Testing 600 fish

We claim battery life of 600 fish on our BlueTipz transmitter. What does that even mean? Batteries aren’t measured in fish. So how did we get that number, WHY did we get that number, and how did we test it?

An early prototype being tested in the freezer to make sure it worked in cold temperatures and high humidity. Yep.

An early prototype being tested in the freezer to make sure it worked in cold temperatures. Yep. I was not about to stand in front of the freezer with the door open for 24 hours, though. My mom raised me right.

First, we started with math. We know what the rated life of the battery is (220mAH. This means it can source 220mA of current for 1 hour, or 1mA for 220 hours). We also know that our LED consumes 20mA, but it’s only on for 1/8 of a second every second. That averages to 2.5mA continuous. And we know the consumption of the Bluetooth module, which transmits for a given amount of time every fraction of a second. If all we were running was the LED, we would have enough battery for 88 hours of operation with the LED flashing for 1/8 of a second every second. The Bluetooth module adds quite a bit of complexity so we won’t get into it, but we promise we did the math.

It gets even more complex because after a certain period (60 seconds), we stop transmitting as frequently. And after 10 minutes, we turn ourselves off entirely. So how does that play into the equation? If we used our previous math, we’d be able to say “the BlueTipz battery works for 20+ hours of continuous transmission.”

Why number of fish?

“20+ hours of continuous transmission” doesn’t mean anything, and it isn’t right. We don’t continuously transmit for 20+ hours ever, and we don’t want users to think that they’ll have to change their battery every day. If they don’t catch any fish, they won’t use any battery. We wanted to tie battery life to fish activity.

So we said “what if we estimate how long it takes to get the alert, put on coat and boots, and get out to your tip-up to turn it off?” Then we’d know how long the device was on during a normal event. We figured 2 minutes. So we did the math for 2 minutes. That’s 2 minutes of LED flashing at 1/8 second bursts every seconds. That’s 60 seconds of rapid Bluetooth transmission followed by 60 seconds of less rapid transmissions. And we did the math to get about 600. It was higher than that, but we figured the battery voltage would drop off at the end and there’s no way we’d be able to get the full 220mAH out of the battery. So at the end of our math, as responsible engineers, we threw some slop at the end to account for unknowable circumstances or otherwise covering our butts.

Math was boring; let’s do experiments!

Math wasn’t good enough. We needed to test it. And how do you test something that takes 2 minutes 600 times in a row? We were not about to sit in the cold for 20+ hours and tip it up for 2 minutes over and over again. So we built a rig. A super scientific rig that had a little box attached to a servo, and every 2 minutes it would tip down, wait 10 seconds, then tip back up again. Another special Bluetooth Low Energy module listened for the signal, and we had software count the number of tip-ups before it stopped transmitting.

And what did we get? 640. Sweet. Yay math and experimentation!

Letting the experiment run. The cables go into the freezer.

Letting the experiment run. The cables go into the freezer. Why not test outside? We needed repeatable consistent conditions, and it was September outside.

The computer was connected to the inside and running a simple Terminal program that listened for signals from the unit inside.

The computer starts recording battery and count. The temperature wasn't calibrated.

The computer starts recording battery and count. The temperature wasn’t calibrated.

It just kept printing a very simple string, which indicated the battery as reported by the BlueTipz, the temperature (not calibrated), and its count.

Inside the freezer. There's a power supply and USB cable for data. The black box has electronics and a servo to turn the BlueTipz horizontal and vertical every set number of seconds.

Inside the freezer. There’s a power supply and USB cable for data. The black box has electronics and a servo to turn the BlueTipz horizontal and vertical every set number of seconds. The lights are mood lighting, and because we used MakeBlock to build our rig. The saran wrap is to keep humidity out of the sensitive electronics (we didn’t protect the BlueTipz, though).

The unit that runs the show. We used MakeBlock components, and didn’t need to solder anything. There was the MakeBlock Arduino clone, the MakeBlock shield on top of it, and it was connected to a MakeBlock servo module, which powered the servo. A BlueTipz was reprogrammed to act as a receiver and used the Arduino serial library to communicate with the Arduino. The Arduino’s job was to run the servo and listen to the BlueTipz and transmit to the computer outside. It was mostly simple.

Conclusions

Our math was pretty good, and the performance under testing was pretty good, too. We’ll be able to use this rig again for tests in the future and to get repeatable results. Saying the battery is good for 600 fish makes sense to explain how long the battery lasts, without being deceptive or confusing. It lets our customers know that it will last a few seasons, and shows that we’ve tested it out that far.

Changing the battery takes under a minute, and the battery costs 25 cents, so it’s not really that big a deal anyway, but we like to make sure that our customers are getting a quality product.


Cart