My Project: Modern Linear Power Supply

[B0B45]

Hello,

I would like to share and discuss my current project: a modern linear power supply with digital controls and analog regulation.

I started with some plans of an old Elektor project and replaced the old parts with modern components step by step. With some tips from the German mikrocontroller.net forum, I was able to add some tweaks.
At the current stage, the power supply is working as intended but is still in testing.

Specs:

Output: 36V - 5A
Resolution: 1 mV, 1 mA
Error: < 0.05%

Controller: RB Pico2 - RP2350B
Measuring: INA228
Op-amps: ADA4522 (ch1) and OPA182 (ch2) for comparison

Main transformer: 400 VA - 2x 18-0-18 V + 12-0 V
Power transistors: MJL3281A

What I share: Some pics of the building process, the schematic of the power supply PCB, the file for LTspice, and some measurements.
What's coming soon: a schematic of the controller PCB, LTspice for current regulation, and more tests.

Feel free to discuss the plans and my ideas; I bet there is still room for improvement.

Greetings,

BOB


last version on page 2

[inse]

Impressive design!

[d-el]

Think Off-On slew rate is very fast. Combined with long cables and specific loads, this can cause unpleasant effects. Consider slowing down the slew rate.
My power supply used to respond about the same as yours, but then I reduced the slew rate by about a factor of 10 to minimize overshoot.
Transient process with settings 8V, 40mA, 50Ohm load (oscilloscope):


It's more interesting the next.
Measuring transient processes of load current changes, as I could do:
Setup: 8V 4A. One 10 ohm resistor is connected in parallel with the DUT output, and a second 7.3 ohm resistor is connected in parallel with the DUT via a controlled switch. Current is measured using a Lecroy AP015 current probe. Voltage is measured directly at the DUT terminals using a passive high-impedance probe.

[B0B45]

Talking about slew-rates, I did some tests with a 10 Ohm resistor:

Pic1: ch.1 slew-rate from 0V-15V with 10 Ohm CV-mode
Pic2: ch.2 slew-rate from 0V-15V with 10 Ohm CV-mode

Pic3: ch.1 slew-rate from 0-1A with 10 Ohm CC-mode
Pic4: ch.2 slew-rate from 0-1A with 10 Ohm CC-mode


Guess the current of 40mA was a little to low and in the range of the super high slew-rate

It's more interesting the next.
Measuring transient processes of load current changes, as I could do:
Setup: 8V 4A. One 10 ohm resistor is connected in parallel with the DUT output, and a second 7.3 ohm resistor is connected in parallel with the DUT via a controlled switch. Current is measured using a Lecroy AP015 current probe. Voltage is measured directly at the DUT terminals using a passive high-impedance probe.

I'll make you a screenshot of the results later.

[B0B45]

I did your test with a 10 Ohm + 7,3 Ohms resistor.  Current is measured with a CP 35 (30A-100kHz) probe.

I'm afraid the main-relais of ch2. died and is stuck at high level during former tests. Maybe to high inrush current for 16A relais? 

[Kleinstein]

The relais for range switching are a weak point. The issue is the inrush current to charge the filter capacitor. This may cause sticking contacts. With only 2 windings the alternative could be electronic switching or a power stage that does the cross over.

[Fire Doger]

Is this chassis off the shelve or someone had too much fun making it?
It looks indestructible...

[NiHaoMike]

Consider using an automotive relay, which is quite easy to find with very high current ratings.

[AVGresponding]

Consider using an automotive relay, which is quite easy to find with very high current ratings.

Two things which may, or may not, cause problems with using an automotive relay: 1, the coil current tends to be quite high, and (b) the contact resistance might be quite variable

[B0B45]

Hey,

Instead of using a relay, what about using thyristors or triacs? 
I found an interesting idea of using two thyristors instead of a relay and a full-bridge rectifier. (source: https://buebchen.jimdofree.com/selbst-gebaut-diy-homebrew-feito-por-mim/elektronik/labornetzteil/)

If the thyristors are off, the caps get charged with half-waves from both inputs. If the thyristors are on, it works as a full bridge.
(Sim is just an example; no final parts)

What do you think of that idea? High current shouldn't be a problem.

Edit: Screenshot of current from the transformer

[Kleinstein]

The solution looks sensible. The currents may be a bit lower if one includes a fuse and maybe have higher transformer resistance and maybe a little stray inductance.

[temperance]

You can switch the relays when the transformer voltage is below the capacitor voltage.

[voltsandjolts]

Tidy construction.
I like the box heatsink, built for fan mounting - where did you get that?

[B0B45]

Tidy construction.
I like the box heatsink, built for fan mounting - where did you get that?

Thanks,

the heatsink is from aliexpress : https://www.aliexpress.com/store/1100040145

[d-el]

I'm afraid the main-relais of ch2. died and is stuck at high level during former tests. Maybe to high inrush current for 16A relais? 

I think the problem isn't the high current, but rather the sufficient voltage required to generate plasma. What DC voltage is the relay rated for?

To extend the relay life, you can implement the algorithm as follows: { set 0V; delay 5ms; relay_off }.

What's the idea behind using a relay? Factory power supply Agilent typically don't have one relays.

[Kleinstein]

Relays for tap switching are very common the smaller lab supplies (e.g. 30V 3A). The problem here is that with higher power / a larger transformer and plenty of filter capacitance the current pulse from switching up gets quite large and thus a limited lifetime for the relay.
The relays may suffer less with the often smaller steps with 2/3 relays to use more transformer steps. Also a smaller transformer helps with added impedance.

Switching the output off is not really an option, as the relay may have to switch under load without interrupting the output. Ideally there would be no significant glitch at the output.

[B0B45]

The relay I was talking about is K1 in the schematics. It switches the transformer coils from 0V-18V to 0V-36V AC. So NO and NC always get inrush current when turned on or switched to higher voltage.

I know the Agilent schematics; that's why I mention the SCR/thyristor design as an alternative.

[d-el]

Sorry, I thought the relay was at the channel output. I missed the full schematic.

Four voltage levels, a four-level power stage, one output.
The first time I saw this idea was here:
https://www.eevblog.com/forum/projects/diy-bench-power-supply-psl-3604

I used the same idea, but implemented it myself.

[calzap]

Impressive project.  I don’t understand C4 and C3 ... 100 nF cap paralleled with a 47uF cap.  Wouldn’t the smaller one be within the tolerance range of the larger?  Maybe some empirical fine tuning or maybe different types of caps with different charge/discharge timing?

Mike

[Kleinstein]

The parallel combination of capacitors like 100 nF with 47 µF is usually a combination of a film or ceramis with an electrolytic. It is not about the extra capacitance, but a better high frequency behavior and the different position on the PCB. Electrolytic capacitors may not be that good for the very high frequencies (e.g. > 10 MHz).

[B0B45]

Today I disassembled the relay. You can see in the picture, it was welded together.

Guess it's time for a version 1.2. 

My ideas for now:

- Replace relay with SCR

- Replace INA228 with a 4 ch. diff ADC like ADS1219, so I can read the sense voltage correctly.

- Rearrange some parts on the PCB, like the pots.

- Do some tweaking of the regulation

- ....

Any more suggestions?

Happy Holidays

[Kleinstein]

From the photos the spacing around the 230 V to power the auxiliary transformer looks a bit on the low side.
For the regulators one may include a resistor in series to the main FB capacitor. This is a common place for some additional tweaking / adjustment.

[voltsandjolts]

Could perhaps use a small ssr in parallel with the relay. The ssr would only be needed for 40ms or so when switching the relay on/off.

[mariush]

I'd consider doing 4 x 9v-12v AC secondary windings to allow for switching between low voltage / high current or high voltage / default current  (parallel 2 / series 4 )  and also be able to use relays to switch between 1-4 windings to have less heat produced.

There's reasonably cheap solid state relays that can handle up to 7-10A of current and AC voltages which you could use instead of mechanical relays if you're worried about longevity.

[temperance]

You have a microcontroller. You could use the microcontroller to operate a crude buck converter.