Chapter 846 Good secondary structure

This thing was almost completely eliminated later, and only a few places still use it, such as HiFi. Because after this thing is closed, it behaves as a resistor in the circuit with very good linearity and works well in audio.

That's why it's easier to make money from HiFi guys, but harder to make money from junk guys who try to make the best of every penny.

Chief Engineer Cui nodded. This thing is indeed low-cost and can be used if you make some. At least it is much smaller than a relay and cheaper.

"If you want high speed, I have just considered a possibility, using photodiodes as photosensors. Photodiodes have a fast switching speed." Chief Engineer Cui continued to discuss his ideas with Gao Zhendong.

With proper design, the switching speed of photodiodes is extremely fast. If we talk about switching speed, Chief Engineer Cui really can't think of any faster than this.

It has to be said that each of these senior technicians is very capable.

Gao Zhendong nodded: "The speed of photodiodes is indeed fast, but there is a problem with photodiodes, which determines that its scope of application is limited."

Where there are advantages, there are disadvantages. Things without disadvantages will most likely have the disadvantage of being expensive.

"Mr. Gao, what do you mean?"

"The current conversion ratio of a photodiode is too low. If you calculate carefully, you will find that under current technical conditions, it is only a few thousandths, because it is a passive device that relies entirely on the energy of light to generate current. The luminous efficiency, photoelectric efficiency and other reasons all determine that its current conversion ratio cannot be large."

The current conversion ratio is the ratio of the input current at the control end to the output current at the controlled end. In fact, it can be understood as the amplification factor.

Unlike methods such as using triodes, the only energy source of the photodiode is the light beam on the opposite side.

"Do you believe in light?" "Yes! You have to believe in it."

Under this condition, no matter what method is used, the current conversion ratio of the photodiode alone will not be good.

"Chief Engineer Gao, I remember that we also made solar panels, and they also have an efficiency of several percentage points, right?" a comrade asked. Although it was also very low, it was still an order of magnitude different from a few thousandths.

I know this, and I know it very clearly. Gao Zhendong smiled and shook his head: "The structures of solar panels and photodiodes are different. The two cannot be compared. In addition, there is also the issue of luminous efficiency of the light source of the optocoupler."

I understand now. That little bit of electric current is exploited twice during the light-emitting and photosensitizing stages, so the result is naturally not good-looking.

However, Gao Zhendong did not completely deny this idea, because Chief Engineer Cui's idea at least solved half of the problem.

"But what you said does have an advantage. It provides a very fast switching speed foundation. If you add a suitable amplification output circuit behind it and use the output of the photodiode to drive the amplification circuit, you can get good output parameters."

Knowing that everyone is confused about the switching speed of the subsequent stage, Gao Zhendong added.

"Although the switching speed of the optocoupler depends on the switching speed of the amplifier circuit, at least we can focus on solving the problem of the speed of the amplifier. By matching different amplifier circuits, we can get optocouplers with different characteristics. Moreover, these things can be integrated into a silicon chip. Although it is a two-stage structure, the device is actually just a whole chip, which is very beneficial for manufacturing."

Gao Zhendong directly gave the ultimate answer he chose.

Good results can also be achieved by using a photodiode as a switch at the front end of the output and adding an ordinary transistor or other amplification circuit to the back end.

And at this time, as long as the switching speed of the transistor is increased, the speed of the electronic switch will naturally increase. In fact, the ultimate form of optocoupler - solid-state relay has such a structure.

It's just that the realization of solid-state relays requires another thing, but without this thing, Gao Zhendong can also make a usable electronic switch. After all, it can be called a relay, which is different from an ordinary electronic switch, so it's normal to have a little threshold.

Chief Engineer Cui nodded repeatedly: "Well, the structure you proposed is good. It suddenly dawned on me. First, use a diode to provide the basic switching speed, and then use the post-amplification circuit to determine the specific performance. In this way, a problem is broken down. Good, great."

The comrades at Factory 1218 also had their own thoughts. One comrade thought that this was a bit ridiculous. "Chief Engineer Gao, can we use phototransistors directly for photosensitive switches? In this way, the structure of the output electrode can be simpler."

It is definitely not as convenient to drive a transistor with a photodiode as to directly make a phototransistor.

Gao Zhendong was happy. This matter seemed to be true at first glance, but it was actually not that simple.

He did not laugh at this comrade. He was thinking about it, but he just didn't have the ability to see it clearly. This kind of triode structure does exist, but Gao Zhendong did not choose it for a reason.

"In theory, you are right. Phototransistors can indeed achieve this effect. But in practice, it complicates things because phototransistors are more difficult to manufacture than photodiodes and post-transistors. The overall performance they can provide is not as good as this two-stage structure with each performing its own function. I chose this structure based on the technical difficulty and prospects."

Solid-state relays ultimately use this two-stage structure, but with special post-amplification devices, and there is a reason for this.

This comrade was very happy to get Chief Engineer Gao's affirmation in theory. He was extremely pleased. As for the fact that his ideas were not thoughtful enough, that was nothing.

Other comrades had the same idea as him, but they just didn't say it or didn't have the time to say it. However, after listening to Gao Zhendong's explanation, they understood.

The phototransistor brings together photoelectricity and amplification, which makes the design and manufacturing of the phototransistor itself more difficult. It is integrated and structurally fused into one whole, and it is very troublesome to take various parameters into account. Although it seems to be highly integrated, it is actually far less convenient than this two-level structure.

Suitable photodiodes and post-stage amplifier circuits are basically available now, but suitable phototransistors are still a mystery. Adding too many new things will have a great impact on the successful completion of the project.

Gao Zhendong, who has limited resources, directly chose this method which is relatively easy to use but has better actual development prospects, and there is actually not much difference in production costs.

There is not much difference between making a diode + transistor and a phototransistor on a silicon wafer in semiconductor processing. In fact, the phototransistor may even have a higher processing cost at this stage because it has more requirements and greater restrictions.

"Another reason is that the characteristics of the optocoupler with a phototransistor structure are limited by the transistor itself. When a new type of transistor appears, the phototransistor has to be redesigned for this transistor, which is a very troublesome thing. But for the two-stage structure, it is actually a matter of replacing the latter stage, which is much simpler and can obtain the performance of the latest amplifier device at the fastest speed." Gao Zhendong said with a smile.

No matter whether it is an ordinary triode or some special tube, in theory you can just replace the post-stage directly.

President Cui listened to this point and slapped his thigh, "Right! That's right! Chief Engineer Gao, you are not only proficient in theory and have broad thinking, but also have a very good grasp of specific work and research routes, and are very thoughtful."

This is not flattery. Gao Zhendong's speech on the choice of the structure of the optocoupler was incisive and well-connected. It took into account the performance, difficulty and long-term considerations.

In the past, he had only observed Chief Engineer Gao from afar and had not participated in the specific work process, so he had no idea about it. But now, he and Director Pan have truly felt it. Chief Engineer Gao is very famous in the 12th Ministry of Machine Building. His technology, experience, and ideas are truly extraordinary.

Gao Zhendong felt a little embarrassed. He was making choices based on the answers, and he was naturally much better than all of them.

What? The system is your own? The results have brought real benefits to society and the masses? That’s fine, copy it!
Chief Engineer Cui quickly sorted out what his factory would do next in this regard: "Chief Engineer Gao, this is what I think. We will first use LED light sources as the basis to develop photoresistor-type optocouplers and photodiode output post-amplifier optocouplers. For the latter's amplifier stage, we will first use high-speed ordinary triodes. At least we can get it working in principle and get a batch of basically usable devices."

The advantage of using a two-stage structure can be seen from the current arrangement. Now we need high-speed switching, right? Then use the fastest triode as the post-amplifier. As long as the other parameters are acceptable, we can make one and try it out first.

Even Chief Engineer Cui has now seen a broad prospect for this structure. By replacing the back stage, it will be a new device with unique features!
Chief Engineer Gao's method is to get rid of all kinds of optocouplers at once.

Gao Zhendong nodded in agreement: "Well, your idea is good, and you don't have to rush. The target installation equipment of this switch still has a long way to go in research, so we should proceed step by step. By the way, remember, it is best to develop one kind of unidirectional conduction and one kind of bidirectional conduction."

Ordinary triodes are directional and can only conduct in one direction. However, for programmable switches, Gao Zhendong wanted to make them bidirectional in one step, so as to reduce some work restrictions for his comrades at the Institute of Communications.

"Well, that's easy. I can just use two transistors back to back, but this will further deteriorate the linearity. Is there any problem?" Mr. Cui's basic thinking about components is very comprehensive. Speaking of linearity, I think of the photoresistor again, but it's a pity that it's too slow.

(End of this chapter)