DIY Benchtop PSU, what about these.....

[portreathbeach]

I've been looking at building a Benthtop PSU and found that you can buy very cheap DC-DC converters based on the LM2596 chip. They take up to 40v input and can output anything from about 1.5 - 37V at a couple of amps.

Here are a few links:

LM2596 Mini DC-DC Step Down Adjustable Power Supply Module | eBay

LM2596 DC-DC Converter Board Voltage Regulator Stepdown Module Built in LM2596S | eBay

New LM2596 DC Buck Step-Down Voltage Adjustable Converter Power Module Regulator | eBay

This one looks interesting:

LED Voltage and current display LM2596 DC-DC Step Down CC-CV Adjust Power Supply | eBay


OK, now time for my question.....

If I want to make a dual output power supply that I can connect in series to give me a +/- DC voltage, for testing OpAmp circuits etc, can I use two of these DC-DC converters? I was thinking it would work, as long as I use a dual secondary mains transformer and feed each DC-DC converter from a different winding.

Any thoughts?

[portreathbeach]

Just reading a bit more about the first link I posted:

LM2596 Mini DC-DC Step Down Adjustable Power Supply Module | eBay

It goes on mostly about constant current and charging batteries or using it as an LED driver. Could this simply be used as a PSU with current limiting feature. All you would have to do then is buy a digital 7 segment, 3 digit Volt/Ammeter and you have a very cheap PSU (except for the mains transformer that is).

But, can these be series wired if fed from different windings of the mains transformer, I still havn't found the answer on the internet.

[portreathbeach]

Just found this web page:

SMPS, Switching Power Supply, DC Power Supply, Switching Mode Power Supply, User Manual, Service Manual about INSTALLATION, WIRING, CONNECTIONS

In fig 15, it shows 2 switched mode power supplies connected in series and they use a diode across each PSU for reverse protection. It doesn't show a centre tap in the diagram, but I assume you could tap off the centre for 0v and then have +/- volts.

[Elvee]

Originally Posted by portreathbeach

Just found this web page:

SMPS, Switching Power Supply, DC Power Supply, Switching Mode Power Supply, User Manual, Service Manual about INSTALLATION, WIRING, CONNECTIONS

In fig 15, it shows 2 switched mode power supplies connected in series and they use a diode across each PSU for reverse protection. It doesn't show a centre tap in the diagram, but I assume you could tap off the centre for 0v and then have +/- volts.

For experimentation purposes, switching supplies are awful, even the professional ones: they accumulate about every flaw in the book: they generate residues, have suboptimal dynamic responses, need at least some tens of µF at the output making fast current limitation hopeless, etc.

In addition, mains fed ones (this is not the case here) require Y capacitors which are a nightmare from a sensitive electronics point of view.
In short, this kind of supply should be avoided except for power electronics where they are not only useful, but indispensable.

As a lab supply, a traditional linear one is really the best, and is not that more costly compared to cheap and dirty switching types.

[portreathbeach]

Thanks for the reply.

I already have a spare toroidal transformer and rectifiers etc, so would it not be worth getting a couple to make a cheap bench top supply for testing circuits?

You say about 'tens of µF' at the output, but if I wanted to use it purely to set a voltage and test some circuits, this wouldn't be a problem, would it? Currently I use 2x9v batteries and it really is a pain connecting them and regulating them if I need different voltages.

Anyway, could 2 of these be put in series to give -ve, 0 and +ve voltage supply? Is it worth spending a few pounds and making a simple DC power supply?

Thanks again for your help Elvee

[Harleyjon]

I would steer clear of digital power supplies. The main reason is by way of the beast. They are either on or off. When the threshold voltage is reached, there is always an over kill. Use a linear device. They are cheap and variable. Use two windings, that will keep the two supplies separate and then you can series couple them, giving +ve and -ve supplies.
An LM317T is a good start and if more current is required feed a power transistor with the LM317.

[portreathbeach]

Thanks. I thought about using an LM317 as I have a few here, but the problem I thought with using one is the heat that is dissipated. If I feed it with say 30v, regulate it down to 5 and draw 1 amp, the 317 has to dissipate 25 watts. I suppose a large heat sink would do the job.

[sofaspud]

That's one solution. There are others (eg, a Hi/Lo range switch). The example reminds me more of a digital supply, which would probably be better as a separate bench supply anyway (with, say, 12V-5V-3V3 fixed outputs).

[Harleyjon]

On my bench supply. 30 - 0 - 30 at 25Amps, I use a pair of 2SC5200 on a large heat sink per side and it works very well.
A good suggestion is to switch the supply DC to a lower level if you don't use the voltage. That will keep the dissipation down. I did build mine in 1983 and it has had one new voltage selector pot only!

[portreathbeach]

I know you said it was built in 1983, but have you got a schematic?