They brought the ADP-90YD power supply for repair from Asus laptop. Either it charges the laptop or it doesn't. You take it out of the socket, you insert it like normal, maybe something is moving away.
I plug it into the network, I check 19.35 V with a tester, I moved the wires, it began to fall smoothly, as if the capacity was discharging, well, maybe it’s leaving. You need to open the power supply. He inserted the knife into the joint of 2 halves of the body, gently tapped the knife with a hammer, the body opened.
Board in three layers of screens. Soldered everything, took it off. The power supply is plump, and a lot of sealant is also poured.
During a cursory examination, a torn off leg of the filtering inductor was found along the 220 V input circuit. “Here it was that caused such a strange voltage drop,” I thought. I restored the throttle, I check - the result is the same. When 19.35 V is turned on, after 1 second it begins to smoothly fall to zero. Apparently from my swotting with a hammer on the PSU case, the throttle fell off. But here's what I noticed, if you turn off the power supply from the 220 V network, after a few seconds, 19.35 V appears at the output and even the charge light on the laptop lights up, but then the network capacity is completely discharged and the PSU turns off. It is very strange, apparently some kind of protection is triggered and does not allow the power supply to work, but what is the reason ...?
I assembled a small load from 5 watt resistors, the current consumption was only 0.07 A and the power supply started up normally. It’s not clear at all ... but does the laptop’s current consumption mean it’s not enough for him? I didn’t want to, but I’ll have to surf the Internet, remove all the sealant in order to check everything.
I measured the PWM controller, the protection obviously worked there, but the protection turned off when the network capacity started to discharge, but I was not even twitched to check the voltage on it.
An internet search turned up the following:
check the voltage on the mains electrolyte if it is more than 450 V ( and where are there so many?), urgently change 2 film capacitors 474 nF 450 V and you will be happy
Red containers for replacement 
Voltage on the network capacitance.
So it is, the voltage on the mains capacity is 496 V, everything fell into place. Such an idle voltage is very high, the PWM controller sees this and goes into protection, and if the mains voltage is turned off, the capacity is smoothly discharged, reaching normal values and the power supply starts up for a short time. That's where 19 V came from if you turned off 220 V. And when I started the PSU even under a small load, the voltage did not jump up like that and the PWM did not go into protection.
It was possible to finish this, to replace the film containers, with which, as it turned out, there were serious problems.
15% of the capacity lagged behind the first. 
The second retained 68% capacity.
But it became interesting where almost 500 V came from on the hot side of the power supply and where did these two capacities come from. The Internet helped again, I did not want to open the entire BP in search of an answer. The information was found on the forum, everything was explained by the phrase:
It stands there passive power corrector. if the metal-paper capacitors in the corrector circuit fail, and the corrector goes into overdrive, the voltage on the network bank drops above 500 volts. Therefore, if you just replaced the network bank, then it will not work for long. It is necessary to bring the voltage of the corrector back to normal or completely eliminate it.
It remains to buy and replace the containers, but here, too, not everything is so simple.
The Chinese had containers with such a denomination and dimensions, but we don’t. There were only 400 or 600 V. More - no less, but the left capacitance is just 474 nF 600 V, but how to put it in instead of those in the middle. There is not so much space there, and even at 400 V there was not smaller size. Moreover, the sellers assured that in such small dimensions, the Chinese were unlikely to be able to shove a quality one, which is why they failed. I had to choose the size. The right tank was well suited in size, but it was 330 nF 400 V, so I had to install them.
When buying a laptop or netbook, more accurately calculating the budget for this acquisition, we do not take into account further related costs. The laptop itself costs, say, $500, but another bag is $20, a mouse is $10. When replacing a battery (and its warranty life is only a couple of years), it will cost $ 100, and the power supply will cost the same amount if it burns out.
It is about him that the conversation will go here. One not very wealthy friend recently stopped working power supply for acer laptop. You will have to pay almost a hundred dollars for a new one, so it would be quite logical to try to fix it yourself. The PSU itself is a traditional black plastic box with an electronic pulse converter inside, providing a voltage of 19V at a current of 3A. This is the standard for most laptops and the only difference between them is the power plug :). I immediately give here several power supply circuits - click to enlarge.
When you turn on the power supply to the network, nothing happens - the LED does not light up and the voltmeter shows zero at the output. Checking the power cord with an ohmmeter did not give anything. We disassemble the body. Although it’s easier said than done: there are no screws or screws, so we’ll break it! To do this, you need to put a knife on the connecting seam and hit it lightly with a hammer. Look, do not overdo it, otherwise you will cut the board!
After the case slightly diverges, we insert a flat screwdriver into the gap formed and forcefully draw along the contour of the connection of the halves of the case, gently breaking it along the seam.
Having disassembled the case, we check the board and parts for something black and charred.
Continuity of input circuits mains voltage 220V immediately revealed a malfunction - this is a self-restoring fuse, which for some reason did not want to recover when overloaded :)
We replace it with a similar one, or with a simple fusible one with a current of 3 amperes and check the operation of the PSU. Green LED lit up, indicating the presence of a voltage of 19V, but there is still nothing on the connector. More precisely, sometimes something slips, like when a wire is bent.
You will also have to repair the cord connecting the power supply to the laptop. Most often, a break occurs at the point of its entry into the case or at the power connector.
We cut off at the body first - no luck. Now there is no contact again near the plug that is inserted into the laptop!
A hard case is a break somewhere in the middle. The easiest option is to cut the cord in half and leave the working half, and throw out the non-working one. And so he did.
Solder the connectors back and test. Everything worked - the repair is completed.
It remains only to glue the halves of the case with glue "moment" and give the power supply. The entire repair of the PSU took no more than an hour.
ASUS in the computer component market is primarily known as one of the largest manufacturers motherboards- in terms of their supply, it is in the top three along with ECS and Gigabyte. However, recently ASUS has decided to release other products under its own brand, which were previously unusual for it - for example, cooling systems, cases and, which is especially interesting for us in this case, power supplies.Three power supplies from ASUS - A-30F, A-30G and A-30H were tested in our laboratory.
Power supplies
In this article, I beg to differ standard scheme consideration of each power supply separately - the fact is that, as shown visual inspection, all three units have absolutely identical electronics, and differ only in cooling systems.As you know, the classic and most commonly used power supply cooling scheme is active cooling using an 80 mm fan located on the back of the unit and drawing hot air out of it. This scheme is simple, cheap, but, unfortunately, on high power units it is relatively inefficient either in terms of cooling or in terms of noise generated during operation.
The fact is that in any ATX power supply there are four elements that need forced cooling - a group stabilization choke (in the photo below it is marked with the number "1"), a radiator with output diode assemblies (2), a power transformer (3) and a radiator with key transistors (4), on which the transistor of the stabilizer on duty is also often located (the photo shows a power supply not from ASUS, but from Codegen, model 250X1 - due to the lower mounting density, individual components are better seen in its example).
The hottest elements are the group stabilization choke and output rectifiers, however, in the same classic design, they are located just to the side of the main airflow created by the fan (generally speaking, I have seen power supplies in which these elements were located on the same side, as the fan, but these were single copies). Thus, in powerful block power supply, in which, accordingly, a greater amount of heat is released, for acceptable cooling of the entire volume of the block, it is necessary to increase the air flow, that is, the fan power. However, along with the power of the fan, the noise it produces also increases, which does not suit many buyers ...
The younger model, ASUS A-30F, was made according to this scheme.
Pay attention to how the ventilation holes are made in the inner walls of the power supply - they are not located on one wall (usually the back or top), as in most blocks, but are distributed along different walls so that the resulting air flows cool the entire power supply. Separately, small holes are made for cooling the passive PFC inductor.
The simplest and cheapest way out of this situation - installing a second fan on the back of the power supply - is not very effective and is usually used in inexpensive power supplies. The second fan is placed coaxially with the first one (or, at best, with a slight shift to the center) and somewhat improves the air flow of the power transformer and both radiators, since the air flow from it blows directly onto them. The photo below shows the implementation of such a cooling scheme using the Codegen 350X power supply as an example:
In more expensive blocks, both in newer models from Codegen and in those discussed by ASUS, other schemes for improving cooling are used. Firstly, these are blocks with two 80 mm fans that have gained a fair amount of popularity, one of which is located in the usual place, and the other is located on the top wall of the power supply, and usually it is shifted to the center of the lid so that the air flow from it blows not only radiators, but also a group stabilization choke located on the side of them. This, as well as the fact that the flow of cold (relatively, of course - after all, it is taken not from the outside, but from the computer case) air is directed directly to the radiators, allows you to seriously improve the cooling efficiency and, accordingly, use less efficient and quieter fans.
According to this scheme, a more expensive model from ASUS - A-30H is made. Instead of stamped gratings, the fans are now equipped with wire ones, which also has a positive effect on the noise level.
Although, of course, the vents from the top cover have disappeared - now they are replaced by a fan - on the back cover they have been preserved in the same place. A number of holes also remained next to the passive PFC choke.
And, finally, the fourth power supply cooling scheme, which has also gained notable popularity recently, although it is inferior in popularity to the two-fan scheme. In this scheme for top cover a large 120 mm fan is installed, which, firstly, occupies most of the cover, and therefore evenly blows all the components of the power supply that need it, and secondly, at relatively low speeds it gives a fairly powerful air flow. Therefore, there is no need for a fan on the rear wall - in such a block, a perforation is simply made in its place. AT model range ASUS made the A-30G block according to the scheme with one 120 mm fan.
Of course, the back wall of the power supply is now made deaf - it does not require additional air intake, on the contrary, with ventilation holes it would turn out that the hot air from the power supply is blown back into the computer, which is clearly superfluous.
Testing
As I already noted, inside all three blocks are almost identical, so I will describe the contents of one of them (using the example of A-30H), after which I will point out the differences between A-30F and A-30G.
A-30H
The block is made very carefully, which immediately makes a good impression. The inscription on the PCB says that the block is actually manufactured by Enhance Electronics, and as a study of the company's website shows, ASUS A-30F corresponds to the Enhance ATX-1130F model, the A-30G block corresponds to the Enhance ATX-1130G model, and the A-30H block , respectively, is completely similar to Enhance ATX-1130H. The PWM controller chip is also marked - "Enhance 16880A".
At the input of the block, a put LC filter is installed on two chokes, which dampens high-frequency interference from a working PWM stabilizer. The capacitors in the high-voltage rectifier are 680 microfarads, which is enough for a 300-watt power supply. At the output on the + 12V bus, one capacitor with a capacity of 3300 uF is installed, at the + 3.3V output - two 3300 uF each, at the + 5V output - one 2200 uF plus one 3300 uF; all outputs are equipped with chokes.
Radiators are medium thickness, about 2.5 mm - this is more than in most blocks of the lower price category, but less than, say, in models from InWin. Let me remind you that the thickness of the radiator affects its efficiency - the thinner it is, the greater the difference in temperatures between its top and lower parts; in other words, the upper part of a radiator that is too thin will simply not work, since it will not warm up due to insufficient thermal conductivity of the radiator. However, for a small radiator, this thickness is more than enough.
The heatsinks in the A-30H block are T-shaped, however, a noticeable part of the top plate is sawn out so as not to interfere with the installation of the capacitors of the high-voltage rectifier, power transformer and PFC choke.
A-30G
In the A-30G block, despite the absence of PFC, the radiators have exactly the same shape as in the A-30H, but in the single-fan A-30F they are already made in the form of vertical plates with "fingers" at the top. The reason for this is clear - due to the lack of a fan on the top cover of the unit, they can be made taller by using cheaper flat radiators instead of T-shaped ones with the same cooling efficiency.
A-30F
All three units are equipped automatic adjustment fan speed (or fans, in the case of A-30H) with a sensor mounted on a heatsink with diode assemblies. Measurements of the dependence of the fan speed on the load on the power supply, shown in the table below (all measurements were carried out at a room temperature of 21C, after setting the required load power, the power supplies warmed up for 15...20 minutes), showed that the adjustment works quite efficiently .
The twin-fan A-30H turned out to be the quietest unit, but the A-30G could not compete with it - despite the relatively low speed of its 120 mm fan, its impeller at close to maximum speed made a distinctly audible buzz, combined with the noise of a powerful flow air. Of course, the cheaper A-30F could not compete with the A-30H - its fan speed reached almost 3000 rpm.
However, the large fan power in the A-30G unit can be considered both a disadvantage and an advantage - it all depends on the point of view. The fan used in it Adda AD1212MS-A71GL on top speed rotation creates an airflow of about 80 CFM, which is more than double the capacity of the fans in the A-30F (about 38 CFM at maximum speed) and A-30H (about 31 CFM for the fan on the back wall and 22 CFM for the fan on the cover of the unit) . Thus, A-30G will provide excellent cooling not only for itself, but for the entire system unit.
Voltage ripples in all three blocks were observed at two frequencies - at the frequency of the PWM stabilizer, that is, several tens of kilohertz, and at twice the frequency of the supply network, that is, 100 Hz.
Bus +5V, 10 µs/div.
Bus +12V, 10 µs/div.
At the operating frequency of the PWM stabilizer, the oscillation range turned out to be very small - it barely exceeded 15 mV, which at acceptable level 50mV on the +5V rail and 120mV on the +12V rail can be considered negligible.
Bus +5V, 4 ms/div.
Bus +12V, 4 ms/div.
But with oscillations at a frequency of 100 Hz, the situation was somewhat worse - their swing at the maximum reached 40 ... 50 mV on the + 12V bus and 20-25 mV on the + 5V bus. However, these figures are in any case noticeably below the acceptable limit, so there is no reason for concern; this can be explained by the not very successful design of the board or power transformer (third possible reason- lack of capacitance of the capacitors of the high-voltage rectifier - here, obviously, it disappears immediately).
The stability of the output voltages depending on the load was measured in two stages. The fact is that all three models from ASUS differ from the "standard" 300-watt power supplies by an increased current of up to 18A on the +12V bus. This was done in connection with the greatly increased consumption of modern computers on this bus and was done not only in ASUS / Enhance blocks - for example, new models from Zalman with the "B" index (ZM300B or discussed in the last article ZM400B) also have a maximum admissible current on the bus + 12V to 18A. At the same time, the absolute majority of the previously tested 300W power supplies have a maximum allowable current on this bus of 15A, as recommended by the ATX standard; therefore, in order to be able to compare the results of ASUS blocks with the models tested earlier, the first series of measurements was carried out at a maximum load current of about 15A, and in order to evaluate the capabilities of the blocks at maximum load, the second series was carried out with a load already of about 18A. The tables below show the average results of all three units, and the graphs show the results of the A-30H model.
As you can see, the blocks show very nice results both under the "standard" load, and under increased. Unless the voltage spread on the +3.3V bus is relatively high, however, this bus no longer has significant significance in modern computers - most powerful consumers with low-voltage power supplies are equipped with their own stabilizers (for example, CPU and GPU video cards). Moreover, it is worth noting that, despite the artificiality of our tests (such large fluctuations and load imbalance, as at our test bench, do not occur in a real computer, and therefore the spread of voltages produced by the block in it will be significantly less), none of the output voltages block has not gone beyond the limits allowed by the standard (± 5% of the nominal value).
In conclusion, it is worth noting that the blocks are equipped with six power connectors for ATA hard drives or CD-ROMs, two power connectors for SerialATA devices, as well as AUX and ATX12V connectors. The AUX connector uses 16 AWG wires, in all other connectors, except for non-critical to the maximum currents of the drive power connectors - 18 AWG.
The power supplies are shipped in a plain white cardboard box, and only four inch-threaded bolts are included to secure the unit.
Conclusion
As the test results showed, power supplies sold under the ASUS brand are able to take their rightful place in the market due to their high quality workmanship and very good parameters.The presented models in tests showed results on par with products sold under the brands FSP, Zalman, InWin and others that have already won recognition from customers. All three models belong to the middle price category and are not equipped with either gold-plated connectors, or multi-colored fan illumination, or other external paraphernalia, which are very popular lately, but do not affect functionality and quality of work, so they are perfect for people who need a quality unit. power supply, but not willing to overpay for an abundance of blue LEDs or gold-plated fan grilles.
I have to admit that the most interesting model is ASUS A-30H, equipped with two 80mm fans - thanks to high-quality fans and effective fan speed control, the unit turned out to be very quiet.
Unfortunately, the ASUS A-30G with a 120mm fan wasn't quiet, but it provides a very powerful airflow, so it's well suited for those who care more about efficient cooling than silence. However, with a relatively small load, the fan of this unit reduces its speed to a level at which it is very quiet.
The ASUS A-30F model, in turn, belongs to the middle class both in terms of cooling efficiency and silence, however, due to a lower price and exactly the same electrical parameters, like the "older brothers", also has a good chance of success.