Видео

Hi! I think it is delta, but there is no way to prove it. The center point is not wired to anything, even the ground wire. Resistance was around 2-3 Ohms per any pair of contacts. Did not measure the inductance, but can try with another motor if you need the numbers

The oil I used was similar in thickness to diffusion pump oil, so I gave it a try fully understanding that damage most probably will be done. Temperature monitoring confirmed that bearings are not run dry - for around 200 hours temperature was stable 45 or so Celsius. To be honest, until the end I was expecting it to fail in a spectacular fashion and so must anybody else doing this DIY)

@@semiconductorsstuff Diffusionpump oil is also much to thick and also doesn't really have any lubrication properties or any friction reducing additive and similar, so that would have actually even been worse. Temperature could absolutely look just fine on the bearing, but then all of a sudden the cage could break fex after being hit with high velocity heavy oil for many hours. I'm just saying this so also people that will read the comment understand what really should be used and what could happen if not. A friend was in a corridor around 20 meter from a lab where they used a similar sized turbopump. All of sudden he heard a sound like a bomb went off in the corridor but it was shown that the turbopump had catastrophically failed. He had his door close but it was open at the lab. You don't want to be even in the same room when these blows up.

Yes, I agree. I did it this way because I was not sure if either the controller was good. Closing off the chamber would mean very little air on the low side of the pump, which is easier on both pump and the controller. The pressure in the main chamber was kept at acceptable fore-vacuum pressure for the pump at all times. With this manual machines it is indeed important to know the rules and limits of the equipment

Thanks! Your comment made me check the manual) The ultimate pressure is 10^-8Pa per datasheet. Probably it is due to the nature of the systems used and lack of experience of my part, but I have never seen any apparatus go this low. At least with TMP alone. This made me treat this pump as 10^-4Pa system and not think too much about it. Makes me want to make a flange that does not leak through the bolts, now that I know that this TMP can do more xD

@@semiconductorsstuff The ultimate pressure is taken at fab with a vacuum gauge bolted right on top of a CF flange that in turn is bolted directly to the pump and then pump is heated for a couple of hours to drive off moisture and let the pump go for 24h and then read the pressure. ISO pumps are tested the same but without heating. You would normally never get that low in a system but close if done right and have the right pump and so on based on the size of the chamber and amount of rubber gaskets and so on. You would only be able to get down to this exact CF pump ultimate pressure if it is a complete CF-flanged system, which this is not. But I expected at least a little better but there could also be a lot of moisture in the chamber or many gaskets. If you could quite quickly get down to 10^-4Pa I would personally have been happy for a big system like this and you really don't need to go down that far anyway based on what you do since you handling plasma as I understand, only to get rid of any unwanted gases.

@@dtiydr Even with heating and fully CF flange system, I am impressed at how low can this pump get even in theory. Cleanest I have personally seen was probably 10^-7 at Auger Electron Spectroscope, but that was another level of monster. Ion pumps, load lock, heating, plasma cleaning, 24/7/365 operation and all administrative stuff. As for this system - it is not the moisture, sadly. The reason is that I had failed to make a conversion flange (CF to o-ring) with blind holes, so threads on the bolts (even filled with vacuum grease) provide a leak path into the chamber. Was pretty sad when I noticed that oversight, but for sputtering oxides at 0.5Pa it will do, probably. One day someone will get curious enough to TIG weld those bolts and get to 10^-4Pa or so is my hope.

@@semiconductorsstuff Oh yes I see that picture now on the screws, no that is not that good and just the very fact that you using a en rubber gasket doesn't help either¨but that is all that you can do. Turbopumps goes down low but the one you have, as I mentioned before, have a little bad blade design compared to modern ones and thus don't have exactly the same pumping capacity but it shouldn't matter that much. At Max4 I saw 10^-8 Pa on a vacuum gauge, and that was held there since they needed that level.

Those are special lint-free wipes for cleanrooms (Kimwipes). This method is indeed too barbaric for IC chips with complex structures already on them, but for bare wafers before cleaning it is acceptable. It is a good time-performance tradeoff in my experience

Thanks! Spin-coating and basic lithography will be next, I hope) Cameras/phones are surprisingly OK to bring in. If anything sheds dust or lint - it is a no-no, for example ordinary paper or pencils are forbidden. One more problem is leaving traces, thats where my mistakes were

Hi! This was filmed in Japan.

I agree, that was traumatic to perform for me as well, hence the disclaimer) There are a few things I would do differently now, when I know how it is held together. However, it now has more than 100 hours on it at this point and pumps dirty and leaky sputtering chamber to 3*10^-3Pa. Probably will upload the final video on it this month if everything goes as planned

I said that 10.3mm is non-standard because few catalogs had only 10mm ones in them, also did not show for any acceptable price on marketplaces. Usually - yes, it is not possible to change the bearing thickness without spacers, it was jamming severely on Pfeiffer magnetically levitated pumps. However here, the base surface of the rotor shaft (part contacting the bearing) and the bearing bed on the stator are in the same plane and bearing is held there by spring tension. Meaning that as long the inner ring and the outer ring of the bearing are the same width - it will work. Completely agree with the statement that the original bearing is better. I would never put the pump with $5 bearing into mission critical or production hardware. However, for applications where reliability is not that important (something with low runtime and pressure feedback loops) - this will do. For heat-fitting the bearing - really appreciate the input, on mission critical stuff will definitely do that. On this pump rotor is held captive from the top with the top bearing, so the forces will push it in the bottom bearing, not pull it out, so I am pretty confident in it. The oil is actually a test. It would save a lot of money long term if we could use rotary pump oil, so after controller for the pump is found - a lot of monitoring will take place. Re-balancing comes together with proper bearings and oil and in the end - I am just too poor for that) The whole goal was to do a cheap&dirty hack and make it work for another 1000 hours. I plan to run this pump at 30k rpm, which is 6k rpm below what was in the bearing datasheet. If it explodes in a horrible fashion - I solemnly swear to update this video.

@@semiconductorsstuff Yes I know you mentioned 10.3, I just made a statement to why the size. "However here, the base surface of the rotor shaft (part contacting the bearing) and the bearing bed on the stator are in the same plane and bearing is held there by spring tension." In this you are then lucky because in some turbos you cant put another thickness bearing in without a spacer but to get the oil system to work on some others you need to have the distance below the bottom of the bearing and then you also need to make a distance for the inner race of the bearing as well. Had to do that for a Edwards Next 240 turbo that had a magnetic bearing in top and it had to have exactly the perfect tension as well other wise did the rotor blades hit the stator so the baring I used also needed to be able to handle that, it also had to be adjusted within 0.2 mm, pain in the ass to get right. For a turbopump reliability is always number one! No matter its for hobby or pro. So as minimum buy a hybrid bearing because the balls in these can never seize, then take one stated for 75 - 90k rpm, cage of nylon or more normally phenol that is brown, steel cage is not recommended and rarely never goes up that high. Hybrids are not extremely expensive at all but will hold and be safe for a long time as long as they are being lubricated by the oil or anmy greese. I don't know the precision on the bearing on this old pump but for newer like the NEXT series the bearings inner race are made within 1/1000 of a specific millimeter. So when you change the bearings to the original ones in this NEXT series you should check what size the axle is, which is stated with a single digit, and that digit fex a 4 means how many thousands + from 8.000 mm the axle is and thus what bearing should be ordered. With the right bearing the bearing just slides on with little friction and here heat should never be used. In your case it could, but never open flames or similar but oil and the same oil that is used to lubricate it. I would strongly not recommend to use rotary pump oil because its to thick to use for these very high speed, it can actually destroy the bearing cage since it flung around with big force since its to thick and cant get away fast enough like thin turbo pump oil can. You see how thin the oil is that's in it. You can get hold on turbopump oil, and it not that expensive and it will last a long time. If the pump vibrates so you can just feel it then the bearing will be strained which is never good. If it rotates quite much then the rotor could be in danger and I would never use it then. 30k or 36k rpm doesn't matter with a rotor that weigh half a kg is still as dangerous. I just mention all the stuff here for others as well for safety.

@semiconductorsstuff Hello, First of all, thank you for the video you shared, it opened my horizons to good directions @semiconductorsstuff and @dtiydr I want to learn and do TMP repair and maintenance, I cannot find enough resources for this, can you help me on this subject? Thank you in advance for your interest.letkenle

@@semiconductorsstuff Hello, First of all, thank you for the video you shared, it opened my horizons to good directions @semiconductorsstuff and @dtiydr I want to learn and do TMP repair and maintenance, I cannot find enough resources for this, can you help me on this subject? Thank you in advance for your interest.letkenle

@@murataldemir9732 Hi! Thank you for the interest in the subject. I do not know how to find more that I have already mentioned in the description. Some understanding of how turbopump works can be achieved with learning about jet engines. Principle and rotation speeds are really similar, it just has a different power source - electric motor and not chemical reaction. Also I believe there were a few demo videos from Pfeiffer and Leybold - manufacturers of TMPs with simple service tips, that also gave some insight. Aside from that I just took what was destined for garbage and made this video.

That is a pretty specific knowledge on the pump lineup, cool! Are you talking about microparticles being pushed back or gas? Definitely will try enforcing the glove use in the future, sometimes I get too relaxed when working on non-critical stuff.

@@semiconductorsstuff Whatever it will pump so will a small amount of that go into the chamber again. This is because the blades are made by cutting them out from the metal disc and then just bending them. Thus the thickness of the blade material will become a wall at a 45 degree upwards towards the chamber. This is the reason you see these sold for quite cheap since they are not worth fixing for any university or industry and such and they absolutely don't want anything go back into the any chamber with delicate things in. But for DIY they could work if you can get hold of a controller and cables as well, and if the bearing are not destroyed either since original ones hardly exist and are very, very expensive.

Yes, quality was surprisingly good, gotta love old school Japanese manufacturing. You are correct, any generic 7605 voltage regulator will do. I got TEA7605 for a dollar or so in a local store, works fine. Now using it as my tabletop power supply.

@@semiconductorsstuff Nice to hear, thanks for replying.

Yeah, semiconductor device lab in a Japanese university. Did not see much people showing stuff like this, so decided to make a channel

@@semiconductorsstuff I understand, really cool, keep up the good work.

Thanks for the feedback! I want to make a followup, but there is no controller for that pump for now. I plan to make one myself, but it will take a while. Writing a thesis and doing academic work takes a lot of time. If lucky, I will film it until the end of the year. For now there are plans to show semiconductor production process in detail, we will see.

@@semiconductorsstuff Ok, thanks, semiconductor production sound pretty cool, I have subscribed already. By the way removing bearings from a rotor shaft is almost always hard, they are pressed on (I replaced bearings on asynchronous motors so I know), you should get a better bearing puller.

@@Alexelectricalengineering Yeah, my pulling sucked) I should make a nice fitting for this purpose. On the other hand, doing expensive stuff (official overhaul is slightly above 1k) on the cheap is the main goal here. If there is a need to do it again - I will turn some proper spacer on the lathe.

@@semiconductorsstuff I understand, all good 👍. Some time ago I was reconditioning a rotary vane vacuum pump from 1980 I think.