# Connect computer to battery

#### CVMichael

Joined Aug 3, 2007
419
How can I power a motherboard using batteries without going through the conversion to 110 then back to 12 V (and the other voltages)
I mean, right now, I have a UPS system that has a 12V battery, the UPS converts that to 110V, then the computer power supply, converts back to 12V (and the other voltages)

This is the ATX connector for the motherboard, and it shows the voltages it needs.

So, if I have 1 (or more) 12V batteries, how would I connect them to the computer ?

Any help appreciated.
Thanks

#### n9352527

Joined Oct 14, 2005
1,198
The easiest way is to get a picoPSU, as long as your computer doesn't need more than 120W. If you need more power, then try other beefier PSUs, such as the M2-ATX or the PW-200 (look on the left hand side menu on the above page).

#### CVMichael

Joined Aug 3, 2007
419

Actualy, "PicoPSU-120-WI-32V" looks more interesting becuase if I connect a battery to the device, then the battery will discharge, and therefore it won't be able to give 12V anymore.

So I wast thinking that if I put 2, 12V batteries in series, making 24 V, then both batteries can be discharged (up to 6V) to provide the motherboard the 12V neccesary

Also, do you think it can power a P5LD2 motherboard, with ~ 4GHz CPU ?

Also... I live in Canada, do you think there are any stores around here where I could find this device.

Sorry to ask so much... you don't have to answer all my questions...

#### Papabravo

Joined Feb 24, 2006
21,225
You do realize that the voltages on the connector have to be stable and regulated within a couple of percent. How were you planning to generate +5V, and +3.3V and -12V from one or two twelve volt batteries.

BTW +12 does not mean greater than or equal to 12 it means +12V plus or minus the supply tolerance. I suppose you could use a marine deep cycle battery but I wouldn't think it would last too long with a 4 GHz. CPU.

I use one for a 100 Watt HF Transceiver doing contesting and it lasts about ten hours. I would say the transmitter duty cycle is 40%.

#### mrmeval

Joined Jun 30, 2006
833
It depends on the wattage you need. This wiil feed 12VDC to a converter that then creates the voltages you need but it's only 120W
http://emea.kontron.com/index.php?id=226&cat=433&productid=1714

This one takes a good voltage range of 10VDC-25VDC and outputs the needed voltages. I'm not sure if it's available yet.
http://emea.kontron.com/index.php?id=226&cat=433&productid=1752

This is another one but it's intended for a PC104, a cable would have to be made for it to work on an ATX motherboard. You can check the wattages in the data sheet.
http://www.parvus.com/support/Transportation/Board-LevelOEM/PRV-0617/

#### GonzoEngineer

Joined Jul 8, 2007
46
I assume you want to do this for portability.....but 100 watts is still 100 watts!

I would look into a used portable generator.....it will weigh less than a couple of car batteries.

I have seen some pretty cheap ones in the large discount hardware stores.

#### CVMichael

Joined Aug 3, 2007
419
The easiest way is to get a picoPSU, as long as your computer doesn't need more than 120W. If you need more power, then try other beefier PSUs, such as the M2-ATX or the PW-200 (look on the left hand side menu on the above page).
By the way... I went to the local Electronics store, but I could not find any of this, they did not even hear about something like this...

To buy this, I would probably have to order it, and pay for shipping and other charges (tax or something...), it will come to a total that I'm not sure if I can afford...

#### Papabravo

Joined Feb 24, 2006
21,225
Yeah! Working within a limited budget can be a real bear. I guess the answer is to pick a lucrative field, invest time an energy, so this will cease to be a problem. I can tell you that engineering has very limited prospects for achieving that goal.

#### CVMichael

Joined Aug 3, 2007
419
I assume you want to do this for portability.....
Actually, it's because I want to make a robot. But I don't want to make a simple one that just moves in a specific pattern, I want it to make decisions on it's own (on where to move & go), provided that I gave it a "task" to do...

I have a lot more questions I want to ask, but probably I'll start another thread since it's not related to powering up the computer to battery.

By the way... I'm also a member of VBForums.com. I'm a pretty good programmer, that's why I want to make the "AI" using a computer (instead of other programming chips that have limited speed and memory, and I have to write code in ASM (and I don't know ASM)).

#### hgmjr

Joined Jan 28, 2005
9,027
By the way... I'm also a member of VBForums.com. I'm a pretty good programmer, that's why I want to make the "AI" using a computer (instead of other programming chips that have limited speed and memory, and I have to write code in ASM (and I don't know ASM)).
Greeting CVMichael,

The PIC and AVR microcontrollers are fairly fast. The AVR is a Reduced Instruction Set Computer (RISC) with the major portion of its instuction set executing in one cycle of the system clock. The AVR can run at 20 MHz. That equates to very nearly 20 MIPS.

They both have C-language compilers with the AVR C compiler available for free. Also the Intergrate Development Environment (IDE) is free as well. Being on a tight budget, it is hard to beat free.

Also, if a single microcontroller fails to provide the horsepower you need, you can add multiple microcontrollers that talk to each other over a simple SPI bus. The microcontrollers come in a variety of package styles with different features such a timers, A/D converters, Analog Comparators, and the like. These devices cost only a few dollars US.

All that said, I recommend you give PICs or AVRs a serious look as the basis for one of your robots.

hgmjr

#### CVMichael

Joined Aug 3, 2007
419
Actually, yeterday I bought a PIC16F877, it was \$20 CAD, and I also have a programming KIT K8048 to program it.

I want to use it to control the stepper motors on the robot, the robot will have around 20 motors, but only 4 of them working most of the time, the rest are controlling wheel direction, and camera movement (in all directions).

So I want to have the "AI" in the computer, then the computer will talk with the PIC microprocessor, and the microprocessor will give the neccesary input for each stepper motor. I also bought a L298 (Link to PDF), with a L297 for each stepper motor.

The problem is that I don't know ASM, and I can't find a good compiler in C for PIC. Actually, I have a "PIC C Compiler" (PCW.exe) but the interface is sooooo confuzing that I don't even know how to make a basic program. I managed to compile a sample program, but the output HEX file does not seem to be compatible with the K8048 KIT.

From what I saw on the net, the "PIC C Compiler" (PCW) has it's own KIT programmer, but I bought the K8048 first (and I don't want to get a new one, especially when I can't even figure out how to use the interface), and when I saw that it does not come with a C compiler, I found the PCW complier on the net.

Anyways... worst case scenario, I will try to learn ASM...

Actually I wanted to put this in a new thread, but since you started talking about PIC cips...

#### spar59

Joined Aug 4, 2007
64
A few posts ago you mentioned using a couple of 12V batteries in series to enable you to still get 12V when each was down to only 6V. Irrespective of what you feed the batteries into (presuming it can cope with an input varying between 2 fully charged batteries giving 27V and 12V) your batteries will not last long if you treat them like this.

Lead acid cells are not designed for such extreme discharging and many battery charging / control systems actually disconnect the batteries automatically to prevent them being discharged to anywhere near this voltage.

Steve.

#### CVMichael

Joined Aug 3, 2007
419
So then, what is the optimum discharge voltage ? (I did not think that was a problem until now...)

#### spar59

Joined Aug 4, 2007
64
Manufacturers of high reliability batteries used in standby power systems typically specify a discharge endpoint voltage of 1.75V per cell i.e. 10.5 V for a nominally 12V (6 cell) unit. Though I will stand corrected if anyone can point me at a specification sheet I suspect that no lead acid battery will come near 1V per cell.

Most standard lead acid cells are just not designed for lots of cycles of deep discharge followed by recharging. A car battery fits this category, normally it just floats across the alternator, it gives a few hundred amps for starting the engine but only for a few seconds which ampere-hour wise amounts to very little (it is not just a straight multiplication though since at high discharge currents battery efficiency is reduced and the nominal ampere-hour capacity could easily be reduced to 25% of nominal) and then gets topped up again by the alternator in the next few minutes.

Some batteries (deep cycle types - sometimes called leisure type batteries) are designed for multiple charge / discharge cycles.

I am not saying a battery will necessarily be instantly destroyed by a deep discharge if it is recharged soon afterwards but the number of charge-discharge cycles and the battery storage efficiency will be adversly affected if this is a repetitive event.

Actually this is not as bad as it may seem - the battery will maintain a decent output voltage whilst its ampere-hour capacity is used up, and hence there is not much more energy to be extracted by allowing the voltage to fall to 6V.

A quick google to http://www.batterystuff.com/tutorial_battery.html , since I didn't have a chance to hunt through my files in the office yielded the following information.

"State of Charge Specific Gravity Voltage
100% 1.265 12.7
*75% 1.225 12.4
50% 1.190 12.2
25% 1.155 12.0
Discharged 1.120 11.9

*Sulfation of Batteries starts when specific gravity falls below 1.225 or voltage measures less than 12.4 (12v Battery) or 6.2 (6 volt battery). Sulfation hardens the battery plates reducing and eventually destroying the ability of the battery to generate Volts and Amps."

I know that the battery information at the top of this post conflicts with the above but that information is for a specific cell type from Chloride which is used in battery installations within electricity substations - obviously ditching the battery and all the protection and control sytems it feeds would be a last resort - hence the low accepted endpoint voltage. However such a battery is only rarely discharged and spends 99.99+% of its time being eased through life by a high quality float charge system.

Sorry for the length but hope this helps.

Steve.