Proposed Community Project; Lab Power Supply from Notebook Power Adapter

Projects
#1

I have attached a zip file for my latest design, done in 6.0. The idea is to take power from a leftover notebook power adapter and produce a lab power supply with variable output voltage and some degree of current limiting. I have components in hand and boards en route from JLCPCB. I did not ask them to build it because many of the parts have been laying around in my lab, and having them build it means the design must be done around what they have.

Beyond my own boards, I was pondering what else to do with it. The design is heavily biased around skill and components that I have, and the fact that JLCPCB has a particularly good price for a board that is 100 x 100 mm or smaller.

Power is my area; but the digital realm not so much. The control is based on a dual digipot (coarse and fine voltage) controlled by a couple of inexpensive mechanical shaft encoders. I would have used a DAC if I thought I could figure out how to control it. And a digital readout might be nice but this has only the duty cycle of blinking LEDs as a relative indication of where the digipots are in their range.

EDIT: I have cheap commercial lab supplies on my lab bench. My complaint about them is due to the cheap mechanical potentiometers. The output voltage on those cheap commercial units tends to wander in a noisy way due to the potentiometers. That was one reason to try the Digipots.

The power converter is a single or doubled-output SEPIC converter targeted to operate at 300 KHz. I targeted output up to 30W. Depending upon configuration, the output voltage range should be from down to a few volts and up to about 60V.

So there is a lot which could be added or changed if the interest + skill is there. I will be happy to contribute to redesign.

Is there any interest? I am wondering: Does anyone want to contribute design expertise/ideas and get the board design in return (presumably to build a few)? If not, I will climb back into my hole. Wait…I guess I have not stayed put there very well anyway. :wink:
Digipot SEPIC 12-11-2021b.zip (519.8 KB)

3 Likes
#2

Wouldn’t this be more of something for the absolutely wonderful Github/Gitlab scene where you spend half a day logging in/out and checking files in/out before being able to work?
[/sarcasm]

But seriously, I have reservations about this kind of thing on this forum.
I’m sure you have built a wonderful power supply, despite the lousy libraries on KiCAD.
But is this really a thing for this forum?

I thought this place is about KiCAD issues.

#3

I do not understand either github or gitlab. I do not “git” git. But I thought that this might be a fit for either the projects or community category. What do you think of the descriptions of those categories? “Issues” has a connotation of “negative” or problems (but it is not necessarily.) My thought is that collaboration on a KiCad project for a lab instrument might be of general (or at least lieutenant) interest.

I generally use my own symbols and footprints. But any collaborator in this project is welcome to tweak it to their preferences.

#4

Write it up for Hackaday.io, Bob, and post a link here. It’s easier for people to judge whether it interests them by looking at a writeup and viewing photos, even just 3D renders, compared to a Github/lab repo, which in turn is easier to manage and update than a series of zip files.

For example without unpacking the zip file, I have no idea whether you use a linear or switching regulator, what sort of power terminals you use (barrel? banana?). What is the output current range? Does it have an output meter? Is a heat sink required? And lots of other details. I might be interested as I have lots of 19V power bricks rescued from e-waste, but for breadboard projects I use something like this, available for a couple of bucks off the shelf:

breadboard-power-supply
breadboard-power-supply800Ă—800 59.5 KB

#5

hackaday.io is a great idea. eevblog.com has an open source hardware section as well.
But your project is misplaced here IMHO.

#6

I think that if you are not interested enough to unpack the zip file, then you are not interested. I do not mean that against you. But the common element is KiCad and I am looking for collaboration. As I indicated, the design right now suits me but as it is, is probably not in its best form for general interest. I have not built it yet so have no photos. I have no interest in 3D renderings and have never used that part of KiCad. Your post also focuses on details (such as input/output connectors) and of course all of that can be changed as a result of preferences and collaboration. You can see SEPIC in the filenames; that is a type of switching converter which can take the voltage up or down. In fact, it can do a doubled SEPIC which can do a good job of taking the output higher than a normal SEPIC. If desired, we could make a 3x or 4x SEPIC. This is an area of my particular knowledge. But if there is not enough interest in any of that, that is OK. We can let it fizzle.

#7

With all due respect to you, Bob, marketing is not your forte. It’s not mine either, but I recognise that one needs to present more than a zip file to get people to stop and look.

#8

Wow am I glad that is true…

#9

I think the point is a wider audience. Open source works best when the right people get interested. What may fizzle here may sizzle some where else. But nothing to keep someone else from picking up the ball either. You might want to look about putting an appropriate license on your original though.

#10

I am not sure what sort of license. As I say there are a lot of changes to be made for general use. I do not think there is anything patent-able. I know that some circuit boards are sold as kits online, but I think that is probably not a good fit where 0.5 mm pitch SMT ICs are needed. (I do not like them either but not much choice in many cases.)

I think there are a lot of creative electronics hobbyists on this forum. To me, that sounds like an ideal audience. I decided that I would never make money selling a board such as this which was entirely my design. So making money is not my objective at this point.

I posted the zip file to show that I am offering something from the start. Not really looking for “help.”.

#11

Yes. And they’re here because they help and info on KiCAD.
Not because of your hobby projects.

#12

I think that if you look around on this forum, you will see that I have helped many KiCad users with their KiCad issues. In spite of the fact that I am not a KiCad expert.

What is your meaning beyond that? BTW; before you dismiss my hobby projects too much, have a look at my article posted 10 years ago (not exactly the same but it is related.)

https://web.how2power.com/article/2012/may/more-boost-with-less-stress-the-sepic-multiplied-boost-converter.php

1 Like
#13

At the least looking at hackaday.io shows what has already been done. Here’s one of the more impressive ones I’ve seen.

If nothing else, put it on Github with an appropriate license, and someone might take it further. The license is to reassure people that they can freely reuse the design, not about money at all.

2 Likes
#14

Thanks I think your heart is in the right place. When I started with KiCad, I read about git for version control. I could not understand it, and decided that if using KiCad required me to use git I would not use KiCad.

I guess I do not understand why this is not suitable for the projects category. If there is not enough interest then so be it. Also…I have no interest in porting it to other CAD software. So if anyone wants to work with it, they ought to use KiCad and access this forum occasionally.

BTW my design should be able to produce up to about 60V output from a 12V - 20V notebook adapter input, depending on how it is built.

#15

Git is not as scary as it looks. Most of the time I use the same 4 or 5 commands from the CLI: init, status, log diff, add, commit. GUIs are available. Compare that with creating zip backups with dates, etc. And you can backtrack to a finer granularity. To interact with Github/lab you just need a few more commands: push, pull. But I guess SCMSes are second nature to me, coming (actually returning) to hardware from software.

#16

That sounds very easy. If I am updating a file, whether it be a Word document or spreadsheet, I routinely “save as” a new filename including the date. For KiCad, I learned to do that (with a file manager such as Windows Explorer) with the entire project folder including the files. That takes me 1-2 minutes.

I do not know what you are talking about. I really do not.

#17

Personally it isn’t lack of interest on my part. It’s lack of time.

#18

No harm and no foul! Then this thread is not for you…(at your discretion.) I am semi retired. But if I were fully employed right now, I also would not have the time for this.

#19

I think that the contributor ment that, ahh never mind… :smiley:

I am happy to give it a look… will PM soon.

#20

Sounds good. I will be happy to dive into details (which will not be obvious.) I did not want to get into unsolicited details at the getgo.

EDIT: If there is interest, I can do a detailed circuit write-up. A Word document might come up 2-3 pages or maybe more. There are many details to explain. But anyway what I have initially posted is all my design.