Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

So how about a Tablet power supply?

Status
Not open for further replies.

MrAl

Well-Known Member
Most Helpful Member
Hi,


Has anyone here given any thought to creating a power supply for a Tablet computer?

What i mean by power supply is an external battery with a conditioned output like say 5v at 2 amps. That's a good starting point i think. We would want a battery with high capacity like 7AHr or better i think.

One idea would be a 6v battery with a linear 5v, 2 amp regulator, but we might want to add extra over voltage protection and possibly other stuff to protect the valuable Tablet computer we will be powering with it.

Another idea would be a 12v battery (can use the car cigar lighter plug too this way) and a buck converter that puts out 5v at 2 amps, or maybe 6v and then followed by a linear regulator.

More ideas or even some circuits?
 
Last edited:
Had a power blackout recently? ;)

Well really it depends on the voltage required to run the tablet. Small stuff probably uses a 4v LiPO battery or might be two LiPOs in series, you need to sepcify that. Then as the tablet is battery operated it already has all the regulators etc built in so it doesn't need particularly good power.

What might be best if you use an external regulator is to use a fixed voltage in the low end of the range, so the internal regulators in the tablet run cooler (assuming some or all are linear regulators), some testing might be needed for the ideal voltage. That way you could use an external buck SMPS to get good efficiency, and do all that external to the tablet.
 
Hi MrRB,

Yes i did mention the voltage and current, 5 volts and 2 amps. So it has to put out a constant 5 volts and up to 2 amps current level.

I was also thinking about some sort of extra protection in case the primary regulator put out too much voltage due to some sort of fault. For example if it runs off of a 12v lead acid battery and something went wrong, i would not want it to be able to pump the input jack with 12v at whatever amps.
 
... Yes i did mention the voltage and current, 5 volts and 2 amps. So it has to put out a constant 5 volts and up to 2 amps current level.

Yeah I saw that, but probably should have used more words in my reply. The tablet does not run from a 5v regulated battery, and with a 5v input oprobably uses a linear reg to charge the Lithium battery and to supply the 3.3v etc needed to power the tablet.

A bit more info would be good as you might be able to supply it with (say) 4.5v regulated and get less internal heating in the tablet and better battery life if you are using a SMPS. Having a look inside the tablet (or getting a schematic) would help a lot in designing the perfect supply for it.

Manufacturers don't specify a 5v PSU because it is perfect but because they are readily available. A 4.3v PSU etc might be better but they don't make those. :)

...
I was also thinking about some sort of extra protection in case the primary regulator put out too much voltage due to some sort of fault. For example if it runs off of a 12v lead acid battery and something went wrong, i would not want it to be able to pump the input jack with 12v at whatever amps.

You could add extra protection, but I think if you make or get a well designed 12v -> Xvolt buck SMPS it can't turn on the main switch if the output voltage is over X volts, so they are very safe. Your only real threat is if the main switch FET etc blows short circuit, and from a properly designed and cooled 12v buck that would be extremely rare.

If you still wanted protection a typical overvoltage SCR crowbar would do the trick. Again looking inside the tablet would help as you could see what power it requires and can tolerate, it might have it's own internal overvoltage or reverse connect protection.
 
Hi again MrRB,


Thanks for the reply.

For this first project though i want to stick with 5v and up to 2 amps because it's very difficult to take some (maybe all?) tablets apart. It can be done, but it requires ripping off the top panel which they glue on, and that's the clear plastic part that you look through when viewing the screen. I dont want to play around with that i just want to use the external DC jack that allows a plug in of a wall wart. But of course i dont want to use a wall wart either but rather a back up battery with one kind of converter/regulator or another.

And yes some sort of secondary protection just in case something goes wrong with the main circuit. It has to be independent of the primary regulator so as you say maybe a crowbar is the best way to go. The MOSFETs these days can be had with extremely low resistance too as im sure you know, so maybe a secondary low R MOSFET in series with the output that could be opened fast in case of over voltage.
It's mostly a matter of voltage i think because if it drew too much current that would mean something must be wrong with the tablet anyway. So current monitoring isnt mandatory, although maybe an 'extra' add on. For now i'd like to think about the different ways of doing this.

To start, maybe a 6v lead acid with a 5v low dropout linear regulator. That would be the simplest. It would be most incompatible with my other stuff however which i make to run on 12v, so maybe i have to go with a buck converter. To get the run time i'd say the battery would have to be maybe 20AHr or something like that.

If you've never seen one of these guys disassembled you might have a look on the web. You'll see that it's not something you want to do unless it becomes absolutely necessary. I figure if the battery ever goes and i want to replace it i'll have to do it, but until then i dont want to have to dig into it.
Also, with an external battery pack like this it should keep the internal battery from discharging whcih means long life hopefully :)
 
Thanks for the info I understand better now. :)

I would go with a 12v to 5v buck SMPS, that voltage conversion range is particularly efficient. Even a common MC34063 IC with external low-Rds FET would work quite well for 5v 2A out, and also has overcurrent protection built in (with one external resistor). Over voltage protect is already included in the 34063 IC but if the IC or FET fails you could have external overvoltage with a comparator and voltage ref zener/diode and a second series FET as you said.

Re the 5v and running a different output voltage you could do some tests with a variable DC supply, and see if dropping to 4.5 or 4.0v is still functional, and runs the tablet cooler. The last time I used one at a friends place it was on external power jack, and the thing started to cook my lap it was most unpleasant the wattage wasted... Yeah I know some is from the (unneccesarily) high powered CPU but quite a bit was also from the internal linear regulator as it got a lot hotter running on the power jack than it did running from internal battery. Just something to think about. :)
 
Hi again,


Yeah that sounds about what i wanted to do. I should have mentioned also that i understood where you were coming from when you suggested attacking the battery directly (which i agree fully), which i may do some time down the road. That is the best way, directly with another battery, but yeah it's a bit difficult to do and you know what i didnt find a web page that put any of these BACK TOGETHER again (ha ha) all they did was take them apart, which means i'd have to hope i could get the transparent display cover back on good enough (flat enough) to allow the touch screen to work properly again. I fear that it wont work as well as it seems like it should. But i may have to attempt this later down the road i'll just have to hope it doesnt come to that.

It would be nice to know what they are doing with that 5v too yes, all i know for now is that they charge the internal battery AND run the device from that. If i knew what kind of regulator or charge circuit they had in there i could figure out what voltage it would take, but yes testing might be interesting. I also already noticed that the wall wart puts out 4.8v not exactly 5v, so there has to be some tolerance in the guts somehow.
Im waiting for my connectors to come in the mail so i can start experimenting. Just hope it doesnt blow up :)

I will be doing this for someone else too so it has to work good and not blow up their computer.
 
Last edited:
Hi again,


Well, even though the internal battery lasts around 5 hours that's still not enough if a storm hits. So that's what motivates this project more or less. If i can run it off of 12v that would be much better so a converter of some type seems in order, but also with a good secondary protection circuit to protect the valuable device. Makes sense if you ask me :)
 
I've had some fun storms too. One time a few years back I got blacked out and all my rechargeable torches etc were dead. I ended up finding a small 6v SLA battery that I knew was charged, and twisting some wires from it's terminals to a resistor and high brightness white LED (using candlelight of course). The LED uses about 30mA and it lasts for many hours (maybe 24 hours?), it's great as a hallway/bathroom lantern etc where you don't need a lot of light.

And I decided to make a rechargable human powered lantern with a 1W white LED and small 3.3F supercap;
**broken link removed**
Picture can be found here; https://www.romanblack.com/CoolStuff/lantern.jpg


The 3 phase capstan motor has no cogging and is very efficient making current at low cranked RPM, and it is smooth and quiet compared to those nasty geared dc generator crap torches they sell in the stores. Although it needs cranking every couple of minutes it's a great carry around type lantern for blackout jobs where I never have to worry about the internal battery being flat as it is a cap. :)
 
Last edited:
Hi MrRB,


Apparently ETO went down for a while and some data was lost, so we lost a lot of our conversation :)
 
Whoops! That's a bummer, but no big deal. I feel for EM and the amount of work/worry he must have done trying to save and restore all the forum data.

If you still want to talk about blackout related things, I've been playing with the idea of a super high efficiency permanently connected 12v battery charger for my "blackout" car battery. At the moment I just overnight charge it every couple of weeks.

One idea was to make a high effiiciency isolated SMPS for 240vAC->12v battery and put a PIC on the battery side as an intelligent monitor. That way it will come on as needed and keep my 12v battery charged all the time. I figured if I make the PIC on the 12v side of the SMPS drive the primary FET gate using a pulse transformer, then the primary side will use no power from the AC mains when the PIC is not driving it. It's an unusual topology driving th eprimary from the secondary but it makes sense for this app as the secondary has permanent power from the 12v battery.
 
Hello again MrRB,


Yes that's interesting too. I could think about this for my power requirements too as i do them all by hand myself as well.

First thoughts would be to mimic what i've done with the NiCd's charger. That would come on every 24 hours for 30 minutes and then shut off. That was for self discharge. Button for full charge after use. Lots of colored LEDs to indicate what mode it is in at the moment :)

For the lead acid if you want to charge every two weeks (which seems like it might be a lot to me though) all you'd have to do is set the PIC up with a nice long count cycle. As you know, you can get count cycles way into the years without too much trouble. Using the sleep function you could even have it sleep (using micro amps while sleeping) and wake up every 10 seconds and register one count, then when the count reaches two weeks have it go through the charge cycle, then shut down and start over.

The only pitfall i can think of at the moment is that the oscillator during sleep isnt that accurate, so it could be two to three weeks before a charge happens.

The NiCd charger was adjustable using a pot to set the 'on' time, but you probably dont need that with the lead acid unless you want to be able to adjust it for a more accurate cycle time.

Of course included a button to force it to do a full charge too like the NiCd charger. That way after use it will charge up fully again.
I used a pot for the charge time setting although a count button could do all the settings too, store the count in EEPROM. Short push once for one week, twice for two weeks, etc., or use better resolution like once for one day, twice for two days, etc., long press resets the count or reverses it to count down after that. Include lots of colored LEDs for status :)


Just some ideas you can kick around to start with. I should really do something like this myself for my 12v batteries.
 
Last edited:
While you're at it why not use the PIC to integrate load current over a given period (day/week/...) then control the charger to top-up accordingly?
 
Hi alec,

Yeah that would be an interesting addition i think. The unit could also calculate an approximate self discharge over time since it would be 'on' all the time anyway.
Probably have to do a little calibration. Have to add a current sensor to allow monitoring current in and out.
 
...For the lead acid if you want to charge every two weeks (which seems like it might be a lot to me though) all you'd have to do is set the PIC up with a nice long count cycle.
...

At the moment I have a voltmeter permanently connected to the battery and if I see the voltage get below about 12.6 or 12.5v I put it on charge. The battery is an old one pulled from a car that has less than perfect performance but like I said before it's fine for an emergency light even if its original 90Ah is now more like 40Ah... ;)

Alec_t said:
While you're at it why not use the PIC to integrate load current over a given period (day/week/...) then control the charger to top-up accordingly?

Interesting, MrAl is suggesting charging every X amount of time and you are suggesting monitoring load current to control the charging. I was going to use <Y battery voltage as the indicator of when to charge it, using the PIC ADC input.

Mainly I would just like to have it totally automated, so I can forget about it. Ideally I could leave it for 6 or 8 months then when there's another blackout it would be charged and ready to go.

The thing for me was making the AC mains charger efficient, so it does not draw any power at all from the mains when not charging. (I'm not worried bout 1 or 2 mA drawn from the battery as self-discharge of a car battery is >30 mA or so).

First I thought about using a relay so the 12v PIC operates a relay to turn on the 240vAC mains to start charging. But I wanted a SMPS charger, and not a crappy wallwart but something more efficient and more rugged that could be relied upon for years.
 
Hi again,


I was suggesting charge every X amount of time because the self discharge is not measurable so we have to assume something. Like you said MrRB if it is 30ma then we can only calculate the approximate discharge over say two weeks and charge for that amount of time times 1.5 (charge acceptance of LA plus a tiny bit more). So 30mAhr for one day over a 10 day period is 300maHr so charging at 3 amps for 9 minutes would be the calculation. Since there is no current to measure here, that's all we can do unless we want to figure out the self discharge with temperature and monitor the temperature vs time and integrate that, then come up with a better calculation, but is it really that critical...i dont think so, and we'd have to look up or figure out the self discharge with temperature for an aged battery...not that simple to do.

They do make more efficient wall warts these days, the regulated inverter types. They put out more current vs case volume than the typical transformer wall wart. But personally i like the transformer wall warts better as i know exactly what they can and cant do and i know they are very reliable. I've had a regulated inverter type blow out on me already after being plugged in for a few years even though not running anything. I also like the isolation we get from a standard wall wart as opposed to a direct line operated type.

But if you really want a switcher then the National Semiconductor (now TI) simple switchers are the way to go, unless you want a line operated switcher and then you'll have to work a bit more :)

On the original topic, i finally got around to getting a small power inverter that converts 12vdc to 120vac, as i was too lazy to build one myself. I tried it with the soldering gun and it runs it pretty nicely (inductive heating) so i was a bit impressed as i only paid 12 dollars for the inverter on sale. But the thought was to simply power the wall wart that came with the Tablet plugged into the inverter and the inverter of course running off one of my 12v lead acid batteries. That would be simple but i am a little leary of plugging an unknown wall wart into the inverter as it is not a pure sine inverter (single +/- pulse type) and the wall wart is the regulating inverter type.
 
Re the battery charger, I don't mind designing and making an off-line 240vAC->12v SMPS, I'm not scared. ;)

Congrats re the soldering iron and inverter, you can now solder stuff when there's a blackout. Last time I got caught out I was twisting wires. Mind you I do have a 12v soldering iron somewhere, but didn't think of digging through boxes to find it in a blackout. Maybe I'll just buy one of those little "gas" ones powered by lighter fluid?

Re your suggestion of using the 12->120v inverter to power the tablet plugpack, I think that is a bit lazy and will cost you energy efficiency. Neither device is likely much better than 80% efficiency, probably even worse on the inverter as the <10W output is likely to be lower than its peak efficiency range. So overall efficiency might be pretty poor.

I would REALLY like to see what current the setup draws from the 12v battery to supply the 5v at X amps to the tablet. I think for the cost of one chip and a few parts you could make a >90% efficiency regulated 12v->5v SMPS. :)
 
Hello again MrRB,

Well, if you do 240vac to 12vdc off line and something goes wrong it could possibly send the peak of a 240vac rms sine out where 12v is supposed to be. Can you spell "Fry Daddy" :)
I guess you would have to take extra precautions or else us a transformer anyway, but a higher frequency transformer.

Oh yeah the soldering 'gun' i was hoping to use with the car, and also in a blackout. I have a small 12vdc soldering iron which came in handy a few times but that doesnt have enough power for the larger wires and terminals so i like to dig out the gun, and that is only 100 watts but does the trick in most cases. And yeah during a blackout we could very well end up having to solder something and that's when it will really be nice to have. A broken wire or something, an LED or multiple LEDs to provide better lighting, etc. I think at least one iron should be included in the survival box and of course a way to power it.

Yes you're right about the inverter taking more power than a specially designed DC to DC converter just for the job. I just havent gotten that far yet and was giving it some more thought as to what approach i would start with. I have a 15vdc switcher that operates off of 120vac that i think i will convert over to 5v out with say 8 to 25vdc in. I made that for the scanner a long time ago but i dont use that scanner anymore because it's too slow and i have one now that is much faster. It wont be hard to convert it and it can handle up to 3 amps which is plenty. Efficiency around 85 percent so i could do better i guess.

I'll have to measure the current from the 12vdc side when it powers say 5v at 1 amp and at 2 amps at the output. The inverter draws 300ma no load so it is probably not going to be very efficient for that low power load. I am guessing maybe 60 percent which is terrible in this day and age. At least it will serve as a backup method. So i'll have to get to work soon before the next storm hits.

BTW have you ever tried to run a regulated switching wall wart off of a 12v to 120vac power inverter without it being a pure sine (like the 'modified' sine type which is really just a plus and minus 145v pulse) ?
 
Last edited:
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top