Actually what you do to go cheap is you buy a Jet milling machine and an aftermarket kit. But there are complete machines that actually work starting around $6,000 but with a smaller motor.
For aluminum you can use a router with the right bit and lubricant pump. A lot of aluminum can be machined with woodworking tools and special tooling. But real tools are not surprisingly beter.
For aluminum you can use a router with the right bit and lubricant pump. A lot of aluminum can be machined with woodworking tools and special tooling. But real tools are not surprisingly beter.
So you found the wrong welder in the first place, then found a machinist that under quoted to get the work that then raised his prices to recoup his losses. That's quite an amateur business plan to start with.
PS
What was the special welding technique, arc welding
Last edited:
We once built an ultrasonic phased array with 24 class A amplifiers that worked to 20 MHz
at 300 W level. We needed a metal hood for them and had a fine mechanics technican
bend one from brass. Took him the major part of a labor day and looked sh*tty. Yes,
precise to 0.05mm, but nevertheless.
Next day they took an Al brick, milled away everything that was not needed, made it
really thin & lightwight, with ribs to strengthen it, bore holes, threads, logos & polishing effects
and it also took the major part of a man day. The other 2 dozen hoods took an hour each,
nearly unattended. The whole thing was cheaper than bending from a brass sheet.
The removed filings went back to recycling, if you return exactly one type you get a
good price. The master of the workshop was quite proud.
Good & cheap.
They used the phased array to test the inner enclosure of nuclear power plants,
1st level pressure tubes & the like. Also for the propellant of a rocket booster. People felt
uncomfortable when they imagined to focus the ultrasonic output of 24*300W in
the propellant when checking for cracks.
Last edited:
Metalwork was one of my favourite subjects at school. My metalwork teacher was Mr. Cutmore, he had one finger missing due to an accident. He was also an excellent teacher
What happens when you take a lump of metal and you take away almost all the inside of it...
Our greatest concern, besides making the cases from aluminum and relatively thick was to make the exterior seamless. We had to sell these products to fussy people! Hogging it out was the best choice for us. We were too small to shop around for discounts. Of course, those of you with little or no experience, like we were when we started, have big ideas as to how it can be done much cheaper without significant visual compromise. In fact, we made a similar set of boxes, about 1/2 size for a stand alone phono stage. It cost us about $1500 to have it machined from separate parts, designed by a mechanical engineer friend of mine, and when it was done, the machinist said we should just 'hog out' the next ones, because it took just too much added machining to do it without hogging it it. Yes, we looked into it.
Fokker was extensively using machined alum sections from the 60ies. Then Airbus at 1970 and ATR.
Boeing didn’t adopt alum machining till late (15-20 years) and then selectively (*). For aluminum, they insisted on forged parts, chemical milling and riveted sheet metal work.
(*) they did machining on SS and titanium structural parts
Cheaper yes but not more reliable than previous fabrication methods.
Machined alum parts provide less vibration damping than forged or riveted parts, they are prone to exfoliation corrosion and suffer from lack of crack propagation control.
George
Does anyone here have any idea what SOTA mask making costs? Two orders of magnitude more than that just for each mask set, I have no idea what the equipment costs today.
George,
We can talk about the effect of the milling bit template on cracking, the issue shows up in more than aircraft.
But tempering is more of a material rather than a machining issue.
One of the shops I have spent time in was formerly used to make very interesting parts. There was a titanium swarf pile the size of a mid size truck and a tempering oven about 50' long by 22' wide, but only 10' high.
Suspect it wasn't used for commercial aircraft.
A friend did mention to me the method used to make titanium into hulls would be a completely different process.
We can talk about the effect of the milling bit template on cracking, the issue shows up in more than aircraft.
But tempering is more of a material rather than a machining issue.
One of the shops I have spent time in was formerly used to make very interesting parts. There was a titanium swarf pile the size of a mid size truck and a tempering oven about 50' long by 22' wide, but only 10' high.
Suspect it wasn't used for commercial aircraft.
A friend did mention to me the method used to make titanium into hulls would be a completely different process.
No idea, the minimum feature set I've worked with was 4" chrome masks with 2 um features (that spec could go down to 0.5 um). Think that was well over 10k for the device stack. So a couple orders of magnitude away from 8", much less 12" on far finer features. I have to admit, I don't know what sorts of minimum feature sizes precision analog components would garner.
I have priced a couple of 0.7 um High voltage MOSIS runs through and those are not cheap either. Edit to add, they don't have option anymore.
Scott: 1E7 dollars per mask times approx 35 masks for a SOTA digital chip.
No refunds if your flakeazoid photoresist guy says "I just don't like this mask, make me another one."
These are Intel-Xilinx SOTA prices. Not second tier Greenfield Fab In Malaysia, Two Generations Back prices.
No refunds if your flakeazoid photoresist guy says "I just don't like this mask, make me another one."
These are Intel-Xilinx SOTA prices. Not second tier Greenfield Fab In Malaysia, Two Generations Back prices.
Last edited:
I can provide a ballpark as of 2019: one e-beam pattern generator ~20M, one KLA reticle inspection (against the design database) ~30M. A full mask/reticle shop including the clean room, for the 5nm process node, is close to 1B. For the next gen 2021 2nm process node, probably double.
350M per mask set? I think this is grossly over inflated. The whole photomask market in 2018 was 4B in expansion to 5B by 2026 Global $4.97 Billion Photomask Market (2019 to 2026) - Key Players Include Applied Materials, Photronics & LG Innotek Among Others - ResearchAndMarkets.com | Business Wire
Last edited:
Never overlook the way to the egress.
I once attended a wedding on an island. Never again. I once attended a wedding reception on a boat. Never again.