building a prototype

hey guys I´m really interested in building an electron beam gun, but when I downloaded the assembly zip file and used freecad to open the .STP file it just stops responding, could some one help me to build a prototype?

THANKS!!!

See notes at the bottom of files and parts section

'''Freecad''' the above downloadable files only work on the daily development version on the latest Linux ubuntu trusty! (see below instructions)

(The latest version of freecad is required called Freecad-dev (the daily updated version of freecad only runable on the latest ubuntu linux, called linux ubuntu trusty (no windows or OSX supported currently)), install the PPA from [launchpad.net] by adding its package name PPA "ppa:freecad-maintainers/freecad-daily" , simplest method is install synaptic package manager on linux ubuntu trusty and add PPA above to list and update packages, or follow instructions on site[https://launchpad.net/~freecad-maintainers/+archive/ubuntu/freecad-dailyfreecad-dev])

kind regards
Rapatan

hello, thanks for your reply, I noticed that the metalicarap page has been updated, I´m still trying to build a prototype, i have no science or engineering background so im not sure if whats in the page is going to work or not, specifically the power supplies and the electron beam gun, can someone tell me if those 2 elements of the metalicarap are functional?

THANKS!!!

In answer to your question, nobody knows if anything on the wiki will work. Until the guys on the metalicarap team actually build the power supply they've designed, and make an electron beam using it, I'd forget about this project.

If you have any electrical engineering expertise, the wiki might be a good place to start.

Honestly though, I'd not expect this team or specific project to produce a working machine any time soon. I've been following them for 3 years now I think, perhaps even longer. I check in every few months at least. In that time they've updated the wiki once and finished the design of a power supply that hasn't been built or tested and is based on a patent from the late 70s.

Maybe I'm wrong. Perhaps they'be been busy working away every night, and have an almost-working prototype by now! But their web presence doesn't suggest so, neither does their slow progress or lack of physical hardware. And it doesn't inspire confidence when they do stupid things like use a version of freecad that's Linux only, or write a ton of incomprehensible gibberish on their wiki page.

Sorry, bit of a rant there. I'm just getting a little frustrated at the lack of progress from this project, only because they are soliciting donations. I'd have no problem with the slow progress if they were using their own time/money.

... the metalbot forum seems to fall asleep ... no real action in the last months

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
wogglestick
Honestly though, I'd not expect this team or specific project to produce a working machine any time soon. I've been following them for 3 years now I think, perhaps even longer. I check in every few months at least. In that time they've updated the wiki once and finished the design of a power supply that hasn't been built or tested and is based on a patent from the late 70s.

I have to agree. I've been keeping an eye on this project for about 4 years now and there's very little in the way of a physical prototype.

The biggest problem i see is that there trying to run before they walk, building a large device that will cost tens of thousands which very few people will want to pay for. What they should be doing is making a small cheap version ('prototype') that more people can afford and have space in there house to have. I've already suggested reducing the build area to a 5 by 5 cm area, while that might not sound very large there are plenty of things that can be made in that. Would drastically reduce the costs and overall size. The closer distance of the electron gun means there less distortion from the small amount of air in the vacuum meaning instead of have a £2,000 10-4 pump required you could get away with a £100 cheap 10-3 one which, while not giving the best results would also drastically reduce costs. If they sold these with a slight profit at the right price people will buy them, even if the output products of them aren't the best, money which they could then funnel back into making improvements like the bigger version, better parts for upgrading the device. I'm sure someone could come up with a design for a better pump that would reach the 10-4 which owner's could print out and install making there printer's better.

If they did come out with a small version that is cheap i would probably be one of the first in line to get one, i have idea's for a lot of things it could be used for and would probably be wanting a few more in no time, possibly even at some point having use for the larger version for bigger run printing of parts.

... one of the biggest problems they're facing is the complexity of a proper vacuum + e-beam system.

From the beginning on I was recommending the simplicity of laser-sintering at ambient pressure (with or without shielded gas) instead of e-beam melting in a vacuum ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

This guy has done a bit [alangrimmer.blogspot.co.nz]

... look at the volume to evacuate ... and you'll need a pretty good (expensive) vacuum pump and a stiff enough sealed housing to get a vacuum good enough for an e-beam.

With a laser the housing has only to be 'light-tight' ... and you can get a laser powerfull enough to melt powder for under USD500 ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Do you have a link to the $500 laser ? What spec is it? and roughly what would the optics cost for such a high power setup? I searched and searched awhile ago and never found anything that cheap

I've always wondered if you get away with a lower wattage laser in a vacuum, you already have half a chamber for the inert gas with a laser.

... some years ago I've made a WiKi for some laser gear and infos - [reprap.org]

You can find fibercoupled 9Watt-IR-diodes at ebay for roughly USD300 - this power and energy density (best spot diameter is 0.1mm) is enough to melt dark plastic powder and even some metal powders, so good to start with powder bed SLS.

I was selling 9Watt-diodes (which I'm extracting from broken fiber-lasers, where they are used for pumping the resonator fiber) with cleaved fiber-end and an aspheric plastic lens for 120 Euros, drivers/controllers for them for 80 Euros, complete with armour around the fiber and housing for the lens for 300 to 400 €uros ('naked' or in a module housing).

Instead of a NIR-diodelaser you can use 445nm-diodes too - they are available with powers of 2W, 3.5W and 6W, what's more than enough for absorbing plastic powder and can be combined to a more powerfull module too - look here: [forums.reprap.org]

But you can start with a cheap chinese 40 Watt CO2-laser too ... you can find them with Controller-PSU for below USD 500 too ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Oh wow those are small lasers, I've been looking out for 200W's as a minimum to melt metal power and have it fuse into a solid metal part.

Its all well and good melting powder but you need to re-melt the part behind it which is a huge heat sink. I've played with the 405nm 2 W Blueray diodes and they can melt soft powder's but there useless at doing it to a solid metal part.

CO2 laser's are the wrong wavelength for metal's. I think you need around 800-1000 Watts to do what a 200 W fiber of the correct wavelength can do.

Edited 1 time(s). Last edit at 05/16/2016 05:52PM by pyrotronics.

... there are some powders with a metal core and a thin coating of plastic or glass.

The SLS process will only melt the coating, the parts will be later sintered/remelted in an oven with some shrinkage.

Pretty interesting metal powder materials could be low-temp melting alloys like rose's metal (melts at 95 degC) or field's metal (melts at around 70 degC, but more toxic than rose's).

Other common metal fabbing methodes are SLS with wax- or plastic powder and then "lost-cast"-moulding with embedding the parts in sand and pouring molten metal into them ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
VDX
... there are some powders with a metal core and a thin coating of plastic or glass.

The SLS process will only melt the coating, the parts will be later sintered/remelted in an oven with some shrinkage.

Other common metal fabbing methodes are SLS with wax- or plastic powder and then "lost-cast"-moulding with embedding the parts in sand and pouring molten metal into them ...

Thanks for the info! I appreciate what you're saying VDX, and lasers are awesome, but I really don't think the methods you listed here are comparable to an e-beam setup.

First off, those metal powders coated with plastic or wax must be expensive. Then you need a really hot oven, a kiln most likely to remelt the parts. Of which unless you built yourself an industrial high-temp kiln (no small feat) you'd be limited to aluminum, bronze, copper and other lower melting point metals. Steel, iron, titanium are all absolutely out of the question.

Secondly, if you have to go through the hassle of building a kiln to remelt, you might as well just go the lost-wax route to begin with. Wax is way cheaper than the metal powder coated in plastic/wax and recycled Al ingots are like $5/lbs in small quantities. Much cheaper wholesale. Why not save yourself the hassle and expense of a laser sintering setup and use a cheap, readily available FDM printer (or resin printer) and lost-wax your parts?

I'm not really seeing an advantage to the way you described except "pew pew lasers!" Your way certainly sounds more expensive as far as the metal-coated powder goes. More niche in that it requires a laser sintering setup (side note: pretty sure SLS is still under patent so you can't sell machines), and yet still requires the same amount of equipment as a lost-wax setup (a kiln). I guess there is an advantage in that you don't have to pour molten metal, but there's still no way you could do any of this inside your home.

With an ebeam, it uses less power because it's more efficient, is much 'cleaner' in the sense that there's no dirty kilns or molten metal to pour everywhere and can do things like steel and titanium without breaking a sweat. On top of that, an e-beam machine can make parts that are non-porous and 97% density of forged parts. So they're amazingly strong and durable. I wouldn't trust a remelted part on my nightstand, let alone on a bike or car. But I'd happily trust an ebeam part in those applications.

Am I saying it's trivial to build an e-beam machine? Absolutely not! But please don't compare a 9w laser sintering setup with an e-beam machine. They both have their pros and cons, but shouldn't ever be compared apples to apples...

I'm pretty sure that SLS patent has expired late last year.

Also remember that the E-beam is the cost of the vacuum chamber + pump + power supply. The laser cost would be 200W laser + optics + galvo's + power supply + inert gas chamber.

I would love to use a laser, if we found a dirt cheap way to make a 200 W fiber laser?

... another aspect is the overall resolution possible with the different methodes - the smallest spot with a 405nm@500mW-diode (what's more than enough for wax-SLS) is 0.03mm

My last experiments with NdYAG and fiber-lasers with common optical heads gave me a best resolution of 0.02 to 0.03mm ... the smallest spot with a NdYAG-disk-laser and a better optics for the fiber-laser was between 7 and 10 microns, what was good enough, to fuse/melt a 10 micron thick platinum wire to the surface in an additive-wire-melting setup ... but yes, this lasers will be more like 20k€

Now check, which resolutions were common with e-beam-melting -- AFAIK, most of the working (proffesssional) setups states a detail resolution of 2mm!! ... they mill/lathe the parts to the final form ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

... searched a bit for accuracy numbers with e-beam fabbing - here is one more actual datasheet - [www.additively.com]

Quote

- Minimum feature size: 0.1 mm
- Typical tolerance: +/- 0.2 mm (can be improved through machining)"]

It seems, with special, more precise (and expensive) setups, they can get actually around 0.5mm overall accuracy ... compared with 0.05mm with 405nm-laserdiodes in wax or 0.01mm with fiber-lasers ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

... but here again - the same site gives pretty coarse numbers for laser-SLS! - [www.additively.com]

With fine powders (particles between 300nm to 3 microns) I've got around 0.02mm accuracy/resolution with standard optical setup and can go finer with better/finer optics.

And UV-Stereolithography is stated with nearly the same 'coarsiness' - [www.additively.com]

... while with 2-photon-UV-lithography (see Nanoscribe) they print with 25 Nanometers!! resolution ...

Edited 2 time(s). Last edit at 05/17/2016 08:59AM by VDX.

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
pyrotronics
I'm pretty sure that SLS patent has expired late last year.

Also remember that the E-beam is the cost of the vacuum chamber + pump + power supply. The laser cost would be 200W laser + optics + galvo's + power supply + inert gas chamber.

I would love to use a laser, if we found a dirt cheap way to make a 200 W fiber laser?

Good to know, thanks! I don't get your second point... I realize the cost for an ebeam machine is in the ancillary support equipment. Are you saying it's cheaper for fiber lasers? I also would love to use a laser, see my points below though on cost.

Quote
VDX
... another aspect is the overall resolution possible with the different methodes - the smallest spot with a 405nm@500mW-diode (what's more than enough for wax-SLS) is 0.03mm

My last experiments with NdYAG and fiber-lasers with common optical heads gave me a best resolution of 0.02 to 0.03mm ... the smallest spot with a NdYAG-disk-laser and a better optics for the fiber-laser was between 7 and 10 microns, what was good enough, to fuse/melt a 10 micron thick platinum wire to the surface in an additive-wire-melting setup ... but yes, this lasers will be more like 20k€

Now check, which resolutions were common with e-beam-melting -- AFAIK, most of the working (proffesssional) setups states a detail resolution of 2mm!! ... they mill/lathe the parts to the final form ...

Firstly, you didn't answer any of my concerns regarding why bother with remelt parts in the first place, and the expensiveness of the wax coated metal powders all while still needing extra equipment (kiln) and being limited in materials (no steel/titanium).

Yeah, for lasers of that power level I'm not at all surprised they can get spot sizes that fine. But again you're talking about remelt parts, and with a remelted part what is the FINAL tolerance after remelt? Do these parts perfectly, uniformly shrink as they are remelted? No, they don't. Unless you pay a LOT of attention to the way in which the part absorbs heat as it shrinks (which would require some fairly hefty simulations) it's basically a crapshoot with what your final part will look like. I understand your arguments, and I'm with you guys on using DMLS but please don't resort to sales tactics. We all know that lasers are much easier to work with and more accurate, and if we can find a low cost fiber laser, that's the way to go using DMLS.

But low cost fiber lasers don't exist. The cheapest 200W fiber laser I can find is... I have no idea. I found some 50W fiber lasers for $12,000, I can't imagine it gets cheaper as the power quadruples. The only companies I've found that produce fiber lasers are American or European, which means prices to match. So until one of the following happens, we're completely out of luck:

1) The Chinese figure out how to make low cost 200W fiber lasers.
2) We wait 10+ years for the prices to drop.
3) We figure out ourselves how to design and make cheap fiber lasers.

Of all those, 3 is really the only option we have any control over. I have no clue how a fiber laser works. I watched this: [www.youtube.com] and STILL have no clue how it works, let alone how I'd go about building one. I'd be happy to help any way I can, but until someone figures this out at-home metal part creation isn't happening anytime soon.

Edited 1 time(s). Last edit at 05/17/2016 01:27PM by wogglestick.

... why are you so focussed on 200 Watts?

You can evaporate/engrave metal with 10 Watts @NIR on a spot of 30 microns, so a fiber-laser with 20 Watts is more than enough to start with metal powder SLS.

This numbers with 200 Watts are more interesting for high-speed XY-scanning the beam, so the building time per layer is minimized - what's essential in comercial views.

For private use a SLS-job can be ok with some hours needed time, instead of 10 minutes -- if you'll invest only 6000 USD instead of 150000 USD.

My last price lists shows me prices below 5000 USD for a pulsed chinese fiberlaser with 20 Watts ... and 50000 for one with 200 Watts ...

Here is one at ebay for below 4000USD, sold in the US - [www.ebay.com]


For starting with metal powder, cheaper lasers (laser-diodes) and lower resolution than 30 microns you'll need the same energy density - so let's sort some numbers ...

With my fiber-coupled IR-diodes with 9Watts@975nm out of a 0.1mm fiber I'm receiving a spot diameter of 0.1mm best - so, to compare with the fiber-lasers -- averaged roughly 1 Watt on a spot of 30 microns only!

But I can combine several fibers on the same spot! - so, to get 10 Watts on 30 microns (or 100Watts on 0.1mm) I'll expect 12 to 15 fibercoupled diodes with 9 Watts each and a combining head.

With a price of roughly 300 USD per diode and the optical head it's in the same price range like the chinese fiber-lasers ... so a fiber-laser will be much better in terms of accuracy and efficiency ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I'm "so focused on 200W" because that's the absolute minimum I think would be needed to get useful print times. A regular CO2 laser with a 60W tube will barely cut plywood. Now, I understand it's all about focusing the beam enough to get a really small spot size, but let's be realistic, how long does it take to melt a 1cm x 1cm area of metal powder with a 20W laser?

A ridiculous amount of time, that's how long. It wouldn't be practical to make anything larger than a few cubic cm unless you want to spend weeks or months waiting for a print. I can't be bothered to do the math, but I'm assuming 1/10th the power of a 200W laser equates to at least 10 times the print time. And 200W is low to begin with by commercial standards, so a part that would take 1 day for a 200W machine to make would take you 10 days.

I don't mind leaving my printer on over night while I'm in the house, but to leave it on for 10 days with a really powerful laser firing all the time? No thanks. Think of how much inert gas that will use as well! In 10 days, I can 3D print a part, hand sand it/finish it, cast it in silicone, make a bunch of wax copies, set them all in plaster and burn them out in the kiln, then pour a bunch of copies in aluminum using equipment that cost under $1000. If I wanted to lost-wax just a one-off 3D print I could probably do that in maybe a day or two.

I guess I'm getting to the crux of my argument here, which is why? Why would you want to try to sinter parts using a 20W laser when traditional methods are quicker, cheaper and more efficient? It doesn't make any sense. There are no practical benefits, there's nothing that can be done with 20W DMLS that's cheaper or more efficient than other methods. So why would you do it? Please note: I'm not questioning if you can do it, I believe you can, I'm asking why would you?

Literally the only reason I can think of is to have the cool factor of saying you make things with lasers... which is pretty cool. But is it worth all the cost and effort to have a machine that cranks out tiny parts really slowly?

... my personal goal is not printing big parts (this is better made with CNC-milling), but small, high-accuracy parts, not possible to make with milling ... and multi-material-printing.

If you look at the best accuracy numbers from Stratasys, then you can read:

Quote

What level of detail can be obtained with DMLS?
DMLS is available in several resolutions. At its highest resolution, the layer thickness is 0.0008” – 0.0012”
and the X/Y resolution is 0.012” – 0.016”.
The minimum hole diameter is 0.035” – 0.045”.

DMLS GENERAL TOLERANCES
Based on historical results (and application dependent) general tolerances for DMLS parts are ±0.005 inch for the first inch
and ±0.002 inch per inch hereafter (±0.2 percent). The finish of DMLS parts as built are 350 Ra – µinch.

In Millimeters:
- layer thickness = 20 to 30 microns
- X/Y resolution = 300 to 400 microns
- minimum hole diameter = 900 to 1140 microns
- adding tolerances = +/- 127 (50) microns


One of the main "drawbacks" with high powered lasers is the bigger focus, compared to lower laser powers, caused by the bigger resonators and needed beam combinig.

With my resolution tests with spot sizes of 10 microns with a 20Watt-disk-laser (or a 16Watt-fiberlaser with beam-expander) and powder with 2-5micron big particles I've got much finer XY-resolutions of 20 to 30 microns and minimum hole diameters of 50 microns while sintering, and 20 microns with 'drilling' !!


Now compare the power - 200 Watts on a focus diameter of 50 microns is only 40% of the energy-density of a 20 Watt-laser with a 10 microns big focus!

So with a smaller focus I'll need much more time to melt the same volume, but I have much higher accuracy and slightly more power (=energy-density) than with the 200 Watt laser.


Another "drawback" with the common comercial systems is the sheer size or footprint they have or need.

Think about a 'desktop' system with a footprint of maybe 200x300mm - and a building volume of maybe 100x100x100mm - and 10-fold better accuracies, than possible with the actual comercial systems ... and much lower volumes of used (lost) shielding gas too

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Again, totally believe you. I feel like you're not understanding what I'm saying to you. I understand your math and the physics behind it. I understand you can melt metal with an insanely small spot size, nobody is saying it can't be done. But I really think you're down-playing the importance of print times. Sure, a 10 micron spot might make melting a little quicker due to higher power density, but the spot then has to travel at least twice the distance over a 100mm x 100mm bed size. Effectively doubling your print time for each layer. I'm sure a 10 micron spot will not be able to do as large of a z height either, further increasing print times.

You conveniently forgot to address my other related issue, which was that nobody will want a relatively high powered laser left unattended in their house or workshop. Which they will have to do due to long print times.

Look, I'm not trying to be negative here just trying to get you to think about your design and it's uses but it's obvious you already have a good idea of what you want. I'd actually love to see some pics or video of your melting experiments! It sounds awesome. Do you have a blog or something?

... I think, It's the different "viewing angle", what's causing some communication issues

My goal are small, high accurate parts with only some millimeteres to centimeters building sizes -- so the building times are in comparable regions as with common SLS-fabbing.

When developing/using this "micro-SLS" was under heavy NDA's, so no info to share.

In the last years I was collecting the needed gears to repeat this for my private use -- and have now all sorts of needed lasers, optics, electronics and micromechanics now in my basement, so actually developing the basics (controllers, software, housings, ...) and collecting/organizing spare time to build the modules.

I've posted parts of my laser-development in different forums - here is some of the english threads/posts in the RepRap-forum: [forums.reprap.org]

But you can find more of the 'related' infos in the metalbot forum too ...

Edited 1 time(s). Last edit at 05/18/2016 05:48AM by VDX.

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I guess I'm just asking the question, I know its probably impossible, if someone built a dirt cheap 200 W fiber laser I would be interested, otherwise the E-beam wins the price/ power war.

Quote

Good to know, thanks! I don't get your second point.

You also need a high power laser for melting the metals to get a nice grain in the finished part that will be stronger.

Really awesome discussion. Pyrotronics do you think you could build or buy an electron gun system for less than the cost of building or buying a 200w laser? I know SEMs can easily run into the 100k plus range and simply assumed that an electron gun system would register a significant cost as well

... it's not only the "laser_vs_e-beam" price, what's counting in - I think, the lack of progress of the related projects in the metalbot-forum is caused by the high complexity (and costs) of building a vacuum chamber and electromechanical setup, good enough to work with an e-beam in UHV (ultra-high vacuum).

If you're not lucky enough to get hands on a surplus vacuum setup, this can draw some thousands to tenthousands bucks more, so not in the common DIY range.

While for a laser system it's as simple as building a box and find some pneumatic gear to feed the shielding gas into the working area ... and you don't need 200Watts, even for faster and coarser SLS - 60 to 100Watts will be enough -- and can be found for below 20000 USD new for a chinese 60W-fiberlaser, or much cheaper if second-hand or surplus ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
FRC
Really awesome discussion. Pyrotronics do you think you could build or buy an electron gun system for less than the cost of building or buying a 200w laser? I know SEMs can easily run into the 100k plus range and simply assumed that an electron gun system would register a significant cost as well

The cost for the beam really isn't in the actual electron gun itself. It's in the ancillary support equipment. Vacuum chambers aren't cheap. Turbo pumps aren't cheap. High voltage power supplies aren't cheap. So as a whole if you tried to build an EBM printer using off-the-shelf parts, you could easily end up spending $100k. Probably ten times that if you went with a custom-made vacuum chamber.

But each individual piece isn't really THAT expensive when you take it down to the nuts and bolts. A high voltage power supply is maybe $2k in electronic components, cheaper if you bought in bulk. It would be possible to build the design on the wiki using a 3D printer, PCB mill, soldering iron and an angle grinder. Turbo pumps can be acquired from ebay relatively cheaply, you might get a great deal and snag one for $1k with it's power supply. Personally I'd try a really good rotary vane pump that goes down to 10^-4 torr and see if you could get away without a turbo pump first, you can always add one later. As for the vacuum chamber, well... that's by far the most difficult part of the build in my opinion.

Making a vacuum chamber is no easy feat. Specifically the welding, you have to get the welds absolutely perfect, and weld from the inside-out. For the application we want, which is considered high vacuum, buna rubber seals and an aluminium vacuum chamber would probably work. You may also be able to coat the welds with high vacuum epoxy to cover over any imperfections in the welds. But high vacuum epoxy isn't cheap. However, it would be possible to build one with a well equipped shop and an expert welder. You can only use non-magnetic metals btw, so aluminum and stainless are your friends.

Another tricky part is the internal motion. You can't just wang some normal stepper motors in there, because the plastic that coats the wires inside of normal stepper motors will out-gass when put inside the vacuum chamber. Out-gassing is not good, it will ruin the vacuum inside the chamber and make it that much harder for the pump to reach the required pressures. I found some vacuum rated stepper motors once (they were basically steppers that had had their wires replaced with a vacuum-rated coating) for the low low price of $1000 each. Yes, each. So what about rewinding the coils inside of a stepper? Not a particularly great idea. The stepper will lose a lot of its torque when disassembled and put back together again. You could probably convince a Chinese manufacturer to make steppers using PEEK coated wire. Maybe. I have no idea how much they would cost though.

Other bits inside the chamber are probably not that hard, so long as you pay attention to the materials you make them out of. There's a great wikipedia article about materials best suited for use in high-vacuum applications, I'd try and stick to aluminum and vacuum rated plastics (teflon) for your internals.

In all, if you had a very well stocked shop or makerspace with some decent equipment it would be possible to build yourself an EBM machine. It'd take you some time to get right, but you could probably do it for $10-20k if you shopped for bargains and didn't count your time. I've been considering building one for a while, but I don't have access to all the equipment I need (or a spare $20k!).

Edit: VDX just saw your post and I'm going to have to call you out on the Ultra High Vacuum part. UHV is anything at or below 10^-7 torr. E-beams start working at 10^-4 torr, which is considered high vacuum. High vacuum is much easier than UHV. There's tons of epoxies that work at high vacuum, buna rubber will work as a seal instead of copper gaskets, and some rotary vane pumps go down to 10^-4 torr without any help from a turbo pump. So again, much easier. Not saying it's trivial by any means, just giving some perspective.

Edited 1 time(s). Last edit at 05/20/2016 04:30AM by wogglestick.

... so it will be ok without the "Ultra"

I was mentioning UHV because of my own experiences while studying - our group was building all sorts of electron- and ion-sources and UHV-equipment for GSI and CERN research and I had to help/assist with some of the projects (my own work was software related)

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

That's awesome you have experience with UHV stuff! Can you say what you were working on? Or is it classified?

Edited 1 time(s). Last edit at 05/20/2016 02:11PM by wogglestick.