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.

Paralleling two voltage regulators

Status
Not open for further replies.

king.oslo

New Member
Hello there,

I made a post previously, and one of the users kindly pointed out that it is impossible to parallel two voltage regs and expect the circuit to draw equal from them. He told me that one regulator would have a slightly higher voltage than the other, and the result would be that all circuit would be draw from that one, and the other would be left unused.

My question
Is there any way we can draw current from two almost identical voltage regulators, say two L7805CV?

Thanks.

Kind regards,
Marius
 
Last edited:
As with anything else, you could use resistors to sum them together, at the cost of a little regulation.

Or better still, an external power transistor on a single regulator.
 
How would wiring of these look:

- use resistors to sum them together
- an external power transistor

Thanks.

Kind regards,
Marius
 
For the external power transistor, check the datasheet which usually shows it.

Two resistors, one from each output joined together, take the output from where they join.
 
So you are saying this circuit will draw 50% from each voltage regulator?

Is that correct?

Why is the use of resistor different from using straight wire?

Thanks.

Kind regards,
Marius
 

Attachments

  • wireing.jpg
    wireing.jpg
    17.6 KB · Views: 193
So you are saying this circuit will draw 50% from each voltage regulator?

Is that correct?

Why is the use of resistor different from using straight wire?

It won't share 50/50, but as the current is drawn from one regulator the voltage on the output of the resistor will drop due to the voltage dropped across the resistor, this means the lower one will start to provide current as well. In practice, and by careful resistor selection, you get equal load sharing.
 
So you are saying this circuit will draw 50% from each voltage regulator?

Is that correct?
It will be very close to 50, but only if the resistors value matches. Two ±20% resistors won't be any good.

Another disadvance with your chematic is that you won't have a real regulator as the voltage drop over resistors will vary with current draw. Then the regulator don't regulate any more, and it's just wasted.

Is there some limitation of your project you haven't told yet? Is it really neccesary that the load share is exactly 50% ? Why? Do you have two physically separated heatsinks?
If not, it should be pretty easy to make one start conducting when the other one reach a predefined level, say 90 percent. But then again, it's more convinient to use a power transistor in parallell.
 
Last edited:
The percent split will depend upon the load current, the difference in regulated voltage between the two regulators, and the value of the series resistors.

The current required before the regulator with the higher voltage starts to draw current is equal to Vdelta/Rs where Vdelta is the output voltage difference between the two regulators and Rs is the resistor in series with each regulator output.

The worst-case difference in output voltage is the output voltage tolerance given on the regulator data sheet.
 
Last edited:
Use the power transistor approach. For more power u can stack the transistors with sub 1 ohm resistor on the emitters. The rating on the regulators is only for heatsinked operation. You prob could push a 1/2 watt dissipation on a 7805 without a heatsink.
 
Here is a power supply regulator I was designing, before I got side tracked on another project. It's not complete but it does show how one can be done.
 

Attachments

  • POWER SUPPLY.JPG
    POWER SUPPLY.JPG
    14.3 KB · Views: 197
Here is a sim that shows the problem of paralleling two (supposedly) identical LM7805 1A regulators in an attempt to get 2A. The instantiating two LM7805 results in identical behavior in the two instances, so to model the difference that would likely occur if two actual regulators were paralleled, I inserted a small offset voltage in the ground lead of one of the regulators. The plot shows the currents through the two regulators as a function of the offset voltage over the range from -20mV to +20mV. Note that per the LM7805 data sheet, the offset between two randomly selected LM7805s could be as much as 400mV.
 

Attachments

  • Dublreg.png
    Dublreg.png
    28.9 KB · Views: 186
If the need is simplicity, best is, distribute load and let each regulator handle a part load, with ground being common. Small differences wont matter at signal Level.
Of course, the theoretical best is, series pass transistor. But calls for additional components.
 
Last edited:
Here is a power supply regulator I was designing, before I got side tracked on another project. It's not complete but it does show how one can be done.

This way of doing it greatly compromises the voltage regulation. A much better way is shown on the National Data Sheet and Application Notes, where rather than following the regulator with an NPN emitter follower, you use a PNP transistor to add current into the load. This has the advantage that the voltage regulation is not degraded because the current booster is inside the regulator's feedback loop.

The sim shows how this is done. Note that in this example, when the load current is ~3A, 1A is coming through the LM7805, while the other two Amps is coming through the collector of the PNP transistor. Note the voltage out as the load current is varied from 0.1A to 3A.

This is a much better way of boosting the current of an IC regulator, rather than paralleling two or more regulators.
 

Attachments

  • BoostReg.png
    BoostReg.png
    37.9 KB · Views: 234
Last edited:
The sim shows how this is done.

Hee, hee, ha, ha, ho, ho.:p
The 2N2907 is dinky little 600mA old transistor. You need to use a power transistor for 3A.
Also, use less than 1A in the regulator.
 
Oops, I meant to use the 2N2955 . Note that the 2955 has a higher beta at 3A than the 2907 (not surprising), so with the 1.4Ω base resistor, the regulator current is ~0.8A when the load current is 3A.
 

Attachments

  • BoostReg.png
    BoostReg.png
    37 KB · Views: 200
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top