|
|
Post by westom on Feb 7, 2020 18:19:46 GMT -5
Mostly off-topic, but a mechanic at work once dropped a wrench across the output of a 480v transformer, which itself was about as large as an end table. Wrench vaporized. Mechanic deafened for a couple days. Discuss such stories so that others can learn from mistakes. In one facility, the president wanted employees to do some minor electrical work - since everything was only about reducing costs. Employee accidentally lost the cover plate inside that panel. It hit the 440. An explosion blew that plate and a still attached cover plate into his chest. Fortunately others were in the building Saturday - to call an ambulance. He was not hurt. But so upset that they thought he was having a heart attack. Lightning is not a high energy source found in AC utility wires. For example, well over 90% of all trees struck by lightning do not even have appreciable damage. But lightning (just like static electricity) can blow through electronic junctions that are not suppose to conduct current - therefore cause damage. Just another reason why all discussions must include perspective. And that always means numbers. Subjective claims are best ignored as if a lie. |
|
|
|
Post by leonski on Feb 8, 2020 1:30:08 GMT -5
I was working in a place finishing cast aluminum car wheels. Tool was a tracer lathe. Someobody did not secure the 'trace' properly and the tool cut THRU the wheel and the rim flew across the room and buried itself in a wall.
Had someone been in the way? A crushed wherever-it-hit would have been the least of the worries. I've got the OTHER half of the wheel, which was clamped in the lathe, outside, hung and used as a doorbell (one good rap!)
A constant abd visible reminder.
We were forbidden, when I worked on stuff, to NOT resecure a plate / cover / guard with less than the number of fasteners it was designed for. This is particularly vital in Implanters which go to 200KEv and have several other HV
power supplies. Not to mention the Magnet PS which is a HUGE transformer supplying current to the analyzer magnet. Buddy was in violation of the rules, had an RF power supply on the FLOOR being serviced. An operator
walked by and DROPPED his or her pair of Stainless Steel Wafer Tweezers INTO the PS. Than reached in and retrieved them. NO smoke or flames. No death or RF burns.
Wanna hear about the Beckman Meter we charcoaled when the implanter arced THRU the meter to ground? I think the guy was checking the Extraction PS of 20KEv.
i'm not even going to get into the Toxic Gasses (some proprietary, so you only have the MSDS available) and liquid chemicals. Every work with HydroFluoric Acid? Or HCl? Or Sulferic? Entire fab was downed after a worked who
had JUST had the safety orientation (taken seriously) and cut thru a double containment pipe the INSIDE pipe of which had HF flowing. HF labels were 2 or 3 feet away on EITHER side of the cut. He was supposed to be working
on another line which had been emptied and flushed with DI water. Guys life? OVER. Next 5 or 6 months in Hospital. HF attackes Calcium so he got LOTS of shots. Lungs damaged. And more.
The lightning struck treses I've seen had Bark Peeled for 15' to 20' of trunk. A long 'slice'. That was in BOTH Chicago (fall tornado season was fun) and out in Cathedral City.....
|
|
|
|
Post by westom on Feb 9, 2020 14:48:12 GMT -5
i'm not even going to get into the Toxic Gasses (some proprietary, so you only have the MSDS available) and liquid chemicals. Every work with HydroFluoric Acid? Or HCl? Or Sulferic? Entire fab was downed after a worked who had JUST had the safety orientation (taken seriously) and cut thru a double containment pipe the INSIDE pipe of which had HF flowing. HF labels were 2 or 3 feet away on EITHER side of the cut. Those stories are legendary and necessary. If working with this stuff, then you know Gallup Labs. We build equipment to even handle Silane. At one point, designed hardware so that Silane could even be vented directly to air. And still it was exploding inside pipes. Eventually discovered people, who never learned how one must think to be safe, connected air to the Nitrogen lines. And deionized water (cleanest water) to water faucets. That happens when one does not understand basic principles for safety. For those who do not know (or foolishly think cleaner water is safer), had anyone drank from those water fountains, then their esophagus would have disintegrated. Semiconductor plants are chock full of dangerous chemicals including phosphine and phosgene. So I always review the MSDs just to know (refresh) what is in that plant, to know what to watch for, to run up or down, and how fast to run. So we know why something called a surge protector, is essential (and must be properly earthed). And why something called a surge protector (that is completely different if not only ineffective) can also be a threat to humans. |
|
|
|
Post by leonski on Feb 9, 2020 19:33:31 GMT -5
Terrific.
Venting SILANE to the air? One big problem is that since you know what happens with that stuff it's CLOG PRONE. Than the Silance in back of the clog doesn't react until AGAIN exposed to air. All in all, a real hazard and nasty stuff.
The Quartz Powder which aerosols into the air is toxic as hell, too. BAD on the lungs. Some Silane would come PRE-MIXED with some Phosphine gas which produces is less brittle and 'softer' oxide. Easier to etch, too. But the Phospine in the
Silane? AWFUL.
I saw a safety film I THINK made by either Bell Labs (no) or maybe IBM, when they were running East Fishkill as a manufacturing plant. It showed a silane tank in a vented box.....I think they did a simulation of a catastrophic failure of the plumbing. They shot at what in those days was a pretty FAST 400 frames per second, from the look of it on 16mm film. From one frame to the next, the whole thing disappeared.
Are you talking about Dionized water? No Drinkee.....
Plenty more....
How about the nice lady who put a measure of IsoPropyl Alcohol into a hot pot of Sulferic, instead of the required HydrogenPeroxide? Know what happens when you mix a Solvent with an ACID? OOOOOPS! Nobody hurt!
|
|
|
|
Post by westom on Feb 10, 2020 14:48:36 GMT -5
I saw a safety film I THINK made by either Bell Labs (no) or maybe IBM, when they were running East Fishkill as a manufacturing plant. ... They shot at what in those days was a pretty FAST 400 frames per second, from the look of it on 16mm film. From one frame to the next, the whole thing disappeared. Others told me about the IBM test. They measured the speed of a Silane explosion at greater than three times the speed of sound. These stories were constantly retold. It provided me with a right attitude to design software. And is why I redesigned that hardware. All Silane bottles are now scrubbed out before filling. That was not previously done until after people died. And after (this is not routine in all industries) one manufacturer openly shared details with all other manufacturers. A peer once told me his experience. He was installing hardware in the basement of a new Taiwan semiconductor fab. A silane scrubber would alarm. Tech looked inside and saw nothing wrong. So he reset that alarm. After a fourth alarm reset, that scrubber caught fire and burned down almost all of that new fab - just before it was to go into production. He said this. When a Chinese guy comes down, arms flailing, and lots of foreign language words, then drop everything and run. All this stuff is completely safe ... only if one follows the rules. Why did I rewrite software? Because the most dangerous item in a fab is that human. Returning to the topic. Human mistakes (ignoring specification numbers) explains this threat: imgur.com/X4s2tso |
|
|
|
Post by leonski on Feb 10, 2020 16:04:27 GMT -5
Don't get me started on the Chinese. I was working in a fab which was closing down. And was sold to China. Plant manager and head of engineering were BOTH Chinese, so when the
Chinese Reps showed up to evaluate stuff for shipping / packing, they really stretched the law and agreements. A couple guys went around PHOTOGRAPHING stuff including the spec book.
That was NOT part of the deal and company proprietary information. Company made stuff for military applicaitons. Some very exotic stuff including the use of Platinum Ssilicides.
Safety is TRAINING, specs and equipment. I did some pretty wacky stuff a LONG time ago when dealing with an Amonia Cracker used as our Hydrogen source....
|
|
|
|
Post by westom on Feb 11, 2020 11:03:52 GMT -5
I did some pretty wacky stuff a LONG time ago when dealing with an Amonia Cracker used as our Hydrogen source.... Hydrogen fire. Never seen until you start burning (as almost happened to a Space Shuttle crew before Challenger - before safety became important again). And another example of why better companies assign you a union guy. To point out what might be dangerous and where one should not go. General Motors top management called that feather bedding. Blaming unions for making more work. Those top managers never went into a fab and often did not know how the product works. Yes, many top GM executives did not even have driver's licenses. Just another reason why we must be pro-active about our safety. |
|
|
|
Post by leonski on Feb 11, 2020 17:24:36 GMT -5
the engineer I worked for was a top guy. No need for 'union' help, whatever THAT might be. We had multiple hydrogen burn-offs with redundand circuitry. My boss was also NUTS. I mean 'crackers', if you'll pardon the bad pun.
We had the catalyst in cracker changed. It was a MAJOR project to open the vessel. The company we hired for some of the work FAILED to make a leak (Helium Leak Checker) proof weld. My boss LOCKED HIMSELF in his office and
wouldn't come out. After the hired help left, he came out and did the weld himself. This guy had a VERY complete skillset from electronics to mechanics to design. His PAIR of Honda 750 Choppers were Epic with incredible quality,
handbuilt frames and perfect welds, even unfilled. I have a picture around here somewhere of one of the wrecks, being driven by somebody who had NO idea what he had. This was 35 or 40 years AFTER I had last seen the bikes.
Semiconductor processing (a decade later) uses lots of hydrogen It is flowed into a tube with an abundance of Oxygen. The 'source furnace' is typically at 800c or more, which causes the reaction to spontaniously proceed and make
a jet of STEAM. Steam oxidation is a way to rapidly oxidize Silicon Wafers. The color of the flame in the source furnace was unique. It also ran 'oxygen rich' so you consumed ALL the Hydrogen.
In my experience, all the bosses SAVE ONE were typically engineering types. The few times we were requested to basically violate the laws of physics, we were able to correct the problem. I was always attached in some fashion to
the engineering group, so generally, sanity ruled.
|
|
|
|
Post by westom on Feb 12, 2020 12:39:51 GMT -5
Semiconductor processing (a decade later) uses lots of hydrogen Purity of the materials is critical (ie grade 5). For others, that means a gas is 99.999% pure - five nines. To purify hydrogen, it is pushed through a metal sheet (platinum or palladium - I forget). Hydrogen nucleus passes through. Electrons are carries around and reunited on the other side. Then only pure hydrogen exists. Well, the fab needed hydrogen line pressure raised above 100 PSI. Many feared it would be unsafe. So they did leak checking on all hydrogen line. And discovered hydrogen lines on the roof, that distributed all over the building, already had tens of leaks. A reason why many pipes are routed outside (or are coaxial). Discovered was the plant spending much on hydrogen. A rather interesting way to make a gas pure enough for semiconductor work. And why management must come from where the work gets done. As one engineer in a local nuclear power plant told me, there was never a day when he did not learn something new. Bean counters have neither the attitude nor knowledge to do this stuff. Its not just technically challenging. Requires constant attention to numeric parameters. It can be dangerous. Always follow the rules. Rules exist for good reason - called experience that is never found in cost controls. Today, butt welding 316 steel is best done by a computer. Quality workmanship is critical. Previously every weld was video recorded using a camera inside that 5 mm pipe. |
|
|
|
Post by leonski on Feb 12, 2020 18:08:54 GMT -5
Virtually ALL our fittings were attached using what is called an 'Orbital Welder'. Tube and fitting put into a jig. Set up machine and press 'go'. Welding head makes one lap and you are DONE.
When a line is COMPLETE? Push 100PSI of Nitrogen into it and give it 24hrs or more to 'leak down'. It THAT is ok, than pump down using a helium leak checker and confirm No Leaks.
In those days you 'd use either VCR (a metal seal) or VCO (Viton or other elastomer seal) fittings. Swagelock was an option, but generally NOT for toxic or critical process gasses.
All LIQUID chemicals were 'SEMI' grade, meaing semiconductor grade. ALL gas bottles came with a certificate of analysis specifying contents down to PPM or better.
Our Hydrogen was sourced for the reduction furnace) from Amonia then put thru a molecular sieve of some sort. Sieve was BAKED and back streamed with N2 on a daily basis while the OTHER sieve was in use.
Oxidation furnaces used Bottled H2 but control was thru a device called an MFC. A cousinn of this is used in automotive emission systems ot measure airflow and the air/fuel ratio......
|
|
