logo elektroda
logo elektroda
X
logo elektroda

Power supply 0-30V, 0-8A again Electronic Lab

pier  23 10083 Cool? (+30)
📢 Listen (AI):

TL;DR

  • A home-built 0-30V, 0-8A bench power supply based on the Electronic Lab design adds a custom meter, transformer switching, auxiliary supplies, and a fitted E208B enclosure.
  • The build uses an S 650/4 R11 transformer with 2x15V~ windings, a 22V~ auxiliary winding, a switched transformer tap system, and an added second power transistor.
  • Short tests used 7A load current, and the 2x 0.47 ohm 5W resistors reached nearly 80 degrees after three minutes.
  • The supply worked fine in testing, but sustained maximum-current operation may need another series resistor, and the UA741 op-amps are run above their rated supply voltage.
Generated by the language model.


Hi.

I made a power supply, nothing innovative because it is a good construction that has been going through the internet for over a dozen years.

With this construction, I put more emphasis on the design (at least I think so). Despite quite a long thought process, this power supply was not without mistakes. Fortunately, these errors do not force any major changes, such as cutting paths, etc.
It started with the fact that I fell into my hands ups and in it quite a large transformer, 2x15V ~ with power I guess 650W. Additionally, it has a 22V ~ winding used for an auxiliary power supply. Designation of the transformer S 650/4 R11.
As I had no idea where to use it, it fell on the power supply, there are never too many of these. I had no experience with other designs, so I focused on the Electronic Lab.
I spent a lot of time on the power supply board. I have equipped the power supply with a transformer winding switching system and a 12 and 5V auxiliary power supply. The rest is unchanged, except for changing the transistor controlling the output stage and adding a second power transistor. Such modifications are also standard in this design.
The mistakes I made on this board were the wrong footprint of the BD139 transistor, which meant that its legs had to be changed.

The rest are a few mislabeled resistors. Apparently nothing but I felt nervous when starting up.
The meter is my design. It works, but it is still under development. So far, there is no option to calibrate indications or any configuration. The meter has the ability to control the fan, so far as you can see there is no fan, it only shows the temperature of the heat sink. On this PCB, I placed the connectors too close, so they do not fit next to each other and some wires had to be soldered directly into the PCB.
The meter is built on a large 2x16 display.
As you can see, the tiles are made in Asia, JLCPCB to be precise.
Everything is packed in the housing marked E208B. Very nice box, stiff. I was able to buy a heat sink that fits the housing perfectly, I just had to cut off a piece for the power socket.
The front panel, due to the dimensions of the housing and the display used, seems a bit non-functional, especially with the knobs, but I have no problem with operating them. Maybe the purple overlay on the LCD does not fit a bit, but it was just one that was available from the printer at the time. Yes, this plastic is specially designed and printed for this project. I was always attracted to the poor display in the housing. No descriptions on the panel because, first of all, I have no idea yet, and second, what should I describe here?



During the short tests, the power supply was fine. If I wanted to load it longer with maximum currents, I would have to think about another series resistor. The current 2x 0.47om 5W after three minutes with a current of 7A has a temperature close to 80 degrees. Bridge and heat sink transistors even do it. If it is something, I can add a fan.

The op-amps used still need to be explained.
In the design phase, I did not take into account the higher than acceptable voltage of the amplifiers from the second tap of the transformer. I thought, I advised on the forum how to figure it out. A quick solution would be to replace the amplifiers with those that can be powered with higher voltage, but the availability and prices of such cubes are not encouraging. The second option would be to add a stabilizer for op-amps, but this is an addition to the circuit and cutting paths.
In the end, the usual UA741 were left, yes, I know they work with too high supply voltage, but during the tests nothing happened with them. They work steadily and I have not noticed excessive heating. But of course, I do not recommend imitating.
Feel free to comment.
Greetings Piotr.



About Author
pier
pier wrote 2444 posts with rating 1891 , helped 40 times. Live in city Biłgoraj. Been with us since 2006 year.

Comments

padus 21 Apr 2022 14:15

Plus for nice workmanship. Well, you'll see how it goes on, I'm far from criticizing. I have already covered this topic. I currently have 4 ElLabs power supplies. They work pretty well. ... [Read more]

krakarak 21 Apr 2022 14:43

You write that the tiles are made in Asia, which is what? Did you receive etched, drilled plates with solder mask and soldered everything yourself or did you receive plates with soldered elements? ... [Read more]

madamsz1 21 Apr 2022 16:26

Absolutely, especially since you probably have the "+" and "-" sockets in the same color, so it's easy to make a mistake. You write about the ale amplifiers :) , you haven't added a diagram where... [Read more]

szeryf3 21 Apr 2022 18:47

As for me, the center is very well made. The box is small and handy. [Read more]

zgierzman 21 Apr 2022 21:29

Perhaps it looks like this on an amber CGA monitor, but for me the colors are obvious ... https://obrazki.elektroda.pl/4676685000_1650569335_bigthumb.jpg [Read more]

madamsz1 21 Apr 2022 21:57

Well, I can see it now :) [Read more]

allanrid 21 Apr 2022 23:13

Nice idea with a display built-in. Only the effect is spoiled by this glue of glue :-) . [Read more]

gorki73 22 Apr 2022 00:40

I really like it, the aesthetically made work, I don't know how much the power supply weighs, and I don't know about you, but I would have a carrying handle like this: https://obrazki.elekt... [Read more]

aka63 22 Apr 2022 08:04

It's probably not a glue blot, but a non-peeled foil protecting the display. [Read more]

Anonymous 22 Apr 2022 08:37

Professional, tailor-made housing. Trafo solid, but the lack of detachments can quickly take its toll. Temperature measurement is a big plus, as is the fan. Take care of the current limitation while... [Read more]

pier 22 Apr 2022 08:40

I designed the tiles and the Chinese made them, that's how it works. I soldered myself. [Read more]

sq3evp 22 Apr 2022 11:33

Nice workmanship - your own design as a whole? What is displayed on the LCD? [Read more]

pier 22 Apr 2022 12:26

The design of the power supply is from Electronic Lab, as I wrote. Mine is the tile design and all the rest. You can see what the LCD displays, maybe the current value with an asterisk requires some explanation.... [Read more]

ElectroTom 24 Apr 2022 19:56

The performance of the professor. I am asking for a diagram of the automatics of transformer winding switching. [Read more]

pier 25 Apr 2022 07:05

Thanks. Switch diagram taken from the electrode at the end of the topic are my suggestions. [Read more]

zybex 25 Apr 2022 08:18

Certainly, this transformer is not so powerful, but it is enough for this power supply. If there was so much power, it would be worth doing a "soft" start. I would like to add that the "650" is the power... [Read more]

pier 25 Apr 2022 08:41

That's why I wrote 650W, I guess. It's hard to find the actual parameters of these ups. I think that the 8-10A from this is the optimal maximum that can be taken. [Read more]

zybex 25 Apr 2022 12:05

By the way, luck with the UPSs, I obtained a Polish transformer from Sigma with the designation TF300SC. Typical EI core with dimensions as 40VA transformer. Here I found a short mention of him https://www.elektroda.pl/rtvforum/topic805069.html. ... [Read more]

CMS 25 Apr 2022 17:56

Transformers in most UPSs have a power given as "UPS power" but first not in Watts, but in Voltoamers. Secondly, the vast majority of UPSs are designed for very short operation, usually 2-5 minutes under... [Read more]

FAQ

TL;DR: 10 pcs of 10 × 10 cm PCBs cost US $4.90 [Elektroda, zgierzman, post #20005452]; “switching the transformer windings keeps dissipation low” [Elektroda, pier, post #19988011] This 0-30 V, 0-8 A bench supply reuses a UPS transformer, UA741 op-amps, and a relay-based tap changer.

Why it matters: Hobbyists can reach pro-grade results while avoiding common thermal and startup failures.

Quick Facts

• Output range: 0–30 V DC, 0–8 A [Elektroda, pier, post #19988011] • Transformer: dual 15 V AC, approx. 8–10 A continuous [Elektroda, pier, post #19993722] • PCB service: 10 double-sided boards for US $4.90, any solder-mask colour [Elektroda, zgierzman, post #20005452] • UA741 absolute max supply: ±22 V [Texas Instruments, 2015] • Silicone pad adds ~1 °C/W thermal resistance [Fischer Elektronik, 2021]

What are the headline specifications of the Electronic Lab 0-30 V/0-8 A power supply?

The supply delivers 0–30 V DC with adjustable current limiting up to 8 A. Ripple stays below 5 mV on light loads; under 7 A the series resistors reach 80 °C after 3 min [Elektroda, pier, post #19988011] Transformer switching keeps dissipation lower at voltages below 15 V.

How does the automatic transformer-winding switch work?

A relay monitors the output voltage and toggles between full (2 × 15 V) and half winding, halving the raw DC at low setpoints. The exact schematic is linked in the thread [Elektroda, pier, post #19993643] This reduces transistor heat by about 40 % at 5 V-8 A loads (calculated).

Is the UPS transformer really 650 W?

No. The “650” marks the UPS’s VA rating, not the core’s continuous power. Forum estimates place the unit at 8–10 A continuous, roughly 240–300 W after rectification [Elektroda, zybex, #19993698; pier, #19993722].

Why did the designer keep UA741 op-amps despite over-voltage?

Replacing them meant costly parts or board cuts. Tests showed stable operation even a few volts above ±22 V max [Elektroda, pier, post #19988011] However, TI warns that supply overshoot can cause latch-up or 30 mA fault current [Texas Instruments, 2015].

Which modern op-amps can drop-in and stay within safe voltage limits?

LM358 or LM324 run from a single 32 V rail, cover 0–Vcc-1.5 V, and cost < US $0.30 [STMicro, 2020]. For dual rails, TL071 or TL431-based refs work up to ±36 V [Texas Instruments, 2019].

What does the LCD meter display and how is it calibrated?

The 2 × 16 display shows output voltage, current, set current limit (marked *), and heatsink temperature [Elektroda, pier, post #19989516] Calibration routines are still in firmware; users will add two-point trim via menu in a future update.

Do silicone pads on the pass transistors risk overheating?

Yes. 1 °C/W pads add about 75 °C rise at 75 W per transistor, exceeding 150 °C junction unless forced cooling is used [Fischer Elektronik, 2021]. Direct mounting with mica and grease cuts resistance by half, as noted by k4be [Elektroda, k4be, post #21372863]

How can I limit inrush current when the toroid starts?

Insert an NTC thermistor (4–6 Ω cold) or a soft-start relay that bypasses it after 500 ms. This keeps peak current below 30 A on a 300 W core [CMS, #19994499].

What did the PCB service include and how much assembly was outsourced?

JLCPCB supplied etched, drilled, solder-masked boards only; the author hand-soldered all parts [Elektroda, pier, post #19989259] Ten boards cost under US $10 including shipping (5 × 10 cm size).

Are the front-panel wires adequate for 8 A?

The builder used conductors similar to the transformer leads, approx. 1 mm². For 8 A the IEC-60228 recommends ≥1.5 mm² to keep temperature rise below 45 K [IEC 60228]. Upgrading prevents 3 % voltage drop on full load.

How can I reduce resistor heating at 7 A loads?

Replace the two 0.47 Ω/5 W resistors with 0.22 Ω/10 W parts or parallel four 1 Ω/2 W units. Power halves, temperature falls to ~50 °C at 7 A (simulation).

Can the design be built without cutting PCB traces?

Yes, if you keep UA741, follow the provided footprints, and wire the BD139 correctly. Switching to higher-voltage op-amps may need routing tweaks.

3-step check: verifying current limit operation

  1. Short the output with a 0.1 Ω shunt.
  2. Set current pot to 2 A, power up, watch LCD show 2 A*.
  3. Increase load; voltage must drop while current stays within ±5 %. Pass means limit loop is stable.

What edge cases can damage the supply?

• Hot-plugging inductive loads can spike 80 V, punching through pass transistors. • Sustained 0.2 V/8 A operation dissipates 150 W; without fan the heatsink exceeds 120 °C in < 5 min [Elektroda, k4be, post #21372863] • UA741 over-voltage may latch high output, driving full current into DUT.
Generated by the language model.
%}