Building a linear bench power supply
Вставка
- Опубліковано 16 лип 2022
- Let's design a small linear bench power supply with an adjustable (or fixed 5V) output voltage and a current limit. Some calculations, choosing the transformer, drawing the schematic and how does it work, some tests.
The next episode:
• DIY bench power supply...
Please support my channel on Patreon:
/ diodegonewild
Instagram:
/ savage_danyk - Наука та технологія
Down to earth as always! This is why you are the only UA-camr I support on patreon. Please never change 💪🏻
oh, come on... there are many others who do a fantastic job in terms of science and knowledge
@@PHOBOS1708 but diode gone wild is the most efficient to watch for me aswell
@@PHOBOS1708
No one explain like him
@@warrax111 I'm gen-z I'm sorry for their behavior. I really love this channel. His way of teaching is so detailed and has a high level of mastery. Helps me in college.
A much clearer explanation of ripple than I got at school.
No 3D printing, microcontroller, ASMR, PCB sponsor, RGB LED, LCD, WiFi or Bluetooth.... NICE!
Sadly it missed the much needed Arduino with code copied off the internet :)
Every time I think I understand something, like linear power supplies (and I've built a bunch over the years!), I watch one of your videos and discover that there was a huge amount of stuff I didn't know. Thanks for making such great content!
He is a good EE tutor mixing theory with practice and real life applications
This videos are amazing! One of the best UA-camrs nowadays. Keep doing this amazing work!
Brilliant, as always. Even the rant about power plants.
13:12 Random emotional opening about european shitty energy politics in the middle of a lesson on linear powersupplies. I loved it:Dddd Your videos are some of the best on youtube. Keep going👍
You know what makes you stand out from everyone else on YT?...
You make your own tools!
I just love your digressions, would it be on RGB LEDs, your view counts or nuclear power plants... putting the world to rights in front of a bunch of transformers is near perfection
25:00 you can use another transistor of the same type as a diode so that the voltage drops are most closely matched.
This would also mean that they have more less the same temperature drift.
You can then add a trimmer potentiometer to the base of the sensing transistor to make the current limit adjustable.
I've upped my monthly payment to your channel because you speak the truth to our energy needs
Big thanks!!! ;)
I've built a lot of linear power supplies over the years and still managed to learn several really useful things! Thank you very much.
Thank you I really appreciate how are you teaching us. you are one of the best
I really like this video. It explains so good how a linear power supply works. I found my 15 year younger me in it as I also build several power supplies from scratch as I was a teenager.
The voltage drop and regulation in small transformers was informative! The current limit scheme in this supply seems to depend on the transformer's secondary drooping under load. In other supplies the current limit usually attempt to change the feedback divider, or trigger the shutdown of the voltage regulator circuit.
I really liked all the alternate ways you showed of building the power supply.
You are the best Danyk keep it going!
Liked the enlightened social commentary.
I liked your comments about energy. But you have to admit that we could improve our existing use by simply reducing wastage and increasing decentralisation. For example, add simple and efficient generators to our cars and houses. Even better, use appropriate "new" technology for heating, cooking and water collection/purification. Or reduce waste by just stop buying junk.
Our economy is based on programed obsolescence otherwise the factory will put a big padlock on the gates in very short time ,anyone wants to maximize profit and that's a big problem .
Great video showing tradeoffs always needed when designing electronic circuits.
Btw, i think the rule of thumb for smoothing capacitors is 1000uF/Amp.
Thanks :). The 1000uF/Amp rule is a bit too oversimplified. It aims at a particular voltage ripple (about 10V peak to peak theoretical at 50Hz, both halfcycles), but the lower the output voltage, the more significant the ripple gets.
You are right dear cat 20:43 🤣 Thank you Dany for all your great videos and explanations!!! Keep it up, many are learning from you, including me!!
Regulators also require a minimum load, which can vary from piece to piece, on adj versions, divider should draw enough current.
For example LM317 MAY need more current than LM117.
Resistor between output and feedback on 317 is recommended to be 120ohm, yet, most schematics found online still use 240, it will work 90+% of the time, but to unknowing, it may cause confusion.
Here is the text of DGW's remarkable statement on the state of Europe's energy supply, starting at 13:08
"...electricity is probably going to get bloody expensive now. Of course when your coal power plants are converted to natural gas ones, despite having no natural gas in the entire european union, relying on supply from a country with a dodgy leader, and phasing out most of your nuclear power plants without having any reasonable replacement or backup plan.
And of course energy generation can have some health effects, some emissions, some rare accidents, just like vehicles and transportation. It might slightly reduce your life expectancy, but can you imagine how much your life expectancy reduces without all this technology and transportation and power generation, all the logistics, all the comfort and safety and security it creates.
It's easy to criticize the technology for being a bit dirty when you can't imagine the hardship of life without it. Well, just get rid of all the dirty or radioactive technology, we can just go back and live like hunters gatherers in cave or forest. I'm wondering how many modern people would actually survive in this environment with all their medical conditions, allergies, obesity, diabetes, myopia, and so on, and complete lack of survival skills. And their horrible genetics after many generations with no survival selection.
Criticizing non sustainable technology when humans are not sustainable in the first place. Keep technology running before [until] you have a better replacement, and I mean a realistic replacement, that works in real life on a large scale and is economically viable, not just some dream or some fake kickstarter just to run away with investors money.
And of course the very reason seven billion people can survive on this planet is all the technology, engines, energy generation. Before the industrial revolution for a very long time there was a steady one billion people population, because the excess would just always die of hunger."
Great philosophical arguments about the electricity and the technology.
Really appreciate your work. You are humble and a really knowledgeable man. Respect.
Another great video, and a great dissertation on our energy situation right now.
Great stuff, I can't believe how much you're teaching us about even a simple linear power supply! I had no idea transformers could be so wasteful under no load.
Nuclear energy is among the cleanest energy available, but people are too afraid it might explode like chernobyl.
Personally i believe that chernobyl is no longer relevant when discussing nuclear power safety because technology has advanced so much.
Yes, technology has advanced a lot (unlike people).
the Democrats trash everything they touch
Chernobyl was never relevant in a western context because such hazardous uncontained power plant designs were never permitted in the west. It isn't like we just know better now, we all knew better then too. the Soviets just though they could cheat and operate a dangerous reactor with more caution, which doesn't work once you introduce humans, nevermind their system of management.
Chernobyl wasn't even relevant back in the 80s. It was a very old design military reactor that was stable enough for most of its life until someone had the bright idea to turn off all the safeties and do an "experiment" that had no real purpose and no safe outcome.
Meanwhile coal power plants generate more radioactivity than chernobyl ever did simply because there's always naturally-occurring radioisotopes in coal. Very cool, thanks watermelons.
Great transformator collection.
Thank You So much 😭
I always wanted this video
I found your work fantastic, understanding is always the best base for a better world - the series of old phase "clocks" and how you explained the diff was awesome and now a very good point, without tech we will reduce our life, not mention not everyone can afford to not have transportation, or light, or heat because their heath can't handle. Keep up with great work
Excelente, basic but very useful to remember concepts from the beginning, keep going 👍👍👍
Thank You so much
I always wanted this video
Awesome big thanks to
DANYk ❤️
Beautiful as always. Funny as hell too 🤓 But it needs more 😺
Awesome work !...cheers.
great portion of information, thanks
Thank you! Looking forward to the next part!
your cat is very wise.
thanks for forgoing view and providing excellent information
Really appreciated 🙏🙏
Love u from India. 🤓
Stay safe and be happy sir.
DGW, this two part video series is some of your best work so far. Extremely good explanations. Your drawings go perfectly with your explanations. Please keep more videos like this coming, especially those having to do with RF. Do you own a NANOVNA? I would love to see you cover its use and build some filters, measure some transmission lines, test some antennas!
Liked the video very much. Looking forward to seeing current regulation with an Op amp.
PCB transformer, nice to meet you ! Thanks!
Very nice and informative.
Some of things you said in this video will be useful for the linear laser diode psu I am building rn:D
Super-excellent information = thank you!!!
Best content on youtube hands down, also did you review the adjustable usb tester you used in this video??
I didn’t know you were a philosopher too
Love the discussion on technology
Thanks so much love your theory. Awesome and entertaining as always.
Your dog gets no love. Lol 😂 I miss you entertaining cat.
Just love your videos. Thank you.
I built a battery eliminator once in school. I couldn't add the voltage regulator at the time. Now I want to add it so much, and copy your MOSFET current limit circuit! ☺️
I guess, I'll buy the whole kit and build it again!
was like 1 diode ( DR25 was popular or BY127) before there were in4007 + capacitor
@@mohinderkaur6671 BY127 was that fat green diode with rusty iron terminals?
I used it in one of the projects!
Very thorough explanation. 19:10 the limitation of secondary current of a transformer is usually done by means of loose magnetic coupling. Much like the transformer of an old electric welder or a neon transformer.
A tiny mains transformer limits the current mostly resistively. It's different from the big ones.
Thank you.
Please include how to choose a rectifier output capacitor capacitance.
Just recently bought a 60V 5A SMP. I got it because if my charger fir my ebike ever blows I can charge it in an emergency. I happy with it coz recently too iv been asked to fix a few e-scooters. So it comes in handy for that too. €65 on Banggood. Good buy I think.
One pitfall you didn't mention is the fact that the transformer current is significantly higher than the load current because the diodes only conduct for a small part of the cycle. That makes the resistive voltage drop of the windings a lot higher than one might imagine.
WTF super video!!! 😍😍😍
You're one of the greatest teachers, my friend! Thanks a lot.
P.S I think that the european obsession with the russian gas supply must end, for God's sake. Enough! Diversify energy supplies!
Yeah, change it to Russian uranium. ;)
@@atdzsny of course, hungarian putin lover. 1956. we can repeat it. easily.
You should always consider the rectifier diode current after inserting a smoothing capacitor. Maybe introduce an inductor?
I like using a pi network. Filter cap - inductor - filter cap, it does help reduce ripple a little better than just the sum of the capacitors.
Most rectifier diodes are designed and built to cope with the much higher capacitor charge current. Look at the details in their data sheet.
Nice.
I think you can get a sharp current cutoff with a TLV431 (1,25 V) based circuit.
There also was NCP-something IC with a 0,6 V reference, but not available anymore it seems.
Good job man, lots is learned from this
Could you please later on make addons like OCP OVP UVP OTP RPP OPP OLP NLO
Thanks
You must add RGB lighting on that PSU, it is a necessity these days haha...
If some people would enjoy a ASMR edition of this vid, why not? Drop it 😉
I like this kind of videos too, like "watch repair channel"...
And with the click-money, you could buy more travel adapters 😀
14:16 I love your digession hahaha
Wasn't there also something that if you make the smoothing cap too big, it's bad as well? Because the ripple is lower with a larger cap, the same amount of energy (the same, because it's defined by the connected load) has to be pumped in in a shorter time, stressing the transformer, the diodes and the capacitor. Is there a "sweet spot" for the ripple?
Yes, the diodes will see a higher peak current with a larger capacitor, and the capacitor will also see a higher peak charging current. The average power will however still be the same (if we ignore the slightly higher forward voltage diode and all the resistive losses). So you should still check that the diodes are rated for the peak current, and that the capacitor is also suitably rated (it’s ripple current).
In practice, with a small transformer, it’s winding resistance may limit the current enough. For larger, higher current systems, a suitable series resistor may be needed, or a choke.
The lowest ripple voltage only has to exceed the combination of the output voltage plus the drop out voltage, the voltage regulator eliminates nearly all of the ripple on it’s output. So as long as the voltage does not drop too far, you don’t need a very large smoothing capacitor.
Compare that to a PSU that does not use a voltage regulator, just diodes, capacitors and resistors. Now you need much larger values for the capacitors to keep the amount of ripple low, even at modest output currents.
Dodge leader... This made my day. Very educational, good work!
that was the only thing wrong in this video....everything else....amazing information ..
you are so correct, in the energy thing, have something to replace what you are getting rid of. but governments have no common sense.
I stopped watching the video when I found out the power supply wasn’t going to be voice activated by Alexa and conneced to my Apple Watch and an OLED display on my Christmas tree.
As I understand the tricky thing about using a power MOSFET as a linear device is finding one still in production with a DC SOA line/rating.
Well, you're right, MOSFET datasheets mostly don't show a DC area in their SOA, just pulses up to about 10ms. But for something that's never going to dissipate more than 2W and drop more than 12V, I didn't even bother to look into the datasheet and I tried IRF630 and IRF540 and it worked :).
... and a lot of MOSFETs actually have a DC SOA rating. At least one quarter. Mostly Infineon, IR, Fairchild, NXP, and the ones with the largest DC SOA area are from Alpha & Omega semiconductor (and yet cheap). Not sure if they are better, or just more thoroughly tested, or this maker has the balls to give them this specification :). Also Onsemi ones have quite a large DC SOA. (like FQP32N20C)
The calculation (mathematical model) for pulsating DC after the bridge without smoothing will give Vpk / SQRT 2. However, if you use a true RMS DMM on AC volts you will get a lower value than RMS when measuring the output of the bridge because the waveform is not AC but pulsating DC. So, a better measurement reference when using a DMM would be on DC volts, where you would be measuring the average voltage and this should match the math derivation for average DC after the bridge with a very light load. In looking at output voltage it is best done with the capacitor in circuit under load, obviously on DC volts, but here again you are really measuring the average DC, because ripple is just another type of pulsating DC.
If you are lucky enough to have a scope that can do math controlled by cursors you can select the AC pulse (1/2 the sinewave) at zero cross, set the math to volts rms, and the scope calculated value should be very close to the AC rms.
For current limit as graduated in terms of power, the one technique made popular by the LM723 regulator IC is called current foldback.
Where's the 723 when you need one? Ha. That was a classic chip, and the metal case version (e.g. MAA723) is a thing of beauty and a joy for ever.
Got a 2x20V 3A transformer I'll use for a dual-tracking linear PSU on a 723 or 317+337 with an additional transistor (preferably TO3 e.g. KD502) in series. I could go with switching mode for more output power, but I'd really love to have a real deal regulated and current-limited low noise unit for testing analog circuits.
BTW. Got a few of them old TO3 78xx regulators too! I just love this package too much.
Loving the big-ass fuse holder too. And these potentiometers. Wait, do they come in tube-like boxes? Nice. It's gonna be an awesome build.
Yes, MAA723 power supplies are an absolute classic :) One day I might build one in this channel too. Yes, these Tesla potentiometers came in boxes that look like tube boxes :).
A old chip LM10 is very useful and has everything you need.
An OpAmp and a voltage reference in one package! :) Quite handy... but it also seems to be hard to find or bloody expensive (for a tiny chip).
@@DiodeGoneWild The National Semiconductor datasheet of LM10 has all types of power supply schematcis. Its not that expensive on ebay! But ofcourse you can do it quite easily with reclaimed transistors etc. a reverse biased silicon transistor base emitter junction is around 6 v zener. works good as reference. PNP pass transistor + differential amplifier makes a good regulator. Elektor did it many times in its old 80s issues
I love this power supply. But for me I need more powerfull one. few questions. Can I use parallel transistors/mosfets in current limiting section. Is an current limiting transistor works in liniar reagion or as a switch? Can I crate voltage control using transistors and make it in parallel to achieve more power. I understand that I need much powerfull transformer but I think that almost 0.5kW transformer from old audio amplifier could be enough to achieve 12V @ 10A. Sorry If I do not understand all from Your video, I tried my best.
The electronic specialist from Transilvania
What about using a toroidal transformer for these small loads and no load voltage behaviour? What power level do toroidal transformers become efficient? These are things that are not taught in school! Thanks
LM350 can handle twice the current of the LM317 and it's adjustable.
I was curious if we can convert a flyback smps supply from 0-12 volt.. The minimum i got was 3.3 volt but i need 0-12 volt... Is it possible i can give external supply to main switching control ic ... Will tl431 and other circuit work with it ??? Share your views.... It would highly appreciated... Thanks.
@@kennmossman8701 If somebody already tried it that it will give me a lot of information and knowledge that he gain after many hours of trial...
do you have any guidelines on the current rating of the transformer? Say you want a total load current of 1A, should your transformer secondary be rated for 1A, 1.5A? 2A? i appreciate the impedance of the transformer will drop the voltage of the output, but surely the copper itself is limited to a current before it starts to heatup and damage any coating (extreme case)
1.5A at least.
@@DiodeGoneWild so a general rule of 1.5 times the load current?
@@Taylor_26GE93 yaaaaa it depends on the rectifier/transformer arrangement. sadly it is not a topic that is well discussed anywhere except in in-depth power supply manuals. I believe it is square root of 2 times load current; 1.4 times load current
I have this problem, everyrime I build something, I fall into the rabbit hole of overthinking how many things I could add during the design phase I get stuck in overthinking and I delay making anything till I get all the research done. The transformer for example, it's a nightmare for me, always hot and I fear using it because of thermal dissipation. I implemented a 6V 1VA transformer to step down the mains voltage, then do envelope detection after a precision rectifier to feed into a microcontroller. Anyhow there is little or no load on this transformer and it get really hot to the touch. I wonder how these never burn up in enclosures.
I miss ur cat :D ♡
nice video do you make something video about tesla coil or psu?
He did some time back on a sstc or a solid state tesla coil
@@309electronics5 ok but maybe he do something new
I mostly use usb powersuply for 5v. Sometimes,if i need Higher voltages, i also use an qc3 compatible one controlled by an Arduino. It can go as low as 2.3v and up to 20v. Also its easy and cheap to use. (Also i use an small Display, current and voltage Sensor and also an rotating encoder for easy control)
13:12 Regarding your comments about electricity getting _bloody expensive_ now, I believe that will be for a shortish time. Once the country with the dodgy leader is removed from Ukraine, they, Ukraine, has enough natural gas under it's territory to supply *all* of the EU and all other remaining European countries. When Ukraine eventually becomes a full EU member the gas supplies will be guaranteed until there is sufficient renewable power sources available.
In the meantime I shall keep enjoying your videos and I wish I was in a financial position where I could contribute to your channel. *Thank You* for the videos. They are greatly appreciated. 👍
I like your optimism! I hope you will be proven right! Because some other, unfortunately possible options, like devastation of modern civilization and the death of billions, is not quite pleasing...
The only thing removed from Ukraine will be the criminal us puppet government and the nadzees with them.
You should put the current limiter after the regulator otherwise the current limit won’t be consistent with different output voltages, you can get the voltage feedback from before the current limiter to help ensure stability.
Putting the current limit after the voltage regulator would make the voltage very inconsistent :). I believe it's better to put it before LM317. The voltage divider of it passes a constant current, it doesn't really skew the limit much. And of course, in most applications the voltage accuracy is much more important than the current limit accuracy. But for the best accuracy of both, it would be best to use just one power transistor and control it such a way that it both regulates the voltage and limits the current.
@@DiodeGoneWild the voltage will collapse when the current limit is reached regardless of placement, putting it after the regulator will give more precise control of current for the circuit.
@@DiodeGoneWild Will You make a video about that? The single power transistor controlling both voltage and current?
dodgy leader xDDD
12:48 This is why I'm here.
Can I replace the mosfet or transistor with an IGBT transistor?
IGBT has a high C-E voltage drop. It only makes sense in circuits with hundreds of volts. For low voltage circuits, MOSFETs are much better.
12:50, You are really making education videos, 100% sure of that
Great educational value as all your videos! Maybe you should tell us about your use of the linear PSU and why you choose to build it as linear PSU. Even small switch mode supplies can be cured resp. to ripple with some coils and capicitors or some regulation via transistors or mosfets unless you need something with no ripple and harmonics at all. So it would be good to know when to choose what kind of PSU for what purpose :-)
Thank for for an excellent video - very educational. I lnow you are being careful but I hope novices don't try to copy the way you handle live wires when testing the transformers.
12:36 #ASMR! Omg
That.may save us
But why the current limiter is set before regulator and not after it ?
The transformers are short circuit proof due to high resistance! why do you need additional current limiting circuit?
Maybe because he wants some over current protection for the circuit fed by this PSU?
Co kdybyste nám řekli, jak vyrobit desetkrát citlivější "wattmetr"?
ua-cam.com/video/SVP3on0oHHs/v-deo.html
20:39 Where is the Kitty ?
Mreooww meowww !
199K! Check that with the AVO or Fluke. Just in case..
I love the videos that he is making. Nowadays he is always ranting somebody or something. I love it. But he has the point, everything that somebody is building has to be connected to wifi or has to have rgb or things like that but those things makes everything more complicated. How would beginners learn about the basics when they just see really complicated 32-bit microcontroller projects. At first you need to build some shitty stuff because that's the way you learn. If you want to build a circuit that can turn water into wine, you will not succeed at the first time.
this is more complicated than a ready circuit
6:14 ... K, bye 🤣🤣🤣🤣🤣🤣
takovy veci nerikej 😁
Transformer first thing after fuse in old tv's
Mr Diode, about four days ago I made a comment on here thanking you for for some information you published, (not on youtube). I also went on to provide details of what I used the information for. (switching psu). My comment stayed for a few hours and then was deleted. I am sure that it would not be you that deleted it, rather it would be you-tube filters somehow not liking it. Is it possible for you to have a look what happened? I believe that interaction and exchange of ideas are very important and I feel that you-tube is growing to suppress this exchange. The comment was entirely positive and would not have been deleted by any human viewer. Chris.
It was probably deleted by youtube. This generaly tends to happen to comments with links in them. I'm looking into comments "Held for review ", but there's nothing.
@@DiodeGoneWild Thanks for looking. My comment was to the effect that I found your extensive list of ATX power supply circuit diagrams listed on Danyk very useful. They helped me to design a PSU of my own for a special purpose 1000v DC input to 12v 35A output. Thanks for answering my comment and even bigger thanks for listing all the info on Danyk. Chris UK.
You're solid as fuck. I have to ask, what do you do for a living?
Some dark mood, eh? People have no idea how hard it is to make a "simple" vacuum tube in a post apocalyptic world (imagine a voltage regulator), most think that things are just as easy as spewing out some stupid ideas like some billionaires do these days and to hell with reality... On a side note, I surely am in that list that would be dead a long time ago without technology. Thanks for the video, awesome!
Dark or not, this is reality. Being aware of the reality isn't a mood, it's a permanent state for me.
@@DiodeGoneWild Totally agree. It's funny how people will call you radical when you are not delusional.
The lm78xx series is pretty linear( low noise) even at a 2A power draw, compared to a professional parts designed to have low noise, it can be used if you are entering the high precision low noise but wouldn't recommend to use it as a primary part of the circuit.
A good comparison ua-cam.com/video/XpbDMo8an5w/v-deo.html
This is a DIY P$U for hobbyists
@@igrewold I know, I just brought up one of its non intentional features
Hello sir, i am one of your subscribers. i have watching your videos. i am new to inverters. i trying to make inverter high frequency inverter with ferrite transformer and the Mosfet keeps heating up under no load condition. i want to know to tweak to solve this challenge. i could really use your help sir.