Thread Tools
Sep 27, 2004, 01:01 PM
Registered User
Dan Baldwin's Avatar
My mistake. With my color blind eyes, it looked like the resistor shown in the photo just below the capacitor has a color code of brown black black which would be 10 ohms, so I had somebody else take a look at it and confirmed that it is brown black red which is 1k. Sorry.

I should have realized and mentioned that the 120 ohm resistor used to limit the current through the LED is too low a value for your 8.3 volt shunt regulator. It should be about 220 ohms.

150 ohms for the resistor that goes between the TL431 and the base of the TIP137 should be fine.

If you want to use a resistor to make the transistor run a little cooler, that would be fine, in fact it would also help limit current in the event that the TIP137 shorts out. I don't like the idea of putting the transistor right across the battery without any protection at all.


Dan
Sign up now
to remove ads between posts
Sep 27, 2004, 07:31 PM
cpp
cpp
Registered User
Fortunately I have a black jumper wire going from the TIP137 collector to ground, so I can cut that jumper and easily insert a low ohm high watt resistor. I'm going to do that because, like you point out, I'm a little uneasy just grounding that collector.

On the diode, I have a 330 ohm resistor from the collector of the transistor to the diode. The diode seems to operate fine so I guess I'll leave it instead of changing it to a 220 ohm resistor.

Congratulations on your SS experiment with the autonomous flight with the gyro. I'm planning on doing much the same thing with my solar plane, since who wants to actually fly it for hours. You can just "park" it in the sky if you can get the stability worked out and then get a small computer that reads a GPS output and then controls the rudder.
Sep 27, 2004, 07:45 PM
Registered User
Dan Baldwin's Avatar
330 ohms on the LED is fine. Too little is better than too much current.

I think that the tricky part will be parsing the GPS data, and turning it into useful information. I'm not looking forward to playing with that code. I don't think it would be too bad on a PC, but it's a whole different thing on a microcontroller.

Do you have a use for this plane once you get it flying for an hour, or is it just for fun?

Dan
Sep 28, 2004, 12:54 PM
cpp
cpp
Registered User
I'm also into aerial photography, so what I'd like to do eventually is get a camera, GPS receiver, and small computer on board. Then I'd like to be able to tell it to go to lattitude X, longitude Y, with height H and take pictures to the N, S, E, and W and then come back home.

It requires a very stable plane and also the computer has to control the rudder, throttle, and camera. The throttle control is for climbing or descending, but also to turn the motor off when actually taking the pictures.

My current plane could probably support all this but it would be a little heavy and I might only have an hour's duration in the sunlight. I have the camera working and the plane is pretty stable so next is the GPS and computer. One advantage I might have is that I have been programming all my life so I look forward to the software part of the project.
Sep 28, 2004, 01:07 PM
Registered User
Dan Baldwin's Avatar
I have done some digital photography with my slow stick, and you're right, I had to kill the motor when I took the pictures, or I ended up with waves in the picture. Another guy at the field uses his SS for taking pictures, and he gets paid for it. He has some OUTSTANDING pictures. When he showed me his, I didn't go running to the car to show him mine. As I have said before, I don't have a use for an autonomous plane, so maybe I'll just adopt yours. Sounds like a good one.

Once I get to that point, I'm going to see if I can use PICs to do all the math on the data from the GPS, but I have serious doubts that I'll be able to pull it off.

Dan
Oct 18, 2004, 08:49 AM
KG7TTQ - Las Vegas
mark_q's Avatar
Dan,

Any chance you will produce a pcb for this little gem (the balancer)? I'd certainley pay $5 a pop for the added cleanliness this would add. Even better if you could make a 4S, 5S or nS board with snap off sections.

Mark
Latest blog entry: 390 Degree Geared Pan and Tilt
Oct 22, 2004, 12:13 PM
Registered User
Dan Baldwin's Avatar
Sorry for the delay asnwering Mark. Yes, I can make PC boards in small quantities. Dunno how much I would have to get for them. probably wouldn't be too much.

Dan
Nov 12, 2004, 11:32 AM
Registered User
Hi Dan ! IŽam confused about some changes here. Can you post the correct schematic circuit with the values of the components? (4,2V).

Thanks for your information!

Sergio TH.
Nov 12, 2004, 12:06 PM
Registered User
Dan Baldwin's Avatar
All the schematics that I have posted in this thread should be correct. The schematic and parts list shown in post #1 is the latest version of the simple balancer. The schematic in post #128 should also work fine, but it has not been tested yet. It has another LED that shows when high balancing current is being drawn.

Dan
Nov 16, 2004, 04:23 PM
Got shenpa?
flieslikeabeagle's Avatar
Dan, many thanks for the circuit!

I have one suggestion to add: it seems to me that when using your charge balancers, there is no longer any need to use a dedicated Lipo charger as the power source. All that is really needed is a current source, set to the 1C charge rate of the cells being charged. The current source will ensure that the peak charge current stays within the 1C spec, and your charge balancers will see to it that the charge termination voltage stays within the 4.2V max spec!

I'm attaching a quick schematic of my idea, using an LM 350 3-pin voltage regulator and a resistor to create the current source. The max charge current is set by Rs = 1.25/Imax. So for instance, a 1.2 ohm resistor will result in a charge current of about 1 A. The blocks marked "CB" are your charge balancers.

Many inexpensive DC power supplies should be suitable for powering something like this; All Electronics has a 20V, 3A switching power supply that should do nicely for 2S and 3S packs (with appropriate heatsinking on the LM 350 for larger currents): http://www.allelectronics.com/cgi-bi...231&type=store

-Flieslikeabeagle
Nov 16, 2004, 05:12 PM
Registered User
Dan Baldwin's Avatar
Yes, it would be possible to make a charger using just the balancers and a current source (I've used a light bulb to limit current), in fact that has been discussed in this thread, but it's not something I suggest. The balancers would be sinking 4.2 watts at 1 amp once the battery reaches full charge, so you would have to have big heat sinks for all the balancers as well as the current source, and there is no indication that the pack is charged. If you're going to go to the trouble of making a current source anyway, you might as well make a simple charger like the one I have posted here (https://www.rcgroups.com/forums/showthread.php?t=270573), or one of Scott Henion's.

Dan
Nov 16, 2004, 08:21 PM
Got shenpa?
flieslikeabeagle's Avatar
Thanks for the reply, Dan. I missed the mention of a constant current source in the thread (found the thread today, didn't wade through all 159 posts).

I'd like to discuss the issue of the balancers dissipating full power when the batteries are fully charged a little more. It seems to me that if the charger (usual CC/CV lipo charger) is set to a higher charge termination voltage than the sum of the balancer voltages, even by a few mV, then the charge balancers will end up dealing with the full current limit of the charger at charge termination anyway. This will happen because the charge balancers are trying to clamp the voltage to a value lower than the CC/CV regulator, so with its high internal gain and feedback it will go into the CC mode, putting out the full current it is designed to supply. To be precise, the output current will try to attain the value (Vcharger - Vbalancers)/(Rcharger+Rbalancers) , where Rcharger and Rbalancers are the Thevenin resistance (output resistance) of the charger in CV mode and of the charge balancer circuits respectively. Since Rcharger and Rbalancers are very low by design (good load regulation), this current value will be large, so the current will rise till clamped at the CC value designed into the lipo charger. Of course the cells themselves will draw negligible current since they're charged, so the charge balancers have to deal with the full current.

I was thinking of this scenario, thats why I suggested a plain old CC source would work as well as the CC/CV charger circuits.

If, on the other hand, the CC/CV chargers charge termination voltage is set slightly lower than the voltage at which the charge balancers kick in - by even a few mV - this situation does not arise. The CC/CV charger stays in CV mode at charge termination, and the charge balancers stay off, dissipating only quiescient power.

I guess the proper procedure is to set the charger to terminate at, say, 4.15 V/cell, and the charge balancers for 4.2 V/cell. If the charger is commercial and not user adjustable, it means the trip voltage for the charge balancers needs to be set to suit the charger (slightly higher than the chargers charge termination voltage), rather than to 4.2 V as you recommend.

Is all this going to work over the normal range of operating temperatures? Both the charger and the charge balancer are going to have some temperature coefficient. And there is so little margin of error with the Lipo cells - only 50 mV/cell between 4.15 V and 4.2 V limits.

-Flieslikeabeagle
Last edited by flieslikeabeagl; Nov 16, 2004 at 08:25 PM.
Nov 16, 2004, 10:21 PM
Registered User
Dan Baldwin's Avatar
Yes, the balancers need to be set a little over the voltage of the charger to make sure that the balancer doesn't end up sinking the entire charger current. I suggested to Larry that the best way to calibrate the balancers would be set them to a voltage that just allows the charger to come to a full charge and drop out. If all the balancers come on at the same time, they are set too low.

To be perfectly honest, as I stated at the beginning of this thread, I haven't used any of these to balance a pack, so I can't tell you imperically that they will work over the entire normal temperature range with commercial chargers. According to the spec sheet the reference voltage of the TL431 would change about .01 volts over the entire temperature range, which would equate to about .017 volts (17 MV/cell) change in balancer voltage. If the balancer were set .03 volts per cell over the voltage of the charge, it should be sufficient.

Dan
Nov 17, 2004, 01:33 AM
Got shenpa?
flieslikeabeagle's Avatar
Quote:
If the balancer were set .03 volts per cell over the voltage of the charge, it should be sufficient.
...but the charger will have its own temperature coeff, too. For instance your charger design, using the same TL431 IC, will probably end up with about the same temperature coeff of about 17 mV/cell. Since the temperature of the charger and charge balancer do not track each other, one could potentially have up to 34 mV/cell temp coeff (worst case).

I was all ready to forget about the parallel charge/series discharge approach and use charge balancers instead. But now I am surprised by the precision needed to do the charge balancing! By contrast the parallel charge/series discharge method is relatively simple and robust - only one 4.2V CC/CV source to get right. Too bad that approach involves finding six-pin connectors (for 3S) or using multiple connectors with fewer pins.

Out of curiosity, have you been keeping an eye on your multi-cell packs and finding they do not need balancing, or are you using the parallel charge/series discharge method to eliminate charge imbalance problems?

-Flieslikeabeagle
Nov 17, 2004, 10:42 AM
Registered User
Dan Baldwin's Avatar
At present I don't have any battery packs with taps, so I can't even check to see if they are out of balance. I would never use parallel charging and series discharge because the only way to do that is to build your own packs, and I won't do that. The pack manufacturers do do some cell matching, which I can't do at home, and it's just too easy to break off a tab and scrap out a cell. I'll leave battery building to others.

Any packs I buy in the future will have taps, and if I ever have to balance them, I'll use balancers. I don't think that the temperature coefficient is as big a problem as you do. If I had to put together a balancing system I would just read the voltage on my batteries just before the charger quits, divide by 3 (if 3 cells) and set my balancers for that voltage plus maybe 10 or 20 MV. It doesn't matter that my meter might be out of calibration because I'm using the same meter for both readings. If I do ever end up with all 3 balancers turned on at the same time for some reason, then I would bump the voltage on the balancers up a little. If the temperature coefficient does end up being a problem, all you would have to do is put a very small resistor (maybe just long skinny leads) on the balancers to limit the current at very low out of balance voltages.

The PolyQuest charge guard uses 4.35 volts as the overvoltage setting, and it is considered safe. My understanding is that the 4.2 volts typically used by chargers leaves a good safety margin, although some say that if you only charge to 4.1 volts you will get much more life out of your batteries.

Dan


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
How to avoid high-current interference? swami Beginner Training Area (Aircraft-Electric) 8 Jul 25, 2001 04:07 PM
High current performance, RC3000 NiCads? BobStewart Electric Plane Talk 21 Jul 15, 2001 10:50 PM
High current fuses/fuse holders Paul Electric Plane Talk 13 Jun 26, 2001 08:02 AM
High current sub C question. DavidWillis Power Systems 5 Jun 19, 2001 07:00 AM
Vendors for high current/capacity single cells? Paul Electric Plane Talk 2 Jun 06, 2001 11:02 PM