![\"\"](\"https:\/\/www.dclabs.lt\/dainiaus\/wp-content\/uploads\/2022\/09\/image-2.png\")
<\/a><\/figure>\n<\/div>\n\n\nThe internals of the power supply are the same and looks like this: <\/p>\n\n\n\n Pictures takes from the power supply I got for repair, exact model is KPS1505D, one of other models, sold by this „manufacturer”:<\/p>\n\n\n The power supply had burned main fuse (3A), so it was not long to find out that power transistors and diodes gave their souls to the god of electronics in form of smoke. And we all know – all electronics is made of smoke, when it goes out, electronics stops working. Well, before repairing, lets study this board for a while. By the way – as this power supply is so typical, I am not explaining about its functionality, controls and basic things like that.<\/p>\n\n\n\n The main PCB is marked POWER-ZB-02 2022.01.10:<\/p>\n\n\n The white connector on the bottom right is for 230AC input and all this bottom part along the PCB is for input filtering, rectifying and stabilizing. So right to left – yellow input capacitor, common mode (CM) filter, which, with its \u2018current compensated\u2019 windings, provides high impedance to CM noise without saturation from line current, two small blue caps and one more yellow capacitor. Then a small diode bridge, yes, yes, the fuuuuuuull bridge rectifier (DB107S). This rectifier, together with the nearby electrolytic capacitor, that strange IC with 7 pins (TNY280GN<\/a>) and JKL-10-12 transformer (skipping some minor SMDs) forms low power supply for logic and control ICs, LED indicators, etc. <\/p>\n\n\n The bigger rectifier (KBL610) and two big electrolytes are for the power section of the device. <\/p>\n\n\n And on the left side of the PCB a place for missing connector, it wasn’t there, but on the other boards the main power selection switch (115VAC\/230VAC) is connected. This version of device has no connector on the PCB and switch on the back panel, so, if you want to use it with low voltage power system just connect those two pins on PCB or put a switch – when pins shorted use 115VAC, when open use 230VAC. The rest of the board is, basically used to hold the main controller (TL494C<\/a>). The TL494 or its clones or variants are used in all sorts and kings of PWM power supplies and, probably, in 95% of their population. <\/p>\n\n\n As this IC is so popular, there are plenty of schematics with it on the Internet, which would help you to make the repairs. I was surprised, but this version of board has all electrolytic capacitors marked as low ESR, which is good (hoping, that its not just marking One of three output caps is also Changxin 2200 \u00b5F 35V Low ESR, other two Sanyo 1000 \u00b5F 50V, together with a big and small choke, they form the output filter (on the top of PCB picture). And thick copper wires on the negative trace are shunts for current measuring.<\/p>\n\n\n\n In the middle part of the PCB the main component is TL494, couple of op-amps, one of them turns on the fan when aluminum board with transistors and diodes get hot. The temperature is sensed by 1N4148 diode, drowned into thermal paste. The last things are power transistors and diodes:<\/p>\n\n\n\n The big thing is a dual Schottky diode with common cathode – MBR30200PT<\/a>. But the two transistors are blank, not marked And yes, it’s not recommended for new designs, most probably you wouldn’t find them for purchase eather. Unless fake AliExpress clones. Luckily, we have me and I already selected proper substitute – AOTF15S60L<\/a>. Why particularly this one ? Because at the date and time of repair it suits all requirements: <\/p>\n\n\n\n This transistor a slightly better that original, is almost cheapest (~2,5\u20ac\/piece) from the selection and is available for order – perfect :).<\/p>\n\n\n\n Parts received and soldered in, surprisingly nothing exploded, and power supply shows the set voltage. But the test under load was not successful – connected 0,62 \u2126 resistor and raised voltage to raise the current. At some point heard a clap and everything turned off. The current was not high, I think it was about 6-7 A, but can’t say for sure – I did not expected it to explode, so did not looked at the display all the time and also display refresh rate is not instant. Nevertheless, fuse blown, most probably transistors and diodes also. Let’s disassemble it again and see what happened. But it worked with some small current, liek couple of amps, so it might be, that TL494 is counterfeit and cannot work normally under bigger loads, while the idea of this power supply is to be fearless when speaking about currents up to 10 A or short circuit. Luckily, I still have some transistors and diodes left To be continued after parts received and soldered in. 2023.10.12<\/p>\n\n\n\n Parts replaced and that did not helped, so we need to look deeper into the schematics. First of all desoldered the main transformer – the big yellow thingy, because its my first suspect (maybe the coil wires are damaged and shorts the transistors, so they pop and brings together the fuse). After desoldering tried the output part of the schematics for faults:<\/p>\n\n\n\n The idea is simple, but lets draw a small schematics, so it would be even more understandable.<\/p>\n\n\n When transformer T1 is removed, it is OK to connect the power from separate power supply, positive to points marked with red arrows and negative to point, marked with black arrow. This imitates (except its DC instead of AC, but its OK) the transformer and puts voltage through the circuit to the output pins. On the way we have a choke L1, couple of low ESR caps for smoothing, small choke L2, couple of shunt resistors R9 and R10 (to measure current), final output capacitor C9 and thats it. So now I can confirm, that output circuit is OK. Next – lets try to test the input circuitry, also without transformer.<\/p>\n\n\n\n The primary part, judging from the drawn schematics, should be designed using Half-Bridge topology.<\/p>\n\n\n Half-Bridge topology is quite simple, two big capacitors forming capacitive divider, two transistors and one more addition, polypropylene capacitor C22. This capacitor is rarely found in cheap power supplies (AFAIK, might be wrong), its primary purpose to block DC currents to prevent transformer core saturation. There are some facts that we should take into consideration, before diving deeper, as those might be the cause of the problem:<\/p>\n\n\n\n And this leads to our second suspect – is the high side FET really driven by floating driver ?<\/p>\n\n\n\n 2024.04.04<\/p>\n\n\n\n Another spare minute, spent on further diagnostics. Regarding the floating driver – should be fine, transformer is present and its not the first part I would suspect in the SMPS. FETs are proper voltage, so it should be also OK. Ringing is a bit more complicated thing, need oscilloscope for checking, even if I have it, I wouldn’t suspect ringing in a first place, so lets keep it as a plan B. Or C. Or next to what ever letter we are at now If I have not made mistakes finding traces on PCB and transistor part numbers on the Internet (transistor marking BA), we can see that there Q5 and Q4, along with couple of resistors and diode are controlling the Gate of Q2 and those BJTs are controller from transformer T2, which makes it floating. This part of PCB looks like this:<\/p>\n\n\n Ale as we are here, lets quickly check transistors Q4, Q5, Q6, Q7 and diodes D2 and D3. And here lies the fault – transistors Q4 and Q5 are shorted. That means they do not control the Q2 transistor, so it can be in some uncertain state, if its fully open, then its a short circuit every time when Q3 opens and everything blows. If its semi-open (is there such a word at all 2024.07.12<\/p>\n\n\n\n Transistors arrived and replaced, I bet cant even tell the difference I picked the transistor with BA marking by EVVOSEMI, which is 2SA1015 transistor (datasheet<\/a>). It might be incorrect, because manufacturers could mark some different transistor with same letters and the original letters are a bit different shape and spacing, but lets think that its because of new manufacturing batch and I did not made a mistake. <\/p>\n\n\n\n Damn, the transformer ferrite fell down and now its broken. Does that mean I will have to make a new transformer ? Or maybe somebody from my readers could send me the faulty board to take the transformer out for testing For OktO:<\/p>\n\n\n\n My guess we should concentrate towards dead time of the TL494, probably it needs to be adjusted. Dead time ensures, that both transistors would never be open at the same time, creating direct path for current to flow from positive to negative, thus effectively shorting the power supply on the high voltage side. Pity cant check with oscilloscope, how the transistors are switched and if their working time overlap because Is still have no transformer Feel free to correct me if you find a mistake :). U8 opamp is inside of the TL494. In theory, current limit is set by this opamp, pins 15 and 16. Normally, if voltage drop on shunt resistor R9\/R10 is more than preset value, opamp starts adjusting output voltage. Was surprised, but it seems, that pin 15 is connected between two shunt resistors. I think that is because one shunt is used to overcurrent control and other shunt is used for indication on the current. Also we have second opamp in SOIC-8 housing LM358 for some reason.<\/p>\n\n\n\n Why I suspect dead time ? In my power supply, after I replaced both burned transistors and diode, it was working to some extent. I was slowly raising voltage to raise current and it popped again at about 6 Amps. So must be related to control, because power handling part was working. To kill both transistors you have to open them both, making a very nice and shiny short circuit on the high voltage side, which instantly damages transistors and eventually fuse. As the control was working, maybe the dead time is not enough for this transistor and while one is still closing, other is already open, whilst load is increasing and here we have magic smoke coming out. <\/p>\n\n\n\n Started to analyze the schematics from TL494 side, at the same time checking the power transistors control signals. First connection is made at TL494 pins 8 and 11. This confirms, that dead time exists for transistors Q9, Q10, Q11 and Q12 (but not the power transistors!).<\/p>\n\n\n Brown wires – power supply (+12V and GND), white wires are control outputs. The signals on the oscilloscope are not so bad:<\/p>\n\n\n The switching on channels has some delay, not happening at the same time, but that delay is about 1,8 \u00b5s. But lets follow the controlling signal path and see whats happening there.<\/p>\n\n\n\n Removed the isolation transformer for better access:<\/p>\n\n\n And updated the schematics:<\/p>\n\n\n Remark – transistors, marked BA are 2SC1015-H<\/a>, marked HF are C1815<\/a>. Four BA’s (Q4, Q5, Q6 and Q7) already replaced, because they were burned, before continuing I will also replace both HF’s and BA’s in push-pull transformer driver circuit. Just to be sure. Logic simple, chinese power supply, probably parts are taken from faulty manufacturing batches, so just to be sure, lets replace all, as OktO did. It seems, that C1815 is not available, but according to their really shitty and uninformative datasheet I managed to pick a substitute, which I hope will work in this schematics – Toshiba 2SC2712-GR<\/a>.<\/p>\n\n\n\n Signal from TL494 is Ground, which is pulled up by R14 and R15 resistors to 12V. Signal at the transformer input:<\/p>\n\n\n\n <\/p>\n\n\n\n To be continued…<\/p>\n\n\n\n <\/p>\n","protected":false},"excerpt":{"rendered":" The names of those power supplies are different, but inside they are all the same, with veeery slight differences. The brand names are Wanptek, Hanmatek, Naweisz, Sky TopPower, Circuitspecialists, Eventek and bunch of other brands and even without any brand, they all look pretty much the same and are priced at about 70-80 \u20ac. The…<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[],"class_list":["post-4584","post","type-post","status-publish","format-standard","hentry","category-elektronika"],"_links":{"self":[{"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/posts\/4584","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/comments?post=4584"}],"version-history":[{"count":70,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/posts\/4584\/revisions"}],"predecessor-version":[{"id":6026,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/posts\/4584\/revisions\/6026"}],"wp:attachment":[{"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/media?parent=4584"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/categories?post=4584"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dclabs.lt\/dainiaus\/wp-json\/wp\/v2\/tags?post=4584"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}<\/a><\/figure>\n\n\n\n<\/a><\/figure>\n<\/figure>\n\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n). Even those big input filtering caps (470 \u00b5F 250V) are low ESR, which is not really needed for their job.<\/p>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n\n\n\n<\/a><\/figure>\n\n\n\n<\/a><\/figure>\n<\/figure>\n\n\n\n, guess what are they. Ok ok, I tell you, its 2SK3569<\/a>. <\/p>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n\n
.<\/p>\n\n\n\n
<\/p>\n\n\n\n<\/a><\/figure>\n\n\n\n<\/a><\/figure>\n\n\n\n<\/a><\/figure>\n<\/figure>\n\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n\n
. Lets pick our next suspect – FETs in most cases are controlled by something, either directly by TL494 or using some sort of gate drivers. Lets update our schematics:<\/p>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n?) or closed then power supply could even somehow work, which I noticed during previous test after repair, on low current power supply was working, but when current raised – all popped again. Lets wait till new transistors arrive and replace them .<\/p>\n\n\n\n.<\/p>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n ?<\/p>\n\n\n\n. Bet lets start from the current regulation, which you mentioned. Started to draw some schematics, it might still contain errors and surely is not complete, but it will give some insights whats going on.<\/p>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n<\/a><\/figure>\n<\/div>\n\n\n