Обслуживание контроллера BLDC. Повышение мощности контроллера электро самоката.
Hello guys! Andrey is with you again. And we continue to talk about the best type of transport - electric transport. And about the best type of electric transport - electric scooters. Today we will discuss such an important topic as controller maintenance, zero maintenance, its cooling and preparation for overclocking. Or as it is also called - forcing. This service has already been carried out based on my experience of driving electric vehicles and experience with other controllers. That is, these are some new, more advanced technologies that I
applied in 2024 on the latest controllers. Out of a number of controllers that I had, and there were five or six of them, I still decided to settle on this YunLi controller. It is written that it is 50 plus or minus 1A. Well, this is all a gainful thing. Let's talk about this a little further. And, accordingly, the complete trigger for it is QS S4, with analog control and a six- wire connection. This controller turned out to be the most optimal of those I tested. The one that most satisfies my requirements for a controller. For the requirements, see
the previous video on the channel, it’s called “How to choose a controller.” And so, since I settled on it, that is, it suited me the most. But again, this is a budget controller. It is not ideal, and accordingly it is necessary to bring it to mind so that it travels for a long time, far, reliably, and without failures. And how to do this - we will discuss with you today in this video. I already bought this controller as a spare, let it sit. Sooner or later, his time will come. And the serviced controller is already in the scooter. I drive it safely, I’m very happy and
I don’t know any troubles or sorrows. I drove it in stock, for 1000 or 2000 kilometers. I see that it suits me, and accordingly I need to disassemble and service it. And this, let’s say, is a rather long process, not easy. And everything must be done correctly. Because even assembling the battery was not as difficult as servicing the controller. That is, it was long, but in terms of complexity it was slightly lower. Or not a little bit more than servicing the controller. Then, in descending order of complexity, there were some adjustments to the frame, finishing touches. And probably the simplest thing is
engine maintenance. But with the controller a lot of difficult, non-standard problems arose. And they had to be solved and dealt with. And therefore, all this must be done very thoughtfully, without rushing. Because if you make one mistake, you can basically throw away the controller. The first step when servicing the controller is to discharge the capacitors before disassembling it. So that, first of all, they don’t shock you. Because at 70V they already bite a little. And secondly,
so that they don’t burn something on the board. For example, a processor that is powered by, well, only 2V-3V, new types of processors. This means that the capacitors are discharged in two stages. The first step is you turn off the power to the battery while the trigger is on. It is advisable to do this with Smart BMS so that there are no sparks on the contacts. And after a second or two, the trigger takes over the power, goes out, and what remains on the capacitors on the board is no longer 70V, but not zero either. I took measurements, there remained
somewhere around 12-13 volts. And, if you somehow miraculously short-circuit the tracks somewhere with a screwdriver or some wire, and 13V gets to the processor... Well, it will quickly burn out, you understand. Therefore, the second stage of discharging condensates is to take an ordinary light bulb. Well, 60 W, 100 W, incandescent lamp... The more powerful, the better. If there are 200 to 300 - generally great. The lamp is in the socket, there are wires coming from the socket. And here, on the controller connector, you place the wires and
hold for 10 seconds. The capacitors through the lamp are additionally slowly discharged. And as a result, according to the measurements that I made on the board, the positive and negative power paths - the voltage remains below Volt. That is, there is somewhere around 0.8-0.9V, after 10 seconds of exposure to the lamp. This must be done and is necessary before any
maintenance work on the controller. Because if you don’t do this, there may be trouble later. Then comes the actual disassembly of the controller, the removal of this stupid plate, which no one needs for nothing. And after removing this plate, we can take out the board, see what and how it is arranged, and what it is. The first thing I did, of course, was photograph the board. AND, To be honest, it was not easy to do from the phone. But I set the maximum resolution, turned on the most powerful light, installed a diffuser, and got these two wonderful photographs that you now see on the screen. These pictures gave me the opportunity
to see how it works there, what details, some inscriptions on the parts. And be able to send these pictures to my colleagues for consultation so that they can suggest something to me. And in the future it will be more efficient to maintain this controller and work with it. Naturally, immediately after I took a photo of the board on both
sides, I photographed the transistors and looked at the markings of the transistors. Because this is a very important point, you need to understand what you can count on. These are the photos of transistors. I looked at them, and accordingly the next step was to go online and look at their specifications. So, what do we see from the passport? We see a tolerance of up to 100V,
a resistance of 3 mOhm, and a current of one key of 120 amperes. The performance is certainly not as remarkable as that of the best SMD transistors. But still, these are not SMD transistors. That is, the output comes on three legs, firstly. And secondly, this is a budget controller. Therefore, I believe
that these characteristics are more than enough. 100V is a good voltage margin for reverse surges from the battery. Let's say when the electronic brake is working. 3 mOhm is also not a very bad indicator for resistance. And 120A is, in principle, a fairly decent current indicator.
If we are talking about the controller from an old scooter, and about those mosfets that were in it... In my opinion, the voltage tolerance was only up to 80V, the currents were something in the region of 100A, not 120, and the resistance was 5.5 mOhm. This is a 6-feet controller with six keys. And here we already have 12 keys. And accordingly, in each phase there are four keys. Two keys operate in the same direction of engine rotation. two keys on the other. Four don't work, two work. But,
two keys, what does this give us? Two switches are no longer 3 mOhm resistance, this must be divided by two, this is already 1.5 mOhm, yes, instead of 5.5. And the current is no longer 120A, but 240. Since two keys in a pair work, and 240 A is no longer 100, yes, it’s logical... Therefore, in principle, for this level of controller, I think that the keys are more than quality. Then I looked at the graphs of resistance versus current. It is almost horizontal. Here's the first graph. And this, as it were, cannot but rejoice. That is, when the current changes from 10 to 100A, R of the switch practically does not increase. Well, that's according to the passport. How true this is is difficult to say. Well, you see the data from the factory
on the screen. And the characteristics of the key resistance at a current of 50 amperes depending on temperature. And we see that the lower the temperature, the lower the resistance of the key. Naturally, it heats up less and the efficiency of the controller increases. Which also cannot but rejoice. The question is - how to ensure a decrease in this temperature? Because in stock,
if we are talking about such a box, the controller on the old scooter heated up to a peak of 140 degrees, as I saw on the sensor. This one is somewhere around 105-110. How to do 50... constantly... so that the electronics don’t strain - that’s another question. And in this video we look at just that. In this screenshot, we will pay attention to the tenth graph, and we see that the manufacturer stated that the current characteristics begin to drop when the temperature reaches 100°. And remembering the passport for the previous controller from the old scooter - there the current characteristics went down after 60 degrees. That is, 60° for those keys was already the limit. And it was generally undesirable to go over 60. I’m not even talking about 140 anymore. There is a supply of up to a hundred, but again, this is a passport from the factory. How true it is, I don’t
know. Therefore, the lower the temperature on the keys, and the more stable it is, the better. While connecting a Chinese wheel motor with a bad Chinese winding, the motor jerked for a long time, and I had to turn on the training mode eight times... And the diodes on the board that are involved in the phase bridges, in the operation of the phase bridges, failed. And one diode burned out, you can now see its photo on the screen. Naturally, this diode had to be changed somehow. It's good that not all three burned out. Two remained intact, and with great difficulty on these
tiny details, two millimeters, from the phone, enlarging the image 10 times, I was able to read their markings. Then find, order, and accordingly I resoldered this diode. Attaching this terrible plate of keys directly to the body did not suit me at all. This technology was made through ass... And from the start I decided to do it the right way . I ordered several cases like this in advance for attaching the thermal buffer with bolts from the outside. And accordingly, here is this side, it is not convex, but was originally intended for attaching keys. It is flat, and accordingly, here I can already put the buffer that I consider effective. And separately attach the mosfets to the buffer,
then insert them with the buffer into the case, and attach the buffer itself to the case. I found about four of these boxes . There is already one with such a controller in the scooter, it works wonderfully. And I still have three more of these left. Well, naturally, these buildings must be found, this is a rarity today. Secondly, the board itself must be clearly adjusted to the dimensions so that it does not move anywhere or dangle. Squeeze it into this key slot. And, accordingly,
in the stock controller, where this bulge for the keys goes, it is not there. So we need to make a buffer here . Well, naturally, I never planned to make this buffer from aluminum. The copper one stood up , stood up very well, and I’m pleased with the results. Well, here we only talked about the body, and about the buffer, respectively, and how it looks, how it stood up - we will talk a little later. Here, for now, I’ll just note that since the body is different, its holes are different, the sizes of the covers are different, then, accordingly, it was necessary to resharpen and make different covers for this body. That is, other lids were taken, holes were drilled in them, and they were adjusted
to fit so that they would fit normally. Well, and accordingly, holes were also made in these covers in order to place fans on the covers for flow-through cooling, to quickly remove heat from the thermal buffer. This is what the already bored cover looks like, prepared, with holes of the same size, and with a circle for installing fans. Naturally, zero maintenance of the controller is impossible without preparing the trigger for driving. And this must be done. Because if the trigger is serviced, then at some point you might run into trouble))). As for the trigger... Well, naturally, it was disassembled and examined inside. All these four
nuts were glued with super glue. so that they don't hang out. Next, which is very important... Here in this place inside the trigger there is a magnet glued into the trigger. But it's glued on with some kind of Chinese crap glue. And this glue over time, say in the summer at a temperature of +40, it becomes softer, and this magnet falls out. This story doesn't suit me at all. Therefore, when removing the trigger, naturally I glued this magnet, sanded the glue, and now this magnet is held tightly in the trigger. Inside the trigger itself, I adjusted the spring stiffness. Because in stock it is very hard. It is difficult to pull the trigger;
you can exceed the force and the throttle will be sharp. So I bent the spring. So that, firstly, she does not jump out of her seat. This is very important. It can jump out - and the scooter will move with your participation. Secondly, I moved the bend lock to a different place to make the trigger softer. Now I’m reviewing my archive, unfortunately I didn’t take any photos while servicing the trigger... now I regret it (((. Therefore, I have to talk about the service a little on the fingers. What is also important to note is that you need to fix this
bolt. Because it is not fixed in anything in the stock, and at some point... the trigger is constantly rotating... this bolt can get unscrewed, and the trigger will simply fly off onto the asphalt. Therefore, when servicing the trigger, you need to put this bolt on the lock so that it stays there Well, of course, the entry of the parts into each other, all these places in a circle... they are covered with a very, very thin layer of silicone, neutral sealant RTV-904. So that there is such a rubber gasket, and there is no water access there. I threw out the useless USB port from the board, broke it off along with
the wiring, and also put it on 505 glue. So that they do not break off from the vibrations. So that they are securely fixed there, not only with wire, but so that the glue also captures the insulation, so that it is reliable. Well, of course, these places... I don’t know why this nonsense is here, round. This is the place for the USB output... I went around these two points with super glue. That is,
he made them airtight so that water wouldn’t get in there either. What else do I often encounter when servicing triggers? You see, you get this joint, a clamp, tightening the nuts with screws... It turns out that when you have already squeezed it all out, the trigger still rotates on the steering wheel. That is, the twenty-second steering wheel, but it happens that the screen is designed for the twenty-fourth.
That is, this clip is not enough to squeeze out on the steering wheel. And often you have to grind down this area. You see, no matter how much the nuts are squeezed out, the plastic itself seems to be at an angle. And if you squeeze it, it has already rested in the opposite direction. And it turns out that
in order to have some kind of reserve, in order to compress the twenty-second steering wheel, you have to take... well, a file won’t fit here... it turns out that I first took a utility knife, I cut off the excess plastic on both sides of the trigger . And it was a little uneven, the file wouldn’t fit there. I took diamond files, I have a set, I recommend you buy it too)) And with these diamond files I already brought it level so that it would be beautiful. And after that, then this circle... I don’t know, I think it’s designed for 24 diameter... Then it already normally compresses the 22nd size pipe. Well, in order not to wrap the entire steering wheel with electrical tape, so that it would be somehow more or less human... These are the basic things
that came to mind... But definitely not all)) About servicing the trigger. Here we talked about the main aspects of servicing the controller and preparing for its servicing. Well, in the next video, at a more expert level, I will talk about methods for implementing our plans. And what results all this brought. For today I don’t say goodbye to you, but say “goodbye.” And - wait for the release of the next video. Bye bye!
2025-02-20 21:04
