This is indeed from some of my previous power supply projects, but closely related to the LED before disassembly records. Now we all go out and buy summer\'s little lace lamps, which are powered by solar energy during the day, charged during the day, and turned on at night, and they become a side garden lamp. Of course, their life is limited, because cheap imported products, in the old weather in Britain, battery packs fail, sometimes even just solar panels fail. Usually, you buy four or more packages of these things, and the light source is a single, low-power, inexpensive variety of LEDs. Once they die, we throw them into the dustbin and then into the landfill. Well, let me think, why not convert it into a 10-watt LED power supply? But it must be safe and free from the weather, and it must be cheap. I wonder if this can be done? Will 10W be too much? You can see from the picture that the light source is a stainless steel tube design with a diameter of about 60 mm and a plastic diffuser. Plus another tubular lid, which is mounted on top with solar panels inside. The first thing I did was to remove the white LED and square solar panels from the roof. The idea is to install the LED on a panel fixed to the radiator, which passes upward through the opening of the solar panel. Recently, I purchased some 10W single-core COB LEDs. I want to know if I can use single-core LEDs to use switching power directly from the power supply. [240V without insulation] The candidate is a Buck switch mode power driver chip FL7701 and sensor 1. 4mH coilcraft. Unfortunately, the FW converted from 240V to COB block[12V] It is not easy to work because the current required to pass COB is much larger than 10W that can be processed by the driver chip. The chip can process 0. The 5A with a forward voltage of 12V will only make you reach about 5W. You can do this by using a forward converter switch mode with isolation, but the cost is starting to soar. After all, all this is considered cheap and pleasant. So how can I get 10W with only 0? 5 A of current. Considering the energy conservation theory, the only way to increase the wattage is to increase the voltage, and the only thing I can do is to increase the forward voltage of the LED by using more than one of them. If you look at my LED disassembly indicator, you will see why they did it in that design. On eBay, I can easily find some 1W LEDs with a forward voltage of 0F 3V@330mA. Now, if I run them at 266ma using 10 and below, I end up with 10 x 3 x 0. 266A=8W. . . close enough. There is a bipolar approach to under-load operation. . . Keep the heat down and conserve or prolong life. Lower intersection temperatures mean happy lights. Looking at the photos of the garden lights, what we need is a way to install these LEDs. Of course, if they sink 266 mA, we need to eliminate the 8W energy between them, which requires a radiator. The inner diameter of stainless steel pipe is slightly less than 57 mm, so if I can install any electronic component in a sealed plastic pipe and install it in the pipe. Then I can install the LED board downward on the top of the housing and illuminate the diffuser. So, how do we arrange the LED? First, I cut a 46. Use a circular saw to cut a 5 mm aluminum circle through the central hole. [see pic] Cover one side with double-sided heat dissipation tape. You can buy this tape on eBay. It\'s quite inexpensive. It\'s usually used for radiator accessories. Look at the picture. Aluminum is an old power housing, but you may be able to buy it on eBay. I used a 2 mm thick piece. You need to cover and insulate the metal from the base LED, but still have good thermal conductivity. Use double-layer vertical hot tape. This will change the thermal conductivity. We lose another 20 degrees Celsius at the intersection, but that\'s what we need. I\'ll revisit this later, maybe I\'ll look at a completely sealed water-soluble solution, but not now. Then I use AutoCAD to position the LED on the base. Please refer to the attached PDF picture. I printed the design plan proportionally and used the punch to make the installation template for the layout as a rough guide. I put this on the sticky bottom and drew the outline of the circle on the tape. Next, I arranged the LED so that I could get the location of some copper tapes, which I would use to connect the LED to the surface of the insulating tape. In order to ensure that no copper bands invade the bottom of the slug, I welded them together. Of course, you need to make sure that the cathode goes to the anode. You can stick them down and use some wiring between the pins, although using copper tape helps to dissipate some heat into the tape. In the case of heat, these generate a lot of heat, so a considerable radiator is needed. I chose a 40 x 40 x 30 hour radiator to keep the floor around 58. -60 degrees C. In this case, his size matches the removal of solar chips, so that the heat at the junction between the LED housing and the housing is about 4 degrees Celsius/Watt, and the heat from one panel to another is 1 degrees Celsius/Watt, which means that the temperature at the junction should be 1 degrees Celsius/Watt. （8x1)+4= approx. 60 + 12 degrees centigrade = 72 degrees centigrade, which is reasonable. The total voltage of the whole LED will be about 10 x 3 V, so the current passing through them will be tested in the next stage. The attached PDF has an outline that can be used as a template, but you can always design it yourself. Check easam attachments, and you can download eviewer to perusewe. We\'ll use a fl7701 driver chip and play with the Xcel spreadsheet designer to come up with a set of numbers that might work. The key of Buck converter is to reduce the ripple to a reasonable level given the RMS value we need. Ripples directly affect the size and frequency of inductors, and indirectly. Therefore, if we increase the ripple, we must increase the inductance size. The only way to reduce the required inductance is to increase the frequency. See the attached picture, which lists what I\'m iterating on and is the key to the values on the schematic. This is solder LED placed on my template before pasting. Please pay attention to the use of the radiator. The radiator plate and the installed LED card are at the bottom. By adjusting the peak current to 500 mA and increasing the current to 266 mA rms, the voltage on the LED is set to slightly higher than 30V, which means that if we have 10 LEDs, the voltage is actually close to 3V forward. Note that the calculation is expected to be 286 ma, while in fact we only managed 266. The frequency should be 101kHz, but it seems a little low in scope. I\'ll discuss schematics, drivers, and waveforms in the next step. So plugging in the power lights up the floor like a Christmas tree. Caution! This is a non-isolated design, so all devices that can be upgraded to the power level need to be thoroughly grounded. This will include heat sinks. If you look carefully, several holes need to be automatically tapered to the heat sinks through the grounding tag, stainless steel metal parts and input power grounding. Pay attention to the wiring of the LED to ensure that there is no short circuit between the LED and the ground. If this is the case, there will be voltages larger than the design voltage on the LED and they will be destroyed soon. I have a test device that has a main isolation transformer, but when connected directly to the main power supply, one side of the inductor is at the main power supply potential. If it is connected to any piece of isolated metal, it will be a danger. So let\'s look back and see what we need to drive the LED. We\'ve already said that we need to support around 266 mA, so we\'ve done that. Reference schematic notes are as follows: through the fuse 1 into the bridge rectifier, and then with two C filter inductors. D1 is a recovery diode and a method of reducing inductance current. Q1 gate is driven by pin 2 of FL7701 through R3, and D2 assists in clearing charge from the gate during negative stroke of FL7701. The output frequency is set by R5/R4. The two pins have some decoupling and CS pins. . Pin1 is a current induction, which monitors the voltage and therefore the current through r6. Refer to the peak current of R6 0. 5a, which will lead to IC reset and prepare for the next turn-on stage. Notice what\'s missing in this circuit. The input end does not need a large DC cover of the rectifier. FL7701 handles input changes skillfully internally. Considering that this is usually an expensive part, it helps to save costs. Once the printed circuit board was filled, I checked the ripple. The current probe is used on the cathode of the LED module, and the ripple is 150 mA. The average current measured by the instrument is about 1. 5 mA. 260mA. For LEDs, the maximum current is reduced by 100 mA, making them cooler, thus prolonging their life. Frequency measurements were 81 kHz and the slope decreased to 1. 71us. This is 13% of the chip/sensor capability, so it should be good. The starting point of the whole design is to use 1. The four-hour coil process showed that there were some errors on the prototype board. I corrected the layout of the newly uploaded printed circuit board. Note the jumper above to bypass some incorrect fixtures. . . . doh. This caused some explosions before I realized the mistake. . . Must be tired! There are two upper parts and one lower part. So here it\'s sewn together. I will enclose a list of all the parts I need at the end. Some things need attention. I grounded the top of the radiator and sent it through the device to a grounding point at the bottom. Then it is grounded to the power supply. Watch it. Finally, the cathode voltage of the LED is less than 310V and the peak power supply voltage is about 30V. This can damage if contacted, so isolation and grounding of any metal parts that may be touched are needed to ensure that the fault current path is clear. Attention should be paid to the top and bottom of the cable seal sleeve to prevent water from entering the electronic equipment. The grounding screw at the bottom acts as a stop for the \"cannister\" of the power supply and has a drainage hole to prevent moisture from entering. This is not a waterproof container, but the main pipe is blocked out of the finger and the drain hole is very high from the ground. The top radiator needs to be sealed around the top, which needs to be completed. I\'m going to put this in the garden in summer and add some more later.