Cheap bench power supply tear-down (it's bad!)
Вставка
- Опубліковано 5 жов 2024
- Although I've been using this cheap MCH-K305D 30V, 5A 150W switching bench supply for quite a while, the current limiting has never worked so I figured I'd tear it down and take a look inside. It's awful - but what do you expect for US$60 I guess?
Stay tuned because eventually I will trace the fault and repair this unit.
> Stay tuned because eventually I will trace the fault and repair this unit
so....did you?
I'm not an Electrician or anything but i do find these Video's really interesting due to the Way Bruce explains things in way that regular folk can understand.
You really don't know what you are talking about. An adjustable switching power supply is by no means trivial.
The build quality of the pcb is actually quite good for such a cheap device, all the IC seem to be made by premium manufacturers. The trim pots are multiturn variants which allows for repeatable adjustment.
The mechanical work is of course dodgy - but Wat do you expect from something as cheap as this?
+bongolongoallthetime He has no idea what hes on about, hes basically seen EEVBlog do a PSU teardown, and attempted to copy him with his own commentary...
Oh look, a component with an adjustment, it must be a pot. (The blue multiturn trimmers ARE pots, however the rotary encoders are definitely not)
I don't think he found anything to actually complain about...rather he just wanted a sensational title for the video, and to seem like he knows what he is on about...
I especially love how he didnt even check the caps, then about halfway through he realizes they might still hold a kick...Then proceeds to not discharge them STILL and continue manhandling it!
+Tyson Popynick LOL... I've been a professional electronics engineer for over 40 years -- but hey, if you say I have no idea what I'm on about then you must be right :-)
+xjet Are you really a professional electronics engineer? Or maybe just a plain engineer/hobbyist? I ask because you said in the video you did not know what a TL084 is! Really!?!? That would be like a Star Wars fan not knowing who Darth Vader is! After the uA741 op amp the TLC070/TLC080 series of op amps must be the most known and common ones! If you were 20 years old I might just believe you never saw one at uni but I really would doubt it, and would most certainly call BS for someone over 20! Maybe it was elecrical engineering and not electronics?
Judging people's competence based on their static/learned knowledge overlap to your own is not a good guide. Given the massive diversity of parts available even from the 1980's and that many people specialise in to certain fields it's entirely not unusual that a majority of part codes will be foreign to most people in the industry. It's entirely plausible that he's never been involved in a project that needed such things. The more you specialise the more stupid/lacking-of-knowledge you appear to others.
+xjet probably has forgotten more than most people making comments have yet learned (and sadly I hate to say as you approach and pass 50 that becomes horribly true)
+Paul Daniels Nope. There was no "massive diversity of parts available" 40 years ago, at least not OpAmps, and the TL084 is what you would call a Jelly Bean part, common and popular world wide! It was impossible to go through college (anywhere) and not work with this TI opamp series several times, and you would find them in everything, from cheap radios to studio recording equipment, from industrial control boards to expensive reaserch laboratory equipment, everywhere! If you do not recognize a TL082 or a TL084 or a TL071 then you are NOT even an experienced electronics serviceman let alone an electronics engineer! And that's not the only clue on this video!
One thing I have noticed with these cheap ebay power supplies (I have had a few of them) is that they say for instance 30V 5A which makes you thing it will put out 150W but they do not. They will usually only put out about 3-4V @ 5A which is only about 17W. BIG difference in output power. At 30V you will only get about 0.5A before it will start to limit the voltage. So what they mean is that they are 30V OR 5A at an unknown wattage. They can be useful for some things on a budget but don't trust the deceptive output.
Love to see how you track down the problem. Dave Jones style :)
+one4stevo he sounds more like a total Dave wannabe
he dose sound a bit like him too
Actually ,scary he really does ..the little nuances in expression ..
Nah, I think it's just the way us Aussies and New Zealander's talk when we get a bit excited. But really, I don't think it's all as bad as he's going on about. Yeah it's got problems, but what do you expect for $60 bucks from China? As for the soldering, yeah the board probably was fixed at some point. Wouldn't be the first time an SMD component fell off a board on the flow solder machine and certainly wouldn't be the last. These issues would be picked up during visual inspection and if not, during testing and fixed before leaving the factory. They're only paid in rice so I wouldn't expect too much effort on their behalf. I think one question you should ask is "Can I make a PS with these features for under $60"? I can make a power supply, but not for under $60 with these features - repairs or not. If it's been bought of eeeebay, it's probably not a factory unit and more likely a reject that's been discarded (because it's cheaper to do that than fix), repaired by someone else and then resold. I reckon a lot of stuff from China's ebay fit the description.
Hi,
The green cap at 13:19 had zero solder on the right leg. .. Thanks for sharing.
Hi Bruce ! take a look at the legs of the green capacitor you filmed.. one of the legs isn't solderd at all..NIce tear-down !
+Erwin Scheringa
I caught that as well... looks to be C12; it's a through hole.
+rwbishop yes C12
+Erwin Scheringa Hah! You beat me to it!!
+rwbishop You beat me as well. :¬)
+Erwin Scheringa It is soldered, but it did not flow to the top side. The whole board looks like the holes were not plated through. Which is a bit of an issue itself. I guess they were counting on their wave process to just make up for it, but didn't have the belt speed right.
I keep expecting this guy to outburst "And Bob's your uncle!"
I really hope you soon make the video explaining your process of troubleshooting things like this. I'm learning so much from these teardown videos!
I have a 30v 10amp version of something like this but the inside is another world. I modded mine to quieten down a bit of switching noise and it's been great. I run it along a pair of tti dual, quad mode linear psus when I need the amps. Funnily enough the current control didn't work on mine but a quick look found one of the pins on the lead from the pot to the board had pushed out when it had been built. The board on mine is mounted top and bottom rigidly. Pretty well designed and functional for 40 UK quid.
You had me worried prodding around those caps 😊
I'd like to see it be fixed.
Got myself a PS-305D which seems pretty solid. Same basic spec, but all forms of limiting work fine. Used both forms to my advantage all the time. Is more chunky like the second one you showed but I've never torn it down as it just works. Cost me around 80 AUD including local delivery. Worth every cent.
Fault-finding video - yes please! These tear-downs and technical exposés are great.
Hang on Bruce. At 13.02 I see the right hand leg of that green component appears to go through the board with no solder on it. And the left leg looks badly soldered too.
At £28 it's a hunk of junk.
Its so refreshing to see Kiwis doing teardowns on youtube. Great content! Hello from Dunedin!
Nice video. Maybe the max current has been set too low. This is how I adjust the max current on my similar PSU 1) adjust the voltage to 3v 2) use the button to turn output off 3) short out leads 4) turn output on to create short circuit 5) disconnect leads. The maximum current has been set.
I have a similar power supply and to adjust the current you need to short out the positive and negative.
Dave, is it you? :)
Nope, two octaves too low.
@maj Blagor nama :)
To be fair, I've recently bought the 30V 10A version of this and not surprisingly, within the first 10 minutes, it stopped working. Opened a case with the seller and received a refund. However, being the techie that I am, opened it up and found the output rectifying diode had shorted. Replaced the rectifier with a brand new 'sample' spare that I had, good as new and works a treat. Haven't taken it up to the full 10a, as that's a lot. It is however built much, much better than the one shown here although I'd like to see the board secured at the bottom of the case rather than just sat in slots. I also don't like how the ground terminal in the middle actually goes to the MAINS earth! Don't use it if you've got an RCD in your house as it'll just trip the whole house mains.
Xjet bringing us the things we wish Dave Jones would go back to. Bloody oscilloscopes.
I would like to see more track down and fix videos. DIY electronics for rc videos.
Thanks for taking the time to show us this.
Have a Merry Christmas and a better New Year.
Hi Bruce,
13min 20sec in to the video the is a green "blob" without any solder on the right leg, just a naked pin trough a hole in the board.
You should try thinking for yourself instead of mimicking what Dave Jones would say :)
It looks well done with modern quality components, assembly quality aside. The main problem with the design is that it's much too complex. It can be done much much simpler. Also we don't need those tank steel cases. It's not military gear. Done well it can be matchbox size.
Maybe you should remove the pots and re-solder them on the other side of the board as it probably should have been done initially. That way, the board will not be mounted on an angle.
Too harsh Bruce .Emotive terms like dodgy bodgy and crap are exposing your bias..The angled board is intentional ( a solution to an engineering issue..) and is not worthy of your scorn..Fancy being incredulous that "Nichicon " caps were used then immediately dismissing them as possibly being Knockoffs..The Mains earth wiring you described as bodgy is .. fine ..sound..and not an after thought..The Mains bridge rectifier is mechanically sound and would endure the harshest vibration or shock test any military or aerospace lab would expect from this configuration.. ( don't believe every thing Dave has to say about unsupported TO220 packages ) Perhap when you fix the fault.. take some time and tidy up the wiring , any poor solder joints and OWN the unit Make it yours..P.S. what was the temp rating of the mains capacitors 85 or 105 ??
+reburne Possibly knock offs? I can guarantee you that they're knocks offs. I've never seen a Nichicon capacitor with expansion slots that look like that. I've seen plenty of cheap Chinese capacitors with that exact style expansion slot though. "take some time and tidy up the wiring , any poor solder joints and OWN the unit Make it yours" How you can suggest it is in any way reasonable for people to have to completely go through devices and repair them due to slipshod workmanship is beyond me. It should have been theirs when they paid for it! They did not buy a kit. They purchased a fully assembled product. A product with some pretty glaring defects too I might add.
+Paul Frederick Your guarantee is worthless and your assertion is unfounded.. An image search for nichicon proves you wrong .. PS expansion slots will be found on bridges not electrolytic capacitors,,
reburne
You're looking at other counterfeit capacitors if they look like what is shown here. Either that or you're just fucking stupid. Take your pick. I'm OK with either explanation really.
OK I'm leaning more towards the latter if I am being completely candid.
+Paul Frederick Thank you for your candid response..Your "either or" option misses the elephant in the room.. YOU WRONG ..Google image search shows many many caps with that groove pattern with ichicon branding.. Only a deluded conspiracy theorist would see them all as counterfeit.. OR an egotistical pedant grasping at straws.. Answer A or B ..your choice Your format...
reburne
Nichicon is one of the greatest capacitor manufacturers on the planet so of course they get knocked off a lot. They're practically the Gucci of capacitors you moron. Rubycons are actually better though.
I would like to see the fault finding video
Current limiting seems to work pretty well on mine. But you scared me, I feel I have to open it and look inside :)
Thanks! Looking forward to see it fixed.
hey Bruce, great video, I love to see what is going on inside. do more like this, I like it!!!
You have acquired a LOT of Dave's verbal mannerisms.
I for one would like to see how you fix it. Thank you for the informative videos.
I swear you are a slightly less annoying EEVBlog...
Honestly, I got tired of the oohs and aaahs. Like talking to children.
wannabe eev
I thought it was just me. Most of the Aussie vids are similar to this one and EEV's, viz bloody annoying.
@@2000jago dont insult the guy, he is a kiwi
@@debeeriz don't worry about it, those Mexicans can't tell the difference between an Australian or a kiwi
Defiantly would like the fault finding video Bruce
You can dump the power of a STROBE cap with a piece of conductive foam, but DON'T do that if the
board is powered since the foam will catch fire and make a real mess. (theoretically)
I bought the 10A version of this UPS. After 1/2 hour of using 10A something exploded inside. I haven't had a look inside, but will tomorrow. I wish I have seen this vid earlier. I try to get a refund from aliexpress.
really hope you do a fault finding video on this. Could really introduce some useful information. thanks for the vids Bruce
13:02 I'd have a close look at C12- Looks like it might be missing some solder ;) I would just touch up all the through hole stuff as I can see some horrible ones and would only take a few mins to fix that up. Cheers
+Eric William -MKme Tech It''s a double-sided board so it should be soldered OK on the other side. But R5 at 10:00 looks like it should be reworked.
+Eric William -MKme Tech A lot of bench techs will just resolder whole boards before they even begin to try to troubleshoot. Easier than trying to track down every fault. Well potentially faster at any rate. Just give it the good old once over first, you know?
Paul Frederick Yep sure do. I like taking a look and trying to find the culprit first myself. Funny enough there is a video I did on fixing a bench PSU on my channel. Good old poking and prodding found the bad joint then 1 min of soldering (well a bit longer when trying to do it on cam haha). Troubleshooting is the most frustrating fun we can ever do...
After watching one of your videos I went out and bought a Hakko 888D. Love it, thanks Bruce :o)
If you're looking for a new one I'd recommend the Korad KA3005. It's also a cheapish Chinese one but its linear and still light years ahead of this thing.
Tx for showing the inside, you saved me a lot of aggravation. Even when using quality parts the assembly is not to any quality standard. Shame on the the CE Electronic Industry
Hi Bruce happy new year - interesting video - i would like to see you track down the circuitry and fix the current limiter and also a comparison with the linear power supply
I bought a switched "Gophert" supply for 79$, and it sure has its problem such as a capacitor on its output that is not discharged, but it does work well enough for me.
Yes I want to see it fixed please.
I just received a 30v 5A PSU from Lazada, about $45 AU. looks different to yours inside.
It had coarse and fine adjustment for voltages and current but only single turn pots.
i saw a few too many dry joints on that board. I suggest you use some old school lead/tin solder to redo those dry joints like the one on C12 (the greencap).
+BeachSandInspector Technologies I did not see any RoHS markings on that device. The solder looked a bit too shiny to me to be lead free too. Now I'm not saying any of that is a bad thing. I'm just saying is all.
check the cooling fan circuitry as it usualy is connected to the current limiting circuit and usually the fault is there somewhere, atleast they made most of the smps that way
Please do. I have the same supply I bought at Banggood a month ago. It has developed the same fault (I think it worked properly out of the box).
I love these tear down videos!
The pots are wired as rheostats to give a linear response/turn characteristic.
I would I like to see it tested on a oscilloscope to see the signals of the current to see how clean and low it is. I wish they would do this. It’s a good review to show you the insides but need to show the output on test. Just something I would do.
I guess you'd do that anyway when trying to fix the current limiter: replace the quad-opamp. Having two parts near it replaced, it's very probable that the opamp is damaged as well (or it might have been just the opamp all along and whoever tried to fix it didn't know what to check for first).
I think the current limiter works, but you need to short cut the output to regulate it. I know it's silly, any oven knob switch to the regulated units temporary just to tune them on, coming back to the real value after a few seconds, it's so evident to conceive that I don't understand how even expensive brands ask you to shortcut the output to regulate the current!
I rewired a 180 amp stick welder and know how hard it is to fix. But with much few options than you are looking at. So if you want to start more simply work your way back from the power socket into some cool uses of power. That will be interesting. Start with a bar heater for winter.
It looks as if the potentiometers were soldered on the wrong side of the PCB. If they were soldered on the right side then the PCB would probably sit parallel to the case sides.
I recommend looking at the 10 amp version which is built far better.
Yes, follow up please. +1 vote.
So this guy claims to be a professional electronics engineer but does not know what a TL084 is? Wow! That would be like a supposed Star Wars fan not knowing who Darth Vader is! After the uA741 op amp the TLC070/TLC080 series of op amps must be the most known and common ones!
And it's really annoying that he keeps trying to immitate Dave Jones, why doesn't he just do his own thing?
Mia Lara I've known him for 35 years and that is how he has always talked.
I admit when it comes to electronics I know just enough to be dangerous and have always been curious about current limiting as volts time amps is watts, so ya I would like to see the rest, maybe I could learn something.
+Johnny “Exacto” J
Voltage = Current * Resistance.
Current = Voltage / Resistance
Resistance = Voltage / Current
So if we have 1 volt, and 1 ohm of resistance the current will be I = 1 / 1 or 1A.
If we want to limit that current to .5A, we would need to change the relationship.
Current we want = 0.5A
Voltage we are running at = 1V
Resistance we need to achieve this = ??
1V / ? = 0.5
1V / 2 = 0.5
Therefore, if we change the resistance to 2 Ohm we effectively limit the current.
If you want to know how to use the feedback to change the current on the fly, you will want to have a look at the theory, it seems overwhelming but if you just take it one step at a time it will all work out in the end.
I gave you a thumbs down because you are obviously not qualified to analyse this product. I can tell because I've been designing SMPS's for over twenty years and you never mentioned any of the issues that are actually important in a design. You also trivialized the design of high frequency power supplies as if they are a design formality, which is totally incorrect. I agree that the physical build quality is poor but without actually analyzing the schematic and doing a few tests (such as thermals, transient response, cap ripple current etc.) you can't really pass judgement from an engineering perspective.
I'm not trying to be negative but I really couldn't let this pass.
Roger, i kind of get your position. and have the idea of looking for a posting by you.
in case you decide to apply your engineering knowledge and give us a SMPS review
of one of the many cheap ebay offerings.
While you're absolutely right that the electrical efficacy of this power supply wasn't analyzed. But he never made any claims about it. Also, I'm going to go out on a limb and say that if the mechanical design of a device is this shit, there's a super high probability that it's not super high quality in electrical terms as well. And to his credit, he did say that it "worked". I.e. it functions at least insofar as it supplies 30V / 5 A. If you spend the time to measure the transient response, regulation, ripple, etc. and post a video about it. I'd love to watch it, but until then I think it's more than fair to claim that a power supply with missing solder joints, is cheap crap.
As an electronics diagnostic engineer I don't think you need to be a design engineer to do diagnostics at all. They are two different disciplines and obviously if you are a design engineer you know this stuff inside out but as a diagnostic engineer you are not trained in design, and that depends on if you are hunting faults from a production perspective or feild failures which are two totally different mind frames of fault finding.
Could it be that whoever soldered in the 10-turn pots soldered them in on the wrong side of the board? It's hard to tell if you can't hold the PSU in your own hands, but it looks as if the board could be parallel to the case if the pots were on the other side. Since the pins are all in line they holes should still line up.
EDIT: ok, the mounting tabs are in place so the board has to be angled. And the LED "mount" also only fits if the board is angled. That's attrocious. How can anyone with a the smallest bit of engineering mind design something like this? Mounting a board with (relatively) heavy transformer by two tabs?!
It's bad, but the small form factor is nice. They should sell quality psu's in this size. Or they should sell just the enclosure with knobs and display, would be great for a diy psu.
Another vote for wanting to see if you can find the fault
I'd like to see you get the current limiting working as I have the 2A model (MCHK302D) which also has a problem with the current limiting, it never worked. If you can fix yours maybe I can fix mine :)
You can get the 30v 10A Longwei quality power supply on Amazon now for $64.
Don't. I bought it and its basically a noise supply. I don't even have to have the probe connected to it directly to measure some noise.
Strange......i measured mine yesterday, just out of interest, and it only had 5-10mv of noise/ripple. Maybe you got a bad one. Mine has been great for almost a year now. So good that i actually just ordered another. Hope that one is good too!
That is simplest as it can get for variable power supply with current limiting using SG3525A.
SG3525A is old school voltage mode smps IC, and it is used a lot for half/full bridge SMPS and designs with it cant go much wrong (yes it is not latest and greatest, but it is reliable and if you dont need ZVS or resonant technology it is good choice for half bridge SMPS).
That is worst power supply anyone can buy, I was looking for cheap one and that one was worse rated by people who bought it.
Yes, would like to see you fix the issue.
It is good to me as long as it’s cheaper than USD 50. You may go for Agilent and BK Precision but they are too expensive.
I would like the effects on the output load... Does all the bad construction affects the performance of the device?
On your RCModelReviews channel, please may you explain why 'SimonK' ESCs have motor synchronisation problems and 'BLHeli' ones do not, and under what circumstances these problems do occur. Thank you.
Good you found it.
A lot of people are complaining that you've nitpicked but I doubt they've watched the entire video! I admit at the beginning I thought well what do you expect? But even for 60 bux this thing is a mess!
Hey Bruce,
looks like one of the pot terminals has no solder or maybe it's supposed to be like that. Not so sure.Looks like you found it ,guess I should have watched the whole vid before I made a comment.
Regards,
Pete.
fun video, interested to see how you debug it. I'm facing an issue with a SKY TOPPOWER3005 right now, lots of noise!
Be good to see the fault fixed.
I would like to see you fix it, , also tools and supplies ,recommended.
i'm interested in the fix video please
Nice I get a EEVblog PS tear down and a XJET PS tear town in a 24 hour span!
Hey, Yes agreed the way they mounted the board is absolutely crappy. But I like the board.
I wouldn't mind seeing this work again, with some safety upgrades lol.. Give it a burl, Bruce.
Would be interesting to see on the o-scope how much ripple and crap it puts out under load compared to the linear supply.
+pillsnotbills This one is useless to compare to anything (you will get no useful data). All comments on internet about it say it is worst one you can buy.
This one put much more crap on output than any other variable SMPS on market.
Miloš Lazović Bruce wasn't kidding then. You can't expect much for $60. Still, Bruce, give it a go. I'd like to see the measuring stick of badness in action. :)
aside from the mounting and soldering... it's really not that bad. I wouldn't have wires tacked onto the surface of the board, but all in all, it is okay.
IMHO the quality of items coming from China is improving at a spectacular pace. Yes, there is still junk out there, but in many cases, especially the items made for American manufacturers resale, the quality is very good for a very reasonable price. In machining tools, for example, the use of HSS steels, tolerances that are measured in tenths, fit and finish are excellent in most cases. Welders are included, with Tig welders compactly designed or reversed engineered having pulse, HF start, very low amperage stable arcs are priced several thousand dollars less than American manufacturers who have many components if not all components made in China. They are perhaps training the Chinese and they are doing very well. For a hobby machinist they make it possible to have the tools that would be out of my reach.
love to see you fix it !!!!!
'Keithley' was the word you were looking for (ex Agilent) ;)
Keep telling it like it is Bruce :)
yes I would enjoy watching you fix the supply35 years ago i was a bench tech for upi fixing electronic teletypes called extel
they woudlf fail with a lot of drama because they used a switching supply run directly off the 110 v ac@60hz
when something failed it would let all the smoke out and leave a pile of carbon fun stuff
That power switch is switching the mains voltage yet the only insulation from the PCB is a layer of heatshrink?
Love to see u fix it!
You may already visit his videos, but seeing that rigol oscilloscope in the back made me wonder if you visit AvE's channel.
I believe he posted a video on how to upgrade the software to double the range and capture window on one.
Anyhow.
Very interesting!
+MrJgstoner Yes, Dave Jones of the EEVBlog came up with that hack many, many years ago.
ahh. I'm not in the club but I'm still a fan ;) thought i'd try 'n help
Structural LEDs. Nice.
Gee, you have 2 power supplies, both with inop. current adjustment. In a just, fair world, one of the defective powers supplies would have a good voltage adjuster, and the other defective power supply would have a good current adjuster.
If you need a really cheap power adjustable CC-CV power supply get a LTC3780, 3-35v input 0-35 out.
This video reminds me of something Dave of the EEVBlog would do, that's a good thing in my opinion. I'm sure you've seen his videos.
Sir, I have a same bench power supply with no output. I am new to electronics. Could you please guide me through that from where to start checking to find the fault of the same?
I bought the 10A version. Guess I will have to take a peak inside as well :)
Yeah a troubleshooting video would be great
Below the two wires that go to the power switch you will see a joint that was never soldered. It looks like it goes to the current pot. That is probably the issue.
Hi Bruce. I'd like to see you repair this PSU please!
Thanks, Rich.
LEDs as a structural component - clever! ; ) Any possibility of getting the schematic for this rig?
you talk just like that guy from eevblog lol
i always love a repair video, i vote yes
You saved me from buying one of these cheapies.
I would like to see how you go about fixing it