More efficient going from 3.3V to 5V or going from 6.6V to 5v?

Status
Not open for further replies.

taskman

Well-known member
I have some parts in my project that need to run on +-5V. I need to decide if I should use a voltage regulator or a booster. Should I connect the batteries in series and drop the voltage down to 5V or should I connect the batteries in parallel (3.3V) and boost them to 5V? What is more efficient?

My electric motor can run at 6V max, but 5.5V will make it last longer
My sound card can run at 5V down to 4.5V
My screen can run at 5V down to 3.3V
 
The best solution depends on the type (cost) of batteries you have, and also on whether you need maximum efficiency when operating the motor (high power mode ?), or when in standby (low power mode ?). You will also have to consider how you will interface (control) each peripheral from Teensy (3.0 ?).

A simple solution is to use 4 AA's in series and run the motor directly from them (I'll bet it will actually run even when the battery pack is down to 4 V total). Then use an LDO (linear regulator) to regulate down to 5 V to power the Teensy, sound and screen.

You can get DC/DC converters which are more efficient that LDOs, but the difference is relatively small and probably not worth the complexity. If you do do that, you might look at boosting a 2*AA system (2..3 V) to 5.5 V and run everything from that. However, perhaps the motor draws 1 A (especially when starting), and this requires a relatively powerful DC/DC converter (and you would pull over 2.5 A peak from the cells when starting).

So - 1st determine if you care about run-time efficincy or standby efficiency in general (or compute how much of the total battery energy is consumed at high load vs. low load). E.g. say your standby current load (say Teensy's load) is 30 mA and you standby for 24h; also you run the motor (1 A) for 10 minutes. Now the Teensy is 30m*24 = 720 mA.h. The motor is 1 * 10/60 = 167 mA.h. So you can see that the standby power is most important -- so perhaps a DC/DC stepping down to 3.3 V (from 4*AA's) is best, while running the motor & sound directly from the battery.
 
Sorry didn't think of adding the extra info, would have saved time.


The battery I will use is this 3.3V LifePo4, so far, ordered them, but haven't tested yet
http://www.hobbyking.com/hobbyking/..._Lithium_Ion_2300mAh_Single_Cell_Battery.html

The motor is a vibration motor. It will run only when the player takes damage. I believe I have it setup at 150ms at the moment. The player can get hit every 500ms up to every 1500ms. So every 500ms the motor could vibrate for 150ms long
What the manufacturer said in a mail "the Typical Operating Current of the 324-102 is 56 mA and the Maximum Start Current is 170 mA"
https://catalog.precisionmicrodrives.com/order-parts/product/324-102-24mm-vibration-motor-12mm-type
 
This is what happens currently with our laser guns:
We take out the guns from their boxes and switch them on. We lay them on mats on the ground until the players arrive. Players pick up the weapons, shoot each other a bit to figure out which weapons they like the most, swap weapons and play a bit more. When the game starts we normally have some weapons on the ground doing nothing and wasting battery power

I am building standby code into my equipment - the brain, that will be used in my head sensors, guns, respawn boxes and so on. I am reading a tilt sensor every +- 10 seconds and storing the data into an array and if the tilt sensor hasn't moved for a minute or two I will go into lower power mode. Just waiting for the lower power library and then I will be able to drop the power consumption of my Teensy.

Low battery consumption is very important, because sometimes we drop off equipment at a client on a Thursday night. They play a bit the Thursday night, they have a party Friday and play. We pick up the weapons and then we have games the Saturday and Sunday. Our current guns last maybe 10 hours, advertised as 16 hours (7.4V 4500mA NiMh batteries). So I have guns charging at night after a rental, I get up at 1am to change chargers and then charge further so that we can run on a weekend. One of the reason I am creating my own equipment, I want to sleep :)
 
You will need a boost if you really need to run the motor at 5 or 6 V. Remember the LiFePO4's will range from 4.5 V (while charging and just full) to ~ 3.3 V (you can tradeoff lifetime vs. min. voltage).

Given a 4500 mA.h battery, and a motor that is (say) average 100 mA, but for 150/500 ms, it only consumes an average current of 30 mA. so over 10 hours, that's 300 mA.h -- i.e. only 7 % of your system's power -- a Teensy is also about 30 mA, so something else is consuming most of your power -- is it LEDs ? -- you should optimize for that.

Note that if you use a DC/DC (boost) to convert from 3.6 V to supply (say) 100 mA at 6 V, this will consume 100 * 6/3.3 = 182 mA from the 3.3 V source. The DC/DC will only be about 80 % efficient, so it will actually consume 182/0.8 = 227 mA. You'll need to include that in your calculations eventually.
 
I read voltage regulators are also about 75% efficient so there isn't going to be much difference between using a regulator or a dc/dc converter

I wonder if it is a good idea to have two regulators.

One that will convert the parallel pack from 3.3V/4.5V down to 2.6V. I will run the LED's and Teensy from this. I thin Paul said the Teensy 3 can run down to 1.8V. I will see if 24 MHz is fast enough for me. I saw I can get LDO regulators that have a difference of 0.1V between the in and the out. I might be wrong. Since the conversion isn't huge maybe I won't have such huge losses

I can then have a second 5.5V regulator on the pack in series that might be 6.6V to 9V to power the sound, screen and the motor. I am currently using the 1.8'' LCD screen from Adafruit, but considering moving over to the eInk screen. I think I can power the eInk screen from the 2.6V regulator. I won't have the cool graphics that I was planning, but the kids don't care much about that, they just want to see numbers and I want to save battery power. The 1.8'' screen is a bit difficult to read in the sun, but I knew that when I went this route. So if I power the screen also from the 2.6V regulator then I will only have the sound and motor powered from the 5.5V regulator and since the sound and motor aren't constantly on I am not losing that much energy
 
I read voltage regulators are also about 75% efficient so there isn't going to be much difference between using a regulator or a dc/dc converter

I wonder if it is a good idea to have two regulators.

One that will convert the parallel pack from 3.3V/4.5V down to 2.6V. I will run the LED's and Teensy from this. I thin Paul said the Teensy 3 can run down to 1.8V. I will see if 24 MHz is fast enough for me. I saw I can get LDO regulators that have a difference of 0.1V between the in and the out. I might be wrong. Since the conversion isn't huge maybe I won't have such huge losses

I can then have a second 5.5V regulator on the pack in series that might be 6.6V to 9V to power the sound, screen and the motor. I am currently using the 1.8'' LCD screen from Adafruit, but considering moving over to the eInk screen. I think I can power the eInk screen from the 2.6V regulator. I won't have the cool graphics that I was planning, but the kids don't care much about that, they just want to see numbers and I want to save battery power. The 1.8'' screen is a bit difficult to read in the sun, but I knew that when I went this route. So if I power the screen also from the 2.6V regulator then I will only have the sound and motor powered from the 5.5V regulator and since the sound and motor aren't constantly on I am not losing that much energy

Unless you are using extremely low currents ( << 1 mA), linear regulators' efficiency is exactly VOUT/VIN. In fact, Teensy's microcontroller includes one to regulate 3.3 V from the ~ 5 V USB voltage. Your system will be no more efficient if you use that, or bypass it with your own LDO. If you run Teensy down to 1.8 V, it will consume somewhat less current (and 24 MHz also helps), but the real way to take advantage of this would be to use a DC/DC step down -- If T3 consumed 20 mA at 1.8 V, the DC/DC would only consume 20m*1.8/5 * (1 + 15 %) = 8.3 mA from the 5 V battery (the 15 % is an approximation of the efficiency losses).

From my earlier post, it's still not clear where the bulk of your power is going -- if you figure that out, you would have a better idea wher to focus on.
 
I read voltage regulators are also about 75% efficient so there isn't going to be much difference between using a regulator or a dc/dc converter

I wonder if it is a good idea to have two regulators.

One that will convert the parallel pack from 3.3V/4.5V down to 2.6V. I will run the LED's and Teensy from this. I thin Paul said the Teensy 3 can run down to 1.8V. I will see if 24 MHz is fast enough for me. I saw I can get LDO regulators that have a difference of 0.1V between the in and the out. I might be wrong. Since the conversion isn't huge maybe I won't have such huge losses

I can then have a second 5.5V regulator on the pack in series that might be 6.6V to 9V to power the sound, screen and the motor. I am currently using the 1.8'' LCD screen from Adafruit, but considering moving over to the eInk screen. I think I can power the eInk screen from the 2.6V regulator. I won't have the cool graphics that I was planning, but the kids don't care much about that, they just want to see numbers and I want to save battery power. The 1.8'' screen is a bit difficult to read in the sun, but I knew that when I went this route. So if I power the screen also from the 2.6V regulator then I will only have the sound and motor powered from the 5.5V regulator and since the sound and motor aren't constantly on I am not losing that much energy

Unless you are using extremely low currents ( << 1 mA), linear regulators' efficiency is exactly VOUT/VIN. In fact, Teensy's microcontroller includes one to regulate 3.3 V from the ~ 5 V USB voltage. Your system will be no more efficient if you use that, or bypass it with your own LDO. If you run Teensy down to 1.8 V, it will consume somewhat less current (and 24 MHz also helps), but the real way to take advantage of this would be to use a DC/DC step down -- If T3 consumed 20 mA at 1.8 V, the DC/DC would only consume 20m*1.8/5 * (1 + 15 %) = 8.3 mA from the 5 V battery (the 15 % is an approximation of the efficiency losses).

From my earlier post, it's still not clear where the bulk of your power is going -- if you figure that out, you would have a better idea where to optimize the power efficient
 
The guns we are using are from another company. They use +-7V 4500mA NiMh batteries. I plan on using 3.3V 2300 mA LifePo4 batteries in series and/or parallel

So the guns we are using and causing me to get up at 1am to charge :) are not the guns I am working on to build. The guns from the company seem very inefficient and there are so many bad things with them, but I won't go into it. I am building my own that I hope will be more efficient with more features than the guns we have. The help you guys are giving is what will make the guns great
 
Series Li-Ion cells are harder to charge and to balance the charge (if one cell in the pack discharge slightly more than the other, eventually it will become over discharged and lose performance (or become a safety concern)). For simplicity, I'd recommend using cells in parallel if you need more energy than a single one can supply.

A 2300 mA.h cell can supply 230 mA for 10 hours -- I still don't see what part of your system consumes anything near that -- the T3 is ~ 30 mA, the motor is also about 30 mA average. What else is taking power ?
 
Status
Not open for further replies.
Back
Top