I had never really done much in the way of estimating how much power a set of electronics were going to consume, but (as I do with many things electronics) I just gave it a shot in the most logical manner I could come up with. Luckily, I wasn’t looking for any amount of precision and only a fair amount of accuracy. As it turns out, my estimates were quite accurate. Here’s what I did.
I identified which parts of the project were going to consume current.
- 4 idential 7-segment digits for the score
- 1 7-segment digit for the inning
- 1 2-segment digit for the tens place for the inning (driven by borrowed driver pins)
- 1 7 led display for balls strikes and outs
- 1 atMega328p running the show, without any special power conservation code (so far)
- 1 WiFly GSX module
To give you an idea of how I calculated power consumption, let’s take a look at one of the digits. Take a look at the digit design on the Digits page of this website for more information.
Each digit is made up of seven segments. Each segment is made up of four 10mm bright red LEDs in series. The LEDs have a forward voltage of 2 volts and an ideal current of 20 milliamps. Because the LEDs are in series, they consume four times the voltage, but the current remains the same. Additionally, because I am using a constant current driver, I can set the current for the digits very accurately. Based on calculations in the TLC5916 datasheet, the current for my drivers is set to 20mA. That means that if all seven of the segments in a digit are turned on, it consumes 140mA to light the LEDs. I couldn’t find a value for the TLC5916 chip itself, so I just assumed a high value of 20mA.
Using this method and by using data sheets, I calculated estimates for the other parts of the project.
Adding all of these together, I came up with a maximum value of 1320 milliamps, or 1.32 Amps. Now, remember, this is a maximum value. I know that, on average, I will never reach this value based on how I know the scoreboard will be used. For example, I know that the tens digits for the score will mostly remain blank. That’s 280mA right there. Also, the score will vary, but I know for sure that the score will rarely be 8 to 8, so the ones digits for the score will also be taking less current.
I did some number crunching (using some, I’m sure, very poor statistical models) and estimated a value of how many segments would be on for each digit on the scoreboard. Using this method, I can estimate the “average” current required. I calculated that, on average, the scoreboard would draw 742 milliamps. With a 4000 milliamp hour battery, that gives me more than 5 hours of operation. Even with my poor statistical analysis, I am confident that I will likely be able to push well beyond 5 hours of operation.