I use 6 AA batteries to power a standalone arduino, fully charged the pack yields about 8.3 - 8.5 Volts. The batteries are connected to a 7805 voltage regulator by means of a diode to protect it from connecting the batteries the wrong way round. However the diode creates a voltage drop of about 0.7 Volts, so a fully charged pack results in 7.6 - 7.8 Volts for the 7805. Now the 7805 will stop functioning properly when input voltage drops somewhere between 7.0 and 7.5 Volts, and the voltage will drop when the batteries are used: I have the feeling I cannot get my full 2700 mAh from the batteries because the 7805 will stop functioning properly way before the batteries are actually drained.
The do-not-connect-the-wrong-way-round-protection the diode ensures seems a good way to go. Should I have opted for a diode with smaller voltage drop (it needs to be able to sustain 1 Amps), or is there another way of protecting my arduino without the voltage drop? (I use headers which do not indicate in any way how the battery should be connected, they are the common pin headers to come with pcb's, 2.54 mm pitch; using connectors which can only connect one way round could be an option, but I'm looking for others).
CaptainJack:
the 7805 will stop functioning properly when input voltage drops somewhere between 7.0 and 7.5 Volts, and the voltage will drop when the batteries are used: I have the feeling I cannot get my full 2700 mAh from the batteries because the 7805 will stop functioning properly way before the batteries are actually drained.
Yep.
You need a regulator with a lower dropout voltage if you want to run it like this. Look for the keyword "LDO" when you choose one.
Another option (which I use a lot) is to use less batteries and a 5V voltage booster. This makes sure you get the last drop of power from your batteries and is more efficient than turning excess volts to heat via. a regulator.
If you choose this route, get a booster that handles at least twice the amps you really need (apply a factor of two to all eBay components and you'll be happy...)
CaptainJack:
However the diode creates a voltage drop of about 0.7 Volts, so a fully charged pack results in 7.6 - 7.8 Volts for the 7805. Now the 7805 will stop functioning properly when input voltage drops somewhere between 7.0 and 7.5 Volts ...
It specifies a maximum input voltage of 5.5 Volts, I am wondering whether I can connect 4 AA batteries. They would max out to 6 Volts, but in practice they will probably not reach 4.4 V.
This is a wonderful suggestion, I'm going to make some calculations and see how difficult it will be to implement these in my schematics; reducing the number of necessary batteries from 6 to 4 per package will mean a lot, in so many ways: I'm creating a set of taggers for LaserTag, and the batteries make up a substantial part of the money of the tagger, as well as the weight. Reducing to 4 would make building the tagger easier as well.
Though I have learned a lot about electronics in the past two years I have yet to discover many things. I'm glad I posted the question here, and even happier to get these answers.
The PFET protection solution is one that actually crossed my mind, but I hadn't encountered it before.
Now, even while using a DC DC booster, I suppose it will be good practice to implement some sort of protection I suppose.
[quote author=Nick Gammon link=topic=150619.msg1131096#msg1131096 date=1361737286]
Watch this, all is revealed:
[/quote]That's clever. I'll have to check which of my p-channel MOSFETs suit because anything I have made where the source closely matched the voltage required has been unprotected against reversing the polarity up to now. Thanks for pointing that out Nick.
CaptainJack:
reducing the number of necessary batteries from 6 to 4 per package will mean a lot, in so many ways: I'm creating a set of taggers for LaserTag, and the batteries make up a substantial part of the money of the tagger, as well as the weight. Reducing to 4 would make building the tagger easier as well.
If you're going with a booster there's no reason not to drop down to 3 batteries if your power consumption is low.
You could even use a single 18650 3.5V Li-ion rechargeable battery. Not many people will have chargers for those at home though so you might have to include one.
Well, about power consumption: I haven't got a clear picture, because it depends on the user. The tagger uses 0.05 Amps when in rest, but turning on the LCD display backlight will consume 0.3 Amps, turning on the series of leds to indicate you are sending IR packets or you are being tagged uses about 0.3 Amps as well. The leds will not burn continuous, but if you're respawning they will blink until you get to your homebase. So, power consumption will not be low I suppose. One single battery contains around 12 - 14 kJ of energy (depending on exact voltage, somewhere between 1.3 Volts and 1.5 Volts). If my Arduino uses 0.3 Amps @ 5V (this is not the case, but worst case scenario) the power consumption is 1.5 W, so a single battery will keep me running for about 2.5 hours if no heat is dissipated in the dc dc booster. So, yes. Three might work out.
I use 2700 mAh NiMH batteries from GP. I actually did some research on how to charge these batteries (it would be user friendly if you could just connect your tagger to a wall outlet), but charging them doesn't seem trivial; mostly you need an extra IC to regulate the charging. And now I think of it: some clever switching so while charging my arduino doesn't get hurt, or even runs on the charging device while the battery gets charged. I might read some more and then post some over here to discuss my ideas.
Hardly any resistance, can go up to 14 Amps, switches at about 2 - 3 V, so even with 3.3 V you can switch it open. If any of you see any reason not to pick this one I'm much obliged to hear. I still have a fair share of these so if they are good enough, I'll use them.
CaptainJack:
Well, about power consumption: I haven't got a clear picture, because it depends on the user. The tagger uses 0.05 Amps when in rest, but turning on the LCD display backlight will consume 0.3 Amps,
Only for a few seconds, hopefully.
CaptainJack:
turning on the series of leds to indicate you are sending IR packets or you are being tagged uses about 0.3 Amps as well. The leds will not burn continuous, but if you're respawning they will blink until you get to your homebase.
Smart software can help a lot with power consumption.
eg. The blink only needs to be half a second 'on' every three or four seconds. That's 50mA.
Yeah, the backlight is on for two seconds and switches off automatically, it's just so you can read scores in the dark. I'm upgrading to a nokia 3310 display (haha, "upgrading"), still got to measure power consumption of that, but I guess it will be less.
The blinking is done by switching the leds on and off every 0.4 seconds (f = 2.5 Hz). But when you're hit they stay on continuously for 5 seconds, during which you "regenerate your shield", then they switch off. So 5 seconds on for 5 leds, drains quite a bit, but you have to be visible from a distance - the leds are ultrabrights. So my aim there is not low power consumption, that´s why I initially started out with a six pack of AA batteries. But seeing the DC DC booster and the smaller from factor of 3 AA batteries I will order components asap.
Hardly any resistance, can go up to 14 Amps, switches at about 2 - 3 V, so even with 3.3 V you can switch it open. If any of you see any reason not to pick this one I'm much obliged to hear. I still have a fair share of these so if they are good enough, I'll use them.
That's an N-Channel by the look of it. You need a P-Channel.
Ah right, I do need a P-channel. Had to do some reading up, always nice to refresh your knowledge a bit. I do tend to get lost in farnell's product catalogue looking for the right components: say I want to protect against just a single AA battery for reverse voltage protection, the threshold voltage for the N-channel mosfet should be between 0 Volts and say -1 Volts, otherwise it would never open up, even if connected right. I haven't found the right mosfet yet, but they have a few thousand of them in stock, so I will search some more.
Thanks for the suggestion for the booster, I will add it to my list of useful components for the smd-version of my board; the currect prototype is veroboard for educational purposes - it seems more insightful for students to solder all wires as well. It is always nice to hear what other people use, because the choice can be overwhelming.
@Nick: I meant the AA battery in combination with the DC DC booster So it would be nice to find a N-channel mosfet which is compatible with 1, 2 or 3 AA batteries. Or do I miss out on something here ? Granted: one battery will not last endless, but suppose I would like to try smallest package possible it would be nice to know that it works
@Cookies: very elegant. What do they mean with soft start ? I understand the voltage at the gate will rise slowly, so... the fet opens slowly instead of a single switch turn on ? Why would I want this, is it more pleasurable for my IC ?
strykeroz:
Thanks for the extra ideas Cookies. One question though : what does it mean (in the description for the N-channel option) by
If low-side switching will work
What would you need to check to ensure that was the case?
I can't really think of a situation where low-side switching wouldn't work.
CaptainJack: @Cookies: very elegant. What do they mean with soft start ? I understand the voltage at the gate will rise slowly, so... the fet opens slowly instead of a single switch turn on ? Why would I want this, is it more pleasurable for my IC ?
Soft start might used to limit inrush current to capacitors and maybe to reduce popping with audio amplifiers, for microcontroller stuff you want the power to turn on and stabilize as fast as possible, so don't use soft start.
Just wire a jumper across the diode, and double check polarity when connecting
an external power source. Diodes are for babies [just kidding ;-)], but once you
wire up a power source with the correct polarity, you can't plug it in backwards,
after all.
So it would be nice to find a N-channel mosfet which is compatible with 1, 2 or 3 AA batteries. Or do I miss out on something here ? Granted: one battery will not last endless, but suppose I would like to try smallest package possible it would be nice to know that it works