Author Topic: Help with diy bench psu circuit. (Read 12104 times)

gabri.peic · « **on:** January 27, 2022, 11:40:12 am »

So i got my hands on a pretty beafy toroidal transformer for free. Size is 11cm x 4.5cm and i think the power is around 200va since it says 200 on the sticker, it has 2 identical secondary windings and each one is 26v ac unloaded.
Anyways lets get to the point.
I want to build myself dual channel power supply with it so i can tets amplifiers and things that need two power rails.
So ideally i need something that can take around 36v dc input and provide adjustable output preferably with variable current limit.
My plan is to have two adjustable 30v outputs that i can put in series for dual rail or more voltage. Or in parallel for more current.
If anyone knows of any good way to do this pleas let me know.

Note: I know this will require huge heatsink with forced air cooling.

m3vuv · « **Reply #1 on:** January 28, 2022, 04:20:39 pm »

something with 2x lm723 chips maybe,check out the datasheets!

amosborne · « **Reply #2 on:** January 29, 2022, 05:57:29 am »

This is an awesome but deceptively advanced project depending where you take it.

Is this linear or switching?
Step up or step down?
How will you load share? Passive droop or master/slave?
How tight and fast is the regulation?
Constant current mode?
User controllable OVP/OCP?
How much current are we talking?

I would probably look around at commercial supplies and find a set of specs you like. Then try and hit those specs with your design. Saying you want 36V or whatever is the tip of the iceberg.

I’m jealous of your transformer.

ledtester · « **Reply #3 on:** January 29, 2022, 06:56:08 am »

There's been a lot of interest in using transformers with the Riden RD60xx power supplies, e.g.:

https://www.eevblog.com/forum/testgear/ruideng-riden-rd6006-dc-power-supply/msg3881924/#msg3881924

Look in that thread for an alternative firmware provided by user Unisoft which adds a lot of features and enhancements, e.g.:

https://www.eevblog.com/forum/testgear/ruideng-riden-rd6006-dc-power-supply/msg3941395/#msg3941395

gabri.peic · « **Reply #4 on:** January 29, 2022, 10:45:26 am »

Basically i would like to build something based on a linear regulation with constant current/constant voltage mode, range would preferably be 0-30v and 0-4A,
I would like to have two isolated channels since i have two secondary windings. And a switch that switches modes between independent tracking, parallel tracking or series tracking.
Independent tracking would give me two independent power supply outputs, Parallel tracking would allow me to hook them in parallel and get more current, And series tracking would be used to either get more voltage or center tapped supply.
I was planning on something based on this video:

https://youtu.be/_CFIovMkRyg

Obviously I would have to adjust it for my current and i would have to build two but would this be able to operate in series/parallel configuration.
Thanks.

ledtester · « **Reply #5 on:** January 29, 2022, 05:08:07 pm »

gabri.peic · « **Reply #6 on:** January 30, 2022, 11:50:37 am »

Thanks for all the tips.
But i still have some questions.
1 can i put the two separate channels in series
or parallel if i keep them floating?

2 is the circuit from the video good or not ?

3 is there some design already out there that would fit my needs ?

Also i tested the transformer with a dummy load.
I loaded it up to around 4 amps per coil and voltage dropped from 26v no load to 24v fully loaded. It was pulling around 196w from the socket and didnt get hot at all. Just slightly warm. So i feel pretty comfortable pulling 4A from it. But will 24v ( around 33 Vdc after rectification ) be enough to get 30v regulated output or will i have to drop my requirements a few volts down ?

tunk · « **Reply #7 on:** January 30, 2022, 01:01:17 pm »

I think a rule of thumb is to use ~2000uF/A smoothing capacitor(s)
and that this will give you ~4V ripple, in your case 29-33V.

gabri.peic · « **Reply #8 on:** January 30, 2022, 06:45:01 pm »

Hmmm what about building two of these.

http://electronics-diy.com/30v-4a-adjustable-bench-power-supply.php

ledtester · « **Reply #9 on:** January 30, 2022, 07:02:26 pm »

Quote from: gabri.peic on January 30, 2022, 06:45:01 pm

Hmmm what about building two of these.

http://electronics-diy.com/30v-4a-adjustable-bench-power-supply.php

That schematic looks like it is based on the same design as the "30V-3A" supply I mentioned above:

Quote from: ledtester on January 29, 2022, 05:08:07 pm

There's also the infamous 30V/3A cheap power supply kit you can get from ebay/aliexpress and the like for which you'll want to read these threads and blog posts:

- https://www.eevblog.com/forum/beginners/diy-0-30v-0-3a-again/

- https://www.paulvdiyblogs.net/2015/05/tuning-030v-dc-with-03a-psu-diy-kit.html

There are a lot of caveats with that design which you will want to be informed about so I'd read that thread and blog post.

gabri.peic · « **Reply #10 on:** January 30, 2022, 07:34:18 pm »

Quote from: ledtester on January 30, 2022, 07:02:26 pm

Quote from: gabri.peic on January 30, 2022, 06:45:01 pm
Hmmm what about building two of these.

http://electronics-diy.com/30v-4a-adjustable-bench-power-supply.php

That schematic looks like it is based on the same design as the "30V-3A" supply I mentioned above:

Quote from: ledtester on January 29, 2022, 05:08:07 pm
There's also the infamous 30V/3A cheap power supply kit you can get from ebay/aliexpress and the like for which you'll want to read these threads and blog posts:

- https://www.eevblog.com/forum/beginners/diy-0-30v-0-3a-again/

- https://www.paulvdiyblogs.net/2015/05/tuning-030v-dc-with-03a-psu-diy-kit.html

There are a lot of caveats with that design which you will want to be informed about so I'd read that thread and blog post.

I looked at the links you posted. I still didn't read all of them but i did take a look at them.
Some of the designs seem really interesting and are definitely informative but they don't fit my needs.
The 3A 30V one was the one i was interested in but then i found this 4A version.
I will do more reading as soon as i can.

JohanH · « **Reply #11 on:** January 30, 2022, 09:24:35 pm »

Quote from: gabri.peic on January 30, 2022, 07:34:18 pm

The 3A 30V one was the one i was interested in but then i found this 4A version.
I will do more reading as soon as i can.

3A or 4A doesn't matter. It's the same design. And even more, it is claiming 4A with a single 2N3055 transistor. I wouldn't do that. Two or three will be fine. But you still need good cooling. This design originates from an article in Practical Electronics, 10-1978, page 1070: https://worldradiohistory.com/UK/Practical-Electronics/70s/Practical-Electronics-1978-10.pdf Not sure if it can be recommended today, other than for historical reasons. There are simpler and easier to build circuits today.

gabri.peic · « **Reply #12 on:** January 30, 2022, 09:54:38 pm »

OK i did some more reading. As far as i can see the 0-30v,0-3A circuit is not the best but it should work. I am thinking of building version modified by Audioguru. I am probably going to add some things like fan speed control.
If there is something better pleas let me know.
Also any tips on pcb layout would be great.
I am also thinking of adding relays to the outputs so i can connect or disconnect the load with a switch so i can avoid possible spikes when powering off the power supply.

gabri.peic · « **Reply #13 on:** January 30, 2022, 09:56:04 pm »

Quote from: jukk on January 30, 2022, 09:24:35 pm

Quote from: gabri.peic on January 30, 2022, 07:34:18 pm

The 3A 30V one was the one i was interested in but then i found this 4A version.
I will do more reading as soon as i can.

3A or 4A doesn't matter. It's the same design. And even more, it is claiming 4A with a single 2N3055 transistor. I wouldn't do that. Two or three will be fine. But you still need good cooling. This design originates from an article in Practical Electronics, 10-1978, page 1070: https://worldradiohistory.com/UK/Practical-Electronics/70s/Practical-Electronics-1978-10.pdf Not sure if it can be recommended today, other than for historical reasons. There are simpler and easier to build circuits today.

Can you recommend some better circuit?

JohanH · « **Reply #14 on:** January 31, 2022, 07:00:03 am »

Quote from: gabri.peic on January 30, 2022, 09:56:04 pm

Can you recommend some better circuit?

The one I could recommend for less experienced designers is something based on the LM723 regulator (as also mentioned by others). It is an old circuit, but it is cheap and there are things to learn about it. If you search in this forum, you can find several designs. As usual, there will be lots of opinions, one way or the other. The most important thing is to read component datasheets and look at reference designs. A quick googling turns up good information like this: https://electronicprojectsforfun.wordpress.com/power-supplies/a-collection-of-proper-design-practices-using-the-lm723-ic-regulator/ If you read this, you can see there are a lot of drawbacks with this regulator, too.

gabri.peic · « **Reply #15 on:** January 31, 2022, 01:12:55 pm »

Quote from: jukk on January 31, 2022, 07:00:03 am

Quote from: gabri.peic on January 30, 2022, 09:56:04 pm
Can you recommend some better circuit?

The one I could recommend for less experienced designers is something based on the LM723 regulator (as also mentioned by others). It is an old circuit, but it is cheap and there are things to learn about it. If you search in this forum, you can find several designs. As usual, there will be lots of opinions, one way or the other. The most important thing is to read component datasheets and look at reference designs. A quick googling turns up good information like this: https://electronicprojectsforfun.wordpress.com/power-supplies/a-collection-of-proper-design-practices-using-the-lm723-ic-regulator/ If you read this, you can see there are a lot of drawbacks with this regulator, too.

Thanks i will take a look at that too and hopefully learn something.

gabri.peic · « **Reply #16 on:** February 01, 2022, 07:22:38 pm »

Ok i did some research.
I am going to go either with modified kit or
Elektor 1982 design.
What do you think of Elektor 1982 design ?
Only thing is i would need either a separate power transformer or wind some low power windings on my transformer for that 2x12v that is needed for the Elektor design.
What about the pcb? Is there some design already out there?

Neomys Sapiens · « **Reply #17 on:** February 02, 2022, 12:00:17 am »

Quote from: gabri.peic on February 01, 2022, 07:22:38 pm

Ok i did some research.
I am going to go either with modified kit or
Elektor 1982 design.
What do you think of Elektor 1982 design ?
Only thing is i would need either a separate power transformer or wind some low power windings on my transformer for that 2x12v that is needed for the Elektor design.
What about the pcb? Is there some design already out there?

The Elektor 1982 is not a bad circuit. I did build one once and sold it for good money.

You should think right now about some options, like switching ranges for the instruments, sense int/ext, etc.. also, if you get an separate transformer for the auxiliary supply, have another voltage on it for control and instrument lights (if analogue) or separate panelmeter supplies. Do not use those crappy panelmeters that claim to work as well without one!
With a separate transformer, you could use a relay or SSR for the main transformer and switch it only on if the auxiliary voltage(s) is up and good. Also, the circuit lacks thermal safeties, which you could also wire into this circuit.
Place at least 2 thermal switches between the transistors and maybe one with the rectifier and 1..2 on the main transformer. With really simple circuits like a small additional relay you could also have an indicator for overtemperature. It is not among the most efficient ones as it lacks a pre-regulator, but that makes it easy to understand and modify.
Use really good, precision, inductance free resistors in the power circuit (emitter res. too). Be aware that you can easily switch the A-meter range, but not the current adjustment. In that case you would have to use a circuit which does not interrupt the current shunt (ring-shunt switch).
The thread discussing this PS circuit here on the forum sems to have covered some of those options, although in a more electronic implementation. Some other improvements as well.

Not the worst choice.

gabri.peic · « **Reply #18 on:** February 02, 2022, 11:28:21 am »

Ok, now what about the pcb. Is there some design already available or should i design it similar to the original old design.
I saw something mentiomed above in one of the suggested links. It is supposed to have fan controller and thermal protection.
https://www.eevblog.com/forum/beginners/building-the-elektor-1982-power-supply-what-pcb/
So how does that design look ?
Also will there be problems if two of these are connected in series pr parallel. This project is getting expensive and i dont want to blow something up.

coromonadalix · « **Reply #19 on:** February 02, 2022, 11:47:30 am »

Well if things get complicated, and way too expensive just buy two identicals models ... to put them is serial or in parallel

Designing your own power supply can be tricky and dangerous if you put sensitive devices on it ... and can get pretty expensive over the time, you want this option / that option etc ...
You will find my name in a few threads, i did design some psus around the famous lm723, but in the end there was always a missing cv-cc crossover function that i would need

Ended up buying already made reputable brands for the peace of mind, my 2 cents

For a parallel operation, the supply (ies) must have isolation between the input voltage (ac side) and the output, and check if the output ground (negative) is common or not with the case chassis ground, sometimes you need physical isolation between them Or you have a metal shorting bar on the terminals you can remove (or should be removable ??)

You normally need two identicals models to work with in serial or parallel, for a good / equal load share between them

IF NOT you will be limited from the weakest of the two in current capacity

gabri.peic · « **Reply #20 on:** February 02, 2022, 11:52:34 am »

Also i am calculating the prices and i think i am for now going to go with something simpler without constant current. I just cant go with something so expensive and involved for my first power supply project.
But i will come back to this design in the future since it is really interesting.
So i would need something to regulate the voltage. I don't strictly have to go to 0v but it has to be able to be put in series or parallel.
And maybe adding resetable ocp would be a good idea.
Note:i already have small lab supply that has current limit so this big one not having it is not a big deal for now.

coromonadalix · « **Reply #21 on:** February 02, 2022, 11:55:37 am »

check the velleman k7200 psu you have explanations and schematics ... even some pcb files

It is a simple circuit, not very pricey, built around common parts .... should help cut the costs down

Check the manual section for the schematic

gabri.peic · « **Reply #22 on:** February 02, 2022, 12:27:43 pm »

So if i go with just variable voltage supply would something like lm338 (one per channel) work ?

Kleinstein · « **Reply #23 on:** February 02, 2022, 04:34:08 pm »

Quote from: gabri.peic on January 30, 2022, 07:34:18 pm

The 3A 30V one was the one i was interested in but then i found this 4A version.
I will do more reading as soon as i can.

With a 200 VA transformer and 2 x 24 V AC one would get a little over 4 A AC current. Because of the power factor (1/1.6...1/1.

the maximum DC current after a simple rectifier with filter is more like 2 to 2.5 A. So a higher current would start to overload the transformer already. So a circuit meant for 3 A is not bad, when reducing the current a little. More than some 2.2 A would be for 1 channel at a time or short time use.
One may still allow a slightly higher current limit if one has an eye on the transformer temperature.

With some 33 V to start with, a single 2N3055 is good for some 2 A with a good heat sink, maybe a fan. More gets tricky with the SOA and cooling. As the transitors are no longer that expensive I would prefer a 2nd parallel transitors to allow a somewhat smaller heat sink.

gabri.peic · « **Reply #24 on:** February 02, 2022, 05:44:11 pm »

If i go with just voltage regulation without constant current. What would be the best approach for that. I am having trouble finding parts and not to mention prices...
I dont want to spend so much right away...
So just variable voltage regulation would be enough for now.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Help with diy bench psu circuit. (Read 12104 times)

Share me