I have a long time I suddenly want to engage a full-optical coupling, H-bridge high-speed drive circuit, the result is spent, didn't take it out, I haven't tested it, I have been laughed by the netizens. The field, although it is very lost, but it is indeed a lot of things to say. Now, I spend this experience this time. ^ _^
The picture is a little small, everyone is still open, look carefully.
The reason why this circuit can be considered to achieve high-speed optocoupler isolation, in fact, it is to see the high-speed characteristics of the 6N137 (replacement with PS9601 here), because use 6N137 to achieve high-speed characteristics, the result will of course The connection to think that the high-speed H bridge driver is actually not a matter, here I will tell you what questions like this circuit.
First, Although VMOS seems to be a voltage-driven device, it actually needs to have a fairly low-driven internal resistance to make it high speed switch, because VMOS has a very considerable CGS, which is G The equivalent capacitance between the S electrode, the capacity of this capacitor is almost about 2000P, depending on the specific pipe, it is not small, because the existence of this capacitor, the speed of this circuit is never possible. To reach high speed, our M1 pipe in the lower left corner of the graphics card can see that the shutdown is turned off, it may be done to go, but the VGS's rise is slow, because it has an RC Time, almost 10K * 2000P this time is not short. It is conceivable that M1 is slow to say. Let's take a look at the more destined M2. This is a PVMOS power tube. It can be seen that when M2 is turned on, it is controlled by the transistor, so it is still going, that is not going to shut down time. This is really a matter of life. Due to the characteristics of the MOS tube, the shutdown time is much slower than the opening time. The result is that M2 has not shutdked, M1 has begun to turn back, resulting in two pipes, which have a considerable current, can't use a slightly large-scale P-channel M2 this pipe. It will be serious, and slowly, it is overheated. That's why many people say that the H-bridge is easy to burn the P tube.
Second, let's take a look at the control voltage of the VMOS here. Since the working voltage of 6N137 is subject to the working voltage of 6N137, the VGS is only 5V, which is much lower than the fully on-voltage voltage of the IRF640, so that the VMOS is always working in high internal resistance regions, power tubes The heat is severe, so such a circuit is definitely not any practical.
Since the above reasons have been analyzed, how should we correct this problem?
In fact, it is very complicated, simple to say that there is a large circuit after 6N137, and then push the switch and shutdown of the MOS pipe.
You can refer to this circuit of this circuit, but in fact such circuitry is very complicated, it is very uncomfortable, and because the amplifier has a considerable delay, it will eventually destroy us to get high-speed isolation MOS tube drive. It is necessary to use high-speed op amp, or it is very considerable. So do not recommend using
But now it is convenient, it is recommended to use a dedicated MOS tube driving optocoupler, such as Toshiba TLP250 is a good choice.
Everyone can see that the internal has helped us integrate the amplifier and drive circuit, which can simplify the use of peripheral circuits in the case of ensuring basic speed, so that we can simplify the circuit of the H-bridge, strongly recommended.
Say it here again, it is a matter of the UART, which is the UART of the optocoupler, remember that there are many friends who have encountered, just a serial port connected to the normal low-speed optocoupler, and the serial port communication How to debug can't come out, do not use 6N137 such expensive high-speed optocoupler to communicate properly, it is not, here I teach everyone to use ordinary cheap low-speed optocouplers to complete basic serial communication methods
As shown in the figure, everyone can see that this is the market and its common cheap PS2501 optocoupler, I am because of the actual circuit that is called by it, it is basically the most bad because of the basics. The cheapest representative, but it is still very easy to pass the 19200 baud rate.
We can looked at his manual, know that Tr = 3us TF = 5US This speed is much less than the 26US pulse width of 19200 baud, so the device is theoretical.
Let's take a look at why everyone will pick it up but it is not easy to use. Here, give you a reason.
We must know how the optocoupleting speed and optocoupler output triode work current is related. Because the greater the working current, the smaller of R24, the smaller R24, so that the RC charge and discharge time of R24 and transistor itself is short. This RC time is short, the switching speed of the transistor has naturally risen, but the working current of the transistor is determined by the light intensity of the light-emitting diode, and there is a current transmission ratio, so the lighting current of the light emitting diode should also be appropriate. Increased to meet the current demand of the isolation transistor, the general CPU's IO may not provide such a large current, so use an 8550 a primary current amplifier, of course, if your single-chip drive is high enough, this triode can be used . Another point is to see the triode after the tripod is connected, not a common co-emitting pole approach, because the speed of the co-collecting electrode approach is far from the circulation of the total emitting pole. The Q2 in front of the same is also used to improve the speed of optocoupler and reduce delay as much as possible.
It is said that the optocoupler is not limited, it is best to take a maximum operating current in the manual before using the manual, and then calculate the suitable resistance reuse. Otherwise it is easy to burn the optocoupler.
Finally, give everyone a few simulation test pictures.
This is a photocoupler-luminescent diode current and a crystal triode current simulation diagram in the case of 3.3V in 19200.
This picture is really equivalent to the voltage graph under test, you can see that the PS2501 rubbish optocoupler is good in the baud rate of 19200, and there is no problem with complete use ^ _^
The calculation method of R24 and R23 is added, and R24 is equal to (power supply voltage -0.2V) divided by the limit operating current of the transistor on the manual, R23 is equal to the limit operating current of the transistor to remove the current transmission ratio, and how much current is obtained by the light-emitting diode It can make the lateral transistors completely sat.
Then use (supply voltage -1.8V) to divide this current, that is, the value of R23, of course, R23 and R24 can be appropriately increased by R23 and R24 in the case of satisfying the speed.
PS: I estimate that many people who have to listen to this thing. Don't doubt this thing is written on August 6, 2009. I posted my own blog today, I just hope that there is a place to be home. Will not be lost, read full text, original title: High-speed optocoupler isolation h bridge pain!
Article Source: [Micro Signal: FCSDE-SH, WeChat Public Number: Semiconductor PDSA] Welcome to add attention! Please indicate the source of the article.
Our other product:
