precise 0-5V, 0-5A adjustable power supply

[witek]

Hi,
I've looked around internet and this forum for suitable reference designs, but so far in vein.
As there is quite a few wise men in here I hope I am hoping to get some help here 

My background - lots of experience with embedded systems software design, not very confident with electronics as such.

Recently I got hooked on electro-chemistry and need a power supply to carry out some experiments 

Some spec I would love to achieve:
- voltage -> adjustable, as stable as it can get in full range of current
- voltage range ->  < 5V, most of the time around 3V, for some experiments down to 0.15-0.30V
- voltage precision -> 0.1V as bare minimum, 0.05 more desirable, I would love 0.01V

- current - adjustable, stable, precision much less important than voltage
- current range - less than 5A, most of the experiments around 2-3A
- current precision - in range 100mA (maybe even less) should be sufficient

I have built quite a few adjustable supplies on  lm317/lm338, but because of voltage range it is not the way to go forward. I am not confident I can design analog circuit  on opamps and darlingtons from scratch hence my call for help 
Any reference you could post would be much appreciated.

Besides obvious goal of building the power supply I intend to learn something so next time I need to go analog I wont be so lost 


PS. I searched for supplies of the shelf, but first they are expensive (all have range 0-15V or 0-30V, which way too much and I don't fancy paying for sth I don't need) and main reason - if I buy it I wont learn anything.

TIA

[eliocor]

If you are able to found it, I suggest you to get an used Kepco ATE 6-10M linear power supply.
This model (0-6V, 0-10A) is built like a tank and works flawlessly.
I was able to buy it (used) for about 50 Euros (70 USD) and I will never regret the expense...
BTW I also own two ATE 15-3M and two ATE 25-10M: their specifications are oustanding!

If you are able to get one of them, take care that the rear mating connector (PC12) is installed: without it the power supply is more or less useless, but if you really need it, you can overcome this problem soldering some jumpers directly on the contacts... This can be a good excuse to get a really low price on those units!!!

http://www.kepcopower.com/support/2281757.pdf
http://www.kepcopower.com/support/ate-operator-1-quarter-rack-r1.pdf

Please take care that those linear power supplies are professional ones and they are still sold by Kepco, so sometimes eBay sellers have outrageous prices for them!

[witek]

Thanks for reply,
I was looking around for some ready made power supply. They are not cheap for a students budget. I found one in sort low price and good spec, problem is according to manual it can be operated upto few hours and then has to be shut down for cooling. Possibly could be solved by fitting a fan in the case if there is a space for it or rebuilding the case.

Besides if I buy one I wont learn anything that way.

The supply I am using for my embedded projects is too valuable to leave it working next to a bath full of evaporating acid.

I have boxes of components ready to use, so very likely I would have to spend very little if I built one myself. there are several projects around the net but they are either only volt regulated or only amp regulated or upto 1A or range starts from above 1V.

[robrenz]

These are great power supplies.

http://www.ebay.com/itm/POWER-DESIGNS-TRANSISTORIZED-POWER-SUPPLY-Model-No-2050-/130728745996?pt=LH_DefaultDomain_0&hash=item1e700a5c0c

[saturation]

I'd second robrenz recommendation, in addition add "PRECISION" to the search term for Power Designs and you'll find supplies with uV resolution capable of 500mA-2A typically under $100.  See the threads started on eevblog for more about these supplies.

[witek]

Thanks guys,


The one suggested by robrenz has no current regulation.
Electrolytic bath changes its resistance along with temperature, ingredients and their concentration - all these factors change over time when the electrolysis works and could finally lead to excessive current drown from power supply. If that happens the depositing layer will become a sponge instead of shiny metal or even worse, the outcome can change as elements I don't want to plate out beginning to deposit.
Bottom line, current regulation is a must, however it doesn't have to be extremely precise, as I said in a range of 100mA ( +/- 50mA ) would do.


I think I found a candidate for the supply -> http://www.electronics-lab.com/projects/power/001/index.html.

The problem is that when it works on low V and high A settings the heat dissipation on Q2 will be closing on 100W. Having that on constantly for days would mean I could grow bananas in my shed during British winter.

Few questions:
1) would I have replace components if I replaced the transformer for one with 10V secondary ?
2) I have number of 2N5886 in TO-3 floating around the workshop, could it directly replace 2n3055 ?


Edit: corrected the URL to project

[markus_b]

If power dissipation is a concern to you, why don't you go for a switching power supply ?

Something like this: switching lab power supply on Ebay

You will have to set up you measurement / data-logging up separately, but this is the same with the power supply you pointer out above.

[witek]

price of $69.70 is not bad, shame the shipping to UK is $85 on top of that

[T4P]

You can find it on totobay on cheap. Or wtotoy on aliexpress