I absolutely LOVE cheap chinese electronics. it means I can order loads of stuff from ebay and banggood and aliexpress etc and play with it. if china hadnt revolutionised manufacturing I couldnt afford my hobby.
I agree, I couldn't afford to play hobby electronics otherwise. I think a more fair "teardown" would be to compare Apples to Apples instead of Apples to Oranges. Spend either $8 or $85 from BOTH China and Japan and see what you get. An $8 24 volt power supply from Japan gets you what? The empty box it comes in? Then compare the $8 Japanese PSU to an $85 Chinese PSU. Next compare two new automobiles where one costs $20,000 and the other costs $212,500. Feel free to mix or match the countries of origin. Just saying, the comparison was flawed from the beginning. If you want to compare value then start with two different products that cost the same. The only exception would be to purchase two identical items at different prices, that makes the maths for "value" very easy to calculate.
The fact the $8 Chinese outperformed the $85 Japanese unit in ripple, so what's left for the Japanese to be smug about? Do they really want to humiliate themselves against a $85 Chinese unit? Of course, I won't use a $8 PSU for my $2000 PC, but for DIY hobbyist use it's is more than good enough.
@@visionstills3700 well you get what you pay for but keep in mind that China had swallowed the world in the electronic industry and mostly all the household goods electronics are all made in China nevertheless a lot of famous firms have factories in China so that they keep the price of their goods compatible.
For 8 dollars you get a choke tranformer, rtc's for inrush current limiting, tinned pcb tracks and even pcb gaps and a optocoupler. I have to say for 8 dollars its pretty decent.
I worked for a power supply company for 30 years and we went out of business because we just could compete with the imported supplies. Up to the early 80's most of our business was with the military but as that dried up we did more and more custom commercial work. I taken a few Chinese supplies apart and been impressed at the quality vs cost. I've seen some that would not pass UL requirements but they all worked well.
dell177 hello you seem to be very knowledgeable about adapters. I need a little help I have LED lights that the adapter keeps burning out the specs are input: 100-240v-1.5a and the output 24v 6A what should I get to hold better. The LED are 120 ft long. I will appreciate any advice
@@betopizarro2003 Your LED string is probably too long and drawing too much power. How many LEDs are on the string (How are they broken up into sets of series LEDs)? How much power is each LED drawing and at what voltage? If they are 1.8v LEDs that means 12 LEDs in series gives about 24v. So you would want your lights to be broken into sets of 12 in series and a bunch of these sets in parallel. If you are using Christmas lights they may all be in series instead of being broken into sets of parallel LEDs. Christmas lights run at mains voltage so they have to drop down 120-240v So you can put a bunch of them in a row. As long as your LEDs are not burning out they are probably configured correctly or they would either burn out or be dimmer than expected. Then you need to know how many of milliamps each LED is drawing and multiply by the voltage to get the watt usage. So subtract 148W (24v X 6A) from the watts the LEDs are using (# of LEDs X milliwatts). If you have a positive number that is what your LEDs are over capacity. If you get a negative number everything is ok. Or subtract the watts the LEDs are using from 148 in which gives you how much more power your adapter can supply than what the LEDs are using.
If I see how our business is going, I must say. It is not cheap! But more regular price then overpriced like second one with huge margins.. Where 30% price is only because of old suits man:) Chinese are good in optimal balance between price and quality, these days. However its not always True.
Its seems like whoever crunched the numbers while designing the $8 PSU did a great job. For 1/10th the price you really can't go wrong... Super-educational video, thank you for sharing! :)
Not totally impossible they have ripped it of another expensive one or they just used the schematic example for the controller ic, very common in china to do, the cheaper the better no matter if its safe or not. And there is one thing on the boards bottom side that wasn't noticed. There was one component leg that was folded very badly so it was very very close to touch another bare trace which it could one day and totally ruin the entire PSU and even start a fire worst case, that's what you get if china made not in fex a Cosel. And the reason these are so cheap is just because the components are made in china so the just walk over to their neighbor, which is the component factory next door, and get extremely cheap components not even heard of outside china. You can find very very cheap components on fex Ebay. But if you the other hand buy them directly from the factories you would not belive the prices you get then since they are much much cheaper. And the other reason its so cheap is because the poor chinese workers hardly get any pay for their job.
It depends on how you define "go wrong". If the $8 power supply fails and in doing so, damages a $1000 piece of equipment, that seems like it's going wrong. If the $8 PSU fails and holds up a core function of a business, either in production output, client loss, or just idle employee salaries, thousands of dollars can be lost. The Chinese PSU is "acceptable" for some consumer users. It is a reason to fire whoever selects it for a business application that fails.
@@stinkycheese804If you can afford to by $1000 equipment, then you can afford the expensive SMPS unit! Some of us old pensioner are on a budget here in GB. And just use these cheep SMPS plugged into RCCB's here in Britain! Not burnt my house down yet, using quite a few of these cheep units, and still alive and kicking at this moment with no big bangs!
@@josearrighi RCCB means Residual Current Circuit Breaker -- A residual. current device (RCD), or residual current. circuit breaker (RCCB), is an electrical wiring device that. disconnects a circuit whenever it detects that the electric. current is not balanced between the phase ("hot") conductor.
Simple decision really. If you are powering something worth thousands of dollars or otherwise critical, buy the cosel one and you will not destroy the expensive load after a lightning strike or something like that. If you are powering some christmas lights around the workshop for fun, buy 3 of the chinese one, check each one before use for loose joints and stuff and make sure the grounding is proper before plugging it in... when it blows install a new one and, most fun part, debug/repair the fault (yay!) :-) These are not really competing products at all, both do their intended jobs very well.
The Cosel PSUs output ripple seems "dirtier", but most probably it's because dithering was applied purposefully. Dithering will reduce the EMI from the device and it's actually a good practice, overall.
+DigGil3 It seemed like a good practice, but today, after most chips got the feature, its dark side has emerged: The supplies are noisy in the audible spectrum. I'd choose a "traditional" switching unit that I wouldn't hear at night any time over these "EMI-friendly" wall-warts.
The real concern is about safety. As far as I can see, the Chinese SMPS has a standart 1nF 1kV suppression condenser between primary and secondary side instead of a safety self-healing one. This is a death trap, because it tends to fail dead short at overvoltage condition. Short kV spikes are quite common in mains - tens to hundreds a day - so if you have conected anything referenced to ground (even capacitive coupling is enough) to the output of such SMPS, the cap breaks down, leaving your secondary circuit galvanically connected to Live wire. The SMPS does not stop working, everything seems fine, only the trap is armed and waiting for it´s prey. The funny thing is that it seems that those Chinese "engineers" use proper safety caps in input filtering, where failure of one only leads to blown fuse and not lost live(s). They saved two cents on the wrong place. Life saving rule: Take your Megger and Hi-Pot test every single mains connected device from China! The nice side effect: you will get less concerned about efficiency, output ripple and pulse response very soon.
I really liked this teardown because you took the time to really trace out the circuit. I feel like I basically know how it works. I also thought the Chinese board was worth every penny of the $8 it cost, and it had a lot of effort put into the design and it did have all the key parts. Those cheap electrolytic caps might give it a shortened lifespan - but you can replace em or replace the whole board several times and still be far ahead of what the Japanese board cost...(which, admittedly, is a work of art and I wish we lived in a world where everything was made to that level of quality)
Yes, I think it would be interesting to run both at (near) full load for a week or two, continuous, and see if they both still work! I wonder how well the isolation would hold up, too. We have 240V AC mains, here, and I am VERY wary of cheap PSUs. It's hard to see that the Cosel is ten times better (unless the cheaper one only runs for a few tens of hours...)
I do a lot of high power LED projects and I find the cheap Chinese power supplies are OK when not using at full capacity and also not running too hot. For hot, full-load applications, I used the brand names.
I have used Chinese power supplies in numerous commercial applications where cost was a concern. Stepping away from the basic Ebay units (which I have used successfully on home projects) in my business, I find the Mean Well brand to be particularly reliable. If I were working in a medical or scientific environment I would probably use a German or Japanese supplier but for typical commercial and light industrial applications the less well documented Chinese brands can work quite effectively at a significant cost savings. Note: This video did not check for line voltage isolation which I always check on Chinese power products. Visual inspection doesn't protect you from faulty transformer isolation or suppression capacitors.
The ferrite beads are a nice touch. I'd prefer if they glued them down though, because I have a keen ear to any rattling going on inside of equipment, and have to make sure it's not something that could short out. The Cosel power supply doesn't cut any corners, and has lots of additional safety. But for 10 times the price, it's not worth it, unless if your equipment has to meet certain standards for use in certain fields of industry. The Cosel had a more noisy output, but it looked to be at a much higher frequency, and that would be easier to filter out than the cheap Chinese power supply (using a choke or capacitor) -- if your design permitted that... but then... you could always use a linear regulator too :-p
Yeah, I was looking at them just left there, no silicon to hold them. About cutting corners, I saw no MOVs, the fuse quality appears to be on the lower side.
I think the cheap China made power supply would be fine for anything except aerospace, military, medical, nuke plants, offshore oil, refineries or chemical plants. In other words if human life was at stake. For longevity at $8 I would buy a spare so you would still have spent 5 times less for 2 and you would have a backup on the shelf just in case. The fact that the ripple is cleaner on the simpler is shocking. I have always admired Japanese workmanship but many Japanese engineering designs always seemed needlessly complex for some reason (cultural?)
Rob C, but you would not just have the $8 cost of the backup on the shelf. You would also have the down time (which can be much more expensive by itself) until someone comes and replaces it and the bill for doing so, or else you retain a technician on-site capable of it so an entire salary. It is safe to say that most uses where the function of the powered equipment is tied to productivity which is tied to profit, would benefit from buying a higher quality PSU. As always where is a middle ground here, silly humans often thing in terms of opposite extremes of expense and quality instead of the middle ground. In other words something can be better, even "good enough" without being the most expensive one you can find.
@@stinkycheese804 You're talking about downtime in commercial applications, where (I hope) nobody uses those cheapos. This SMPS is for those driving LED strips at home and not willing to pay 10x more. And to confirm your theory with my personal experience, the cheap ones last no longer than an year before they fail, sometimes just a few months.
nice review though I'm pretty sure those bus bars were simply a necessity due to the diode being in the way of traces going through the PCB. there are other places that the came current capacity goes through traces on the board.
I design electronic equipment in the US and what I see from Chinese manufacturers are products that may look and perform similar to name brand products, but are usually not the same quality. Many factors go into shaving costs of design and manufacturing: poor wages, copying designs, minimal testing, counterfeit components, technology licensing, government subsidies, trade inequity, poor quality control and certification. It's easy to make something work, but much harder to make it work well and do it honestly. Usually this becomes apparent when you test multiple units picked randomly from production with a broader range of tests to see if they are consistent and reliable.
+bruce honnigford Let's admit it, the "brand" products are often so overengineered and bullied by the law that they aren't too good no matter how much money is thrown at them. These Chinese power supplies are ingenious in their own way and they would need little to become as good as their overpriced competition.
+bruce honnigford Yes, reliability is what the 80 odd dollars pay for. But the price factor is interesting in cases when reliability is less important. While chinese factories have made huge progress on the safety and sturdiness of their products, as a whole, getting a UL listing and having the possibility to track components can also be a determining factor. Design trade-offs are what our job is made of. Having the choice of cheap medium-quality components, that can also be over-sized a bit at low cost (that is always good for a power supply), is a big plus, I'd say.
Paul Frederick That's wrong. Sony's aggressive contracts conquer all. It has little to do with price. I honestly wish for Sony to go under. They have damaged the market for us all.
The chinese one is a flyback, not a forward. the 2 output diodes are in parallel and the inductor is extra output filtering, that's why it seems quite small. The Cosel has no extra output filtering. The Cosel does have 2 stages of input filtering compared to the chinese one's single stage. The design of the Cosel seems quite strange. They probably could have moved the layout around to get the tracks shorter on the secondary and not needed those bus bars and added some small post filtering. It's strange they didn't add extra filtering since they added those ferrite beads everywhere. As someone mentioned, the input bus cap being 200V doesn't fit with a typical boost PFC or a design without a PFC, so they may have done a different PFC topology, presumably because they want PFC but a single switch forward or flyback doesn't cope well with the high bus voltage of boost PFC preregulators. Buck PFC doesn't 100% fit in with the 200V cap either since the output has to be lower than the lowest input voltage, but they are probably using a controller that tracks the output voltage to input voltage. Buck PFCs aren't brilliant though and I've seen plenty of single switch or 2 switch forwards following a boost PFC, the single switch ones manage it with an RCD soft voltage clamp on the drain. I've designed 2 switch forwards that run after a boost PFC and that seems simpler than a buck PFC. The diodes/transistors mounded to the side of the case is just as valid as mounting on the base since its the same bent piece of metal and that could make for a better layout, so it's strange the Cosel didn't do that at least on the secondary to get the high current secondary circuit smaller. The ferrite beads on the secondary diodes is also unexpected, since they're adding extra inductance in series with the leakage inductance which is a source of ringing with the diode capacitance, so they might be just lowering the ringing frequency, whereas the more typical RC snubbers actually snub this noise. The chinese supply did use plastic TO-220 packages and I wonder if those junctions get too hot, especially on the secondary and that might be just to save cost. I got a similar supply from ebay to look inside a while ago and I too was pleasantly suprised by the construction. I was expecting no input filtering or protection and insufficient creepage and clearance distances, but this was not the case. You could test the step response to load as well as power factor for a more detailed test.
You are right, at 11:36 its clear thet the the 2 anodes of the diodes are connected so they are in parallel to share the current and reduce the losses...
The video could be shorter. Interesting though, the big one is doing it wrong, just on price/perf. But it's much worse than that. 24V 2A is nothing. A super tiny circuit could do that and well. Custom busbars for 2A?? let's call that massive over engineering. Beautiful, if it wasn't so incompetent.
I was amazed by those beefy bus bars too, but maybe it has something to do with proper output current measurement considering that thick shunt before those bus bars. Seems to me they use the same design and hardware for higher output currents too, hence the huge 680 uF cap, unreasonably capable (for 2A) power MosFet 0.5 Ohm, unpopulated pads for bigger Shottky diode at the output, huge shunt and bus-bars.
I think those rails you mentioned are a result of a design restriction, since they decided to bend the MOSFET and output diode towards the bottom plate, loosing quite a bit of PCB real estate. It could be interesting to hear the reasoning behind this from the engineers that designed it.
+Kallenator1988 Maybe, but the PCB is still larger than its competitor and they could have squeezed on more if needed. I think they purposefully overengineered it.
The Japanese PSU has active Power Factor Correction (PFC). Did you not notice that the large high voltage capacitor was only 200 volt rated and that the controller IC had 24 pins? The extra "stuff" on the board is PFC control and cycle by cycle control of both output voltage and current at the switchmode frequency. There are 4 (four) transformers, not including the dual common mode input chokes and the output choke on the PCB. That should have alerted you to the superior design and quality of the Japanese unit. Unfortunately you compared a modern car to a horse and cart. Please Google "power factor" and learn a bit more about it before recommending cheap Chinese SMPS PSUs that do not have active or even passive PFC. Wikipedia has a very good article on the subject and with modern smart power meters, you get to pay for the wasted power.
You need to test booth with a active programmable load. I know programmable loads are very expensive... I did notice that the switching frequency seemed higher on the Japanese one. THANKS!!!
I notice that the diodes and many of the resistors are standing off from the PCB. Nice touch. I suppose for cooing?. Shouldn't you be using a 10X probe for the ripple measurements though?
Why is the 'bus' capacitor in the Cosal only rated at 200v when at the maximum rated input of 265 volts AC, it will have 375v across it? I was waiting for the bang...
Whilst this video is interesting, as some have mentioned, it is not a true reflection of how it would be used at full load or under actual usage conditions in it's target application. The Cosel unit is more expensive, but what this video doesn't tell you is that the PBA50F-24 unit has an efficiency rating of 84%, which the China one won't have, optional circuity to allow remote on/off, AC3000V isolation, wide operating temperature, OVP, OCP, 5 year warranty, long life etc etc. All this comes at a cost. Additionally the Mouser price is an inflated price for small volumes. Manufacturing price is much lower. Also, the Cosel unit is designed for Industrial applications and therefore for an amateur home use project, is way over spec'd. Whilst the China product is very cheap, as with most things, the cheapest solution is usually the poorest quality. This means it would have been designed with very low cost components and PCB material (Cosel uses FR4) and will have a very short warranty as well as may not even pass the necessary global EMC requirements. Cosel is known in the world as having the highest quality power supplies backed up with facts and figures so whilst this is an interesting video, it's hardly fair to say that comparing the price of a Ferrari against a Fiat makes the Fiat as good!
What if the customer wants an 80% efficiency and no optional circuitry for 1/10th the price? High end devices are very nice but not every customer needs or wants to pay for that bundle of features. Companies that make what they want rather than what the customer wants imperil their futures. I could have paid 10x for a beautifully made billet block engine that makes 2x the power for my pickup truck, but I stuck with the normal output OEM one. I did not want or need the benefits the 10x engine provides.
Why dose the Cosel Power Supply have a 200 volt rated bus capacitor on rectified 264 volt AC? (correct me if I'm wrong but a bridge rectifier multiples the voltage by 1.414 converting it to DC) So 264 volt AC * 1.414 = 373 volt DC witch is a large over volt for that 200 volt capacitor
Excellent review, I have just received some 5 amp 12 volt power supplies that look very similar to the 24 volt Chinese one you reviewed and from wiring it up with 240 AC, it seems to work OK. Small trimpot adjusts the output voltage with no problems. Sad to say my now ancient oscilloscope would probably not work any longer as it has not been turned on for years. One question, would it be worth installing a cooling fan for a power supply of this type ?
Very instructive video, specially for whoever is studying SMPS, altogether you should put a caveat in the end (nowadays UA-cam has this option, to add content to an already uploaded video), stating that as you have not tested both power supply regarding "real world" loads, the bottom line could be misleading, as it seems to make the case that the Chinese power supply is an overall more cost benefit solution. But a good video anyway, keep 'em foming!
This video is very informative, i'm going to buy a similar chinese power supply (12v 5A) for a little homemade NAS based on a Banana PI, i thought that for 8-9$ they were crappy inside but turns out they are quite good it seems! The sanded DIP chip is kinda shady though...i would really like to see what IC they are using there!
Nice review, but I am wondering how well Over Voltage Protection, Current Limit and Short Circuit protection was implemented. It looks like the Cosel has remote V Sense too which is not bad for a 2A converter.
I have a bunch of low-end Cosel power supply (Cosel is crazy expensive). The ripple measurement of SMPS is a bit tricky. You can not use ground clip as it will introduce lots of noise. The test point is best at output capacitor. If you did it correctly, the Cosel should be much better.
The Chinese one was very surprising. Construction was a bit shoddy, but that can be tidied up with a bit of spare time and a soldering iron. I'd be tempted to replace the caps with decent ones if it were going to be used for any extended period so add a few dollars extra for the caps and some TLC to tidy it up and it'd be a decent unit.
The two supplies performed well to my believe although the cheap power supply did have some more ripple on the output. Comparing the price paid for the cheap psu I would not even dream of making it for that price.($8) You get what you pay for but the cheaper supply is really good value for money. Compared the cost is 10 times more for the more reliable one. One other thing that puzzled me was the voltage rating of the 680uf capacitor being only 200v when the mains voltage at 230v is rectified will rise up to more than 300vdc.
The China PSU was and is a basic rendition without too much complexity such as the Japanese PSU. However, I would have liked to have learnt what improvement could have been made to the China PSU. Such as inducter and larger caps would have on regulation. I noticed a drop on the voltage toward the higher end, that does also effect amps of the unit, but where the loss is/was causing this. I assume the MOSFET and caps would be the issue? But for the cost of a few replacement parts which is best value, caps and part being only a few dollars does the brand warrant the costs?.
I have been thinking about buying one of these inexpensive Chinese PSUs. Has there been any test of time experiments done with this comparison since this video was made?
Why the big capacitor is only 200V?? Anyway, the more expensive shows so much ripple so I would say it's a shame. I don't event understand how they managed to do so. They used much more expensive components everywhere, yet the result.. you saw it. So I would say the Chinese one is more prospective. Replace caps, transistors and it is a quite good for the price. Of course Japanise one insures more confidence in reliability, no doubts, but.. I expected more for the price and schematics.
I agree, if that capacitor is straight across bridge rectified mains at the maximum rated voltage, 265V AC, it's going to see 375V DC on no load. Something wrong here. But see comments below. IMHO the expensive supply is over engineered yet is no better performance than the cheap one. However it may last longer and be more rugged. Engineer - someone who can do for £1 what any fool can do for £5.
You should be testing them on load to see how they hold up to specs. Also it depends what your going to use them for too. This was not a real test only a input test and not a output test on full load to see how they held up. It would be nice to see how hot they got with a flir gun to see points inside what got hot. There was a lot of things missing in that test. Sorry not how I would of done it.
Looking at the input capacitor, the cosel supply has a 200V rated voltage. I have learned always that after the bridge rectifier and an elco the voltage will be Vin x 1.41 so this is 230 x 1.41 = 324 volts... So how did that happen...
Bought a couple of the 24V supplies from Ebay and after a short time a cap failed and I had 27V ac leakage coming out of the DC supply but still giving 24V dc. Buyer beware.
All of my cheap ebay china power supplies are working very well. I only load them no more than 50% of Max current. When they arrive from china I tear them down, look for any burnt components. Look at all of the mounted to chassis components to see if there is proper heat sink compound on them. If no compound I unscrew them and add the compound. To decrease ripple and noise I put a 4,700uF, 100V capacitor on the output. One last thing. If buying from ebay. Make sure you set up an account. If the suppliers have a Positive feedback of over 99% rating. I will buy. If its faulty you are guaranteed ebay money back policy. And with a rating of 99% and over means the customers are very happy and will trust them.
dell177 hello you seem to be very knowledgeable about adapters. I need a little help I have LED lights that the adapter keeps burning out the specs are input: 100-240v-1.5a and the output 24v 6A what should I get to hold better. The LED are 120 ft long. I will appreciate any advice
I think its time to stop buying branded parts as you will be paying for brand names, as far as Chinese products well they learnt to work like others and can produce good quality items with the lowest prices ever. Looking at a lot of items nowadays they are made in China and are from famous brand names
Well the cheap Chinese brand cap's are likely to fail that is a given. That said it is durability that would interest me first. Second would be where the birdies are since I primarily use switching power supplies for rf radio hobby. Given the low cost for used server power supplies I find it odd that anyone would spend $85 for a Cosel when you could get 2-4 server supplies for that price and they are built much better. Even if you had to recap them you would still be ahead over buying a Cosel. Same thing for the cheap Chinese $8 switcher you could upgrade the caps fairly cheaper and make minor improvements if need be for far less than the $85 Cosel. In fact even if needed a small fan to keep it cool if that is Chinese as well it is cheap enough. I think where the Cosel comes in is if you are building something for a client and your reputation is on the line!
I'd liked to see the construction of the transformers. Even if a PSU works fine, it can still kill you if the transformer construction is questionable.
For audio I would use the 8 dollar cheap one, less noise and ripple, in fact I would buy 10 and see if the Japanese model last as long as all 10 of the cheapies being it's the same price difference. I'd bet the components don't have the same quality, but the cheapie looks good on the scope, explain that? When the cheap one craps out you could always replace components with high quality, if the board is still good then you're all set. Maybe the Japanese model is a little over done. They're known for high quality and honor, but seemed to overlook the noise and ripple, hmmm. Electrical Pollution.
Why the jeanease one had a 200 v main cap?It sould be 400 or 450 v also usually the x2 capacitor must be rated for 275 v not 250 v but this ine could not be a big deal.
Thanks for the comments. I was getting a lot of questions on the forums to compare a more reasonably priced but good quality SMPS, so I've done a review for a $18 Delta SMPS here: ua-cam.com/video/QWD1Rb55h-k/v-deo.html
Hey! I have a core 2 duo processor,4 gb ddr3 RAM, 1TB HDD,1 optical drive and nVIDIA GT 730 graphic card which requires 49W power consumption. But, my power supply unit (smps) is very cheap around 8$,but provides 450W supply.So,would a cheap power supply like this be good for my dedicated GPU ?
first time i plugged my ebay 48v power supply from china thunders came out from switching transistors and burned of one of their legs. i wonder what was wrong in the circuit. Didn't see any obvious fails of other components
It also a question of cost when power supply fails not in voltage lost mode... If you supply some pricey equpiement by $8 it quickly can be very costly. Also about ripples you didn't show the frequency of them, peak peak voltage is not whole story.
If voltage regulated by tl431 ic you can change a resistor around it and output capacitor with higher voltage but you should learn more about electronics to change the output voltage
***** Of course, but voltage is hard enough in the mains so that you can just multiply by it as a constant. Having a regular meter with true RMS is sufficient, if less user-friendly. And $15 will not buy you any good instrument, the values will be way off, especially with poor power factor!
Co do czasu wytrzymałości współcześnie produkowanych markowych sprzętów nie jestem pewien. Dla przykładu telewizor z 2016 roku ma być wymieniony po 2-ch latach i ma nie nadawać się do użytku po tym okresie. Mówię tu o drogich modelach znanych marek. drogich, znaczy z najwyższej półki jakości. :) Ale video fajne :)
I'm Probably making a fool of my self as I have not watched beyond 13 min yet. If that is what a cheep PSU is supposed to look like, I realy got ripped off. chip, what chip. Try a two transistor JK flip-flop with one transistor being the driver and the other a tiny smt thing. What you have there looks incredible compared to what I got. The 5V 6A unit I have blew up on the bench with only a 100mA load.
I absolutely LOVE cheap chinese electronics. it means I can order loads of stuff from ebay and banggood and aliexpress etc and play with it. if china hadnt revolutionised manufacturing I couldnt afford my hobby.
I agree, I couldn't afford to play hobby electronics otherwise.
I think a more fair "teardown" would be to compare Apples to Apples instead of Apples to Oranges. Spend either $8 or $85 from BOTH China and Japan and see what you get.
An $8 24 volt power supply from Japan gets you what? The empty box it comes in?
Then compare the $8 Japanese PSU to an $85 Chinese PSU.
Next compare two new automobiles where one costs $20,000 and the other costs $212,500. Feel free to mix or match the countries of origin.
Just saying, the comparison was flawed from the beginning. If you want to compare value then start with two different products that cost the same. The only exception would be to purchase two identical items at different prices, that makes the maths for "value" very easy to calculate.
The fact the $8 Chinese outperformed the $85 Japanese unit in ripple, so what's left for the Japanese to be smug about? Do they really want to humiliate themselves against a $85 Chinese unit?
Of course, I won't use a $8 PSU for my $2000 PC, but for DIY hobbyist use it's is more than good enough.
@@nomoredamnnamestouse It's great for hobby use and nothing more. Smug? Interesting choice of words. Are you a Chinese nationalist or something?
@@visionstills3700 well you get what you pay for but keep in mind that China had swallowed the world in the electronic industry and mostly all the household goods electronics are all made in China nevertheless a lot of famous firms have factories in China so that they keep the price of their goods compatible.
For 8 dollars you get a choke tranformer, rtc's for inrush current limiting, tinned pcb tracks and even pcb gaps and a optocoupler. I have to say for 8 dollars its pretty decent.
Dantali0n until it burns your house down, or fails; that logic does not apply to power electronics, safety does not have a price
@@FooBar89 Just plug the cheepy unit into a MCB and your house will be fine!
@@davidteale4270 Please wat's a MCB?
MCB stands for - Miniature circuit breaker
@@josearrighi "wat" ? Lol
I worked for a power supply company for 30 years and we went out of business because we just could compete with the imported supplies. Up to the early 80's most of our business was with the military but as that dried up we did more and more custom commercial work.
I taken a few Chinese supplies apart and been impressed at the quality vs cost. I've seen some that would not pass UL requirements but they all worked well.
dell177 hello you seem to be very knowledgeable about adapters. I need a little help I have LED lights that the adapter keeps burning out the specs are input: 100-240v-1.5a and the output 24v 6A what should I get to hold better. The LED are 120 ft long. I will appreciate any advice
@@betopizarro2003 Your LED string is probably too long and drawing too much power. How many LEDs are on the string (How are they broken up into sets of series LEDs)? How much power is each LED drawing and at what voltage? If they are 1.8v LEDs that means 12 LEDs in series gives about 24v. So you would want your lights to be broken into sets of 12 in series and a bunch of these sets in parallel. If you are using Christmas lights they may all be in series instead of being broken into sets of parallel LEDs. Christmas lights run at mains voltage so they have to drop down 120-240v So you can put a bunch of them in a row. As long as your LEDs are not burning out they are probably configured correctly or they would either burn out or be dimmer than expected.
Then you need to know how many of milliamps each LED is drawing and multiply by the voltage to get the watt usage. So subtract 148W (24v X 6A) from the watts the LEDs are using (# of LEDs X milliwatts). If you have a positive number that is what your LEDs are over capacity. If you get a negative number everything is ok. Or subtract the watts the LEDs are using from 148 in which gives you how much more power your adapter can supply than what the LEDs are using.
If I see how our business is going, I must say. It is not cheap! But more regular price then overpriced like second one with huge margins.. Where 30% price is only because of old suits man:) Chinese are good in optimal balance between price and quality, these days. However its not always True.
The Japanese made was almost a work of art. Looks so durable.
Its seems like whoever crunched the numbers while designing the $8 PSU did a great job. For 1/10th the price you really can't go wrong... Super-educational video, thank you for sharing! :)
Not totally impossible they have ripped it of another expensive one or they just used the schematic example for the controller ic, very common in china to do, the cheaper the better no matter if its safe or not.
And there is one thing on the boards bottom side that wasn't noticed. There was one component leg that was folded very badly so it was very very close to touch another bare trace which it could one day and totally ruin the entire PSU and even start a fire worst case, that's what you get if china made not in fex a Cosel.
And the reason these are so cheap is just because the components are made in china so the just walk over to their neighbor, which is the component factory next door, and get extremely cheap components not even heard of outside china. You can find very very cheap components on fex Ebay.
But if you the other hand buy them directly from the factories you would not belive the prices you get then since they are much much cheaper.
And the other reason its so cheap is because the poor chinese workers hardly get any pay for their job.
It depends on how you define "go wrong". If the $8 power supply fails and in doing so, damages a $1000 piece of equipment, that seems like it's going wrong. If the $8 PSU fails and holds up a core function of a business, either in production output, client loss, or just idle employee salaries, thousands of dollars can be lost. The Chinese PSU is "acceptable" for some consumer users. It is a reason to fire whoever selects it for a business application that fails.
@@stinkycheese804If you can afford to by $1000 equipment, then you can afford the expensive SMPS unit! Some of us old pensioner are on a budget here in GB. And just use these cheep SMPS plugged into RCCB's here in Britain! Not burnt my house down yet, using quite a few of these cheep units, and still alive and kicking at this moment with no big bangs!
@@davidteale4270 Please, wat's a RCCB?
@@josearrighi RCCB means Residual Current Circuit Breaker -- A residual. current device (RCD), or residual current. circuit breaker (RCCB), is an electrical wiring device that. disconnects a circuit whenever it detects that the electric. current is not balanced between the phase ("hot") conductor.
Simple decision really. If you are powering something worth thousands of dollars or otherwise critical, buy the cosel one and you will not destroy the expensive load after a lightning strike or something like that.
If you are powering some christmas lights around the workshop for fun, buy 3 of the chinese one, check each one before use for loose joints and stuff and make sure the grounding is proper before plugging it in... when it blows install a new one and, most fun part, debug/repair the fault (yay!) :-)
These are not really competing products at all, both do their intended jobs very well.
But its still nice to see a cheap supply that anyone who knows electronics would buy for any purpose.
The Cosel PSUs output ripple seems "dirtier", but most probably it's because dithering was applied purposefully. Dithering will reduce the EMI from the device and it's actually a good practice, overall.
+DigGil3 It seemed like a good practice, but today, after most chips got the feature, its dark side has emerged: The supplies are noisy in the audible spectrum. I'd choose a "traditional" switching unit that I wouldn't hear at night any time over these "EMI-friendly" wall-warts.
The real concern is about safety. As far as I can see, the Chinese SMPS has a standart 1nF 1kV suppression condenser between primary and secondary side instead of a safety self-healing one. This is a death trap, because it tends to fail dead short at overvoltage condition. Short kV spikes are quite common in mains - tens to hundreds a day - so if you have conected anything referenced to ground (even capacitive coupling is enough) to the output of such SMPS, the cap breaks down, leaving your secondary circuit galvanically connected to Live wire. The SMPS does not stop working, everything seems fine, only the trap is armed and waiting for it´s prey.
The funny thing is that it seems that those Chinese "engineers" use proper safety caps in input filtering, where failure of one only leads to blown fuse and not lost live(s). They saved two cents on the wrong place.
Life saving rule: Take your Megger and Hi-Pot test every single mains connected device from China! The nice side effect: you will get less concerned about efficiency, output ripple and pulse response very soon.
I really liked this teardown because you took the time to really trace out the circuit. I feel like I basically know how it works. I also thought the Chinese board was worth every penny of the $8 it cost, and it had a lot of effort put into the design and it did have all the key parts. Those cheap electrolytic caps might give it a shortened lifespan - but you can replace em or replace the whole board several times and still be far ahead of what the Japanese board cost...(which, admittedly, is a work of art and I wish we lived in a world where everything was made to that level of quality)
Yes, I think it would be interesting to run both at (near) full load for a week or two, continuous, and see if they both still work! I wonder how well the isolation would hold up, too. We have 240V AC mains, here, and I am VERY wary of cheap PSUs. It's hard to see that the Cosel is ten times better (unless the cheaper one only runs for a few tens of hours...)
I do a lot of high power LED projects and I find the cheap Chinese power supplies are OK when not using at full capacity and also not running too hot. For hot, full-load applications, I used the brand names.
I have used Chinese power supplies in numerous commercial applications where cost was a concern. Stepping away from the basic Ebay units (which I have used successfully on home projects) in my business, I find the Mean Well brand to be particularly reliable. If I were working in a medical or scientific environment I would probably use a German or Japanese supplier but for typical commercial and light industrial applications the less well documented Chinese brands can work quite effectively at a significant cost savings.
Note: This video did not check for line voltage isolation which I always check on Chinese power products. Visual inspection doesn't protect you from faulty transformer isolation or suppression capacitors.
The ferrite beads are a nice touch. I'd prefer if they glued them down though, because I have a keen ear to any rattling going on inside of equipment, and have to make sure it's not something that could short out. The Cosel power supply doesn't cut any corners, and has lots of additional safety. But for 10 times the price, it's not worth it, unless if your equipment has to meet certain standards for use in certain fields of industry. The Cosel had a more noisy output, but it looked to be at a much higher frequency, and that would be easier to filter out than the cheap Chinese power supply (using a choke or capacitor) -- if your design permitted that... but then... you could always use a linear regulator too :-p
Yeah, I was looking at them just left there, no silicon to hold them. About cutting corners, I saw no MOVs, the fuse quality appears to be on the lower side.
I enjoyed your video. you gave both a fair and honest assessment. unlike some other "crappy" China supply reviews.
I think the cheap China made power supply would be fine for anything except aerospace, military, medical, nuke plants, offshore oil, refineries or chemical plants. In other words if human life was at stake. For longevity at $8 I would buy a spare so you would still have spent 5 times less for 2 and you would have a backup on the shelf just in case. The fact that the ripple is cleaner on the simpler is shocking. I have always admired Japanese workmanship but many Japanese engineering designs always seemed needlessly complex for some reason (cultural?)
Rob C, but you would not just have the $8 cost of the backup on the shelf. You would also have the down time (which can be much more expensive by itself) until someone comes and replaces it and the bill for doing so, or else you retain a technician on-site capable of it so an entire salary. It is safe to say that most uses where the function of the powered equipment is tied to productivity which is tied to profit, would benefit from buying a higher quality PSU. As always where is a middle ground here, silly humans often thing in terms of opposite extremes of expense and quality instead of the middle ground. In other words something can be better, even "good enough" without being the most expensive one you can find.
@@stinkycheese804 You're talking about downtime in commercial applications, where (I hope) nobody uses those cheapos. This SMPS is for those driving LED strips at home and not willing to pay 10x more.
And to confirm your theory with my personal experience, the cheap ones last no longer than an year before they fail, sometimes just a few months.
You know, I can hear the essence of Singapore in your voice :D
nice review though I'm pretty sure those bus bars were simply a necessity due to the diode being in the way of traces going through the PCB. there are other places that the came current capacity goes through traces on the board.
The chineye converter must be a flyback as a forward converter requires an second inductor. The secondary inductor is probably a filter.
I design electronic equipment in the US and what I see from Chinese manufacturers are products that may look and perform similar to name brand products, but are usually not the same quality. Many factors go into shaving costs of design and manufacturing: poor wages, copying designs, minimal testing, counterfeit components, technology licensing, government subsidies, trade inequity, poor quality control and certification. It's easy to make something work, but much harder to make it work well and do it honestly. Usually this becomes apparent when you test multiple units picked randomly from production with a broader range of tests to see if they are consistent and reliable.
+bruce honnigford Let's admit it, the "brand" products are often so overengineered and bullied by the law that they aren't too good no matter how much money is thrown at them. These Chinese power supplies are ingenious in their own way and they would need little to become as good as their overpriced competition.
+bruce honnigford But how often do we actually buy "quality electronics" that are just rebranded Chinese stuff!
+bruce honnigford Yes, reliability is what the 80 odd dollars pay for. But the price factor is interesting in cases when reliability is less important. While chinese factories have made huge progress on the safety and sturdiness of their products, as a whole, getting a UL listing and having the possibility to track components can also be a determining factor.
Design trade-offs are what our job is made of. Having the choice of cheap medium-quality components, that can also be over-sized a bit at low cost (that is always good for a power supply), is a big plus, I'd say.
Didn't anyone learn anything from the Betamax VHS war? It doesn't matter how much better one is over the other. Low price conquers all!
Paul Frederick
That's wrong. Sony's aggressive contracts conquer all. It has little to do with price. I honestly wish for Sony to go under. They have damaged the market for us all.
The chinese one is a flyback, not a forward. the 2 output diodes are in parallel and the inductor is extra output filtering, that's why it seems quite small. The Cosel has no extra output filtering. The Cosel does have 2 stages of input filtering compared to the chinese one's single stage.
The design of the Cosel seems quite strange. They probably could have moved the layout around to get the tracks shorter on the secondary and not needed those bus bars and added some small post filtering. It's strange they didn't add extra filtering since they added those ferrite beads everywhere. As someone mentioned, the input bus cap being 200V doesn't fit with a typical boost PFC or a design without a PFC, so they may have done a different PFC topology, presumably because they want PFC but a single switch forward or flyback doesn't cope well with the high bus voltage of boost PFC preregulators. Buck PFC doesn't 100% fit in with the 200V cap either since the output has to be lower than the lowest input voltage, but they are probably using a controller that tracks the output voltage to input voltage. Buck PFCs aren't brilliant though and I've seen plenty of single switch or 2 switch forwards following a boost PFC, the single switch ones manage it with an RCD soft voltage clamp on the drain. I've designed 2 switch forwards that run after a boost PFC and that seems simpler than a buck PFC. The diodes/transistors mounded to the side of the case is just as valid as mounting on the base since its the same bent piece of metal and that could make for a better layout, so it's strange the Cosel didn't do that at least on the secondary to get the high current secondary circuit smaller. The ferrite beads on the secondary diodes is also unexpected, since they're adding extra inductance in series with the leakage inductance which is a source of ringing with the diode capacitance, so they might be just lowering the ringing frequency, whereas the more typical RC snubbers actually snub this noise. The chinese supply did use plastic TO-220 packages and I wonder if those junctions get too hot, especially on the secondary and that might be just to save cost.
I got a similar supply from ebay to look inside a while ago and I too was pleasantly suprised by the construction. I was expecting no input filtering or protection and insufficient creepage and clearance distances, but this was not the case.
You could test the step response to load as well as power factor for a more detailed test.
You are right, at 11:36 its clear thet the the 2 anodes of the diodes are connected so they are in parallel to share the current and reduce the losses...
The video could be shorter.
Interesting though, the big one is doing it wrong, just on price/perf. But it's much worse than that. 24V 2A is nothing. A super tiny circuit could do that and well. Custom busbars for 2A?? let's call that massive over engineering. Beautiful, if it wasn't so incompetent.
I was amazed by those beefy bus bars too, but maybe it has something to do with proper output current measurement considering that thick shunt before those bus bars.
Seems to me they use the same design and hardware for higher output currents too, hence the huge 680 uF cap, unreasonably capable (for 2A) power MosFet 0.5 Ohm, unpopulated pads for bigger Shottky diode at the output, huge shunt and bus-bars.
Very educational and it's enjoyable to see a quality picture.
I think those rails you mentioned are a result of a design restriction, since they decided to bend the MOSFET and output diode towards the bottom plate, loosing quite a bit of PCB real estate. It could be interesting to hear the reasoning behind this from the engineers that designed it.
+Kallenator1988 Maybe, but the PCB is still larger than its competitor and they could have squeezed on more if needed. I think they purposefully overengineered it.
2015 was 8 years ago?. I remember watching this back in 2015 . For some reason got this on recommendation once again 😂
Thats far more than I expected for 8 bucks!
The Japanese PSU has active Power Factor Correction (PFC). Did you not notice that the large high voltage capacitor was only 200 volt rated and that the controller IC had 24 pins? The extra "stuff" on the board is PFC control and cycle by cycle control of both output voltage and current at the switchmode frequency. There are 4 (four) transformers, not including the dual common mode input chokes and the output choke on the PCB. That should have alerted you to the superior design and quality of the Japanese unit. Unfortunately you compared a modern car to a horse and cart. Please Google "power factor" and learn a bit more about it before recommending cheap Chinese SMPS PSUs that do not have active or even passive PFC.
Wikipedia has a very good article on the subject and with modern smart power meters, you get to pay for the wasted power.
You need to test booth with a active programmable load.
I know programmable loads are very expensive...
I did notice that the switching frequency seemed higher on the Japanese
one. THANKS!!!
Thank you for objective review.
I notice that the diodes and many of the resistors are standing off from the PCB. Nice touch. I suppose for cooing?. Shouldn't you be using a 10X probe for the ripple measurements though?
Why is the 'bus' capacitor in the Cosal only rated at 200v when at the maximum rated input of 265 volts AC, it will have 375v across it? I was waiting for the bang...
Alan Cordwell -- Your question was already answered here ua-cam.com/video/71NWT5vrbk8/v-deo.html&lc=z13pezwxnkq2u31vg231c3qpmuutipinw
Whilst this video is interesting, as some have mentioned, it is not a true reflection of how it would be used at full load or under actual usage conditions in it's target application. The Cosel unit is more expensive, but what this video doesn't tell you is that the PBA50F-24 unit has an efficiency rating of 84%, which the China one won't have, optional circuity to allow remote on/off, AC3000V isolation, wide operating temperature, OVP, OCP, 5 year warranty, long life etc etc. All this comes at a cost. Additionally the Mouser price is an inflated price for small volumes. Manufacturing price is much lower. Also, the Cosel unit is designed for Industrial applications and therefore for an amateur home use project, is way over spec'd. Whilst the China product is very cheap, as with most things, the cheapest solution is usually the poorest quality. This means it would have been designed with very low cost components and PCB material (Cosel uses FR4) and will have a very short warranty as well as may not even pass the necessary global EMC requirements. Cosel is known in the world as having the highest quality power supplies backed up with facts and figures so whilst this is an interesting video, it's hardly fair to say that comparing the price of a Ferrari against a Fiat makes the Fiat as good!
+Nick Theodoris That's excellent feedback, thanks for sharing! In the future will be keeping these in mind; it's great learning together :)
What if the customer wants an 80% efficiency and no optional circuitry for 1/10th the price? High end devices are very nice but not every customer needs or wants to pay for that bundle of features. Companies that make what they want rather than what the customer wants imperil their futures.
I could have paid 10x for a beautifully made billet block engine that makes 2x the power for my pickup truck, but I stuck with the normal output OEM one. I did not want or need the benefits the 10x engine provides.
Why dose the Cosel Power Supply have a 200 volt rated bus capacitor on rectified 264 volt AC?
(correct me if I'm wrong but a bridge rectifier multiples the voltage by 1.414 converting it to DC)
So 264 volt AC * 1.414 = 373 volt DC witch is a large over volt for that 200 volt capacitor
Excellent review, I have just received some 5 amp 12 volt power supplies that look very similar to the 24 volt Chinese one you reviewed and from wiring it up with 240 AC, it seems to work OK. Small trimpot adjusts the output voltage with no problems. Sad to say my now ancient oscilloscope would probably not work any longer as it has not been turned on for years. One question, would it be worth installing a cooling fan for a power supply of this type ?
+loneoceans
What does the bus bar do? Why does the Cosel PCB have that plastic on top?
Very instructive video, specially for whoever is studying SMPS, altogether you should put a caveat in the end (nowadays UA-cam has this option, to add content to an already uploaded video), stating that as you have not tested both power supply regarding "real world" loads, the bottom line could be misleading, as it seems to make the case that the Chinese power supply is an overall more cost benefit solution. But a good video anyway, keep 'em foming!
This video is very informative, i'm going to buy a similar chinese power supply (12v 5A) for a little homemade NAS based on a Banana PI, i thought that for 8-9$ they were crappy inside but turns out they are quite good it seems! The sanded DIP chip is kinda shady though...i would really like to see what IC they are using there!
Excellent methodology of testing overall excellent.
Very interesting and comprehensive test, thank you.
He uses screwdriver as a pointer.
*I trust him*
Nice review, but I am wondering how well Over Voltage Protection, Current Limit and Short Circuit protection was implemented.
It looks like the Cosel has remote V Sense too which is not bad for a 2A converter.
I have a bunch of low-end Cosel power supply (Cosel is crazy expensive). The ripple measurement of SMPS is a bit tricky. You can not use ground clip as it will introduce lots of noise. The test point is best at output capacitor. If you did it correctly, the Cosel should be much better.
Very nice video but I would love to see you abuse the hell out of these supplies to destruction like Joe Smiths videos on multimeters.
"RTC" is for inrush current limiting
+mikeselectricstuff Rubbish temperature coefficient!
+Michael Fearnley On the contrary: Reasonable Temperature Coefficient ;)
Or NTC
The Chinese one was very surprising. Construction was a bit shoddy, but that can be tidied up with a bit of spare time and a soldering iron. I'd be tempted to replace the caps with decent ones if it were going to be used for any extended period so add a few dollars extra for the caps and some TLC to tidy it up and it'd be a decent unit.
The two supplies performed well to my believe although the cheap power supply did have some more ripple on the output. Comparing the price paid for the cheap psu I would not even dream of making it for that price.($8) You get what you pay for but the cheaper supply is really good value for money. Compared the cost is 10 times more for the more reliable one. One other thing that puzzled me was the voltage rating of the 680uf capacitor being only 200v when the mains voltage at 230v is rectified will rise up to more than 300vdc.
I think I could get 10 of the Chinese supplies, and a few beers for what the Japanese one cost. Although the Japanese one looked real nice inside.
How does a cheap $8 PSU compares to a branded Chinese one by Meanwell or Delta for $50?
The China PSU was and is a basic rendition without too much complexity such as the Japanese PSU.
However, I would have liked to have learnt what improvement could have been made to the China PSU.
Such as inducter and larger caps would have on regulation.
I noticed a drop on the voltage toward the higher end, that does also effect amps of the unit, but where the loss is/was causing this. I assume the MOSFET and caps would be the issue?
But for the cost of a few replacement parts which is best value, caps and part being only a few dollars does the brand warrant the costs?.
nice comparison, very informative
I have been thinking about buying one of these inexpensive Chinese PSUs.
Has there been any test of time experiments done with this comparison since this video was made?
Why is the main capacitor on the Cosel only rated at 200V? Surely that isn't high enough for rectified mains.
You sound Singaporean. :P
Anyway, that China PSU board looks like it could be phenolic (FR3) rather than 'lower cost' fiberglass (FR4).
What would happen if you supply a voltage between 125 and 190 V on the Chinese one?
It would probably just work fine...
It will still work
Very instructive video.
Why the big capacitor is only 200V?? Anyway, the more expensive shows so much ripple so I would say it's a shame. I don't event understand how they managed to do so. They used much more expensive components everywhere, yet the result.. you saw it. So I would say the Chinese one is more prospective. Replace caps, transistors and it is a quite good for the price. Of course Japanise one insures more confidence in reliability, no doubts, but.. I expected more for the price and schematics.
I agree, if that capacitor is straight across bridge rectified mains at the maximum rated voltage, 265V AC, it's going to see 375V DC on no load. Something wrong here. But see comments below.
IMHO the expensive supply is over engineered yet is no better performance than the cheap one. However it may last longer and be more rugged.
Engineer - someone who can do for £1 what any fool can do for £5.
There is another interesting parameter - conducted emissions into the input power lines
You should be testing them on load to see how they hold up to specs. Also it depends what your going to use them for too. This was not a real test only a input test and not a output test on full load to see how they held up. It would be nice to see how hot they got with a flir gun to see points inside what got hot. There was a lot of things missing in that test. Sorry not how I would of done it.
Michael Mitchell - let us know when you post your superior test video.
Even to add Good Quality Nichicon,Rubycon,Elna ,or other Quality Capacitors you still are not going to spend much to UPGRADE the unit.
It's not important, but I am curious what's the controller chip in the smaller one. I like it's simplicity and I would like to study it more :)
It is most likely uc3842 or its copy or something works similar, a very good peak current mode controller
Looking at the input capacitor, the cosel supply has a 200V rated voltage. I have learned always that after the bridge rectifier and an elco the voltage will be Vin x 1.41 so this is 230 x 1.41 = 324 volts... So how did that happen...
Bought a couple of the 24V supplies from Ebay and after a short time a cap failed and I had 27V ac leakage coming out of the DC supply but still giving 24V dc. Buyer beware.
All of my cheap ebay china power supplies are working very well. I only load them no more than 50% of Max current. When they arrive from china I tear them down, look for any burnt components. Look at all of the mounted to chassis components to see if there is proper heat sink compound on them. If no compound I unscrew them and add the compound. To decrease ripple and noise I put a 4,700uF, 100V capacitor on the output. One last thing. If buying from ebay. Make sure you set up an account. If the suppliers have a Positive feedback of over 99% rating. I will buy. If its faulty you are guaranteed ebay money back policy. And with a rating of 99% and over means the customers are very happy and will trust them.
dell177 hello you seem to be very knowledgeable about adapters. I need a little help I have LED lights that the adapter keeps burning out the specs are input: 100-240v-1.5a and the output 24v 6A what should I get to hold better. The LED are 120 ft long. I will appreciate any advice
I think its time to stop buying branded parts as you will be paying for brand names, as far as Chinese products well they learnt to work like others and can produce good quality items with the lowest prices ever. Looking at a lot of items nowadays they are made in China and are from famous brand names
love your vids dude. btw how old are ya?
Proves that you do not always get what you pay for!
I'm sad your follow-up video stated in the description leads to : "This video is unavailable." is it just a bad link ?
Thanks! I've fixed the link. Appeared to have been truncated.
Well the cheap Chinese brand cap's are likely to fail that is a given. That said it is durability that would interest me first. Second would be where the birdies are since I primarily use switching power supplies for rf radio hobby. Given the low cost for used server power supplies I find it odd that anyone would spend $85 for a Cosel when you could get 2-4 server supplies for that price and they are built much better. Even if you had to recap them you would still be ahead over buying a Cosel. Same thing for the cheap Chinese $8 switcher you could upgrade the caps fairly cheaper and make minor improvements if need be for far less than the $85 Cosel. In fact even if needed a small fan to keep it cool if that is Chinese as well it is cheap enough. I think where the Cosel comes in is if you are building something for a client and your reputation is on the line!
That was a surprise.
I'd liked to see the construction of the transformers. Even if a PSU works fine, it can still kill you if the transformer construction is questionable.
For audio I would use the 8 dollar cheap one, less noise and ripple, in fact I would buy 10 and see if the Japanese model last as long as all 10 of the cheapies being it's the same price difference. I'd bet the components don't have the same quality, but the cheapie looks good on the scope, explain that? When the cheap one craps out you could always replace components with high quality, if the board is still good then you're all set. Maybe the Japanese model is a little over done. They're known for high quality and honor, but seemed to overlook the noise and ripple, hmmm. Electrical Pollution.
Why the jeanease one had a 200 v main cap?It sould be 400 or 450 v also usually the x2 capacitor must be rated for 275 v not 250 v but this ine could not be a big deal.
I bought the exact same China made power supply on ebay for $9. I enjoyed the video!!
Thanks for the comments. I was getting a lot of questions on the forums to compare a more reasonably priced but good quality SMPS, so I've done a review for a $18 Delta SMPS here: ua-cam.com/video/QWD1Rb55h-k/v-deo.html
Great Video!
"Mean Well" makes rather inexpensive (20eur-ish for 24v 2a) power supplies like this that are for all I know reasonable quality.
Be aware that there are many fake mean well out there
So the Cosel unit does not have thermal paste between the active components and the heat sink back? That's strange!
+SteelBlueVision Maybe there were silpads and he just set them aside during disassembly.
Why would they go through the trouble of removing the part number for the IC?
great video.
2 amps thats hardly worth getting a switch mode supply .
Does the Cosel supply have 2 optocouplers?
Yes, looks like one for voltage regulation the other for protection
Interesting and surprising
good job, thanks
Hey!
I have a core 2 duo processor,4 gb ddr3 RAM, 1TB HDD,1 optical drive and nVIDIA GT 730 graphic card which requires 49W power consumption.
But, my power supply unit (smps) is very cheap around 8$,but provides 450W supply.So,would a cheap power supply like this be good for my dedicated GPU ?
first time i plugged my ebay 48v power supply from china thunders came out from switching transistors and burned of one of their legs. i wonder what was wrong in the circuit. Didn't see any obvious fails of other components
mateusz1945 -- Maybe you were running it on 220/240 volts when it was set up for 110/120 volt operation?
CE made in China ? the bigger power supply is marked CE! China Export?
+Todd Hagist It's confusing but there's a difference! www.icomuk.co.uk/News_Article/3794/17103/
thanks
Todd Hagist Yes, CE has two meanings
It also a question of cost when power supply fails not in voltage lost mode... If you supply some pricey equpiement by $8 it quickly can be very costly.
Also about ripples you didn't show the frequency of them, peak peak voltage is not whole story.
Have you load tested these ?
i would have the $8 one no heat paste on dear one ,cheap one will do job its not art work
I have a SMPS led driver with a 5v ref, could the voltage divider be modified to increase output?
If voltage regulated by tl431 ic you can change a resistor around it and output capacitor with higher voltage but you should learn more about electronics to change the output voltage
There is nothing to complain about USD 8,00 power supply. It is meant for people who cannot afford.
Chinese make any quality tier of PSU for corresponding price but japanese can only make the expensive tier.
how about efficiency?
+tomato totato Hard to measure, you'd need a true-RMS amp meter to account for the distorted consumption of the PSU.
zwz • zdenek rms meter cant ,mesure watt. you need watt meter. $15 at ebay.
***** Of course, but voltage is hard enough in the mains so that you can just multiply by it as a constant. Having a regular meter with true RMS is sufficient, if less user-friendly. And $15 will not buy you any good instrument, the values will be way off, especially with poor power factor!
Exactly, over say 10 years efficiency will be a major factor.
85 $ and it's still a PWM PSU ....don't wanna know the cost of a Linear one
Co do czasu wytrzymałości współcześnie produkowanych markowych sprzętów nie jestem pewien. Dla przykładu telewizor z 2016 roku ma być wymieniony po 2-ch latach i ma nie nadawać się do użytku po tym okresie. Mówię tu o drogich modelach znanych marek. drogich, znaczy z najwyższej półki jakości. :) Ale video fajne :)
I'm Probably making a fool of my self as I have not watched beyond 13 min yet. If that is what a cheep PSU is supposed to look like, I realy got ripped off. chip, what chip. Try a two transistor JK flip-flop with one transistor being the driver and the other a tiny smt thing. What you have there looks incredible compared to what I got. The 5V 6A unit I have blew up on the bench with only a 100mA load.
Nice!!
It seems like the cheap power supply was well designed but carelessly built.
The Cosel was very well thought out, though! Much better!
Sometimes people cant afford 80.00 and if a 8.00 job does the job fine people what they are buying.
Chinese transformer might be CCA windings.
china good ideal for cut pcb but japan good use bus conductor