EEVblog
Vložit
- čas přidán 16. 07. 2012
- UPDATE: Triosmartcal have withdrawn it from sale, and Korad are investigating...
UPDATE2: Korad have admitted that the output transistors were not the correct type in this early batch unit.
UPDATE3: Follow-up video: • Korad Power Supply Fol...
Forum Topic: www.eevblog.com/forum/blog-spe...
Watch the Korad KA3005P FAIL!
The review of course is not complete because there was nothing left working to finish it.
EEVblog Main Web Site:
www.eevblog.com
EEVblog Amazon Store:
astore.amazon.com/eevblogstore-20
Donations:
www.eevblog.com/donations/
Projects:
www.eevblog.com/projects/
Electronics Info Wiki:
www.eevblog.com/wiki/ - Věda a technologie
How to access to Memory 5 (no possible with buttons)
To use memory 5, proceed as follows:
1. Press memory key 4 and turn the adjustment knob until indicator M5 lights.
2. Set the voltage and current as desired. Wait until the display stops flashing. The settings are
automatically stored in memory 5.
3. To recall the settings, press memory key 4 and turn the adjustment knob again until indictor M5
lights.
WTF! Not so user friendly if you ask me...
Alternatively you can put it in M4. Summon a spirit and then if the stars align you get into m5
Yup I have that one for 3+ years now. The buttons and display are great, not even the beeper is annoying. I have 4 other power supplys from that price-range and the Korad is a favorite
I now have had this powersupply for a little over 2 years (fixed version) and it hasnt failed on me once. I think after the fix it is most certainly reccomendable!
tHaH4x0r What was the problem with the old one ? :O
Flofy387 See the video above...
tHaH4x0r
Yeah but is it now really short circuit proof ?
Flofy387 I have shortcircuited it often enough and it still works fine. Also, watch the followup video in the discription.
tHaH4x0r
Okay. Because I've seen a video where the KA3005P was rapidly shorted, and then it was fucked up :D
For the 0.2V on constant power problem: I think the load is just fighting with the power supply. The load wants more than 5A sometimes on power up, but the power supply will limit to 5A and stay there.
As for the failure, I'm guessing it's the pass transistors too. Sounds like they failed short, sending the full rectified transformer voltage to the output.
Interesting. This review is from 2012. I have three of these babies (two 30V/5A and one 60V/3A KORAD), but I bought them significantly more recently than 2012. I think all of them were bought after 2018 or so. I haven't had any issues with mine with several years of use on the oldest one.
I got them because the voltage and amperage setting displays the V/A setting instead of the current V/A while you are editing it. Even when the power supply is turned on and the settings are being modified live. Just about all of the other cheap bench supplies out there have digital displays, but analog knobs that are hard to precisely set the limits for, and don't show the actual amperage limit setting unless you short the outputs together.
-Matt
They fixed these issues rather quickly. Dave even has an video on the changes.
He broke it. Unbelievable!
He did not broke it. It only give it a negative feedback =))
19:57 That might not be the PSU at fault, connect any current limited supply to a constant power load and it can get stuck in a low voltage current-limited state. Practical example would be a solar panel feeding a switching regulator.
boriseng That’s not acceptable on a power supply. They design the control law wrong on these things, and it’s very much a noob mistake. Depending on how you set it up, it can get “stuck” at a low voltage, or at the target voltage. The resistance of the interconnection causes this. When you do a solution on paper to a constant power load driven by a cc/cv source, there’s one solution at highest voltage as long as interconnection resistance is
Я тоже считаю, что блок питания тут непричём. Он нагружает его интеллектуальной нагрузкой, которая меняет своё сопротивление управляясь своим процессором. Для чистоты эксперимента он должен был подключить резистор необходимого номинала, и я уверен, что никаких выбросов и ступенек там бы не оказалось. То что он видит на экране осциллографа - это то, как два независимых микроконтроллера подстраиваются друг под друга.
You can acces m5 by pressing M 4 and scroll to the right with the wheel
It was working fine after a tore it down. It failed because it didn't like a particular load.
Love it. Was hoping for a review after the tear down this morning!
I reckon it's feedback oscillation, since both the power supply and load are in CC mode.
My guess would be the output section. I've seen this before - power transistors just fried. If you would fix it, you should stick a thermocouple to the main heatsink to see if it overheats. Something tells me that it's not enough to cool down the transistors.
Thumbs up, great demo.
While you were reading the manual at 11:50, the fifth LED was lit. Part of the failure, or did you hit on activating it?
Hello, always give the polymer a lower voltage.
So, do you have to calibrate the power supply?
Which power supply do you recommend for a double price with the same values
but no dead end. ?
Hi there,
I have the same supply and it has gone in the same direction!!
Seems my alligator clips shorted together on the test bench one too many times!
turns out one of the output transistors failed, I replaced with tip3055's (all I had)
Was once again working but only for about 3min, then went faulty again! I am now waiting on the proper transistors but dont have too much confidence.
Be great if you could go through the repair so others like myself can get it back up and working.
It's important to know if it was failure in the control circuit or in the power circuit, because there are a similar PSU with the same power circuit but simpler control circuit, the cheaper KORAD KD3005D.
Hi Dave. Thanks a lot for your help on avoiding me buying this kind of power supply. Thanks!!!!
Hi! I would like a Video, where you repair the supply (I'm assuming one od the two main transistors shorted), and test if it broke because of bad design (to small transistor), or if the design was correct and the transistors didn't meet spec...
I know this is a super old video but maybe it will still help. I had the exact same problems you were seeing at 19:30. The electronic load you are using is the problem I believe. In constant power mode it gets into a positive feedback runaway type thing and basically just drops a short on the output...
Josef de Joanelli
I know this is a super old comment but.. :D
Yes, you are correct, the problem was electronic load going basically short. I'm suprised how Dave did not get this. PSU went into CC showing 5A. Electronic load was also showing 5A. So there was nothing wrong with that.
@@vladan_kovacevic Dave is not that bright at all if you ask me
@@TheMechatronicEngineer I'll keep up the tradition of replying to another old comment, lol. I think he's thinking quick on his feet while on camera, feeling pressure to have an explanation instantly. So he forms opinions very quickly based on a small subset of data. I think maybe his biggest fault is that once he's formed an opinion, he very stubbornly holds on to that opinion even when presented with additional and/or new data that shows that opinion was erroneous. So he's quick to form an opinion and slow to change it.
This is normally forgivable, except when he occasionally attacks or belittles someone for having an opinion that differs from his own without recognizing that his position may be in the wrong because he arrived at it quickly with limited exposure to the question. I think he's fairly guilty of this on the podcast.
@@KapitanWalnut Agreed. I remember an obnoxious review of kiCad based on the fact that... he just couldn't use it not knowing it at al... embarassing.
were those ripples on full load powerup the AC ripple on the filter cap, and the step where the relay clicked in?
Similar to Hantek:
I have the power supply PPS2116A and I like some of the functions - especially the memories. Unfortunately there are some bad sides of the PPS2116A. The rotation knob is made with low quality encoder and after a few months/years it is really not reliable - very difficult to operate. But the worst is that the PPS2116A can easily damage your circuits. It already destroyed a few of my electronics, mobile phone, AM meters fuse etc. How it is possible? When you turn off the PPS2116A and you do not DISCONNECT your circuit, the PPS will send cca 30V for 200ms to the outputs. They can be ON or OFF it does not matter, the 30V is sent to the outputs. Realy bad design!!!
I can send a osciloscope screenshot for all the cycle: Power on - out On - Power off. There the pulses are clearly seen.
Yeah, makes my DMM's go haywire if I leave them in circuit with the PSU after turning it off.
This is worrying actually and makes this a no go if they still act like this.
the output transistor shorted hard it looks, which is why it won't turn off. the voltage output by the supply is probably dictated by the tap selected on the transformer. I am wondering if that heatsink they used was way too small for the transistors, since it sure looked puny for the power level the supply can push. The most stress would be constant current mode at the highest possible output voltage; the transformer taps cannot be used to lower dissipation so the transistors dissipate it.
hello ,
what happens at 23 minutes50 , the video is cut off; What is the voltage and the charging intensity?thank y
Hi Dave, could you please do some overshoot tests of the new version of KA3005P? Unfortunately the Charles' follow-up video does not cover this topic. I think there are many people out there who want to know if this problem has been solved in the new version of this device. It would be great if you could manage this test. Thank you in advance!
I'm using it about 3-4 times a week for a year now and I'm really happy :). No problems ;). But that's just me.
I noticed the same thing. When Dave hard switched the power supply off the voltage was only decaying on the (1 Ohm load) very slowly. Made me think the BK was doing something wonky there.
You can get to M5 by pressing M4 then turning the rotary encoder clockwise.
If you look at the teardown video you can see that there is a capacitor between the binding posts.
I need some help if possible.
I have a power source equal to that present in the video.
I bought this power supply to work with an arc to cut foam.
The wire is chrome nickel and has a diameter of 0.05mm and 700mm lenght.
I hook the Power Source to the arc, but not heat the wire.
What is the reason?
Thanks
There's a Off/On switch under UVP switch which you have to turn n manually.
Dave, hoping you'll diagnose and post that followup .
Hello Have you reviewed a more recent model?
G'Day Dave, Just thought id let you know. I discovered your videos a couple of weeks back, And all I can say is AWESOME. I always learn something watching your Vids and have a laugh doing it. Your the first person I have subscribed to ever on anything :)
Two Thumbs UP :)
For a low cost power supply, maybe a BK precision 9110 or ITech 6720 (they are the same thing, ITech is the OEM of BK9110) will be MUCH better. I used to use IT6720 as my bench power supply, and tortured it a lot, yet it is still alive.
They are around 250 USD on eBay, cheaper than most high-end branded devices, while providing very exceptional performance.
I've been looking for a decent power supply in the 100 dollar or less range for some time now and it seem that all the ones out there just don't put out good power.
Is there model in this 100 dollar or less group that will actually give out good power?
Could it be that both the power supply and the load are active control systems? If they're digital then they have sample rates and that trace could be the two control systems interacting?
Richard Corfield Not if they were designed competently. A bandwidth limited discrete time controller will act identically to its time-continuous counterpart. Since it’s hard to do perfect band-limiting, usually you oversample enough so that say, a 4th order filter is close enough. Or, in modern circuits, use a sigma-delta converter, since those have very good filters built-in. If you can tell it apart from an ideal analog controller just by looking at the traces, you’re not doing it right.
For inspiration on what it means to do it correctly, look no farther than the phone network. Most modems faster than say 4800bps do their work digitally, so there’s one ad/da conversion at the modem, another at the interconnection to the phone network, another at the other interconnection, and another at the other modem. A signal from one modem to another goes through two DACs and two ADCs, and through four two/four wire conversions (those are often done digitally, by duplicating the adcs and dacs to control line voltage and current independently). And it works just fine, and these modems can fully exploit the theoretical bandwidth of the voice circuit. So, uh, it’s only a problem if you do worse than the POTS network does, and that seems like a very low bar to pass.
Be nice if you teared it down to find out exactly what it was.
I'm guessing it's the MOSFET's due to those rending voltage/current.
Did you also measure the output without the BK connected?
thanks for the forum link. Much easier now.
You can adjust the current "all the way down to 1 millivolt"?
Have you tested your load against a (electrochemical) battery? You'd know if the load had some kind of regulation transient response.
Iv seen a review and repair of a Agilent supply that had an issue like this and it turned out to be A zenor diode and it cause near enough the same issues that cheep bit of crap is having. I would assume the agilent one had taken years of abuse though not 20 mins worth of light testing.
Lesson is to never hook competing control circuits together. Also, is the power supply rated at the full 30V and 5A together or is there a lower limit on power?
For those living in Europe. These are sold under Tenma brand. The 30V/3A non programmable version goes for £58.00 on Amazon.
I was just going to say that. Unfortunately the UI is horrible.
Ventsislav Simonov
That's exactly what I've got.
I've never had any dramas with it.
For sale on ebay for US$89, description says "Korad Technology is a top Quality instrument manufacturer in China and is now available in the US." also says that the unit is "Switchable between 110v and 220v"
I'm thinking that funky waveform at switch-on is contact bounce in the relay(s). Maybe?
I think ya shorted the pass-transistors Dave.
And I'm happy with mine.
rageagainstthebath I got the version with the fix too, and it just works. :) I did change the fan though, it was driving me insane.
Для того чтобы вентилятор был тише я не менял его как все, а обнаружил что питание на него идёт с помехой и он от этого жужжит сильней. Я просто поставил параллельно вентилятору конденсатор 4700мкф и жужжание пропало) Ещё я в разрыв поставил 50 ом сопротивление и он стал ещё тише что не слышно. Удачи!
Hi Dave, can you explain ground loops? PC has USB grounded, so power supply will create ground loop by grounding usb connector? Keep up the good work!
I think the strange buzzer might be related to the strange 1kF capacitor in the teardown :P - it was right next to the buzzer after all...
I wonder if there would be possibility to mout in like delay output switch relay, let's say, when you turn on the power supply, the output gets switched on after , for example. 1 second , so the load doesn't get higher voltage spikes. and the same for power supply turn off, the output switches off 1 second before supply switches off....
Great job Dave. You may have inadvertently found yourself yet another job: namely finding a decent low-cost programmable PSU. You blow 'em up so we don't have to. Thanks.
Could one of the voltage swiching relays gotten stuck, causing 2 of the voltage taps to be connected at the same time ?
10:18 - I don't think this is true. The supply turns the output off whenever you press one of the memory buttons. As to M5, you get there by pressing M4 and then turning the knob. Not very intuitive, but I believe it's mentioned in the instructions.
@EEVblog I am shopping for a Lab grade power supply, which one would you recommend me? Thank you.
Which one did you got? Thank you!
Thanks for the video you save me some problems. Any recommends for DC power supply unit around 50-60 USD (I m from Europe) ?
there is a 6A power supply on Ali Express now with a graphic display (like a speedometer etc) for rather cheap.
Brilliant, amazing video that I could not do myself. For newbies for me, could you possibly do a video showing how you would improve on the power on overshoot please with sugar ontop? :o)
25:40 when measuring the outputs I would expect any other load to be disconnected. Otherwise how could we believe that voltage is not coming from somewhere outside of the power supply?
Good Review! Thanks
at first it looked almost to good to be true, and it was to good to be true.
M5 - press M4 and turn adjust to the right site.
Perfect one to troubleshoot Dave. They always say to build a PS as a first project so why not troubleshoot one.
Thanks for the review Dave! I nearly bought one off eBay yesterday, I won't be buying one now! Still using my cheap PS3025 :)
This thing looks exactly like the Velleman PS3005P, on that one "M5" is "Locked", so I think that's just a silkscreen typo, in newer versions it might me "Locked" :P
Vellemans Supply is the same thing, repackaged for double the price, so this one is pretty good :D
***** I'm getting a Velleman LABPS3005SM, 0-30V 5A. Are they good? It's only 100 bucks, and it's the only PSU you can get for that price here in Sweden :3
SnOpeK Domowei It is the exact same powersupply. I bought it and it is the updated version of the one above. It works a treat and hasnt failed in over two and a half years of service, most certainly worth the money because it is a cheaper powersupply that functions well enough and is linear instead of switching. (lower noise than switchers in the same price class)
M5 is probably accessible. Press M4 and immediately afterwards turn the knob clockwise. Works on clones named KORAD and AXIOMET.
Methinks that looks like "pop goes the transistors". Even with a fan, I didn't like how small the heatsink was in the teardown video. Certainly not for continuous duty at maybe 50 watts or more depending on what the output current and voltage were set to. Crack it open and see if those transistors are cooked
Does anyone know if they have released any new revisions since this video? Looking to buy one on ebay. Thanks
no they haven't, you need to replace some parts to make it bullet proof.
If you power the PPS2116A on and touch a button befor it initialize - it start to be totally messed. Only way how to recover is to turn off and then on again. Alway it is necessary to wait to boot...
I think the strange turn-on behaviour is simply DAC quantization and a less-than-tight control loop in the microcontroller.
Is it the same as the similarly looking TENMA?
millivolt resolution on the current selection?
I hope they'll fix this buggy behaviour. At least front panel of this PSU looks pretty cool.
Ah that overshoot is unfortunate. Maybe put permanent inline switch situation, close circuit once unit is already switched on.
How did you program that KA3005P?
Hey, it still shows the voltage and current when you turn the output on, at least something still works! :p
thanks for the advice
Thanks. You only have several hundred to catch up on :-D
Damn i never saw a PSU die so fast!
Oh and the weird stuff you got when giving it a constant power load makes sense if you think about it. As you introduced the large load the supply voltage dropped and the electronic load tried to pull more current. If the current got big enough for the PSU to start current limiting it then the voltage dropped even more and the electronic load tried to pull even more current from it until it all hit almost 0V where both of them give up.
Drop a falcon feather and this 'thing' from a 5 story building and lets see which one hits the ground first.
PS don't forget to select a location with a perfect vacuum... and don't forget to make a video :)
Does it multiply said voltage by sqrt(2)? =P
And what might that be? The BK is acting as a supply?
Actually that would be even more interesting to see because if that happened a secondary winding would get shorted out and make the transformer start smoking after a while
Distrelec group is now selling this under their own "RND" Brand. 320-KA3005P
I think it's the result of a simplistic slope compensation routine.
My mind and mouth are not necessarily connected properly.
when you switch off ovp ocp that means no protection. you overloaded the transistor!
Hi Dave, I have found on the internet, that the fifth preset is avaiable when you hold the M4 button and turn the wheel to the right simultaneously. But I don't know, whether is it true. I don't have this one to try myself.
It's good to be careful with an electronic load set to constant power. When the supply is off, the voltage is 0 and the current is therefore I = P / 0V = infinity, unless the e-load has some other current limit. Then you ask the power supply to turn on into a short. This type of load would never be found in any real application, it's just a testing convenience. Anyway, just something to watch out for.
David Erickson There are the connecting wires, and even a constant power load can have a minimum resistance set (at least good ones do). So no dead shorts :)
Another good reason to design and build my own - at least that way it will work
I have a very stupid question, I’m in North America and we are 110v, so if this is 30v isn’t almost everything 110v that comes from the wall, cause it’s only a quarter of which comes outta the wall, I’m sorry I’m very new at this and wanna get into testing electric motors and dodads from appliances I just received my Korad 3005d and really thinking I made a boooobooooooo., please help.
Wow, that's potentially dangerous!!! Thanks for doing this, so we less competent people don't have too!! ;o)
too bad there wasn't any magic smoke.
I know. It's an old habit.
any tell me it is buy it worth or not ? please
I wish they put a "Lock" indicator light on the display (for example use the M5 LED light). Because now you don't see if the "Lock" is/has been activated or not ;-)
The 5th memory is used, you hold the 4th and toggle the wheel to access 5.
so what failed on that thing?
These things are popping up for cheap at aliexpress. Seeing as Korad has updated them, would you be willing to give it another go? I'm sure a lot of hobbyists would like such a cheap power supply as long as it is safe.
Ah, turns out he did a followup: czcams.com/video/_HrvuHSywms/video.html
Please investigate, I wonder what has gone wrong with it. And I'd love another video of you debugging something. :)
Looks like it has relay bounce on start... Funny business indeed, skimped a capacitor there. The 0.01V looks like all relays disconnected?
Also, didn't I notice 220V AC input there? :) That's not 230V.
Yeah, it was pretty disappointing. It failed at failing!
you cannot use it for 5Amps for more than 5mins
The biggest problem with these 1HL brands is the counterfeit pass transistors. They can barely supply 1/3 their rated current due to a severely reduced SOA. I routinely buy them - they're a steal - and replace the pass transistors with the genuine parts. I never, EVER have problems again. I am very hard on power supplies and do everything I can to blow them up. :)