Technological power starts with upgrading lenses

In fact, this incident had a huge impact on Wu Lihong, and even gave him the idea of ​​​​retreating to the second line again.

But he was still unwilling to settle the photoresist problem.

Now that Kang Chi is willing to help, it is naturally the best thing for him.

But he was also afraid that he would betray Kang Chi's trust and let his money go down the drain, so he was hesitant.

"Although I know that you are somewhat successful now and have made a lot of money, after all, money does not come from strong winds. Aren't you afraid that it will be wasted?"

"I'm afraid, but am I afraid that I won't engage in research and development?" Kang Chi immediately replied, "And I'm not giving money to your institute for free. I also share in the results. In other words, this is to let Tangtang Academician Wu came to help me with my work. Others may not envy him, but I am worried that you are not willing to do so because of your face."

"..."

Kang Chi's words successfully amused Wu Lihong, and his worries became much less.

"Okay, I'll discuss it with the hospital later and try to help you find a better cooperation plan. Thank you for your trust in me, an old man who is about to die."

"Don't say such things." Kang Chi said quickly, "In addition to photoresist, there are many etching materials waiting for you to show off your skills... You have been researching for so long, so you can distribute the benefits however you want. , there’s no need to be partial, I’m not here to take advantage of you.”

"In addition, when I come here this time, I would also like to borrow your laboratory to make photosensitive chips, is it okay?"

"Photosensitive chip?" Wu Lihong said doubtfully, "Why are you suddenly interested in photosensitive chips again?"

"I want to help Daqin Optics build cameras."

"..."

Sure enough, he has to be a young man,

There is just so much energy.

Wu Lihong couldn't help but sigh in his heart.

Coupled with the oil-mining insect project, he is already fighting on a third front, right?

And every line is very difficult.

"Are you too busy to come here?"

"It's okay. I'll be fine once I get through this period of time." Kang Chi smiled, "It's not a very urgent matter. Just take your time. Once this period of work is over, I can rest for a few more days." .”

"Well, although you are still young, you should pay more attention to your health. There is still a long way to go, so don't worry, it will take a while..."

"Don't worry, I'm measured."

——

After the visit, Kang Chi followed Wang Haowen directly to the Jiangsu Microelectronics Research Institute, planning to borrow the equipment here to rub a simple photosensitive chip.

For Kangchi, cameras should be relatively familiar things. There are actually two core technologies in them, one is the cmos photosensitive chip and the other is the imaging system.

Although the remaining mechanical and electronic structures are also very precise, he was already familiar with them when he opened a repair shop. Now that he has accumulated so much professional knowledge, it is not difficult to reverse engineer a camera.

The only thing that restricted him from making cameras was the lack of a complete set of chip manufacturing equipment.

While wearing protective clothing with Kang Chi, Wang Haowen finally couldn't help but ask: "Junior brother Kang, you don't seem to have prepared anything, so you just start doing it right away?"

"has a problem?"

"It's a big problem, okay..."

Wang Haowen strongly doubted whether Kang Chi knew how to make chips...

Kang Chi just smiled and explained: "I am actually just going through the process briefly this time to verify the principle. Even if I create a CMOS with only one hundred pixels, it does not require such complicated steps."

"One hundred pixels?!"

Wang Haowen was immediately shocked.

However, when Kang Chi first started working on chips, didn’t he just send him a 2D accelerator card to verify his so-called ‘principle’.

can only say,

The research methods of great gods are different from those of ordinary people like you.

Of course, one hundred pixels is just an exaggeration by Kangchi. The specific situation must be based on the actual situation here and try to create a photosensitive chip with better performance.

Although he currently has no shortage of general experience for upgrading,

But save it if you can, because you never know when you will need it.

The structure and principle of cmos photosensitive chips are actually simpler than ordinary chips. The technical difficulties are mainly reflected in the manufacturing process and materials.

From a design point of view, the photosensitive chip can be said to have no difficulty at all. It is a chip type with a very high drawing repetition rate.

For example, a photosensitive chip with 20 million pixels actually has the same chip structure for each pixel.

Just create uniformly arranged photodiodes on a silicon wafer substrate through photolithography, and then use an etching process on these diodes to create color filters.

The specific working principle is actually not difficult.

We can imagine a color filter as a window with four square pieces of glass. The colors of these four pieces of glass are red, green, blue, and green.

Then when the outside light passes through this glass and shines into the house, the four photodiodes arranged in the house can identify the light intensity in the four areas and restore the color and brightness of the light that shines in.

In other words, a 20 million pixel CMOS requires 80 million photodiodes and color filters.

And in order to ensure that the light conditions for each pixel to absorb light are the same, they can only be arranged flatly on the silicon wafer, and cannot be stacked into complex urban structures through etching like other chips.

This is why it is very difficult for photosensitive chips to achieve high pixels.

It really requires too much process.

If a 90nm process photolithography machine is used to manufacture photodiodes, because of the 2X2 principle of one pixel, it is equivalent to a 180nm process.

And because it can only be tiled, given the chip size limit, it can only reduce pixels.

So some mobile phone manufacturers brag that their mobile phones have 100 million pixels.

In fact, just listen to it.

Think about it and you will know that with a lens the size of a mobile phone,

How big is the area of ​​the CMOS chip inside?

Even if it is made with the most advanced EUV lithography machine, the photodiodes inside will never reach the level of 400 million tiles.

The actual estimate is not even 10 million.

The remaining pixels are fake pixels calculated by the processor to make people happy.

Moreover, the smaller the process, the more difficult it is to etch and manufacture color filters. The performance requirements for the material used for dyeing during etching: color photoresist, are even more stringent.

However, Kangchi’s current requirements are not very high, so he can do as many pixels as he wants.

There are also no requirements for the performance of the color photoresist. Wang Haowen can be directly asked to mix it. The requirement is just that the red, green and blue cannot be distinguished...

Although he said he was just dropping by for a massage, in fact, Kang Chi stayed in Lin'an for almost a week.

The main reason is that the production of the mask is relatively time-consuming. At the same time, because it is only a one-time use, in order to pursue cost-effectiveness, Kangchi had to lower its accuracy requirements.

[Item: Photosensitive chip]

[Manufacturer: Kangchi]

[Item Level: 1]

[Experience: 0/500]

[Item status: intact]

[Item parameters: 500,000 pixels]

[Analysis items: no analysis items]

[General experience: 50314]

[Mastery Points: 12.3]

Looking at the low-power chip in his hand, which was as big as a 35mm photosensitive chip but had only 500,000 pixels, Kang Chi couldn't help but sigh again, the chip is really expensive.

This tapeout cost Kangchi nearly 2 million.

And this is already considered saving money,

If we just build a set of mid-range process masks, tens of millions will be wasted, and more time will be wasted.

When Wang Haowen saw Kang Chi directly put the manufactured chip into the container, as if he was preparing to dodge, he couldn't help but asked in confusion: "Don't you need to test it?"

"test?"

The system panel has popped up, what else can I test?

Kang Chi chuckled: "No, I've already tested it in my mind."

Wang Haowen: "???"