He must enjoy it, because I as had electrotechnics on highschool and we had plastic template.. makes your life much easier, there was all the marks for resistors, diode, capacitors...
Extremely useful and clear description of a switch mode power supply. I like how even though you didn't get your expected conclusion with this video, you continued it anyway and shared it with the world. Thank you.
Time traveling from 5 years in the future... thanks for this. It's instructive. I liked the primer on switching power supplies. I only had experience with old-school versions. And thanks for sharing your "failures" as well as your successes. I think we learn more from failed projects than successes. I am in the middle of building my ATX bench power supply. If I ever need a variable supply, I know to just buy one! Best wishes and continued success!
This is EXACTLY what I was looking for. Excellent freehand drawings and great detective work looking at all the different IC's and messing with the potentiometers. This is something I'd been wondering about and wanting to do. Now that I see what the limits are, I'm glad you sacrificed your power supplies to show us. This definitely wins my subscription and thumbs up!
You take great pride in your Penmanship. This shows every time you put pen to paper. The same is said with the quality of your projects. Every detail, every component, every finishing touch. This speaks volumes and it's something that so many have lost. Everything we do is like signing your name, this is the lesson I try to teach my son. I am glad to see you trying to show so many others the value of this very important lesson.
The decision to abandon the experiment on the realization it requires disabling all the built in redundant safety features is not only "the right way to do it", but also quite praiseworthy! Great vid
He wants to imitate a chewing gum munching Ami. This complex can be found more often in Germany and it gives me chills. To meet someone who has trained himself to speak clean English, that would be something. But I commend his videos for being informative, accurate and innovative.
Great GreatScott! Thank you! Small notation: please never left unconnected side pin of variable resistor (potentiometer). 04:45 and 05:55 Short it to central pin always! It will keep your load (sometimes very expensive) safe of short voltage spikes out of range.
I use a few of these for lab power supplies. They run from about 5v to 25v. I tend to remove most of the stuff on the secondary side of the board, and just leave the 12v winding and circuit. The ATX uses the spare fet and tiny transformer for sleep mode to maintain 12v to the KA7500 ( TL494 ). If we then lift pins 15 and 16 above ground, we can sense the current via a very low R resistor to ground from center tap, and use that to use the [pin 15/16 op amp to control current, while using the pins 1 and 2 for voltage control. Then we have a current controlled ( up to 12 amps or more) and voltage controlled 9 up to 25v or so) power supply, that will power very low impedance loads ( motors etc). This requires all of the monitoring stuff to be wiped from the board, or at least disconnected from the pins 1,2,3,4,15,16. So we use the 5v reference from the 7500 to go to both of the 7500 op amps, then we can simply use a divider for the voltage to the first op amp, another divider from the CTap of the tranny to ground to the second op amp, some feedback from pin 3 to both, and we have a totally stable power supply, that is completely controlled. It is actually simpler than it sounds really.
Hey Jeremy, The problem is that your playing with a chip that uses 2 comparators. These are way more complicated. However that KA7500 chip is a replica/update of the old TL494/TL594. You can redesign the circuit around that IC to create a voltage and current feedback loop and therefore a variable voltage/current supply, (good luck building that in a 1 week project though ;) I've been looking into this circuit, but it gets super complicated IMO to make it variable over a wide range. The FET duty cycle and core saturation math gets a bit blurry for me. All the examples I've found attest to the potential operating range of the circuit, but they don't really explain the details of how to design specifically for an variable output. Where and when to compromise specs is the challenge. The best I can tell you build for about a 90% duty cycle at your maximum power output, while trying to maximize the timing frequency at that output, but I'm not 100% clear on that one. I've found some interesting circuits to try and will probably do a video at some point. There is a way to use a power PNP BJT with a capacitor on the base to reduce the ripple effects across a wide bandwidth. Bill Herd over on HackaDay's YT channel hinted ages ago he was going to create a variable SMPS that used a linear design at the output. I've half built a project on that idea, but that is about 3-4 down my list. If you see this post and want to contact feel free Jeremy. I'm using YT to try to connect to more of a community. YT also keeps me committed to completing my projects despite being partially disabled and stuck laying around most of my day. My channel is still not monetized and I'm not trying to copy anyone. I'm just here to keep it interesting ;) -Jake.
If you do any work towards this lab psu thing, please let me know. Seems like an interesting project to keep up with. A couple years ago I have done the same modifications he shows in this video, and I consider it a success. I was able to get the voltage range I wanted. But now I'm looking at doing something with higher range, so I would need to do more intrusive modifications than just adding a potentiometer.
I just converted an ATX power supply into a variable bench power supply by connecting a 1k Ohm potenciometer from 12V to ground, with the middle pin being the output voltage. It doesn't give me more than 12V but hey, it works!
I once had a friend coming to me with his old, potentially broken, computer PSU. We decided we'll have fun with it by modifying the thing to make a lab psu. It was a really cheap looking PSU, the components inside are poor too, but we just kinda go with it. We turn that on and it didn't show any living sign, then I saw a switch which I thought was the 110V/220V fuse thing, then I switch that to the other side. Suddenly, a huge capacitor blows up followed by some diode and ultimately the whole circuit was in fire. We panicked and unplug the power, throw the firey thing outside, and a black-blue smoke is coming from what seems to be a bigger capacitor. That day, we learned our lesson to not mess with any shitty old electronics.
I also had a power supply with that switch the previous owner had set it to 110v I plugged it in 230v it blew up. I got another power supply with that switch again but this time I switched to 230v and permanently glued that switch
Hi Scott. A few years ago I have done the same modification as you did, and I consider the result a full success. Although the voltage range is limited, it varies enough to make the bench PSU useful for my purposes. However, I understand why you disencourage your readers to do the same. The risks of dealing with 240vac is immense and you don't want to be responsible for people risking their lives, specially when forgetting to discharge high voltage capacitors is a common error even among professionals. You are right and I agree with your stance, BUT for those who actually have the skills and knowledge and responsibility to safely do this modification, I think it's totally worth it. It's a nice project, easy to do, fun and rewarding to see it working in the end.
The first power supply that you were modifying, that style works great. The high wattage ones with the UC384x chips are usually 2-switch forward converters and regulate very well. When you're modifying the feedback circuits in these, you must not modify the circuit around the UC384x chip. Since the feedback is with an opto-coupler, the voltage division and comparison is done on the secondary side with a TL431 programmable shunt (or sometimes but rarely a zener) and that sends a simple on/off signal through the opto-coupler back to the UC384x chip. You only need to modify the divider going into the TL431. The TL431 is usually a TO92 package and looks like a small transistor.
the mark of puting the "Great" in Scott, admitting when it wont work, then sharing your failure, and then trying again. Again, I like the variety and creativity of this video. Cheers.
Au contraire, very expected conclusion Everywhere, anglosaxon culture people try to exaggerate dangers and to explain why you should NOT do it instead of teaching HOW TO do it correctly and safely The conclusion that it's better to spend USD$53 (+shipping) than RECYCLING a device that can be obtained almost for free is not only ludicrous, it's the perfect example of what should NOT be done
The BEST VIDEO on youtube.. Thank you so much for breaking the Power Supply down into Blocks with the expected Wave Form and Voltage. Please DONT STOP this method of explaining Electronics as You are a TEACHER really.. Thank you for taking the time to help us un-informed dummies. May the Force be with you.........
4:00 - And by measuring this voltage, while the power supply is turned on, I'm waiting for the point where you accidentally shorted out something with the probe and got a huge spark and a faceful of vaporized electrolyte
I tried this same project about a year ago and became fed up. So I just ended up just buying one of the DIY kits off of Aliexpress/Banggood for around $9.00 US. Love your videos GreatScott! been a fan for 2 or 3 years now.
I created a lead-acid battery charger using a PC power supply with the infamous TL494. I didn't remove the divider resistors though. I simply added a parallel variable resistor to the pulldown resistors to get higher voltage on the 12V rail. Replaced the capacitors to match higher voltages on all rails. you can find the mod on the internet though. Good work Scott.
Thanks for sharing, I was wondering when the warning about beeing killed by the power supply might come and was surprised you mention it at the end and not at the beginning of your video ^^
Thank you for trying to mod an ATX power supply for variable voltages. I have wondered how easy it would be to do so and now know thanks to you. Now it is time for me to look for a proper lab power supply. Take care and keep the videos coming.
wow that looked so simple... My approach to large circuits like that is to immediately salvage everything off it and not even consider using it. Well done
Maybe a video walking your subscribers through the different stages of caps, how many Joules of energy it contains, how much can kill a person, and proper PPE/protocols for dealing with these caps?
I'll just add my .02USD here that with videos like this a warning at the start would be a good idea because man did I cringe when you pulled that PSU board out with your fingers that close to those big caps. That said I always look forward to your videos because you make them so even a dummy like me can understand them.
I am so proud that you did not recommend this, because I made myself a DIY ATX powersupply just yesterday!!! I would be mad if this actually worked. one question, why didn't you put fuses in your DIY ATX power supply? wouldn't shorts be dangerous or kill the powersupply?
It has to have protection features because only one output rail is monitored in the feedback loop. The other voltage rails are usually regulated simply through their turns ratio on the same transformer core as the regulated rail. This is why the primary rail carries the highest current rating and is designed to carry most of the load. The other magnetically coupled rails have a much larger output variance or to put it another way, they have poor output regulation. This is especially the case if you draw more current from them than the main rail. The auxillary rails must first influence the main rail before the SMPS controller can detect the change and compensate. Basically an ATX system is designed as simple as possible for creating multiple voltages at high current. It is far easier to modify a SMPS circuit that is designed with a peak current controller instead of a PWM controller. These circuits only contain 1 comparator and therefore 1 feedback voltage reference. I've uploaded a video on modding one of these (none monotized/cc license/no affiliate marketing). Peak mode controllers are found in everything from car type phone chargers to laptop/printer power supplies. The most ubiquitous and ancient chip is the MC34063. It's a good place to start but it is limited to 500-700mA and was designed before power mosfets were standard so it's terrible for high output.
No mention of the dangers of messing with the mains side of power supplies, or about discharging the input caps... Well, maybe we'll get rid of some people :) Glad you came to your senses.
2:49 aaahhh...so that is how they maintain stable voltage with different current draw, Thank you GreatScott, this step by step explanation with flow chart help me a lot to understand about power supply
Buy a cheap chinese LED supply -- the kind with the perforated metal case. The feedback supply is much simpler, and the 48V ones can be adjusted from 0 to ~63V. It's a bit noisy, but an LC filter cleans it up nicely. All you need to do is make sure the r values are low enough to ensure proper bias current.
How can I do that? I would appreciate if you could help me build one! I've disassembled two at home that were still working, but till now, I have no idea on how I could use them.
Victor Ribeiro Find datasheets from all ICs used on board, one of them will be used to provide feedback using resistor voltage divider, you will have to modify that divider to get desired voltage
Yea I just made a custom housing for my ATX power supply and used a buck converter with a voltage/ammeter display. Works great. Gets me 1.33v -11.8v. And of course kept the constant 3.3v, 5v, and 12v banana jacks.
Hi GreatScott, I am your big fan. I like all your video especially the modifications of rasperry pic to game boy, since this is the video makes me start to watching all your videos, keep going to create more video with more interesting things. I have an ideas for you, can you make a AC timer switch with connection to Android phone so it can programme the timer through the smartphone. 😊😊
For my own lab bench power supply I used an industrial switched mode power supply and a buck/boost converter. It cost me around 30€ and I can get up to 32 volts by around 4 amps, which is a pretty good deal I guess :).
I enjoy watching the videos as a guidepost of how much I've learned about electronics. Back in the day I'd have been pretty clueless from the start, but thanks to GreatScott! and others this time I got about halfway through before i had to pause/think and learn. Next time even further.
Hey Scott, I did something like that yesterday, but i ended up rewounding the core to give me 3.3, 5, 12, 18v ground referenced and two 18v floating for a DRSSTC full bridge driver. Had to use few schottky, one more ferrite, few electrolytics, decreased the switching frequency from 65 to 40kHz as my new core was N27 type, increased soft start and tricked the smps that everything is just all right.
There's a great 38v dc-dc variable buck boost converter module with led display available on eBay that is easily added to an old power supply that makes a great variable shop supply with adjustable current limiter.
Yes, or if 10.5 Volt is enough for your lab (arduino, raspberry pi, and what not) , this buck converter is good value for money: www.ebay.com/itm/Constant-Voltage-Current-Power-Supply-Converter-DP30V5A-L-LED-Module-32V-5A-160W-/172661458452 Just hook it up to the 12V of the ATX power supply.
I just subscribed to your channel. Don't know why I have not done so earlier. Your explanations are some of the best on RUclips. Well done on showing why NOT to do this to a ATX supply.
This was awesome! I'm fairly handy with a soldering iron and have a crap ton of old PSUs laying around. At first I was thinking of trying this. After seeing someone who knows a lot more than me fail at it, I think I might just buy one. Thanks for saving me from ruining my own power supplies!
Hi sir , your explanation about the circuits are more complicated to me during past videos you made . Any way during few days i try to learn from your video tutorials and experiments by pausing and rewinding it. I am so excited to comment your video first time . i am picking up your way of explanations and ya i am learning . Any way this is a really freaky awesome video which i understood what is switch mode power supply . i am an electrical and electronics engineering student , i am just completed my one year now , 3 more years to complete my course. i think your videos pave way to success in my engineering life . any way i will catch you next time in the comment box thank you so much for your video tutorials.
And in Germany every bike with a motor over 250 W and 25km/h. And it has to only assist the driver. I mean, you COULD ignore the rules. And nobody would probably stop you. But should you risk it.
Michi Lo One of my earlier collegue, made his own electric bike, with a 700W motor on each wheel, a speedregulator (no assist), topspeed around 50km/h on a good day, there would be 5 traffic violation, on that thing.
Could you make an electric bike? I think this would be interesting. Like when you also want him to do this. (Sorry for my bad english but Iam German xD) ------------------------------- Könntest du ein E-Bike machen? Ich glaube das es ziemlich interessant sein würde xD. Like wenn ihr es auch sehen wollt.
For low current applications, we can use +12V and - 12V which together gives 24V. But the smps ground becomes a Hot terminal. To vary that voltage, a seperate circuit has to be incorporated.
I once made this modification to a KA7500/ tl494 based atx supply successfully. Not only did I modify it for variable voltage but also variable current output. Worked well for lead acid/ lithium rechargeable charging. The atx I used had a protection chip with three voltage inputs 3.3, 5 and 12 volt. If the voltages varied by more than a few percent it would trigger a protective shutdown. I lifted the three input voltage pins of this chip and created a shunt regulated divider from resistors and colored led's. This gave the protection chip the voltages it wanted to see with the added benefit of still protecting the supply from output voltages of more than 16.5 volts on the 12 volt rail. Some load/ voltage ranges caused audible switching noise so I don't know if it would survive long term in these ranges or not. I killed it though by running it off of a generator while I was also using a welder. I had a bad connection in the main power wire to the welder and I think the arcing on the ac power line fried it. As far as hooking things up to the output of the supply, it did very well. Protection chip still did it's job in the higher voltage ranges it was modified to.
i have the exact same ATX power supply, and i have been modding it for an adjustable output. an LM317T adjustable voltage regulator comes in handy after de soldering all those excess wires that take up half the power supply. if you look at it, the wires are grouped by color, and its then you can determine the function as well as its output voltage.
i actually (possibly) take that back because i forgot to account for how much current the lm317t can actually withstand. although not tested yet, im not for sure.
Holy Cow . . . . oops!, I mean 'Great Scott! . . . .what a fabulous video/tutorial. Your presentation, explanation/documentation style is really world class, and leading the viewer though the exploration/education process only to have such an unexpected but wonderfully safety conscious conclusion is a fun thing and I admire you for it. Thanks much.
I have a question (which I hope gets an answer on a 3 year old video), I've seen a number of videos converting ATX PC and Server PSU's into bench supplies etc. I have recently done a conversion myself but this was to simply turn a PC PSU into a multi USB charger using the 5V rail to charge 20 devices at a time @ 1A. What I would like to know is could I convert the 12V rail down to 5V and expand the capacity even more? I realise I could use a buck converter but is there a better / simpler way?
A series of 5 volt regulators. If you search Digi-Key hard enough, you can even find power management ICs meant for usb charging applications (they can detect a fast charge capable device too!). I didn’t look too hard at them, myself, because I was looking for a different chip, but I don’t think the supporting circuitry was too complex. If you go with the 5 volt regulators (lm7805 or equivalent), make sure you give them adequate cooling. Even at 9 volt input, they tend to want to get warm
love your channel dude. you are awesome! love the way you explain everything, even your accent!! teaching myself electronics to maybe start a channel in my language.
Nice video, glad you posted it. Very good breakdown of the mechanics behind the workings of the power supply. I had made an ATX bench power supply and also wanted a variable output. I made a box that plugs into the +12 and -12 (to get 24V) which is a variable DC-DC converter. I used the MC34063 as the basis for the converter. It's a little clunky having an external converter, but it's very useful and I use it all the time.
You can get a higher voltage by utilizing the -12v rail. To get 24 volts, just connect between -12 and +12v, to get 17 volts, just connect between -12v and +5v. Connect between -12v and 3.3v to get 15.3 volts. If it is an older supply with a minus 5 volt rail you can get even more voltages.
Inspired by this video I just did a crude power supply with +5V +12V +3V, -5V, -12V and an adjustable voltage from 1.2 to 11V using a buck converter inside the power supply, as I had room for that. The converter uses a LM2596 and can handle up to 3A, good enough for my experiments. Thanks for the idea!
So, if I wanted to just use that buck-boost converter method, could I replace the pot with something I can put on the outside? And which leads would I connect it to GND and 3.3V / / 5V / 12V?
Thank you so much. That was fascinating. I have always wondered about the guts of a SMPSU. I knew the theory of course, but never had the need to build one.
GreatScott! I'd been thinking of doing this project for a long time. I've already created ATX power supply as I don't have enough sources to buy a lab bench power supply yet. But now I think that it's quite painful to tinker with these ATX power supplies. I'll buy some power supply as soon as I get some money. :P
Just so you know, a standard atx power supply uses a type of SMPS called a forward converter. Forward converters only consists of a controller chip that outputs a PWM signal that is regulated by a feedback signal which drives only one transistor that is on one side of the transformer while the other side is just connected to ground usually. It is called a forward converter, because the current flows only one way through the transformer. Of course this is the bare minimum of a forward converter which does not include protection circuitry and PFC. Modern atx power supplys usually has PFC, which is why most UPS systems will not work.
Good Job trying and even better Job with the warning at the end!!It's real easy to forget how powerful these power supplies are and how one mis-hap could be dangerous. Watching your first video on this, along with the Video on the viable power supply. I created a combined unit.(Picture on my Google page). One side, I have the power Supply with a Volt meter. It uses 12V, 5V, 3.3V, and -12V rails. Than next to the voltmeter, you plug from the outside, any combination of 24V, 17V, 15.3V, 12V, 8.3V, 7V, 5V, 3.3V and 1.1V. Than on the other half of the unit(w/Volt/Ampmeter), I have a 5AMP Buck Converter, powered by a 24V 5 amp (LED) Brick power supply. It is CC/CV adjustable 24V-1.5V. Its very compact and runs Great! I also used Plexiglas, so you can see the internals. All Ideas for this project came from Great Scott!! Of Course, I also have a couple of units, made apart...LOL EDIT-On The variable side of my DIY power supply, below the power switch is 2 banana female plugs(hot/cold). When the LED power supply is Not plugged in, I can direct power from my Power Supply side(with Banana cables). But Unfortunately the most I could use on that side was 24 V 1 amp. Bucked down, the amp draw got bigger, but not enough for the higher powered devices. Reason I also purchased the LED brick. I hardly ever use the LED brick, the Power supply side normally has the power needed. Lots of Great Ideas come for this channel! New to DIY,😊 Quick Q----I noticed you mentioned that the power on the other side was upwards of 325 Volts DC and I am sure a TON Of Amps....Been looking to create a Battery Tab welder, and exploring all options before I decide on the way I will go. How many amps do you think you get on the 325 V side before it will force shut off the unit? Do you think its possible to use one of these, except a MOT. Just an Idea???
Gosh, these videos are making it hard to choose my path of schooling after high school... on one hand Id like to go for paramedic but I also want to go for electronics it just seems so Interesting! ive also considered getting a soldering iron kit and a bench power supply.
variable zener diode TL431 fixes this problem apparently, I haven't done it myself yet but I found instructions on a website, you use it to set a trigger level for the optocoupler, the feedback mechanism switches the pwm on and off based on whether the optocoupler is on or off ( when output is above a certain level defined by a zener diode ), thats how it works in some designs at least.
Another reason why you should not try this, especially with the protection circuitry disabled, can be seen if you watch closely at about 6:45 When the output voltage is set low, the fan slows down. This makes sense because the fan is powered from the 12V rail. Normally this wouldn't be a problem, but if you connect a high current load, such as a long LED strip or a big motor, if the power supply is badly designed, it will overheat and because of the relatively low airflow it might even catch on fire under extreme conditions. Also, if you were to set the voltage to 16V or higher, not only would the capacitor you mentioned explode, the fan would also be running with way more voltage than it's intended for and certainly would not last very long.
you can easily use one transformer,bridge rectifier and lm317 to make variable dc power supply.i know that lm317 can maximum output current 1amp but with bypass transistor you can change that :)
Thanks for talking some sense into my stubborn head. I was positive I could make an adjustable power supply, but if Great Scott the Hacker says don't bother, guess I won't.
Hello great Scott.... Even if we connected the ATX green (stand by) wire to black (ground) wire it would still not work as there is load in the computer power supply...... How did you get around and about that .... Mine is not working and I have cornfimed the color codes for the wire are ok
Very good explanation of the switching power supply. The hack will be good for convert the PS rather to an exactly 12 PS than a substitute for the bench supply. Also a bit over 12 volt could be useful as car or caravan battery charger.
Add some dummy loads to the 3.3v and 5v rails, this will help with regulation as they are normally cross regulated in this type of supply to save cost and space. A few watts load per rail should do the trick.
I have been using old XBox 360 power supplies, as well as Dell dock supplies like their D846D, for some time like this - as adjustable voltage/lab bench supplies, and they work great. They are adjustable over a fairly wide range (with some protection disabling) given what they were built for (12 and 19 volts respectively). This is a great way to get a custom voltage with high current at a cheap price. Great video, as usual.
You missed the output retifier diodes, before the capacitor filter.. But's ok ! Very nice video ! I always wanted to do that, but i never get enough time ! I share this video to my Engineering group. Stay creative, and i will see you next time ! ^^
Thank you for doing this investigation! I've got a 1kW Mackintosh power supply which I am attempting to convert into a bench supply with both a buck and a boost converter. Upon seeing your title I thought there might've been a simpler way! unfortunately not but atleast that warrants the purchase of a 600W Step up converter :D
You burnt the optocoupler of the first PSU, I think it'll work if you replace it. That caused by giving higher voltage with using secondary potentiometer too low. I think second resistor you removed is current limiting for optocouplers led. Thanks for sharing all information 👍
I'm very impressed by your freehand sine function drawings.
I dunno... They always looked a bit fat to me, especially at the antinodes.
He must enjoy it, because I as had electrotechnics on highschool and we had plastic template.. makes your life much easier, there was all the marks for resistors, diode, capacitors...
They looked pretty but they didn't look very accurate. They looked like they were almost vertical at the zeros.
I enjoy watching his videos JUST for his handwriting!! I'm quite envious.
Pretty much all worthwhile in the video
FULL BRIDGE RECTIFIER
NO WAY, i just did the same joke and saw your comment like 1 min later..
Well fuck me and well played sir :D
*rectum frier
Omar's Garage wrong channel man.
Ave
karlpc200, wrong again...
Extremely useful and clear description of a switch mode power supply. I like how even though you didn't get your expected conclusion with this video, you continued it anyway and shared it with the world. Thank you.
Now if only the title wasn't bait maybe that decision would actually be worthy of some respect.
Time traveling from 5 years in the future... thanks for this. It's instructive. I liked the primer on switching power supplies. I only had experience with old-school versions. And thanks for sharing your "failures" as well as your successes. I think we learn more from failed projects than successes. I am in the middle of building my ATX bench power supply. If I ever need a variable supply, I know to just buy one! Best wishes and continued success!
This is EXACTLY what I was looking for. Excellent freehand drawings and great detective work looking at all the different IC's and messing with the potentiometers. This is something I'd been wondering about and wanting to do. Now that I see what the limits are, I'm glad you sacrificed your power supplies to show us. This definitely wins my subscription and thumbs up!
You take great pride in your Penmanship. This shows every time you put pen to paper. The same is said with the quality of your projects. Every detail, every component, every finishing touch.
This speaks volumes and it's something that so many have lost. Everything we do is like signing your name, this is the lesson I try to teach my son. I am glad to see you trying to show so many others the value of this very important lesson.
The decision to abandon the experiment on the realization it requires disabling all the built in redundant safety features is not only "the right way to do it", but also quite praiseworthy! Great vid
He wants to imitate a chewing gum munching Ami. This complex can be found more often in Germany and it gives me chills. To meet someone who has trained himself to speak clean English, that would be something. But I commend his videos for being informative, accurate and innovative.
Great GreatScott! Thank you!
Small notation: please never left unconnected side pin of variable resistor (potentiometer). 04:45 and 05:55
Short it to central pin always!
It will keep your load (sometimes very expensive) safe of short voltage spikes out of range.
I use a few of these for lab power supplies. They run from about 5v to 25v. I tend to remove most of the stuff on the secondary side of the board, and just leave the 12v winding and circuit. The ATX uses the spare fet and tiny transformer for sleep mode to maintain 12v to the KA7500 ( TL494 ).
If we then lift pins 15 and 16 above ground, we can sense the current via a very low R resistor to ground from center tap, and use that to use the [pin 15/16 op amp to control current, while using the pins 1 and 2 for voltage control.
Then we have a current controlled ( up to 12 amps or more) and voltage controlled 9 up to 25v or so) power supply, that will power very low impedance loads ( motors etc).
This requires all of the monitoring stuff to be wiped from the board, or at least disconnected from the pins 1,2,3,4,15,16.
So we use the 5v reference from the 7500 to go to both of the 7500 op amps, then we can simply use a divider for the voltage to the first op amp, another divider from the CTap of the tranny to ground to the second op amp, some feedback from pin 3 to both, and we have a totally stable power supply, that is completely controlled.
It is actually simpler than it sounds really.
Expensive is a relative term my friend.
Hey Jeremy,
The problem is that your playing with a chip that uses 2 comparators. These are way more complicated. However that KA7500 chip is a replica/update of the old TL494/TL594. You can redesign the circuit around that IC to create a voltage and current feedback loop and therefore a variable voltage/current supply, (good luck building that in a 1 week project though ;)
I've been looking into this circuit, but it gets super complicated IMO to make it variable over a wide range. The FET duty cycle and core saturation math gets a bit blurry for me. All the examples I've found attest to the potential operating range of the circuit, but they don't really explain the details of how to design specifically for an variable output. Where and when to compromise specs is the challenge. The best I can tell you build for about a 90% duty cycle at your maximum power output, while trying to maximize the timing frequency at that output, but I'm not 100% clear on that one.
I've found some interesting circuits to try and will probably do a video at some point. There is a way to use a power PNP BJT with a capacitor on the base to reduce the ripple effects across a wide bandwidth.
Bill Herd over on HackaDay's YT channel hinted ages ago he was going to create a variable SMPS that used a linear design at the output. I've half built a project on that idea, but that is about 3-4 down my list.
If you see this post and want to contact feel free Jeremy. I'm using YT to try to connect to more of a community. YT also keeps me committed to completing my projects despite being partially disabled and stuck laying around most of my day. My channel is still not monetized and I'm not trying to copy anyone. I'm just here to keep it interesting ;)
-Jake.
Keep the spirit. :)
If you do any work towards this lab psu thing, please let me know. Seems like an interesting project to keep up with.
A couple years ago I have done the same modifications he shows in this video, and I consider it a success. I was able to get the voltage range I wanted. But now I'm looking at doing something with higher range, so I would need to do more intrusive modifications than just adding a potentiometer.
Your handwriting is so perfect
Edit: THANKS FOR THE 208 LIKES
Oh stop it, you
I agree.
That Sine wave though
AJLION GAMING my hand writing is absolute garbage!
qwerty keyboard GARBAGE DAY
I just converted an ATX power supply into a variable bench power supply by connecting a 1k Ohm potenciometer from 12V to ground, with the middle pin being the output voltage. It doesn't give me more than 12V but hey, it works!
I once had a friend coming to me with his old, potentially broken, computer PSU. We decided we'll have fun with it by modifying the thing to make a lab psu. It was a really cheap looking PSU, the components inside are poor too, but we just kinda go with it.
We turn that on and it didn't show any living sign, then I saw a switch which I thought was the 110V/220V fuse thing, then I switch that to the other side. Suddenly, a huge capacitor blows up followed by some diode and ultimately the whole circuit was in fire.
We panicked and unplug the power, throw the firey thing outside, and a black-blue smoke is coming from what seems to be a bigger capacitor.
That day, we learned our lesson to not mess with any shitty old electronics.
You flipped it to 110v and it'd be a miracle if that didn't blow up
It's like trying to draw and you make one wrong line and ruin the drawing and then throw it all away because you don't want to mess with art
You done it as a dumb, thats why
@@alexstone691 except when you draw the line wrong you get electrocuted and die.
I also had a power supply with that switch the previous owner had set it to 110v I plugged it in 230v it blew up. I got another power supply with that switch again but this time I switched to 230v and permanently glued that switch
Hi Scott. A few years ago I have done the same modification as you did, and I consider the result a full success. Although the voltage range is limited, it varies enough to make the bench PSU useful for my purposes.
However, I understand why you disencourage your readers to do the same. The risks of dealing with 240vac is immense and you don't want to be responsible for people risking their lives, specially when forgetting to discharge high voltage capacitors is a common error even among professionals. You are right and I agree with your stance, BUT for those who actually have the skills and knowledge and responsibility to safely do this modification, I think it's totally worth it. It's a nice project, easy to do, fun and rewarding to see it working in the end.
The first power supply that you were modifying, that style works great. The high wattage ones with the UC384x chips are usually 2-switch forward converters and regulate very well.
When you're modifying the feedback circuits in these, you must not modify the circuit around the UC384x chip. Since the feedback is with an opto-coupler, the voltage division and comparison is done on the secondary side with a TL431 programmable shunt (or sometimes but rarely a zener) and that sends a simple on/off signal through the opto-coupler back to the UC384x chip. You only need to modify the divider going into the TL431. The TL431 is usually a TO92 package and looks like a small transistor.
also you should only modify the lower side resistor of the divider, otherwise it is possible to destablize the feedback loop
the mark of puting the "Great" in Scott, admitting when it wont work, then sharing your failure, and then trying again. Again, I like the variety and creativity of this video. Cheers.
Unexpeced, but very good conclusion. Thanks for another great video.
this is some enjoyable Proffessionnalism
CabinDoor Can you elaborate? Not sure I've got your point...
Au contraire, very expected conclusion
Everywhere, anglosaxon culture people try to exaggerate dangers and to explain why you should NOT do it instead of teaching HOW TO do it correctly and safely
The conclusion that it's better to spend USD$53 (+shipping) than RECYCLING a device that can be obtained almost for free is not only ludicrous, it's the perfect example of what should NOT be done
lazy...
The BEST VIDEO on youtube.. Thank you so much for breaking the Power Supply down into Blocks with the expected Wave Form and Voltage. Please DONT STOP this method of explaining Electronics as You are a TEACHER really.. Thank you for taking the time to help us un-informed dummies. May the Force be with you.........
4:00 - And by measuring this voltage, while the power supply is turned on, I'm waiting for the point where you accidentally shorted out something with the probe and got a huge spark and a faceful of vaporized electrolyte
If this is the content you are looking for, check out Electroboom.
I tried this same project about a year ago and became fed up. So I just ended up just buying one of the DIY kits off of Aliexpress/Banggood for around $9.00 US. Love your videos GreatScott! been a fan for 2 or 3 years now.
Yeah Scott this is what we wanna see
I am awed and humbled every time I watch your videos. Not to mention your beautiful circuit diagrams and handwriting.
this feeling when you dont understand a thing, but you feel so inteligent while watching it xD
+Adiko. Uhm.. i think you two guys got something mixed up, maybe trade him your intelligence for his understanding?
Sounds like a fair trade to me?
en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
also makes us feel dumber knowing we have a fraction of this guys knowledge...huhu
Read that...now I am conflicted as to what side of that effect im on...
Its ok amigo...just feel free to feel dumb like me...we are only human and cat
Finally I found, best of the best channel for my electronic reference. This is just like I'm in electronic class. Thanks
2:15- RIP output rectification,we will miss you.
I created a lead-acid battery charger using a PC power supply with the infamous TL494. I didn't remove the divider resistors though. I simply added a parallel variable resistor to the pulldown resistors to get higher voltage on the 12V rail. Replaced the capacitors to match higher voltages on all rails. you can find the mod on the internet though. Good work Scott.
Thanks for sharing, I was wondering when the warning about beeing killed by the power supply might come and was surprised you mention it at the end and not at the beginning of your video ^^
Thank you for trying to mod an ATX power supply for variable voltages. I have wondered how easy it would be to do so and now know thanks to you. Now it is time for me to look for a proper lab power supply. Take care and keep the videos coming.
Sen bir dahisin!
Helal olsun sana dostum..
wow that looked so simple... My approach to large circuits like that is to immediately salvage everything off it and not even consider using it. Well done
Maybe a video walking your subscribers through the different stages of caps, how many Joules of energy it contains, how much can kill a person, and proper PPE/protocols for dealing with these caps?
That goes against his religion and culture. Educating people is below scaring them from sinful things on the list
OMG this is the first time in my life I was able to completely understand a GreatScott video!
I'll just add my .02USD here that with videos like this a warning at the start would be a good idea because man did I cringe when you pulled that PSU board out with your fingers that close to those big caps. That said I always look forward to your videos because you make them so even a dummy like me can understand them.
I've been looking for EXACTLY a video this clear and concise for making a variable PSU for like a year and a half. Thank you so much.
I am so proud that you did not recommend this, because I made myself a DIY ATX powersupply just yesterday!!! I would be mad if this actually worked.
one question, why didn't you put fuses in your DIY ATX power supply? wouldn't shorts be dangerous or kill the powersupply?
An ATX power supply usually has a short circuit protection. So it turns off automatically.
GreatScott! I know that verry well, it has saved me more than a few times
It has to have protection features because only one output rail is monitored in the feedback loop. The other voltage rails are usually regulated simply through their turns ratio on the same transformer core as the regulated rail. This is why the primary rail carries the highest current rating and is designed to carry most of the load. The other magnetically coupled rails have a much larger output variance or to put it another way, they have poor output regulation. This is especially the case if you draw more current from them than the main rail. The auxillary rails must first influence the main rail before the SMPS controller can detect the change and compensate. Basically an ATX system is designed as simple as possible for creating multiple voltages at high current.
It is far easier to modify a SMPS circuit that is designed with a peak current controller instead of a PWM controller. These circuits only contain 1 comparator and therefore 1 feedback voltage reference. I've uploaded a video on modding one of these (none monotized/cc license/no affiliate marketing).
Peak mode controllers are found in everything from car type phone chargers to laptop/printer power supplies. The most ubiquitous and ancient chip is the MC34063. It's a good place to start but it is limited to 500-700mA and was designed before power mosfets were standard so it's terrible for high output.
Upcycle Electronics Cool man, I checked your channel and the vids are pretty good!
Neki Dzemaili Also probably the wires will block some of the current and won't allow 30+amps to pass.
No mention of the dangers of messing with the mains side of power supplies, or about discharging the input caps... Well, maybe we'll get rid of some people :) Glad you came to your senses.
According to your block diagram, it looks like a FLYBACK converter right?
This is propably forward or halfbridge, not flyback
2:49 aaahhh...so that is how they maintain stable voltage with different current draw, Thank you GreatScott, this step by step explanation with flow chart help me a lot to understand about power supply
Buy a cheap chinese LED supply -- the kind with the perforated metal case. The feedback supply is much simpler, and the 48V ones can be adjusted from 0 to ~63V. It's a bit noisy, but an LC filter cleans it up nicely. All you need to do is make sure the r values are low enough to ensure proper bias current.
How can I do that? I would appreciate if you could help me build one!
I've disassembled two at home that were still working, but till now, I have no idea on how I could use them.
Victor Ribeiro Find datasheets from all ICs used on board, one of them will be used to provide feedback using resistor voltage divider, you will have to modify that divider to get desired voltage
ethanol L@@L
I really like the inclusion of failures in your videos. They're a learning experience a lot of other youtubers completely gloss over.
@5:00 HiVolt surgical gloves. Hooray
Yea I just made a custom housing for my ATX power supply and used a buck converter with a voltage/ammeter display. Works great. Gets me 1.33v -11.8v. And of course kept the constant 3.3v, 5v, and 12v banana jacks.
Hi GreatScott, I am your big fan. I like all your video especially the modifications of rasperry pic to game boy, since this is the video makes me start to watching all your videos, keep going to create more video with more interesting things. I have an ideas for you, can you make a AC timer switch with connection to Android phone so it can programme the timer through the smartphone. 😊😊
Thanks for the feedback. I put it on my to do list
put an IR blaster online trigger it with an app on your smartphone done.
GreatScott! hello my name is michel of Brazil
For my own lab bench power supply I used an industrial switched mode power supply and a buck/boost converter. It cost me around 30€ and I can get up to 32 volts by around 4 amps, which is a pretty good deal I guess :).
Sounds good
@GreatScott!
can you make a Video building a single phase Inverter (VFD) using an arduino?
I enjoy watching the videos as a guidepost of how much I've learned about electronics. Back in the day I'd have been pretty clueless from the start, but thanks to GreatScott! and others this time I got about halfway through before i had to pause/think and learn. Next time even further.
Another great video
Love ur videos
Thanks mate :-)
Hey Scott, I did something like that yesterday, but i ended up rewounding the core to give me 3.3, 5, 12, 18v ground referenced
and two 18v floating for a DRSSTC full bridge driver.
Had to use few schottky, one more ferrite, few electrolytics, decreased the switching frequency from 65 to 40kHz as my new core was N27 type, increased soft start and tricked the smps that everything is just all right.
There's a great 38v dc-dc variable buck boost converter module with led display available on eBay that is easily added to an old power supply that makes a great variable shop supply with adjustable current limiter.
Yes, or if 10.5 Volt is enough for your lab (arduino, raspberry pi, and what not) , this buck converter is good value for money: www.ebay.com/itm/Constant-Voltage-Current-Power-Supply-Converter-DP30V5A-L-LED-Module-32V-5A-160W-/172661458452 Just hook it up to the 12V of the ATX power supply.
I just subscribed to your channel. Don't know why I have not done so earlier. Your explanations are some of the best on RUclips. Well done on showing why NOT to do this to a ATX supply.
7:35 nice advice man😅🤣..Well if it wasn't expensive i wouldn't have ever watched your video...and you wudnt have 7.44 mins of my watchtime...😄
True😂
This was awesome!
I'm fairly handy with a soldering iron and have a crap ton of old PSUs laying around.
At first I was thinking of trying this. After seeing someone who knows a lot more than me fail at it, I think I might just buy one.
Thanks for saving me from ruining my own power supplies!
Top Tip: Lick all the terminals while switched on to deep clean the ends.
MrGareth66.com instructions not clear: now a shadowy guy with a scythe is ringing at my door, and the house stinks of burnt chicken. What should i do?
Hi sir , your explanation about the circuits are more complicated to me during past videos you made .
Any way during few days i try to learn from your video tutorials and experiments by pausing and rewinding it. I am so excited to comment your video first time . i am picking up your way of explanations and ya i am learning .
Any way this is a really freaky awesome video which i understood what is switch mode power supply .
i am an electrical and electronics engineering student , i am just completed my one year now , 3 more years to complete my course. i think your videos pave way to success in my engineering life .
any way i will catch you next time in the comment box thank you
so much for your video tutorials.
make an electric bikee
he made an electric longboard before. it's basicly the same but for a bike would probably need more custom built parts.
And in Germany every bike with a motor over 250 W and 25km/h. And it has to only assist the driver.
I mean, you COULD ignore the rules. And nobody would probably stop you. But should you risk it.
Michi Lo One of my earlier collegue, made his own electric bike, with a 700W motor on each wheel, a speedregulator (no assist), topspeed around 50km/h on a good day, there would be 5 traffic violation, on that thing.
It would seem that you have the gift of explanation/teaching. Things just make so much more sense when you break them down for mere mortals like us :)
I used to be a fan..... Now I'm an air conditioner...
You mean a "van", right
I cant even draw a straight line with ruler. Your handwrite is amazing
Could you make an electric bike? I think this would be interesting. Like when you also want him to do this.
(Sorry for my bad english but Iam German xD)
-------------------------------
Könntest du ein E-Bike machen? Ich glaube das es ziemlich interessant sein würde xD. Like wenn ihr es auch sehen wollt.
PingasTheFourth Not funny :|
Your English is not that bad my friend, Better than my German.
For low current applications, we can use +12V and - 12V which together gives 24V. But the smps ground becomes a Hot terminal. To vary that voltage, a seperate circuit has to be incorporated.
I was thinking about that idea few days ago after i watched your ATX lab supply video and here it is. Thanks
Nice, I like the way you put up your failures as well as your successes, good bit of forensic electronics
I once made this modification to a KA7500/ tl494 based atx supply successfully. Not only did I modify it for variable voltage but also variable current output. Worked well for lead acid/ lithium rechargeable charging. The atx I used had a protection chip with three voltage inputs 3.3, 5 and 12 volt. If the voltages varied by more than a few percent it would trigger a protective shutdown. I lifted the three input voltage pins of this chip and created a shunt regulated divider from resistors and colored led's. This gave the protection chip the voltages it wanted to see with the added benefit of still protecting the supply from output voltages of more than 16.5 volts on the 12 volt rail. Some load/ voltage ranges caused audible switching noise so I don't know if it would survive long term in these ranges or not. I killed it though by running it off of a generator while I was also using a welder. I had a bad connection in the main power wire to the welder and I think the arcing on the ac power line fried it. As far as hooking things up to the output of the supply, it did very well. Protection chip still did it's job in the higher voltage ranges it was modified to.
the quality of your videos is outstanding. keep it up
i have the exact same ATX power supply, and i have been modding it for an adjustable output. an LM317T adjustable voltage regulator comes in handy after de soldering all those excess wires that take up half the power supply. if you look at it, the wires are grouped by color, and its then you can determine the function as well as its output voltage.
i actually (possibly) take that back because i forgot to account for how much current the lm317t can actually withstand. although not tested yet, im not for sure.
So Creative ! I have not seen anyone done such a mod on a power supply!
Holy Cow . . . . oops!, I mean 'Great Scott! . . . .what a fabulous video/tutorial. Your presentation, explanation/documentation style is really world class, and leading the viewer though the exploration/education process only to have such an unexpected but wonderfully safety conscious conclusion is a fun thing and I admire you for it. Thanks much.
I have a question (which I hope gets an answer on a 3 year old video), I've seen a number of videos converting ATX PC and Server PSU's into bench supplies etc. I have recently done a conversion myself but this was to simply turn a PC PSU into a multi USB charger using the 5V rail to charge 20 devices at a time @ 1A. What I would like to know is could I convert the 12V rail down to 5V and expand the capacity even more?
I realise I could use a buck converter but is there a better / simpler way?
A series of 5 volt regulators. If you search Digi-Key hard enough, you can even find power management ICs meant for usb charging applications (they can detect a fast charge capable device too!). I didn’t look too hard at them, myself, because I was looking for a different chip, but I don’t think the supporting circuitry was too complex. If you go with the 5 volt regulators (lm7805 or equivalent), make sure you give them adequate cooling. Even at 9 volt input, they tend to want to get warm
love your channel dude. you are awesome! love the way you explain everything, even your accent!! teaching myself electronics to maybe start a channel in my language.
Nice video, glad you posted it. Very good breakdown of the mechanics behind the workings of the power supply. I had made an ATX bench power supply and also wanted a variable output. I made a box that plugs into the +12 and -12 (to get 24V) which is a variable DC-DC converter. I used the MC34063 as the basis for the converter. It's a little clunky having an external converter, but it's very useful and I use it all the time.
You can get a higher voltage by utilizing the -12v rail. To get 24 volts, just connect between -12 and +12v, to get 17 volts, just connect between -12v and +5v. Connect between -12v and 3.3v to get 15.3 volts. If it is an older supply with a minus 5 volt rail you can get even more voltages.
Inspired by this video I just did a crude power supply with +5V +12V +3V, -5V, -12V and an adjustable voltage from 1.2 to 11V using a buck converter inside the power supply, as I had room for that. The converter uses a LM2596 and can handle up to 3A, good enough for my experiments. Thanks for the idea!
So, if I wanted to just use that buck-boost converter method, could I replace the pot with something I can put on the outside? And which leads would I connect it to GND and 3.3V / / 5V / 12V?
That block diagram is beautiful!
Thank you so much. That was fascinating.
I have always wondered about the guts of a SMPSU. I knew the theory of course, but never had the need to build one.
How can you connect 325V DC to a transistor? Wouldn't it burn?
GreatScott! I'd been thinking of doing this project for a long time. I've already created ATX power supply as I don't have enough sources to buy a lab bench power supply yet. But now I think that it's quite painful to tinker with these ATX power supplies. I'll buy some power supply as soon as I get some money. :P
Just so you know, a standard atx power supply uses a type of SMPS called a forward converter. Forward converters only consists of a controller chip that outputs a PWM signal that is regulated by a
feedback signal which drives only one transistor that is on one side of the transformer while the other side is just connected to ground usually. It is called a forward converter, because the current flows only one way through the transformer. Of course this is the bare minimum of a forward converter which does not include protection circuitry and PFC. Modern atx power supplys usually has PFC, which is why most UPS systems will not work.
Good Job trying and even better Job with the warning at the end!!It's real easy to forget how powerful these power supplies are and how one mis-hap could be dangerous. Watching your first video on this, along with the Video on the viable power supply. I created a combined unit.(Picture on my Google page). One side, I have the power Supply with a Volt meter. It uses 12V, 5V, 3.3V, and -12V rails. Than next to the voltmeter, you plug from the outside, any combination of 24V, 17V, 15.3V, 12V, 8.3V, 7V, 5V, 3.3V and 1.1V. Than on the other half of the unit(w/Volt/Ampmeter), I have a 5AMP Buck Converter, powered by a 24V 5 amp (LED) Brick power supply. It is CC/CV adjustable 24V-1.5V. Its very compact and runs Great! I also used Plexiglas, so you can see the internals. All Ideas for this project came from Great Scott!! Of Course, I also have a couple of units, made apart...LOL
EDIT-On The variable side of my DIY power supply, below the power switch is 2 banana female plugs(hot/cold). When the LED power supply is Not plugged in, I can direct power from my Power Supply side(with Banana cables). But Unfortunately the most I could use on that side was 24 V 1 amp. Bucked down, the amp draw got bigger, but not enough for the higher powered devices. Reason I also purchased the LED brick. I hardly ever use the LED brick, the Power supply side normally has the power needed.
Lots of Great Ideas come for this channel! New to DIY,😊
Quick Q----I noticed you mentioned that the power on the other side was upwards of 325 Volts DC and I am sure a TON Of Amps....Been looking to create a Battery Tab welder, and exploring all options before I decide on the way I will go. How many amps do you think you get on the 325 V side before it will force shut off the unit? Do you think its possible to use one of these, except a MOT. Just an Idea???
Gosh, these videos are making it hard to choose my path of schooling after high school... on one hand Id like to go for paramedic but I also want to go for electronics it just seems so Interesting! ive also considered getting a soldering iron kit and a bench power supply.
variable zener diode TL431 fixes this problem apparently, I haven't done it myself yet but I found instructions on a website, you use it to set a trigger level for the optocoupler, the feedback mechanism switches the pwm on and off based on whether the optocoupler is on or off ( when output is above a certain level defined by a zener diode ), thats how it works in some designs at least.
I still like the down-up boost converter attached to the 12V power side and an external potentiometer !!! I agree your writing is hypnotizing ! Lol
Another reason why you should not try this, especially with the protection circuitry disabled, can be seen if you watch closely at about 6:45
When the output voltage is set low, the fan slows down. This makes sense because the fan is powered from the 12V rail. Normally this wouldn't be a problem, but if you connect a high current load, such as a long LED strip or a big motor, if the power supply is badly designed, it will overheat and because of the relatively low airflow it might even catch on fire under extreme conditions.
Also, if you were to set the voltage to 16V or higher, not only would the capacitor you mentioned explode, the fan would also be running with way more voltage than it's intended for and certainly would not last very long.
you can easily use one transformer,bridge rectifier and lm317 to make variable dc power supply.i know that lm317 can maximum output current 1amp but with bypass transistor you can change that :)
Thanks for talking some sense into my stubborn head. I was positive I could make an adjustable power supply, but if Great Scott the Hacker says don't bother, guess I won't.
I had this idea myself to make such a power supply. I wont bother doing it anymore.thx.
Thank you for your outstanding information. Its not what I wanted to hear, but what I needed to hear. And your conclusion make perfect sense.
if the transformator works on pwm signals, why should need high speed diods for rectification on the output?
Hello great Scott....
Even if we connected the ATX green (stand by) wire to black (ground) wire it would still not work as there is load in the computer power supply......
How did you get around and about that .... Mine is not working and I have cornfimed the color codes for the wire are ok
Very good explanation of the switching power supply. The hack will be good for convert the PS rather to an exactly 12 PS than a substitute for the bench supply. Also a bit over 12 volt could be useful as car or caravan battery charger.
Add some dummy loads to the 3.3v and 5v rails, this will help with regulation as they are normally cross regulated in this type of supply to save cost and space.
A few watts load per rail should do the trick.
Any tips on what to look for in an ebay bench supply? After something a little better than entry quality
I have been using old XBox 360 power supplies, as well as Dell dock supplies like their D846D, for some time like this - as adjustable voltage/lab bench supplies, and they work great. They are adjustable over a fairly wide range (with some protection disabling) given what they were built for (12 and 19 volts respectively). This is a great way to get a custom voltage with high current at a cheap price. Great video, as usual.
Can I not just use a buck boost converter? Surely it is a good option other than doing what you tried?
You missed the output retifier diodes, before the capacitor filter.. But's ok ! Very nice video ! I always wanted to do that, but i never get enough time ! I share this video to my Engineering group. Stay creative, and i will see you next time ! ^^
Thank you for doing this investigation! I've got a 1kW Mackintosh power supply which I am attempting to convert into a bench supply with both a buck and a boost converter. Upon seeing your title I thought there might've been a simpler way! unfortunately not but atleast that warrants the purchase of a 600W Step up converter :D
So how much current can you get out of your new variable supply (same as before?), and can you limit the current to a set value?
You burnt the optocoupler of the first PSU, I think it'll work if you replace it. That caused by giving higher voltage with using secondary potentiometer too low. I think second resistor you removed is current limiting for optocouplers led. Thanks for sharing all information 👍