Question: I only talk about the cold technology that is not related to electricity in the production of power supply. Because I am a layman to the power supply. I know that there are many exotic flowers and plants in the power circle, but because of limited knowledge, I can only look at the flowers and plants outside the door. In groping, there are some small methods of their own, here to show you the tricks. Slightly speaking below, the text is not clear, please forgive me. I only have junior high school culture.
The day before yesterday I chatted with the friends in the group about the temperature of the transformer. A friend said that the temperature is hot, but I don't know how much it is. So I talked about some of the methods I use to measure temperature everyday, but there are actually several. Summing up a post can be regarded as experience sharing. To measure the temperature of the transformer, the book is about the thermal measurement method. According to the resistance value of the copper wire at different temperatures, the resistance of the coil is quickly measured, so the temperature of the coil is calculated. This is okay in theory. In fact, each copper wire has different purity, and the temperature curve and resistance are not exactly the same. Therefore, the combination of various errors can have a difference of 10--20 degrees Celsius. Although the lookup table can be counted. But it is only a rough calculation. In addition, a bridge must be used to quickly measure the thermal resistance of the coil immediately after the power is turned off, so as to avoid the temperature drop affecting the measurement result. Having said that, if you buy 1,000 electric bridges, you don't care about dozens of infrared thermometers.
The picture is a screenshot of the network. According to the diagram, you can see that the relationship between resistance and temperature is almost linear. I have done similar experiments myself, but the heat dissipation is too fast, which leads to inaccurate temperature measurement. Although the figure shows the thermistor. But it also illustrates the feasibility of the bridge measuring temperature. The same can be used for temperature measurement on copper wire.
Let's talk about infrared thermometers, dozens of cheap ones on Taobao and hundreds of expensive ones. The advantage is accurate temperature measurement, and the disadvantage is obvious: only single point temperature measurement. And it is only at a certain distance that the laser point exactly coincides with the temperature measurement point. For temperature measurement of large targets, this is very useful. For example, measuring the temperature of the tin furnace, the temperature of the oven. For the temperature measurement of small parts, it is often impossible to accurately align the target. Always find the highest temperature in a dangling way, to be sure that the hottest part is a certain place. It takes more time, but it is cheaper after all.
The picture comes from the Internet. Please ignore the ads in the picture. As shown in the figure, usually the upper line is the laser line, and the lower red line represents the direction of the temperature measurement area. Usually it will maintain a good consistency in the range of 20--100 cm. Farther or closer, there will be a problem that the laser point and the measured point cannot coincide.
However, inexpensive temperature measurement methods do not accurately show which part has the highest temperature. Is there a cheap and precise method? When I read the newsletter in the 1990s, I used rosin powder to spread evenly on the parts to measure whether the temperature of the parts was too high or abnormal temperature rise. The theoretical melting point of rosin is 110 degrees Celsius, and the temperature in the powder state will be lower. You can see the change of rosin. Therefore, as soon as the rosin powder found in Nasa begins to melt, it means there is a problem in this area. Overhaul will have a direction. In the early years, this method was often used to repair power amplifiers. very effective. Some amplifiers have a large quiescent current. If it is abnormally hot, it will burn out at any time during work. No infrared thermometer has been seen in that era. At most thermocouples.
The picture comes from the Internet. A pot of braised rosin starts to soften at a temperature of 70-80 degrees. 180-190 degrees began to vaporize.
The picture comes from the Internet. Of course not my soldering iron. Nor is it my hand. The smoke is the rosin in the solder wire and some of the flux. Most are rosin.
With the progress of the times, technology has developed. There are now cleaner and simpler test methods than rosin powder. The key is the cheapest. Five dollars of material can be used for a whole year. Unique trick: thermal printing paper. This small white label like an ordinary sticker is usually used to print small tickets in supermarkets. The principle is to irradiate the paper with an instant high temperature light spot. When the local temperature exceeds 65 degrees Celsius, it begins to change color, 80 degrees Celsius begins to become dark gray, and 100 degrees Celsius is close to pure black. Cut a few small pieces of a sticker on the surface of MOS, transformer wire pack, magnetic core, heat sink, PCB. After a dozen minutes to half an hour of aging, after the temperature of each part has stabilized, you can roughly know how much the temperature of all the stickers is. The cost is not even a penny at a time.
A roll of 3--5 bucks, hundreds of sheets, enough for a long time. When I was okay, I used a lighter to burn a black flame.
The method of thermal paper is straightforward. But the speed is slow. I often use it to measure the surface temperature of certain parts of a stable power supply. Only after the power supply has worked to thermal equilibrium, can we know which is the hottest. Which need to continue to optimize. So I have an easier way. Timely and fast. The same is cheap and easy to use. alcohol! Installed in the needle tube, drop a drop on the transformer wire pack after the heat engine. If it bubbles and evaporates immediately, it means that the temperature exceeds 80 degrees. The vaporization temperature of alcohol is 78 degrees, the same can know whether all parts have more than 80 degrees. Direct and efficient. But only a small drop. Two or three drops of measurement in the same place for a short time are equivalent to cooling the parts. The true temperature cannot be measured. Using alcohol, no damage, environmental protection, no residue.
Just like this, the instantaneous vaporization indicates that the temperature is far above 80 degrees. This shocking scene was not made by me.
So, I spent 2600 oceans to buy the cheapest thermal imaging. At this point, all the questions about temperature are no longer a problem. Measure at any time. Even if it is turned on, the temperature on the starting resistance can be measured. Even if the resistance changes by only 0.1 degrees Celsius, it can be significantly different from the temperature of the adjacent PCB board. This is very useful in thermal imaging. I used it to optimize the absorption of RCD, the value of starting resistance, and various other heating issues. But thermal imaging is not a magic weapon. For a glossy aluminum heat sink, the result of the measurement is the temperature of the object with the mirror-inverted color, not the temperature of the heat sink body. Only black heat sinks can be accurately measured. So I will use a marker to mark a large soybean point on the items to be measured, so that I can get almost accurate results. In the same way, I also use it to see whether the transformer's wire wrap heat is still the core heat. Use this to optimize your own transformer.
The principle of thermal imaging actually uses a combination of virtual and real. The current civilian-level thermal imaging uses two lenses, one is an infrared lens, which is used to sense the intensity of infrared radiation in an area. The other is Ordinary optical lens is used to perceive the real optical image in the same area. After the two screens are combined by the software, we can see the infrared radiation intensity map on the real object. Then through the calculation of the software, we can get the temperature of different areas, Displayed on the screen.
Thermal imaging is expensive because the internal is equivalent to a combination of multiple ordinary single-point infrared thermometers. For example, a single-point thermometer is priced at about 50 yuan. An infrared thermal imaging with a resolution of 240 * 240 is equivalent to Integrated 57600 single-point infrared thermometers work at the same time, because of the high degree of integration, we do not need to really spend the price of tens of thousands of thermometers. It only needs two or three thousand in China. Military grade, such as missiles, An infrared heat source needs to be found tens of kilometers away, and the resolution is terrifying. So the price of a missile is usually comparable to that of a luxury car.
The reason why I am very concerned about the temperature, because recently I have made several fanless self-heating power supplies. Various problems during the period caused me to have strict requirements on temperature. When I carefully calculated from the theory and strictly produced a theoretical inductance When the amount of transformer is used, there is a problem during aging.The original design was to work in the DCM mode flyback, but the result is that the waveform has reached the full load state when the load is more than 70%. I think I have calculated the problem. After repeated calculations with different methods several times, the result is normal. So I turned my attention to the state after aging: when 50% power is aging, it basically reaches thermal equilibrium after 30 minutes.At this time, the temperature of the transformer is measured at around 75 degrees. When the waveform looks full, the transformer temperature is about 85 degrees. So I found that ----- the increase in temperature makes the inductance of the transformer larger. Checking the temperature curve of the magnetic core also confirmed this. It can be said that I I bought junk magnetic cores. On the other hand, even genuine magnetic cores will have the problem of temperature affecting the amount of sensing, but it will not be so obvious. In order to verify my idea, I bought the original printed from the Internet. "Magnetic core. After several experiments, it was confirmed that the inductance of the authentic magnetic core is little affected by temperature changes. 25--100 degrees, the inductance changes less than 5%. And the ordinary magnetic core is 10% or more Therefore, when making actual products, it makes sense to design a 10% power margin. You can work at full load under full load \ limit state.
In addition to the problem of increased inductance caused by temperature, there is also the current problem of MOS tubes. Almost all MOS tube specifications give two current values, one is 25 degrees, one is 100 degrees, and some The current values ​​are 25 and 75 degrees. For example, a 10A MOS is 10A at 25 degrees, and only 6A of safe current can flow at 100 degrees. It is also necessary to consider the power supply when we select parts In the limit temperature state, whether all the parts can have enough margin. Not only the transformer and MOS. Our commonly used 431 and resistance diode transistors also have temperature drift problems. How to accurately detect the circuit and use the negative temperature coefficient The circuit to correct the deviation caused by temperature, this is the problem I am facing. But it is also gradually solved.
8 GPU Motherboard In-line BTC Multi-graphics 8 Card Motherboard Mine Board LGA 1150
ETH Large Space Motherboard Mine Board for Desktop Computers
Feature:
1. Good quality: This product is made of high-quality materials, exquisite workmanship, very good quality and long service life.
2. Easy installation: This product is very easy to install, no
additional tools are needed, and the installation process is very
convenient.
3. Match the appearance: This product matches the appearance very well,
and there will be no sense of violation after installation.
4. Stability: The running process of this product is very stable and the stability is very strong.
5. This in-line 8-card multi-graphics motherboard has a large graphics
card spacing of 65mm, without the need to transmit a more stable chip, all-solid capacitors, 100 Gigabit network card, support four or
five generations of CPU, DDR3 notebook memory slot, support 1660 , 2070,
3090, Rx580 and other full series of 10 series, 20 series and 30 series
graphics cards!
Specification:
Main chip: Intel 3865
Network card chip: RTL8111f / 8105
Processor Model:
Support CPU type: Core 2nd and 3rd generation
CPU slot: LGA 1150
Support: 1 CPU
Support memory type: DDR3 1333 / 1600 / 1866 MHz, maximum 16g
Memory slot: 1 x Notebook DDR3 (both high and low voltage)
Expansion slot:
8 x PCI-E full graphics card slot (actual speed is 1 x, slot spacing 65 mm, clearance 55 mm)
5 x SYS Fan socket (90 degrees)
1 x DDR3L memory slot
2 x SATA3 interface
1 x MSATA interface
1 x JPWR Sata1 power supply interface
Input and output port of 16x6p square power supply
1 * 4Pin CPU fan interface (plug in)
I/O interface:
2 x usb 3.0 interface
1 x RJ45 network interface
1 x VGA output
1xpower SW switch
1x1x buzzer
Graphics slot spacing: 65mm
Graphics card support type: common mine cards support, such as RX
series, GTX10 series, GTX20 series, GTX30 series, etc. (except 3060)
Net weight: 840g
Gross weight: about 1300g
Motherboard size: 59.8 * 19cm
Package size: 60cm * 22cm * 7cm
cheap mining motherboard,mining motherboard,best mining motherboard,asus mining motherboard,mining motherboard 2021
Easy Electronic Technology Co.,Ltd , https://www.pcelectronicgroup.com