Chapter 379 PMOS
It’s not that I don’t finish everything at once, but that every time I say a word, you have to interrupt me to express some opinions, Linochka thought to herself.
"Kerosene? The calorific value of kerosene is higher than that of alcohol. Haha, is this the reason?"
Comrade Minister was relieved at this.
It’s not that he must find some reason to make himself happy, but these situations alone may not seem like much, but when they are all combined, they are somewhat surprising.
A larger range, a larger warhead, the same size and weight, which shows that the opponent must have some technology that it does not know.
“Comrade Minister, should we try replacing the alcohol in P-15 with kerosene?”
P-15 is the 544 missile.
"Oh, no! The grassroots Davaris will complain, at least the current P-15 cannot be replaced! They are used to maintaining this kind of missile, aren't they?"
I have to say that as a minister, I am very sympathetic to the comrades at the grassroots level.
We soldiers under the Maoist flag love to protect alcohol fuel!
"Comrade Linochka, if there is any new news over there, you must tell me earlier. By the way, tell the comrades in Lubyanka that they need to make more efforts!" Comrade Minister General I feel like something is wrong, but I don’t know what is wrong.
——
Gao Zhendong didn’t know about Lao Maozi’s entanglement. He was entangled in his own affairs.
In fact, it’s not a struggle, it’s just choosing a suitable path among the complicated processes.
Although MOS technology has been selected, MOS technologies also include PMOS (P-channel MOS), NMOS (N-channel MOS), CMOS (complementary MOS), etc.
There are different options for each process of these technologies.
It's like building blocks. There are different ways to build a park. As the saying goes, everyone has his own way of killing pigs and butts.
However, in terms of the type of MOS, Gao Zhendong has nothing to worry about. PMOS can be used directly. As the earliest MOS technology, it naturally has its advantages.
Guess why it is the earliest? Simple!
PMOS only requires five photolithography steps to complete.
P area lithography, gate lithography, contact lithography, metal lithography, bonding block lithography, and different deposition or oxidation processes are added between each photolithography.
The P area, gate, external contact holes, internal chip wiring and lead-out solder joints for packaging are formed respectively.
Moreover, the mask needs to be strictly aligned for overlaying. Only the second grating photolithography is required. The alignment requirements at other times are not so high. This makes the technical conditions at hand limited and stretched. Gao Zhendong felt very comfortable.
Packaging aside, there are only 12 steps of process from silicon wafer to semi-finished product filled with finished chip cores (DIE). This includes the step of preparing the silicon wafer.
If you don't even want the passivation step, only the first 4 photolithography steps are enough, and the corresponding process is reduced to 10 steps.
It is really simple, and the requirements for materials are not high.
There are only three elements to consider introducing.
——Oxygen that forms the oxide protective layer, boron that forms the gate electrode, and aluminum that forms the metal wire.
Among them, aluminum does not need to consider the issue of doping and diffusion, it is only used for deposition.
A retired old man: It's so fun.jpg.
NMOS and CMOS are much more difficult than PMOS. PMOS, if nothing else, can at least make logic integrated circuits. Moreover, this thing can be made into C8008 using a 10μm process. In fact, it is more difficult to make. High-end manufacturing processes are no problem either.
After selecting PMOS, Gao Zhendong quickly determined the process route.
Do not passivate in the last step, 4 times of photolithography, only until the metal photolithography is completed.
As for why he chose this one, of course it’s because it’s simple, and he started with a relatively rough process, so this level was enough.
Moreover, with this less step, the cost will also decrease.
First take out the logic gate circuit to support the production of DJS-60D, and then build a better one.
Resolve first whether there is any problem, and then solve it.
After selecting these and copying the book, it was quick. Gao Zhendong spent an afternoon quickly copying the book. The next day, he made a phone call to Factory 1274, who was already a little bit disappointed.
More than an hour later, Factory Director Lu and Chief Engineer Lu arrived hand in hand, waiting for this day.
Looking at the process design guidance document given by Gao Zhendong, the two of them were a little confused. Why is this thing thinner than the original one?
Advanced technology must be complicated. Once it is complicated, the data will be very thick. Everyone thinks so.
"General Engineer Gao, this process is done part by part first. What are we going to do in the early stage?" General Engineer Lu asked.
Gao Zhendong was confused. When did I say something and did it part by part: "No, this is all." "All?" The two of them were shocked, how could this be possible?
Gao Zhendong smiled and explained the simple outline of the process to them, and the two fell into a state of "I don't understand, but I am shocked".
Gao Zhendong had no choice but to explain: "Although PMOS technology is newer, it is simpler in terms of process and is most in line with our current situation."
Mr. Lu was overjoyed. He had picked out half of his hair due to the complex process of bipolar transistor semiconductors. He felt that this was the first good news he heard today and also the best news. General Engineer Gao was very talented.
"General Engineer Gao, are you saying that with this technology, with just a few photolithography and dozens of steps, it can complete what bipolar semiconductors require more than a hundred steps to complete?"< br>
Gao Zhendong smiled and said: "The development of technology is actually like this. Advanced does not mean complicated. Some advanced technologies can be so simple that we can hardly imagine."
He even looks super rustic, Gao Zhendong thought to himself.
The two of them still believed in Gao Zhendong. Basically, what he said had never failed.
"General Engineer Gao, this is great, what do we need to do?"
Gao Zhendong said: "I see the current situation in the factory, there is a diffusion furnace, right?"
Director Lu nodded: "Yes, this is one of our basic production equipment, there are many."
"How high is the highest temperature control accuracy? Is it plus or minus 1 degree Celsius?" Gao Zhendong asked. This accuracy involves the diffusion doping effect.
Mr. Lu knew clearly: "This one can't be reached. There is a slight gap. The main problem is the sensor."
The K-type thermocouple they are using now has no problem with the temperature measurement range, but the accuracy is. It's quite a bit different, 2.5 degrees Celsius or 7.5 parts per thousand, which is a long way from this accuracy requirement.
Gao Zhendong thought for a while and did not recommend type B thermocouples. The error of type B's 2.5 thousandths is almost meaningless, and the response speed is slow.
“Add a platinum resistor for indirect measurement as an auxiliary sensor, cooperate with the K-type thermocouple, and add DJS-60D to run the PID control algorithm, it should be controllable.”
The response speed of platinum resistors is also slow, but this can be solved to some extent by algorithmic means. The accuracy of platinum resistors far exceeds the demand, and the best can reach three thousandths of a degree Celsius.
The biggest problem with platinum resistors is the temperature measurement range. Semiconductor production requires a range above 1000 degrees Celsius, which exceeds the temperature measurement range of platinum resistors. However, if the indirect measurement is well designed, the problem is not big. .
So Gao Zhendong came up with the method of indirect measurement of platinum resistance + direct measurement of K-type thermocouple. One can measure accurately, the other can measure quickly. The two swords are combined and invincible in the world.
He doesn’t know how future diffusion furnaces will solve this problem. Anyway, here, he solves it like this.
The reason why this diffusion furnace is so important is that many processes rely on it.
Oxidation, deposition, diffusion, diffusion-oxidation, these all depend on this thing.
There was originally another process of diffusion-oxidation, but Gao Zhendong did not choose it, because choosing that process would involve an additional piece of equipment called an epitaxial reaction system, which would be troublesome.
So Gao Zhendong still chose the "thermal oxidation method" that can be completed by a diffusion furnace.
Gao Zhendong thought for a while and gave another support: "At that time, I will support some of your technicians who specialize in computer control and thermocouple applications, and will be responsible for cooperating with you in this matter."
General Engineer Lu and the others are now relieved. General Engineer Gao has elite soldiers and generals here to support them personally, which will make things much easier.
Gao Zhendong is also very happy. There are three core equipment in the PMOS technology chip manufacturing stage, a photolithography machine and a diffusion furnace, and only one is left.
"Director Lu, what is the current situation of the vacuum evaporation equipment?" This thing was not found in the current situation of Factory 1274, so the situation may not be optimistic.
Director Lu shook his head: "No, we didn't have to use this before."
Vacuum evaporation is an equipment that evaporates metal under vacuum conditions and forms a metal coating on the surface of the workpiece. In chip manufacturing, its upgraded version is sputtering equipment.
However, sputtering equipment is used for metals such as tantalum and niobium, which have good sputtering characteristics. However, the aluminum used in large quantities for wiring within the chip is different and always evaporates.
However, there are several types of evaporation, including resistance heating, which directly uses tungsten and other metal resistance materials to heat high-purity aluminum to form metal vapor to complete evaporation.
There is also electron beam evaporation, which can avoid some of the disadvantages of resistance heating. For example, the heating resistors will evaporate themselves when they come into contact with the oxidizing atmosphere in the process, and they may even react with the evaporated metal. reaction, it becomes even more difficult.
Gao Zhendong thought for a while: "Then let's build a resistance evaporation machine."
Electron beam evaporation relies on a magnetic field to accurately deflect the electron beam and hit the target for heating. Let’s not talk about the source of high-energy electron beams. How this magnetic field comes from is a problem. There is no need to think about permanent magnets. The electromagnetic field is also a problem under current conditions.
On the contrary, resistance heating with backward technology is more feasible now. Every technology always has its reasons for its existence.
Anyway, this is not a very demanding process now, so although the entire integrated circuit process seems to be full of make-do and make-do to Gao Zhendong, at this stage, it is almost the most suitable.
However, the two people at Factory 1274 obviously don’t think this process is makeshift or adequate.
(End of this chapter)