Hi! I'm building a portable project, and am looking to use a Pi2 (I don't own one yet), a TFT screen (http://www.amazon.com/dp/B00IUGW7PM/ref ... UTF8&psc=1), and a Teensy! My question is would I be able to run this off of a 3.7v batt with a powerboost 500C (allows for a 5V output)? For a few hours of use, how large should the battery be?
Thanks!
-
Haiderali360
- Posts: 5
- Joined: Fri Jan 23, 2015 1:35 am
Re: Battery Questions
Haiderali360 wrote:Hi! I'm building a portable project, and am looking to use a Pi2 (I don't own one yet), a TFT screen (http://www.amazon.com/dp/B00IUGW7PM/ref ... UTF8&psc=1), and a Teensy! My question is would I be able to run this off of a 3.7v batt with a powerboost 500C (allows for a 5V output)? For a few hours of use, how large should the battery be?
Thanks!
Running the monitor at 5v may need a modification, some of them will be able to accept between 5v and 24v as input, others will work natively at 5v (so the dc/dc converter will need to be removed, if you can find it / remove it) and others won't work with 5v at all so you will need to supply them with a higher voltage.
A naked Pi2 (no peripherals) runs at 800mA at 5v (which is 4W), the consumption of the screen is listed as 3W so you have a 7W load.
At 3.7V you will need about 2A to provide 7W, so a 2000mAh (2Ah) battery would last about 1 hour.
The battery should be rated for a 2A discharge rate, drawing power from a battery faster than it is designed for will drain the battery much faster than expected.
Doug.
Building Management Systems Engineer.
Building Management Systems Engineer.
-
Laurens-wuyts
- Posts: 716
- Joined: Wed Aug 21, 2013 7:35 pm
- Location: Belgium
- Contact: Website
Re: Battery Questions
Here is a post on how to remove the regulator.
http://www.raspberrypi.org/forums/viewt ... 41&t=17651
Laurens
http://www.raspberrypi.org/forums/viewt ... 41&t=17651
Laurens
Re: Battery Questions
You've got a couple of factors to consider here.
Firstly, as mentioned above, you're going to need to know whether or not the display can be modified to accept 5V power. Many small screens of this type are designed for cars, so will accept a 12V power input but internally they might be regulated down to 5V - I have a similar 3.5" composite LCD and I've already located & bypassed the regulator. Assuming you can power it from 5V you'll also need to know how much current it consumes.
Next you're goint to need to know how much current the Pi2 will consume with the peripherals connected. Alex Eames of RasPi.TV has done some measurements of the Pi2 power consumption under load, and it works out to around 420mA for all 4 cores under full load. You probably won't know the power consuption of the Teensy until you actually build the project, but let's say it draws the full USB2 current of 500mA - this is overkill, but you have to start somewhere.
So let's say that the Pi2+Teensy will draw 920mA under full load - you still have to account for the screen however. Let's say that draws another 500mA - again this is probably overkill. The overall power draw is therefore around 1.5A. The first thing you need to know is whether your regulator board can reliably deliver 1.5A at 5V. You also need to make sure that the battery/cell you intend to use can actually deliver that current - you also need to account for regulator inefficiency, which I'll get into in a moment, but for these figures a cell that can safely deliver 2A is probably a good bet.
Once you know all the power requirements of your circuit you can start to calculate the runtime you can expect to see from the battery. Using the figures above, the entire circuit will draw 7.5W (1.5A * 5V) under full load. Let's say you have a 2200mAh cell - the nominal cell voltage is 3.7v, so the cell has a total capacity of 8.14Wh (3.7V * 2.2Ah). However, you have to account for the fact that the regulator will draw more power from the cell than it delivers to the circuit. You could make an assumption here - 75% wouldn't be a bad benchmark. If you really want to nail down the efficiency rating you'll need to do some research. Good regulators can achieve more than 90% efficiency, but the only way to know for sure is to look up the data sheet for the IC. Most boost converter IC data sheets should have charts to show the efficiency of the IC at various input/output voltages and different current loads. Some of the Texas Instruments ICs I've been looking at will achieve between 95% and 90% efficiency, depeding on the charge level of the cell.
Once you know the effiency rating of the regulator, apply that to the cell capacity. Let's assume it's only 75% efficient: 8.14Wh * 0.75 = 6.105Wh. Once you've got the adjusted battery capacity you can then calculate the approximate runtime. If the Pi2 and all the components draw 7.5W then a 2.2Ah LiPo cell on a 75% efficient regulator will only get 0.8 hours runtime - about 48 minutes.
Of course I pulled a lot of these power consumption figures out of thin air (except for the Pi2), but you can refine the calculations once you do some testing and measurement. Here's a real world example though. For my purposes I'll be using an A+ with a custom GPIO controller and an Adafruit PiTFT screen. I've estimated that the A+ and screen should draw no more than 300mA under full load (120mA for the screen, up to 120mA for the A+ under full load, plus a little bit on top for headroom) - a total of 1.5W under full load. I've been looking at using a 1Ah LiPo cell paired up with a regulator that uses a TI boost converter IC - I've looked through the data sheet for the IC and it should achieve between 93% and 90% efficiency (it's designed specifically with LiPo cells in mind). Therefore I should be able to get up to 3.3Wh out of the battery (3.7V * 1A = 3.7Wh, 3.7Wh * 0.9 = 3.33Wh). That's just over 2hrs of runtime at full load.
Of course even with more accurate measurements this is all still theoretical - the only way to know for sure is to order the bits and test
.
Firstly, as mentioned above, you're going to need to know whether or not the display can be modified to accept 5V power. Many small screens of this type are designed for cars, so will accept a 12V power input but internally they might be regulated down to 5V - I have a similar 3.5" composite LCD and I've already located & bypassed the regulator. Assuming you can power it from 5V you'll also need to know how much current it consumes.
Next you're goint to need to know how much current the Pi2 will consume with the peripherals connected. Alex Eames of RasPi.TV has done some measurements of the Pi2 power consumption under load, and it works out to around 420mA for all 4 cores under full load. You probably won't know the power consuption of the Teensy until you actually build the project, but let's say it draws the full USB2 current of 500mA - this is overkill, but you have to start somewhere.
So let's say that the Pi2+Teensy will draw 920mA under full load - you still have to account for the screen however. Let's say that draws another 500mA - again this is probably overkill. The overall power draw is therefore around 1.5A. The first thing you need to know is whether your regulator board can reliably deliver 1.5A at 5V. You also need to make sure that the battery/cell you intend to use can actually deliver that current - you also need to account for regulator inefficiency, which I'll get into in a moment, but for these figures a cell that can safely deliver 2A is probably a good bet.
Once you know all the power requirements of your circuit you can start to calculate the runtime you can expect to see from the battery. Using the figures above, the entire circuit will draw 7.5W (1.5A * 5V) under full load. Let's say you have a 2200mAh cell - the nominal cell voltage is 3.7v, so the cell has a total capacity of 8.14Wh (3.7V * 2.2Ah). However, you have to account for the fact that the regulator will draw more power from the cell than it delivers to the circuit. You could make an assumption here - 75% wouldn't be a bad benchmark. If you really want to nail down the efficiency rating you'll need to do some research. Good regulators can achieve more than 90% efficiency, but the only way to know for sure is to look up the data sheet for the IC. Most boost converter IC data sheets should have charts to show the efficiency of the IC at various input/output voltages and different current loads. Some of the Texas Instruments ICs I've been looking at will achieve between 95% and 90% efficiency, depeding on the charge level of the cell.
Once you know the effiency rating of the regulator, apply that to the cell capacity. Let's assume it's only 75% efficient: 8.14Wh * 0.75 = 6.105Wh. Once you've got the adjusted battery capacity you can then calculate the approximate runtime. If the Pi2 and all the components draw 7.5W then a 2.2Ah LiPo cell on a 75% efficient regulator will only get 0.8 hours runtime - about 48 minutes.
Of course I pulled a lot of these power consumption figures out of thin air (except for the Pi2), but you can refine the calculations once you do some testing and measurement. Here's a real world example though. For my purposes I'll be using an A+ with a custom GPIO controller and an Adafruit PiTFT screen. I've estimated that the A+ and screen should draw no more than 300mA under full load (120mA for the screen, up to 120mA for the A+ under full load, plus a little bit on top for headroom) - a total of 1.5W under full load. I've been looking at using a 1Ah LiPo cell paired up with a regulator that uses a TI boost converter IC - I've looked through the data sheet for the IC and it should achieve between 93% and 90% efficiency (it's designed specifically with LiPo cells in mind). Therefore I should be able to get up to 3.3Wh out of the battery (3.7V * 1A = 3.7Wh, 3.7Wh * 0.9 = 3.33Wh). That's just over 2hrs of runtime at full load.
Of course even with more accurate measurements this is all still theoretical - the only way to know for sure is to order the bits and test
.