It is not difficult to solve the inherent technical problems of CMOS and truly make it practical, and now it depends on whether Professor Bai Siwen can follow the ideas provided by Wang Zhengguo and whether he can come up with a usable product as soon as possible.

To tell the truth, as long as a qualified sensor can be made, no matter what the picture effect of its imaging is, Wang Zhengguo can sell it, which is really not blowing, someone has long been interested in a big market, and he is ready to rush up and bite it hard.

If you want to successfully produce CMOS graphics sensors, then the premise is that you must first have a reliable R&D team, which can be said to be follow-up R&D and production, and now Professor Bai Siwen is obviously just a newcomer, although he has a solid foundation, but for the field of graphics sensors, maybe he really doesn't have a clear concept.

"Professor, you haven't done detailed research and understanding of the pattern acquisition capability of CMOS sensors, so I can only talk about the problems encountered by the current industry in this regard, and my personal solutions to these problems. "Now that all the issues that should be stated in advance have been said, we are definitely about to enter the golden finger moment.

The questions to be asked are all real difficulties at present, but as long as you pay more attention and have a heart to collect information, or ask professionals to know.

And this so-called "personal solution to the problem" is naturally impossible for an unscrupulous teacher to come up with, and it is the best solution that professionals have come up with after racking their brains more than ten years later.

Speaking of business, the special little black book of the scientific researchers of the Republic was taken out by Professor Bai from somewhere, and he made it look like I listened to your explanation very carefully, which immediately made Wang Zhengguo feel a little embarrassed, although he had long adapted to this situation, but how could a university professor be compared to the students he taught?

Fortunately, it is not too in-depth here, but it is a general talk, and it is just a general point to a scientific research direction, which corrects the mentality and begins to introduce.

"The first is noise, which is the number one issue that affects the performance of CMOS sensors, including graphic noise FPN, dark current noise, and thermal noise. Although it is a common problem for many integrated circuit chips, including CCD sensors, it seems that it is more serious on CMOS graphics sensors. ”

The so-called "noise" in the integrated electronics industry is completely different from the concept that people usually think that the ear can hear the environmental noise, so it must not be confused, and the noise of the integrated electronics industry, it causes far more harm than the environmental noise that the human ear can perceive.

In the field of graphics sensors, the presence of noise ultimately affects the output of image quality. In fact, the same is true in the field of other types of electronic chips, the most typical example is the military industry, phased array radar uses T/R components as a basic power output/reception unit, and affects the efficiency of radar work, of which noise is one of the culprits.

Decades later, when the entire radar industry is working on improving detection through the use of new T/R component materials, there is actually another way to solve the problem very significantly, and that is noise reduction, in the same case, the better the noise reduction of the radar signal, the clearer and farther the detected signal is.

"Noise reduction", which belongs to the basic work of the semiconductor integrated circuit industry, is the same as the processing and assembly accuracy index of the machinery industry, which is a kind of performance index that looks inconspicuous and ordinary, but has a real impact.

Since Professor Bai Siwen is in the electronics industry, he is naturally no stranger to this indicator, and after writing the word "noise reduction" in his notebook, he raised his head and continued to listen.

"To solve the noise of the graphics sensor, I think we can introduce a similar idea of differential pairs to suppress it, first read out the charge integration signal generated by the light on the sensor, and temporarily store it, and then reset the pixel unit, and then read the output signal of the pixel cell, and subtract the two to get the image signal, so although it is necessary to add a temporary data storage unit, but I don't think it is a big problem. ”

The words have been very frank and direct, and the influence of this sentence alone is enough to promote the development of the digital pattern sensor industry for at least three years or so, after all, this technology is not only applicable to CMOS, but also can be used in the use of CCD.

Maybe it's because he didn't have a careful understanding of the graphics sensor before, but after listening to it, Professor Bai didn't change much in his expression, just frowned and thought carefully about it, and after about half a minute, he seemed to figure out the key points inside.

Only then did he put down the notebook and looked at someone with admiration: "It makes sense, the introduction of the concept of differential pairs to suppress the pixel noise problem of the sensor, I think it should be feasible, and the implementation is not demanding, just increase the data storage unit, and then introduce the initial data computing power, it can be completely realized, but the improvement of the sensor image quality it brings is indeed obvious, and it can even be said to be a revolutionary improvement!"

As long as you put the idea out, simple circuit design is not a problem at all, and thanks to the rapid maturity of CMOS circuit-related processes, the processing and manufacturing aspects are completely too simple.

Continuing to talk, there are only two or three problems that really block the practical application of CMOS pattern sensors, and Wang Zhengguo is not very clear about the other two problems, so he can only roughly say that it is absolutely impossible to remember them all.

"The other two problems are mainly to suppress the dark current, it increases rapidly with the increase of the working temperature of the sensor, and some reliable cooling methods can be used to solve it, but I think the cooling method may not be easy to control, this should be studied carefully; ”

After saying a lot, the back is very brief, and Professor Bai Siwen couldn't digest it for a while, after all, this statement is too vague, and the professor himself does not have a very clear understanding of CMOS sensors.

This is only the first to write down everything, and then immediately start the CMOS graphic sensor related knowledge assault learning, hoping to get started as soon as possible, looking at the current situation, it seems that the scientific research fund is really interested in CMOS sensors, it should have been premeditated, it seems that they may really take a good step this time.

"That's pretty much it, that's what I've learned, I think Professor Bai on your side should start researching related issues as soon as possible. In terms of funding, 50,000 yuan will be provided in the early stage as a pre-research need, and then there will be continuous support, and my requirements are not high, as long as I come up with a sensor sample that can successfully display graphics before the first half of next year, it will be considered that the first phase goal has been achieved. ”

Rashly invested in the semiconductor sensor industry, this is an unplanned thing, now all the funds in the fund are not much, relative to the demand of the semiconductor industry, at present, it is better to do a good job of pre-research, the scientific research team is mainly cultivated.