lab power supply

[glinjik]

Thoughts on this as a lab power supply???

it does not need to be super accurate but be robust and reliable

[wizard69]

it does not need to be super accurate but be robust and reliable

I'm not familiar with this schematic so I can only suggest if it fits your use case go for it.   It is simple enough to cost effective for anyone.   

The other question/suggestion is what is your use case?   Sometimes it is better to design a few fixed voltage supplies instead of  a variable if you want very low cost, linear supplies.   This mainly because you can always find salvage.

[MikeK]

I don't know enough to comment on the design, but if you do build it reply back here with your results.

[glinjik]

i think i will build it but i have one question the transformer i intend to use puts our 42v once rectified this is too much i was thinking of building a dc converter to get it down to 30v is that feasible or would it cause vast problems?

[glinjik]

Here is a simplified version of what appears to be the same circuit

[Neomys Sapiens]

I would suggest to use the first version. If I understand correctly, S1 places the current limit into latching action when closed. That, and the indicators for Uconst/ Iconst are useful features in a lab PSU and definitely worth a few components. Also, the thermal limiter is a good thing to have. The 470µF C7 in the second circuit are not helpful if you ever need it as a current source. If you want it to be more robust, you would have to use more pass transistors with emitter resistors. The basic choice would be two, of which one should be able to handle the power/current/voltage/SOA alone, but this has not worked out in every circuit or build specimen.

[glinjik]

s1 a direct translation

stabilized current / triggered protection

[strawberry]

relay switchable transformer windings?
R18 R25 could be better with dedicated REF for future upgrades and stability

[rstofer]

What's the maximum output current?  Select a 1V output at that current and you will be dropping 40V across the pass transistor.  The power loss across the transistor is (41 - 1) volts times the current.  Pick 1V at 3A and you're talking about a 120W heater.

It is far better to have a transformer with selectable taps and have some kind of logic driving relays to select the tap.  Most commercial supplies have this feature and they get it with several taps on a toroidal transformer.

[glinjik]

The transformer I'm planning to use has multiple taps so yes that's a possibility

[sahko123]

Assuming it's a variable bench supply it's a classic time tested and proven design using an error amplifier going into a series pass element. In other words there is a basic reference and a powerful transistor. The opamps make sure the transistor outputs the correct voltage. If it was a set voltage supply I could think of one way of improving the supply but as it stands it would make a decent supply. Just make sure to put the main power transistor on a heatsink cause it will get toasty.

[blackdog]

Hi,

The first scheme should work well.
It has protections and reasonably well chosen phase margin.
I would only double the MJ29055 to make the power supply more robust.
The second scheme is "stupid" the "designer" does not know what he is doing.

He puts two opamps in series for current limitation, that is why comparatively large capacitors are needed for C4 and C5, it is clear to me that the designer does not know what he is doing.

Can you vary the output voltage with this schematic yes!
Can you set the current with this schematic yes!

But because of the enormous inertia caused by C4 and C5, the circuit is very slow to maintain the output at varied loads.
And the circuit is also too slow to protect your load and the MJ2955 transistor.


A different way
Several power supplies have been designed on the Dutch forum "Circuitsonline" and I give here a few links to one of them.

This power supply is designed with normal components that are readily available.
Use a translation machine if you don't know the Dutch language.

I have done many tests on this power supply and it is very good for hobby and professional use if built properly.
I have also made some improvements in a separate topic on that forum, but these are basically not necessary.
The power supply is simple in construction (Harrison Concept) and can easily be built on an experimental board.
The original board is no longer available for this project.

First page page article:
https://www.circuitsonline.net/forum/view/130041/1

Schematic:
http://www.gigawatts.nl/co/projects/CO_Voeding_2016/CO%20PSU%20Guido%202015%20Schema.pdf

Short discryption:
http://www.gigawatts.nl/co/projects/CO_Voeding_2016/Bijsluiter%202016%20co%20psu.pdf

Have fun!

Kind regads,
Bram

[glinjik]

So looking at the first circuit I'm thinking of replacing the mj2955 with П210Б 45V 12A 45W I think 2 in parallel might do the job i have 3 of them

[xavier60]

In the design in the first post, instead of the usual way with diodes, the outputs of the CV and CC opamps are resistively ORed on the assumption that either one being low will prevent the output stage from being driven by the other. This action looks borderline to me, with the given resistor values.

[CaptDon]

O.P., I understand by your flag you are located in the Russian Federation, even so, do you really need to build something from scratch and compromise the design with whatever you can find for parts 'that are close but not exact'. One thing we can find at every radio flea market and surplus test equipment and military surplus markets is power supplies!!! All kinds, low voltage, high voltage, low current, high current, nearly anything the heart desires. Power supplies don't seem to hold value at all. I was digging HP power supplies out of the dumpster at work, cleaning the pots, checking for proper operation and giving them away for free. Surely building from scratch is only for the learning experience?

[glinjik]

Yes I'm located in RF although I'm English there are several reasons for this build first I need it second learning expierience third for pure pleasure the parts are increadibly expensive and difficult to obtain at the moment