How to get 15V 5A DC power supply from 24VDC battery?

Thread Starter

cRash1001

Joined Dec 22, 2015
15
I would like to rephrase it by saying not smps but regulators that are not linear like buck regulators. Could anyone suggest me a good IC i have a slow internet connection.
 

spinnaker

Joined Oct 29, 2009
7,830
Statement is redundant. Buck==step down, boost==step up. A switching regulator can be configured for either.

Thanks Mr Detail. :confused:. As far as I am aware those modules cannot be configured for either. They are either step up or step down. And the LM2596 is only a step down chip.
 

ian field

Joined Oct 27, 2012
6,536
Thanks Mr Detail. :confused:. As far as I am aware those modules cannot be configured for either. They are either step up or step down. And the LM2596 is only a step down chip.
There are various incarnations of buck-boost regulators, and a small selection of off the shelf chips for that job - but the design isn't trivial, I generally design with the 2 functions separate - that is for example a brute force boost, followed by a buck with whatever regulation I need.
 

crutschow

Joined Mar 14, 2008
34,280
If you want to build you own, here's a relatively simple 5A buck regulator circuit with built-in high-current MOSFET switch.
The most critical thing in using such a device is careful layout with short connections and proper circuit grounding as shown in Figure 16. A PCB or vector board with a ground plane is recommended.

There are also many built modules available such as this.
 
Last edited:

dl324

Joined Mar 30, 2015
16,839
Thanks Mr Detail.
If you're going to use electronics speak, you might as well save yourself some embarrassment and try to use the terminology correctly.
As far as I am aware those modules cannot be configured for either. They are either step up or step down. And the LM2596 is only a step down chip.
Simple 3 terminal linear regulators can be converted to buck/boost with the addition of a handful of components.
upload_2015-12-25_11-56-29.png
It wasn't convenient to breadboard an example, so I simulated the above circuit in LTspice. It sucked at simulating this and was really picky about the R2/R3 divider (and the value of R1), but I know from experience that it works and the values aren't really that critical.

MC34063A can be configured as buck or boost.

EDIT: Removed the statement that it could be a boost regulator. It's a buck regulator and max output voltage is limited to Vin minus saturation voltage of the pass transistor.
 
Last edited:

ian field

Joined Oct 27, 2012
6,536
If you're going to use electronics speak, you might as well save yourself some embarrassment and try to use the terminology correctly.
Simple 3 terminal linear regulators can be converted to buck/boost with the addition of a handful of components.
View attachment 97248
It wasn't convenient to breadboard an example, so I simulated the above circuit in LTspice. It sucked at simulating this and was really picky about the R2/R3 divider (and the value of R1), but I know from experience that it works and the values aren't really that critical. With R4=0, it's a buck regulator; otherwise, with an appropriate ratio, it's a boost regulator.

A switching regulator IC marketed as buck only should work in boost mode if it gives you access to it's feedback terminal. MC34063A can be configured as buck or boost.
One thing no one mentioned yet:

A buck regulator puts the control transistor in series with the load, so if it fails short circuit it dumps the full unregulated voltage downstream and can cause damage.

This was a problem with multisync VGA monitors, the buck regulator sometimes failed and caused a lot of damage in the horizontal scan section. The manufacturers became aware of this and started using low voltage main SMPSU of about 60V and using a boost converter to step it up. If the transistor fails in a flyback converter, it shorts out the supply and the main SMPSU went into over current shutdown. All this happened about the time CRTs were going out of fashion, so it was pretty much game over by then.

There's 2 easy ways to protect the load - first is use a buck to step down to less than the required voltage then boost it back up with a flyback. Or just go for the basic buck and protect the load with a crowbar SCR.
 

spinnaker

Joined Oct 29, 2009
7,830
If you're going to use electronics speak, you might as well save yourself some embarrassment and try to use the terminology correctly.
Simple 3 terminal linear regulators can be converted to buck/boost with the addition of a handful of components.
View attachment 97248
It wasn't convenient to breadboard an example, so I simulated the above circuit in LTspice. It sucked at simulating this and was really picky about the R2/R3 divider (and the value of R1), but I know from experience that it works and the values aren't really that critical. With R4=0, it's a buck regulator; otherwise, with an appropriate ratio, it's a boost regulator.

A switching regulator IC marketed as buck only should work in boost mode if it gives you access to it's feedback terminal. MC34063A can be configured as buck or boost.

Where did I mention a linear regulator used as s buck or boost? Save yourself an embarrassment and learn to read what people write.
 

spinnaker

Joined Oct 29, 2009
7,830
One thing no one mentioned yet:

A buck regulator puts the control transistor in series with the load, so if it fails short circuit it dumps the full unregulated voltage downstream and can cause damage.

This was a problem with multisync VGA monitors, the buck regulator sometimes failed and caused a lot of damage in the horizontal scan section. The manufacturers became aware of this and started using low voltage main SMPSU of about 60V and using a boost converter to step it up. If the transistor fails in a flyback converter, it shorts out the supply and the main SMPSU went into over current shutdown. All this happened about the time CRTs were going out of fashion, so it was pretty much game over by then.

There's 2 easy ways to protect the load - first is use a buck to step down to less than the required voltage then boost it back up with a flyback. Or just go for the basic buck and protect the load with a crowbar SCR.

Interesting. What about more sophisticated bucks like the LM2675? It does say this in the datasheet:

Positive-to-Negative Converter for full protection under fault conditions.

Not sure if that means a full failure.
 

ian field

Joined Oct 27, 2012
6,536
Interesting. What about more sophisticated bucks like the LM2675? It does say this in the datasheet:

Positive-to-Negative Converter for full protection under fault conditions.

Not sure if that means a full failure.
A positive to negative converter is often designated "inverter" in this context - they're more in keeping with the flyback topology, and AFAIK: short the unregulated supply if the transistor fails.
 

ian field

Joined Oct 27, 2012
6,536
I don´t see how it could become a boost when it has buck topology.
Never mind - there's time to learn..................

A positive to negative switcher uses the back emf to generate a negative voltage - but it also has a slight similarity to a buck.

The other option is to generate the negative output with a secondary - which is generally flyback unless you go in for a full or half bridge forward converter.
 
Th specs are misleading.

If Boost, they should mention the how much up converting it can do. e.g 12 V in 15 V out.
"Boost" suggest more than. How much more than Vin is the minimum output.
 
Top