3V Unlimited Power Supply

Hi everybody! This time I’m building an unlimited power supply. Bcuz I have a circuit that needs continuous 2.5 V-5V DC supply even the AC power is there or not. So I designed a simple solution for this as below.

See Pic1
**broken link removed**

I used a 6V/300mA transformer for this. I couldn’t find any 3V ones & I’m too lazy to rewind that transformer for 3V. The bridge rectifiers makes the waveform to DC & 470 uF capacitor makes it little smoother as in the normal Power Packs. The DC voltage is directly applied to the LM 317T voltage regulator. I have adjusted the R to make the output as 4.2V.you can see in the diagram.

When the AC power is available the output 4.2V of the regulator is going through D1 as well as through D3 to the output. The D1 is the charging path of the Batteries.

1. When the batteries are fully charged the current is going directly through D3 to the circuit. Not going through D1 & D2 Isn’t it????

2. Is this good charging all the time the batteries????

When the AC lines fails the battery taken over the run. Then Current is flowing through D2. This time the voltage will be (1.2Vx3) – 0.6V = 3.0V.This is enough to hold my circuit until the AC power comes.

But I don’t like much this serial charging method. B’cuz all three batteries are not constant. One battery can fully discharge & the other one can be half charged. So when charging when the first battery is fully charged it won’t allow current to flow to the other batteries. If I have a single battery this wont be a problem. But I have 1.2V batteries.

I cannot add below method to this circuit. The parallel charging method (one by one).

See Pic2
**broken link removed**


I need some advice from you about this batteries serial charging. Is my cct having a problem?? Any modifications to be done etc….

I like to do this with two AA batteries but the D2 diode will drop 0.6V so the output will be 1.8V.

The batteries I’m using GP batteries AA (Gold Peak) 1.2V, 2550 mAh, Standard charge 16hrs, 250mA.

Your description isn't good enough to paint an acurate picture of the circuit; could you please upload a schematic?

I believe you mean "Uninterruptible Power Supply", there's no such thing as an "unlimited" power supply, unless you believe all the free energy lunatics

hi Gayan,

A picture is worth a thousand words.

EricG

2. Is this good charging all the time the batteries????

Click to expand...

In general, no. SLA or Sealed Lead Acid do well with a constant voltage trickle charge method but most other types do not tolerate this too well.

The batteries I’m using GP batteries AA (Gold Peak) 1.2V, 2550 mAh, Standard charge 16hrs, 250mA.

Click to expand...

These sound like NIMH batteries which do NOT work well when on constant trickle charge.
Change them to NICAD type and limit the charge current at around 0.02C - 0.05C. So if the batteries are rated for 1000mAh then trickle charge them between 20-50ma... If you already have the NIMH, then you can use them but expect a shorter lifespan compared to the NICADs in this configuration.
Or use a supercap if the standby time/current is low enough.

sorry for the attachement noe I have uploaded the schematics.

Really nice art work!!!

Hello is anybody home??????

hi Gayan,

Are you knocking!.

I have been looking thru your drawings for the psu/charger.

You will understand its not always easy to follow what the OP is trying to achieve.

We have do a little 'reverse engineering' which takes time.

Is it possible you could post a summary of the 'required' performance and what you expect the psu to do,
rather than telling us what dosn't work?

Also confirm the type of battery you are trying to charge.

Regards
Eric

As per your diagram; the problem with charging the batteries through D1 has several problems:
1: The current is not limited. Well it is, but it's at the max rating of the LM317!
2: 4.2V - 0.7V (diode) = 3.5V which means that the batteries will never fully charge because the voltage accross a NIMH/NICAD usually needs to go up to apx 1.6V per cell which would be 4.8V for 3 cells.

But I don’t like much this serial charging method. B’cuz all three batteries are not constant. One battery can fully discharge & the other one can be half charged.

Click to expand...

There is nothing wrong with charging the batteries in series. With a trickle charge, the batteries will equalize over time.
Below is how I would do it. You would choose R1 to limit the trickle charge rate to 0.02C - 0.05C. D1 would be a schottky diode and the regulator a Low Drop Out type to minimize voltage drop. You'd probably be better off using 4 NICAD cells and a slightly higher AC input voltage to give you more room for line voltage variations, etc.

kchriste said:

As per your diagram; the problem with charging the batteries through D1 has several problems:
1: The current is not limited. Well it is, but it's at the max rating of the LM317!
2: 4.2V - 0.7V (diode) = 3.5V which means that the batteries will never fully charge because the voltage accross a NIMH/NICAD usually needs to go up to apx 1.6V per cell which would be 4.8V for 3 cells.

There is nothing wrong with charging the batteries in series. With a trickle charge, the batteries will equalize over time.
Below is how I would do it. You would choose R1 to limit the trickle charge rate to 0.02C - 0.05C. D1 would be a schottky diode and the regulator a Low Drop Out type to minimize voltage drop. You'd probably be better off using 4 NICAD cells and a slightly higher AC input voltage to give you more room for line voltage variations, etc.

Click to expand...

Hi kchriste thanks for your explanation. I went through your diagram. Earlier I planed to put the regulator in the output side as in your diagram. But I couldn’t find any LDO type regulators in my place all I have cheapy LM317T & 78xx series. So I planned to put that in the input side.
Your schottky diode is very usefull in this case with low drop forward voltages.I must give input voltage about 1.6Vx3=4.8V as you said & also add a R1 resistor to limit the current.

ericgibbs said:

hi Gayan,

Are you knocking!.

I have been looking thru your drawings for the psu/charger.

You will understand its not always easy to follow what the OP is trying to achieve.

We have do a little 'reverse engineering' which takes time.

Is it possible you could post a summary of the 'required' performance and what you expect the psu to do,
rather than telling us what dosn't work?

Also confirm the type of battery you are trying to charge.

Regards
Eric

Click to expand...

Hi Eric. A little more about my circuit….

I have a PIC based a small sensor circuit which needs 3v & 200mA will be quit enough to run that. It’s running all 24 hrs. And to the output I have connected a LED. Later on I’m going to give an input to my PC with this output. When somebody moving through the sensor the LED lights up. Same like PIR sensor circuit. The whole cct is running from 3v/200mA.now I’m giving power through a 6v/300mA transformer regulating to 3v without no battery backup. So when the AC lines break the cct will shuts down. So I have design the PSU with the battery backup as mentioned above.

Battery Details
I’m using NIMH batteries.
GP batteries AA (Gold Peak) 1.2V, 2550 mAh, Standard charge 16hrs, 250mA.

hi Gayan,

The circuit that Kchriste posted will work fine.
BUT,
you are using NIMH batteries which require an 'intelligent' charging circuit.

I would suggest, as you cannot source a LVD reg, that you change your battery pack to NiCd, say 5 * 1.2V, giving a nominal 6/5.3v into the Vreg.
Also remember that you can configure a +3V regulator using discreet components.

You say that your sensor requires 3V @ 200mA and the transformer is rated at 6Vrms @300mA, so after rectification the input to the Vreg will be greater than the voltage of the battery pack, so there will some current charging the batteries.
Choose the series charging resistor to suit the battey spec.

The correct definition of your circuit is 'battery backup supply' not unlimited.

Regards
Eric

Hi…..I refreshed my basic knowledge to replace this LDO type regulator. Instead of the regulator I made it with a transistor as in the diagram below & I checked in my breadboard. It worked well giving output to 3V.

**broken link removed**

kchriste said:

1: The current is not limited. Well it is, but it's at the max rating of the LM317!

Click to expand...

Look again. It's set up as a 4.6mA current source. It won't work as a UPS, but it is current limited.

Look again. It's set up as a 4.6mA current source. It won't work as a UPS, but it is current limited.

Click to expand...

I noticed that today. It won't work very well as a charger for a 2550mAh battery either then. Not to mention the 200mA required by the "PIC based small sensor circuit", which would drain the battery with no chance of it charging again, so I let it go as a drawing error....

Hi…..I refreshed my basic knowledge to replace this LDO type regulator. Instead of the regulator I made it with a transistor as in the diagram below & I checked in my breadboard. It worked well giving output to 3V.

Click to expand...

Now draw 200ma out of it and see how low the input voltage can go before the output voltage drops below 2.7V Maybe change the 1K resistor to 100hm: @ 1/2watt and you should be able to get a lower dropout voltage. I suspect that you won't be able to get the performance that you want from this circuit.

hi gayan,

Have you considered using a 'shunt' regulator, ref attached drawing.

The values may need adjusting to suit your transformer and NiCd battery.

Eric