Review of DP50V5A & DPS3012 bench top power supply modules, best bang for the buck around!

Any power supply with an output of less then 60V will work fine, it naturally needs to handle the wattage you plan on using as well.

I use a 48V 480W power supply with mine.

To be clear, any power supply with an output a couple volts higher than what you want out of the 5015 will work. In other words, if you don’t need more than (say) 15 volts, you could use a 24 volt supply. This is cool because old computer power supplies are often scroungable for a little bit of effort. So you can save up for that nice 48 Volt supply you have your eyes on.

Correct, it needs a few volts of overhead for the buck to operate.

Although most computer power supplies around here are 12V, servers have 24 and 48V supplies though. I am using the DELL server power supply posted either earlier in this thread or the other thread on the dps5015.

I do not think that RD Tech reads this board anymore, I doubt that you will get a reply anytime soon.

You could send a message to them on aliexpress and they should reply.

For the price the DPS series can not be beat, naturally for more money there are better options.

I will say that while I like the 5015 unless you need the 50V and 15A capability it does have more ripple voltage then my older 3012 and my 5A version has basically zero ripple (around 10mv IIRC). For most things 30v and 12A is more then enough, heck the 30V 5A version is even more compact and enough for a lot of things, although not enough to really test LED’s now days.

Here is the Design for the Acrylic case I made in case anyone wants to make one. It is cut out with a laser cutter from standard hardware store .08” acrylic sheet (“plexiglass”). Thicker acrylic would be stronger but costs a lot more and is much harder to find.

Proper acrylic glue will be stronger and look better but hot glue or super glue works just fine as well.

Which power supply is best to use with DSP5015?
AC 110V/220V TO DC48V 8.3A 400W Regulated Transformer Power Supply For LED Strip or
AC 110V/220V TO DC 48V 400W 8.3A Single Output Regulated Switching Power Supply

It´s almost the same price.

Got a link to them? They sounds like the same thing.

Although either should work, all that matters is that you get a DC voltage out and a 48V output is a great number, you rarely need those last few volts.

http://www.ebay.com/itm/222273922467?\_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Here is some explanation, but I really don´t know which one is the best for the DSP5015. More heat but cleaner current with transformer, less efficient.

I am going to go out on a limb and say that they are both the same and the translation is the thing that is different.

Either will work fine for what you want though, so just go with whatever seller you feel more comfortable with.

The two you link to is both switch mode supplies.
Old style transformer based supplied will usual last considerable longer than cheap smps, they have less efficiency and weights a lot.

Ok, I pick one of them and hope for the best. Thanks!

I used the MeanWell (I got the 48V, maybe ~7A iirc) that the supplier of the DPS5015 recommended, works perfectly and seems of very good quality for an aliexpress item.

I was testing some lights today and figured I would mention this to those that have one of these or want to see the kind of things they can be useful for (they also work great for diy colloidal silver among other things as well).

I use a current shunt for testing tailcap currents. I have tried the clamp meter and while the readings are good I find it a pain to use and you still have to use a shunt anyways in order for it to work.

I figure cut out the middle man, I find using the shunt directly to measure current to be far simpler and easier with only 2 hands.

A shunt works very simply, it is simply a very low ohm resistor. When you pass a current through it there will be a voltage drop according to ohms law. If you measure this voltage drop you can then figure out the current. This is the same thing that happens inside your multimeter, just on a much smaller scale.

You can just connect your DMM to the shunt directly but it will be impossible to get consistent readings, so soldering some “signal wires” to the shunt is much better and easier.

The issues for most is figuring out how to make/buy a shunt for what we need. It turns out it is really VERY simple. Just got to home depot and buy yourself 6 inches of 14 gauge solid core wire, generally used for household wiring. The guy just gave it to me for free since it was so small.

This is your shunt, the only thing you need to figure out now is how far apart to connect your signal wires in order to get a voltage drop that equals 1mv per amp.

In order to illustrate what I am talking about here are some pictures of mine:

SkyRay King version:

Now you can calculate the exact length you need with outline resistance calculators, and they are quite close when I tried it in the real world but it needed some fine tuning, mostly due to the difference in the solder resistance.

So the far easier method is to simple connect the shunt to the power supply and feed say 1A into the shunt (or 10A if your leads can handle that, mine can so thats what I used).

Then you will simply move the wires closer or further apart until the multimeter is reading 1mv with 1A flowing through the wire (or10mv for 10A if you did it that way).

You will have to solder the wire in place a few times and it is easiest to simple heat the solder and move the wire up or down the shunt with the tip of the iron. Once you get it right though it is very accurate and simple.

I simply put some alligator clips on the signal wires and connect them to my multimeter leads when taking current readings. This gives me 2 hands free to hold the shunt in place, and you need both of them.

It is also nice since you do not have to worry about shorting out something due to the meter being in current reading mode and/or blowing your meter / fuse.

This 14awg shunt is good for hundreds of amps should you desire to measure that high.

You can also use a regulated flashlight that will be unaffected by voltage drop in the wires to calibrate it. Simple measure the current with your meter directly and then with the shunt and adjust till they match. For example a 105C could be used for this as long as your cell was charged enough to maintain the full output through the meter wires.

Just be aware that copper will change resistance with temperature (About 4% for 10C).

Very useful in a pinch TA, also when not having a proper size shunt when hooking up to a scope.

It should be noted that copper (as most common metals) has a terrible temperature coefficient. Something like 0.4% /degC. So make sure the shunt stays cool and expect even a seasonal effect.

EDIT: HKJ got the ball :slight_smile:

True it will vary some with temperature although not enough to matter for our uses IMHO.

At 10A when testing it on the power supply it only changed the reading by a small amount even when hooked up for several minutes. I think it ended up with about a ~5% change in the reading.

Seeing as where you place the wire in a clamp meter can have that much effect or more I am not worried about it.

Although it is good to keep in mind that it is not a 1% tolerance shunt by any means, just good enough for what we do and cheap and easy.

Since you are in here anyways, since seeing your test of the DPS5015 I got mine running (was using the 3012 before) and I noticed similar results to your ripple voltage.

Turns out that the larger units have a higher ripple (makes sense, just not real clear in the specs). Testing my older/smaller DP50V5A the voltage ripple is only ~13mv.

I wish I knew where the midrange models fell between those. The 3012 for example was more stable but I forgot exactly how much more so.

Just some interesting datapoints.

Same values that HKJ got, 100-140mv IIRC depending on the output.

Interestingly it was more stable when in CV mode then in CC mode. Think it was under 90mv when in CV mode.

I measured it with my cheap scope, it is limited but works well for basic things.

Is there anything we can do to lower the ripple? Better and bigger capacitors?