Most of the power supplies around my shack and on my bench are homebrew affairs built with junk box parts. I have a few commercial supplies that I’ve picked up over the years, but did not have one that really met all my needs. I wanted one that could handle at least 3 amps up to at least 28 volts. I also wanted dual tracking because of the analog circuit work I do. An I wanted meters–preferably individual meters for each voltage and each current.
I seriously considered building another, but after reading lots of reviews, reviewing lots of designs, and coming to the realization that “metering” was going to be a challenge if I built one from scratch, I finally decide to buy a Tekpower TP-3005D-3 power supply off of Amazon.
There has been a recent chance and the Tekpower supply that arrived actually had LEDs to display the voltage and current instead of the earlier LCD displays. The LEDs are bright, large, and easy to read. Voltage is displayed in green and current in red. The earlier LCD displays had received positive reviews, but I can’t imagine anyone not being just as happy with the new LED display. (I actually prefer LEDs to LCDs in this type of application.)
I carefully read the reviews for this and several other supplies before making my choice and there is some good information out there. But let me run through what I did and what I found as I put the supply through its paces.
Estimated shipping time was two weeks and it actually arrived several days early. (In time to have a weekend to play with it!) It was double boxed. The interior box was the original factory box and the supply was wrapped in a large, heavy plastic bag and held in place with a pair of foam forms. The outer box had taken a beating during shipping and showed it, but the interior box looked fine and the supply itself was in perfect shape when unboxed.
In the box was 1) the power supply itself, 2) a power cord, 3) one pair of polarized (one black, one red) banana connector to alligator clip cables, 4) a second pair of polarized banana connector to alligator clip cables that used two wires each (for higher currents), 5) a short connecting cable with banana plugs on both ends for series operation of the supply and 6) a short printed manual.
As I discuss the supply, I will give some readings or measurements in places. When I do, I will give the right / master side reading first and the left / slave side reading second.
I powered it up and found 32.5 volts and 32.4 volts on the outputs. I confirmed these with my bench meter to verify the accuracy of the supply’s meter. If I tweaked, I could get them one digit out of synch, but the meters were definitely within the plus or minus 2.5% specified for both the voltage and current reading I took at various times.
Initially I connected the supply to a 50 ohm dummy antenna because that is the thing I have setting around that can dissipate significant power without straining. I was kind of surprised that the output voltages stayed the same when a load was connected—basically I was drawing 0.65 amps per side and the voltage stayed the same as the no-load voltage with the regulator control fully clockwise.
There are two buttons in the middle and they are marked with the following configurations:
Up / Up is “Independent” and you can vary the voltage and current limiting individuals for the two variable supplies.
Up / Down is “Series”. This did not exactly work like I expected. The manual indicates this is the “tracking” mode, but it isn’t. More on that in a bit.
Down / Up is “Parallel” and the two supplies are connected in parallel with the voltage of the slave supply tracking the voltage you set for the master supply and the current being evenly split between the sections.
There is a cable that lets you jumper between the two supplies so they can operate in what I would describe as an “independent” series mode where the voltage of each supply can be adjusted independently.
Now to the mode that isn’t listed on the panel:
Down / Up seems to be the true “series tracking” mode. With the buttons in this configuration, the supplies track each other and the included jumper is not needed—the positive of the slave supply and the negative of the master supply are connected automatically and internally.
So basically, I think that the mode most people would want when operating in series (tracking) is actually Down / Up, not Up / Down, and that the jumper is unnecessary.
Let me come back to how the protection and current limiting operate in modes other than “Independent” in a bit.
With a 1 ohm power resistor, I was able to see where the current maxed out and it was 5.08 amps on the master side and 5.24 amps on the slave side. I just don’t have enough power resistors around to see if I could get 5 amps at 30 volts, but from everything I saw I don’t have any reason to doubt I could.
I did pull out a 12 volt motor that drew around 0.6 amps. I ran the voltage up and down a bit and things worked fine.
I do a lot of RF work and for me the “gold standard” for those applications is being able to power a direct conversion receiver. Until now, virtually every commercial supply I had used had to be “RF proofed” with bypass capacitors on the rectifiers and a number of other modifications. With this extra effort, most power supplies will induce home (by modulating the carrier from the local oscillator in the direct conversion receiver that “leaks” out). A lot of times this forces QRP operators with simple rigs to operate from battery instead of a power supply. The Tekpower supply worked fine as it came from the factory. That should be enough to impress any amateur radio operator who has struggled with these issues before.
I usually separate a direct conversion receiver from the power supply by at least three feet anyway, but found that I could set my FOXX-3 transceiver (which is in an Altoids tin and uses a direct conversion receiver) on top of the supply and there was no induced hum.
Part of that may be because of the use of a toroidal transformer. This type of transformer construction is going to offer more self-shielding than a conventional transformer with open windings.
And the transformer in the supply is impressive. It is about 5 inches across and 2.5 inches high and bolted into the middle of the chassis on the bottom. It looks fully capable of handling the 300+ watts the supply can provide.
Okay, by saying that I have to admit I “popped the hood” and took a look inside.
I was impressed. Construction is neat and professional. They made liberal use of heat shrink to cover connections. The boards use a mix of surface mount and through-hole components. (Most of the power components such as diodes and the filter capacitors are through-hole.) There were small touches like a dab of glue on the filter capacitors (which are on a board that is mounted vertically) that most manufacturers probably would not have done.
For each of the two independent supplies, there are four pass transistor and eight rectifier diodes. For a five amp supply, that should offer some healthy headroom and I doubt anything is being pushed particularly hard.
The physical construction is also impressive. The top of the supply have a conventional “U-shaped” cover. It is held in place by twelve screws. Two of these also hold the handle in place. The handle is secured to a cross brace that runs front to back. The handle is a soft rubber / plastic and has a metal strap in its center. There are also two other cross braces—one on each side of the supply.
Bottom line on physical construction is that it is solid, uses about twice as many screws as they could probably have gotten away with, and the handle is solidly mounted. I move the supply from my workbench to my ham shack (both in the same room, but about ten feet apart) on occasion, so it is nice not to have to worry when I grab the handle.
Also, there are no sharp edges. The heat sink is inside the enclosure and there are two fans that indicate they kick on at around 50 degrees (I assume that is Celsius). I hooked motors, resistors, and everything else I reasonably could to load the supply down. Even after letting things set and run for an hour the heat sink was barely warm and the fans never kicked on. It is rated for operations up to 40 degrees (Celsius) ambient temperature which is pretty hot—with those kinds of external temperatures you might see the fans operating.
A lot of other supplies have the heat sink as part of the physical enclosure on the back and the heat sink fins often have sharp edges that can give you a nasty cut if you reach behind the supply without thinking. Having them internal is a nice touch.
There appeared to be surge suppression on the input line and there was a line safety filter cap which may be another part of what helped it perform so well with a direct conversion receiver.
One thing that jumped out at me was that the filter capacitors are not huge monstrosities. At 6800 uF capacitors rated at 63 volts they are decent sized. I have seen a lot of “cheap” supplies try to make up for a lack of “iron” (i.e., a transformer that is being push a bit harder than it should) by adding larger capacitors.
Why does this matter? Well what made me first notice this was that when I shorted one of the outputs there was not even a tiny spark or spit. With a short, you essentially have to drain all the energy off the filter capacitor. With a large filter capacitor this can give you the small spark you often seen when a supply is shorted. The bad thing with that is that if the filter capacitors are after the regulator, that energy is not throttled down by the overcurrent protection. If you design things properly and don’t try to push every component to the limit while shaving a few dollars in parts, you don’t have that issue. And this supply definitely does not have that issue.
Okay, let’s talk about over current protection. I skipped earlier it because this takes a bit.
In the series, parallel, and the series tracking mode, the current limiting when the supplies kicked into constant current mode was a bit unpredictable. In parallel mode you should be able to get 10 amps in constant current mode. I don’t have anything that can sink that amount of current at any reasonable voltage on the bench so I didn’t explore it much. For the series and series tracking mode, things seem to work best if you crank current limiting on the slave all the way up and just used the current limiting adjustment on the master side.
If you think about it, this isn’t unreasonable. Protection does not drop into some type of voltage foldback on overcurrent with this supply, instead the supply switches to a constant current mode. And if one side of the supply is in constant current mode while the other is trying to stay in constant voltage mode they are not necessarily going to play well together.
I rarely use constant current mode supplies in any of the work I do. Honestly, I would rather have voltage foldback instead of constant current. But at the same time, if I was the designer, I can see where the constant current mode makes the supply much more flexible in the types of applications it can handle.
The bottom line is that the over current protection scheme is fine and was a good design decision, you just need to understand how it is working. The fact that there are indicators that light green for a supply in constant voltage mode and another that lights red if it switches to constant current mode gives you a quick indication of what is going on. For most of my work, either of the red lights coming on is an indication that something is wrong. And I can still set the current limit to a position that protects the circuit I’m working on, but the circuit will still be receiving current even if it has shorted out.
I am a big believer in spending the time to understand how your test gear really works both when you are using it like you should as well as when you encounter something unexpected. This supply is intuitive enough that ten or so minutes of playing around with it and a power resistor is more than enough to get a solid sense of what how it behaves.
There is a 5 volt, 3 amp supply that is independent of the others. It has a green LED that is labeled “Over Load”. Really, if the light is on things are fine and if you overload the supply the light goes out.
I haven’t tried it yet, but with the ability to get a bit over 60 volts DC out and an independent 5 volts DC, this supply might even be able to power one of my small homebrew vacuum tube transmitters or receivers. The 5 volts is a bit lower than normal for operating the heaters, but I suspect you could make things work. Anyway, that’s just a somewhat random thought.
So to wrap up…
There are little things I would change if I was designing one from scratch like the use of voltage foldback instead of dropping into constant current mode and maybe a fixed 12 volt output instead of (or, even better, in addition to) the fixed 5 volt output. But those are nitpicks and design decisions the engineers made and not really shortcomings of this particular supply or design.
The only two things I could legitimately ding the supply on are two labels. Really, I think the “Series” mode should show using the two switches as Down / Up instead of Up / Down and the LED for the 5 volt supply says “Over Load” when actually the opposite is true (the light is on when there is no overload and goes off when it is overloaded). You are instantly reminded of the first one because the voltages don’t track when you use the labeled series setting for the switches. The second one is also fairly intuitive.
Overall I am very impressed with the supply’s performance and construction. It seems like the components are not being pushed to anywhere near their limits. The supply is solidly built and not going to buckle or bend. You can grab it by the handle and not have to worry about the handle snapping off. There is no external heatsink that is going to snag things or that could give you a burn if they get warm. The LED display is bright and easy to read. There are indictor LEDs for other things that keep you aware of how the supply is performing.
This supply is a great value for the price and I don’t hesitate to recommend it. I suspect this one will still be going strong a few decades down the road.
W4JBM 