Comments on: Arduino battery capacity tester http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/ That which fades into the ether. Mon, 06 Feb 2012 17:47:06 +0000 hourly 1 http://wordpress.org/?v=3.1.4 By: Andy http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5568 Andy Thu, 01 Dec 2011 15:26:51 +0000 http://www.foobert.com/blog/?p=471#comment-5568 Hi Scott; I too am interesting in creating a constant-current discharger/logger. Would the TL431-based constant current sink work with 1.2v nmh batteries? Would using a MosFet in place of the BJT shown on the TI page work better? Hi Scott;

I too am interesting in creating a constant-current discharger/logger. Would the TL431-based constant current sink work with 1.2v nmh batteries? Would using a MosFet in place of the BJT shown on the TI page work better?

]]> By: Andy http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5564 Andy Sun, 27 Nov 2011 22:50:22 +0000 http://www.foobert.com/blog/?p=471#comment-5564 Thanks for the info. I do hobby electronics (passives, transistors, op-amps, etc.) , but I'm not really up on MosFets. I want to take your idea (and thank you for posting it) and modify it to make the load constant-current by measuring the voltage drop across the resistor and varying the gate voltage to keep the current constant, regardless of the voltage. And of course, shut off when the voltage on the battery gets to the safe minimum. Saw a circuit using an Op-Amp to drive the gate and keep the current constant, but not sure if I need the op-amp since I can use 2 of the A/D inputs to measure the voltage drop directly and calculate the current draw. I'd use one of the PWM outputs and a cap to send a voltage to the gate. Any suggestions on how I would go about choosing a suitable MosFet? Thanks for your help. :) Thanks for the info. I do hobby electronics (passives, transistors, op-amps, etc.) , but I’m not really up on MosFets. I want to take your idea (and thank you for posting it) and modify it to make the load constant-current by measuring the voltage drop across the resistor and varying the gate voltage to keep the current constant, regardless of the voltage. And of course, shut off when the voltage on the battery gets to the safe minimum.

Saw a circuit using an Op-Amp to drive the gate and keep the current constant, but not sure if I need the op-amp since I can use 2 of the A/D inputs to measure the voltage drop directly and calculate the current draw. I’d use one of the PWM outputs and a cap to send a voltage to the gate.

Any suggestions on how I would go about choosing a suitable MosFet?

Thanks for your help. :)

]]> By: john http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5563 john Sun, 27 Nov 2011 02:18:46 +0000 http://www.foobert.com/blog/?p=471#comment-5563 I didn't exactly engineer it and just pieced it together to get the discharge current (drain-source ON resistance) I was looking for. I had a hand-full of unspecified low-current transistors sitting around and I just put them all in parallel to minimize the resistance. Really, most any N-channel FET that's solidly saturated when Vgs>=4 V should work with a continuous current rating of >500mA. You'll want to account for the resistance of the FET and adjust accordingly to get an accurate measurement. Perhaps <a href="http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00174698.pdf" rel="nofollow">this one</a> would work fine for you. Another way to do it is to use a relay, but that's more costly. Alternatively, you could use nothing and just be ready to pull the battery out when it hits the low-voltage threshold. I didn’t exactly engineer it and just pieced it together to get the discharge current (drain-source ON resistance) I was looking for. I had a hand-full of unspecified low-current transistors sitting around and I just put them all in parallel to minimize the resistance.

Really, most any N-channel FET that’s solidly saturated when Vgs>=4 V should work with a continuous current rating of >500mA. You’ll want to account for the resistance of the FET and adjust accordingly to get an accurate measurement.

Perhaps this one would work fine for you.

Another way to do it is to use a relay, but that’s more costly. Alternatively, you could use nothing and just be ready to pull the battery out when it hits the low-voltage threshold.

]]> By: Andy http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5562 Andy Sun, 27 Nov 2011 00:44:44 +0000 http://www.foobert.com/blog/?p=471#comment-5562 Thanks for the schematic. :) What did you use for Q1? What are the required specs for Q1? Thanks for the schematic. :)

What did you use for Q1? What are the required specs for Q1?

]]> By: john http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5561 john Sat, 26 Nov 2011 20:51:12 +0000 http://www.foobert.com/blog/?p=471#comment-5561 Updated! The usage of the FET to stop discharging probably isn't entirely obvious to everyone. Thanks for motivating me to actually draw it out. Updated!

The usage of the FET to stop discharging probably isn’t entirely obvious to everyone. Thanks for motivating me to actually draw it out.

]]> By: Andy http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5560 Andy Sat, 26 Nov 2011 15:44:26 +0000 http://www.foobert.com/blog/?p=471#comment-5560 Thanks for posting this. Did I miss the link to the parts list and schematic? Thanks for posting this.
Did I miss the link to the parts list and schematic?

]]> By: John http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5388 John Wed, 22 Dec 2010 16:07:05 +0000 http://www.foobert.com/blog/?p=471#comment-5388 @Scott: It's good point, especially for LION where the discharge profile has the voltage steadily declining throughout the discharge cycle. NiMH has a fairly stable voltage profile that falls off a "cliff" quite rapidly upon depletion. Thus, I suspect you'd see minimal difference in the measured capacity for the two methods. Constant <em>power</em> discharge would probably be the best method for applications that use nonlinear voltage regulation. @Scott: It’s good point, especially for LION where the discharge profile has the voltage steadily declining throughout the discharge cycle. NiMH has a fairly stable voltage profile that falls off a “cliff” quite rapidly upon depletion. Thus, I suspect you’d see minimal difference in the measured capacity for the two methods.

Constant power discharge would probably be the best method for applications that use nonlinear voltage regulation.

]]> By: Scott Vitale http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5387 Scott Vitale Wed, 22 Dec 2010 11:03:51 +0000 http://www.foobert.com/blog/?p=471#comment-5387 Hi John: I was looking over your code/project. You put a lot of thought into the features ... and its a good application for an ARDUINO. I'm currently working on a project that is similar to yours but is acting as a chart recorder. I am testing Li-Ion charge/discharge characteristics for another project. I had thought about using a simple resistor load but in using that method, the actual load current diminishes as the cell voltage decreases. In my humble opinion, sinking constant current increases the accuracy of your measurments and is a "more correct" way to determine discharge ratings of a battery. In my application, I used a constant current sink. I chose a TL431 "Adjustable Precision Shunt Regulator". On page 29 of the Texas Instruments data sheet, there is a schematic for a "Precision Constant Current Sink". You can power the TL431 from the ARDUINO 5V supply rail and attach the collector of the sink transistor to your battery. Since the TL431's "REF" pin has a constant 2.495V (+/-%), it could also be used as your A/D reference instead of the resistor divider you used. The "Standard Grade" part has a +/- 2% tolerance, while the "B Grade" part sports +/- 0.5% tolerance. Thanks for posting your work. Best regards, Scott Hi John:
I was looking over your code/project. You put a lot of thought into the features … and its a good application for an ARDUINO.

I’m currently working on a project that is similar to yours but is acting as a chart recorder. I am testing Li-Ion charge/discharge characteristics for another project. I had thought about using a simple resistor load but in using that method, the actual load current diminishes as the cell voltage decreases. In my humble opinion, sinking constant current increases the accuracy of your measurments and is a “more correct” way to determine discharge ratings of a battery.

In my application, I used a constant current sink. I chose a TL431 “Adjustable Precision Shunt Regulator”. On page 29 of the Texas Instruments data sheet, there is a schematic for a “Precision Constant Current Sink”. You can power the TL431 from the ARDUINO 5V supply rail and attach the collector of the sink transistor to your battery. Since the TL431′s “REF” pin has a constant 2.495V (+/-%), it could also be used as your A/D reference instead of the resistor divider you used. The “Standard Grade” part has a +/- 2% tolerance, while the “B Grade” part sports +/- 0.5% tolerance.

Thanks for posting your work.

Best regards,
Scott

]]> By: Hector Albornoz http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5286 Hector Albornoz Fri, 23 Jul 2010 12:55:29 +0000 http://www.foobert.com/blog/?p=471#comment-5286 Thank you for this exciting project. I have to test the batteries of 16V or 12V. I could indicate how you made the connections or a Schematic of tester? Thank you for this exciting project.
I have to test the batteries of 16V or 12V.
I could indicate how you made the connections or a Schematic of tester?

]]> By: MichaelDallas http://www.foobert.com/blog/2009/11/08/arduino-battery-capacity/comment-page-1/#comment-5227 MichaelDallas Wed, 24 Mar 2010 03:23:45 +0000 http://www.foobert.com/blog/?p=471#comment-5227 I've constructed a version of this. I noticed your comments about not trusting the VOM on the low R (ohm) readings and to measure the current and calculate backwards. You may want to reconsider. I assume you said this because you had problems making the calibration numbers balance (V/I=R). My calculation seemed to indicate an additional .7R to .8R in the system (in addition to the 2.1R resistors). I finally decided to measure the VOM which was measuring current with another VOM for any R that the (first) VOM might be adding. Sure enough, the VOM measuring current added .7R to .8R. While .7R doesn't mean much in a high impedence system, it makes a big difference when added to 2R of the load resistors. Thanks again for posting your example. I had independently thought of building this gadget. Your code saved me a ton of time. I’ve constructed a version of this. I noticed your comments about not trusting the VOM on the low R (ohm) readings and to measure the current and calculate backwards. You may want to reconsider. I assume you said this because you had problems making the calibration numbers balance (V/I=R). My calculation seemed to indicate an additional .7R to .8R in the system (in addition to the 2.1R resistors). I finally decided to measure the VOM which was measuring current with another VOM for any R that the (first) VOM might be adding. Sure enough, the VOM measuring current added .7R to .8R. While .7R doesn’t mean much in a high impedence system, it makes a big difference when added to 2R of the load resistors.

Thanks again for posting your example. I had independently thought of building this gadget. Your code saved me a ton of time.

]]>