know how high the resistance increases as it nears full charge
If you've got a failing cell it could be infinite. :shock: Don't try and calculate resistance, monitor voltage and control current, but also have a failsafe in case a cell goes open circuit, that switches off the charger.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

hybridsae
Posts: 6
Joined: Fri Jan 06, 2012 1:10 am

Re: resistor size for bypass circuit ?

Postby hybridsae » Fri Mar 02, 2012 2:38 am

Yes I totally plan on monitoring each individual battery voltage. I have designed and tested some analog HCNR200 optoisolators into a circuit that can portray an analog range of voltages to the microcontroller depending on the voltage of each cell. The only thing I am concerned with is whether it will measure the voltage upon charge AND discharge. You haven't had any issues measuring across a cell on charge as well as discharge? I know the current direction changes, but I guess the voltage should remain the same when measuring across the terminals of the battery correct? With the battery voltage remaining constant some of my fellow group members think the current will stay constant through this portion of the circuit, but I have a gut feeling that the current might fluctuate in any circuitry connected in parallel with the battery when charging or discharging.... I guess I'll find out eventually, but I was just wondering if you might know.

Also, do you measure the battery cell voltage constantly? This is what I had intended on doing (like have the analog optoisolators connected permanently in parallel (along with a fairly large resistor) with the batteries and then on the isolated side use a multiplexer to gather the information. However, the current will be constantly running through the battery side LED on the optoisolator. Is this dangerous because it might allow the battery to discharge through this or will it be fine if I use a rather high resistor? I can only have approximately 16mA flowing through the LED on the isolator, and I know this will blow if one of the battery's open circuit, but.... Maybe you have a suggestion for monitoring the battery voltage or I guess I could just simply have a switch on each cell and go through the switches to measure voltage. It just might be a lot more expensive that way. Also, I asked above but will ask again if the circuitry will be the same for voltage monitoring on charge as well as discharge?

Finally, does a battery become open circuited from too much voltage on it or discharging too much or both?

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: resistor size for bypass circuit ?

Postby GregsGarage » Fri Mar 02, 2012 11:39 pm

Have a look at the Lee Hart Battery Balancer. Originally designed for lead acid, it can be used with lithium as well. It takes a different approach than the BMS that Peter, myself and others have taken with the system on this site. But some of the ideas you mentioned are similar to Lee's system.

Any load left connected across the cells will discharge the cell to destruction. The time taken will depend on cell capacity and discharge current. The LHBB gets around this problem by using relays to connect to cells. When not in use, there is no drain.

We check cell voltage once a second or every 5 seconds while charging. Have a look through our bms threads for more details.

A lot of your questions will be answered by actually doing some charging and discharging of cells. Get a small rc pack to experiment on, you will learn a lot by seeing what happens during a complete charge/discharge cycle.

Finally, does a battery become open circuited from too much voltage on it or discharging too much or both?
Sometimes for no reason at all. :shock: Sh*t Happens! :mrgreen:
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

dbecker
Posts: 2
Joined: Sat Feb 18, 2012 7:48 am

Re: resistor size for bypass circuit ?

Postby dbecker » Tue Mar 06, 2012 6:47 am

hybridsae wrote:What you said confused me a bit. I will rephrase what you said and ask questions in between in order to see if I understand correctly. First off, is there a difference between target full charge voltage and target CV voltage?


They are close, but may not be the same voltage. The target CV voltage is the average voltage where you start to reduce charging current. A charger may not be able to taper the current to zero.

hybridsae wrote:Also, I have no idea how to keep track of coulomb count nor what that really even is.


It's a amp-hour accumulator -- a circuit that integrates the charging current to track the total charge. It implies a very accurate current measurement and summation over time. An accurate coulomb counter is more difficult to implement than you might expect.

hybridsae wrote: So you’re suggesting that my microcontroller will sample voltage from the cells during charge (how often?) and keep track of the voltage levels over the entire charge time? Then when one cell reaches target full charge voltage (say 3.65V??) the bms goes into constant voltage and lowers the current. Then what I think you are saying is that as the first cell goes from 3.65V (after constant current) to the target CV voltage which is maximum charge termination voltage (say 3.8V??) with NO bypassing the bms should then (and only then) balance the cells in this region?


Correct. The BMS should completely stop the charging and only then remove the excess charge from the high voltage cells. If needed it may then continue to do a partial balance.

A few BMS systems use a charge pump or boost converter to move the excess charge to adjacent cells. It's appealing to not waste the energy, but the savings will be minimal or trivial with the typical mostly-balanced pack. The circuit complexity and management software to effectively move the charge across the pack adds risk and perhaps cycle wear.

A dissipative balance (e.g. a resistor) is simpler and easier to get right. It's also easier to implement a fail-safe to prevent a complete discharge by putting a zener or LED in series.

Keep in mind that a BMS doesn't need to balance the pack automatically. It may default to stopping the charge short of a 100% charge, with a occasionally-used balance-check mode.


Return to “All things battery related”

Who is online

Users browsing this forum: No registered users and 37 guests