Another Alternate Power Solution.

[rinsewin]

Hi there,

I  just bought a fantastic box of tricks from Radio Shack.  Here are some of the highlights off the box:

  "Multi-Voltage Power Bank.  Power Adapter, Battery Charger"
  "Powers products requiring 3, 4.5, 6, 7.5 or 9VDC and upto 1000mA current"
  "international 90/240VAC auto sensing input - for US and foreign travel"
  Product code: 273-1900
  http://www.radioshack.com/product.asp?cata...t%5Fid=273-1900
  Currently selling for 15 bucks in Oakland, CA.

This little box is great.  Its about the same size as a clamshell zaurus, it has a fold away US style plug and it takes 4 NiMH batteries.  You can use it in different modes: battery charger (using 90 or 240V), AC->DC power adapter, DC power adpater (from the NiNHbatteries).  You can change the output voltage from 3 to 9V and it provides 1A of power.

After a little voltmeter testing I plugged it into my SL-C860 using 4 freshly charged 2500mAh Energizer Ni-NH batteries, and set it to the  4.5V output setting (note: I needed a B style radio shack adpater plug, with the tip set to positive).

The power adapter sustained my zaurus for 4 hours before turning itself off and letting the Zaurus automatically switch to its own power supply (sweat!!!), I ran the following contiously for the 4 hours:

  - Screen on full brightness.
  - Pretec wireless card, with Opera pointing to USGS CA Earthquake website (it auto updates every
    15 or so minutes).
  - MP3 player running in continous loop using head phones.
  - Qtopia clock app
  - note the screen saver popped up every now and again.

This is a little risky since you're charging your zaurus using 4.5V rather than the recommend 5V, but I expect there is voltage regulator (or zener diode)  in the zaurus which drops the voltage down to 3.3V (or something) and these guys normally require that you have at least a 1V delta between your in voltage and output voltage so this is probably right on the threshold.   When I get a bit more time I'm planning on putting together a simple 5V zener diode circuit to drop the 6V adapter output to 5V which will definitely satisfy Sharp's power requirements (but eat up a lttle more power).

Anyway I thought I'd share since the 4.5V output appears to work quite will and gives you an extra 4 hours of high power fun.

Carl C.
Ps. I dont gurantee your results, so please make up your own minds on whether you're happy with the 4.5V or not.... but this is a cool box of tricks!

[ran]

You could probably get by with setting the RS box to 6V,  and using a series diode to drop it down to 5.3V:  iirc,  the official 5500 wall wart is spec'ed at 5.2V out,  so 5.3V is well within tolerance.  

If the RS box happens to be at the upper range of its specified tolerance,  you might wind up feeding about 5.6V to the Z,  which is worth worrying about.  But,  given the number of people using unregulated 4-cell NiMH packs directly without reported problems,  you probably don't need to worry too much.

One definite advantage of your Zener regulator,  though,  is that it'll prevent damage if you use the RS to run some 9V gadget,  and forget to change the voltage before connecting to the Z.

From the picture on the webpage,  it appears that the output power cord connects to the supply with its own coaxial jack.  If that's the case,  what I'll do is buy a couple of connectors,  and make a custom cable for the Z with the diode inline and covered with heatshrink.

Thanks for the pointer:  I've been wanting something like this,  and had no idea it was already available so cheaply.

Ran

[rinsewin]

Hi Ran,

Can you see any issues with the power rating of the diodes you're going to use.  It looks like its easy to get hold of 1A silicon/zener diodes but they only have a power rating of 1W.

According to the sl-c860 zaurus spec its power consumption is 2.7W (I calculated a 5W requirement for the zener diode circuit!!) do you have a similar problem with the 5500?  If, so what diode specs are you planning to use or dont you see this as an issue?

Rinsewin.

[rinsewin]

Hi Ran,

Can you see any issues with the power rating of the diodes you're going to use.  It looks like its easy to get hold of 1A silicon/zener diodes but they only have a power rating of 1W.

According to the sl-c860 zaurus spec its power consumption is 2.7W (I calculated a 5W requirement for the zener diode circuit!!) do you have a similar problem with the 5500?  If, so what diode specs are you planning to use or dont you see this as an issue?

Rinsewin.

[ran]

I'm not experienced with designing Zener shunt regulators,  but I did a quick check on the web,  and it appears that your calculation is right.  Which surprised me,  because I thought the Zener only had to dissipate the "excess" energy (i.e.,  between 5V and the input voltage),  much  like a 3-terminal regulator like a 7805 would.  That would be only about 1W (less whatever is dissipated by the series resistor),  because the RS box specs say it'll only put ot a maximum of 1A at 6V.

If the Zener really does need to dissipate 5W,  you can buy 5W diodes,  but that would be a horrible waste of battery capacity.  I think the formula I found on those websites is probably wrong,  but you shouldn't trust that,  because I don't have the experience to back up my belief.

Otoh,  I do have experience using the series dropping diode method.  A 1A diode will do,  but a 2A or higher would be better,  because it'll have a larger package,  which will dissipate the heat over a larger surface area,  so it won't get as warm.  Remember that it's expected that the diode will be tucked away where you won't be touching it,  so it's considered okay for it to get uncomfortably (for you) hot.

The 2.7W rating on your Z tells you that it draws about half an Amp at 5V.  But the series diode won't "see" the whole 5V:  just the .7V that drops acoss it.  So,  at half an Amp,  it dissipates less than half a Watt,  and its 1W rating is okay.

Ran

[rinsewin]

Hi Ran,

Yep.  When in doubt use Ohm's law!  You've just sparked a few memories from my youth.

I think the calculations for the zener diode configuration are correct, since you're running the load in parallel with the zener it essentially becomes a voltage divider circuit, with the zener sinking any unwanted current.  But, I think the 5W calculation is a worst case scenario that essentially says this is the maximum power this diode will have to deal with if your load goes away and it has to deal with the full 1A of current.

Apparently a better configuration is to have the zener turn on and off a transister in series with a resister.  In this circuit when the zener reaches its threshold of 5V it only has to deal with a few mA of current to turn on the transister, which in turn sits in parallel with the load (or something like that).

For me the zener thing is turning into a pain in the rear so I'm probably going to go with either a 7805 voltage regulator with a small heatsink strapped to the power adapter running on the 7.5V setting, or the straight forward inline 2A silicon diode  that you described.

I need to do a little more testing to find out what I'm comfortable with..... my Z is the backbone of getting me to meetings on time and killing time in those meetings so I want to keep her healthy for as long as possible.

Thanks for your help,

rinsewin.

[ran]

this is the maximum power this diode will have to deal with if your load goes away and it has to deal with the full 1A of current.

That implies that the circuit will always draw the maximum available current,  and anything not used by the Z would simply be thrown away by the Zener.  Which is to say that the system would literally suck  

I'd like to think that's not true,  but I really don't know.  I'll have to try an experiment someday when I have the breadboard set up.

If you really want to do the 3-terminal regulator thing,   check out the LM2940T-5.0 from National:  it's very similar to the 7805,  but only needs .5V,  instead of 2V of extra input voltage to work properly.  You could run that off the 6V output from the RS box,  and waste a lot less energy.

You can get them for only $1.25 from www.futurlec.com,  which has a lot of other interesting parts that are sometimes hard for hobbyists to get at decent prices.  They were a little slow in shipping when I ordered from them (it took almost a month for the parts to arrive),  but that might not be typical.  I found it worth the wait because they had their own versions of some of the small 2-to-5-chip-size protoboards that Radio Shack has discontinued.  Unfortunately,  it looks like Futurlec has since discontinued most of them,  too    

It's too bad RS didn't include a 5V output in their design,  considering the number of gadgets that require it these days.  Maybe it's on clearance because they have a new version coming out?

Ran

[omega]

ran is correct. Zener diode is stupid, as it will not work. Series diode with 6v is also stupid as the voltage can go up to above z tolerances. Use a low dropout 5v regulator, 2A and you'll be grand.