How does one aqcuire a scanning electron microscope (SEM)?

[MindBender]

Hi all,

I've been watching Richard's YouTube movie about operating his SEM in his living room a long time ago:

And I've been following Ben's quest to build his own on his Applied Science YouTube channel, and his adventures on eventually buying and modifying one:
https://www.youtube.com/user/bkraz333
And I've been watching Sam's videos on basically anything he's been doing:
https://www.youtube.com/user/szeloof

And I've been wondering: How does one source a SEM? 5 Years ago I had the same question, and by pure coincidence I've found one on a Dutch craigslist-like site: A Philips 525-M SEM/REM/EDX, fully operational. We never talked price, but I found the same instrument on lab equipment broker site too, for just a measly €24.900. That, and the fact that it would need my entire garage made me think twice.

I have found a trader in used equipement offering SEMs in the EU:
https://www.machinio.com/cat/scanning-electron-microscopes#results
The Hitachi looks nice, but I'm nowhere near spending that kind of money. And I honestly wonder who does. Companies doing high-tech work requiring a SEM usually have higher budgets than this. And the man-hours to get a 20...30 year old instrument installed, maintained and keeping it running make it still an expensive machine.

For an extra 10k€ extra, this 19 year old instrument can be bought:
https://www.ebay.nl/itm/Scanning-Electron-Microscope-TopCon-SM-200/153228494215
Still a boat anchor, but not one that needs its own room.

Doe anybody know how this works? Aren't the instruments listed basically written off? Or do they still have some professional use in them left?

[MadTux]

Basically the same with all used equipment that is more expensive and rather heavy.
Go there, inspect, pay someone with a truck or rent a trailer, get it home and get it back into working order.
But I would read some theory before, so to not end up with a garbage instrument.

That 19year old thing in the last link looks like one of those budget instruments with shit image quality.
Any 30-40 year old field emitter (like Jeol JSM 6400F) will outperform that one in therms of resolution.
Had access to a rather nice Hitachi S4800 several years ago, that thing was way ahead in image quality compared to a new FEI desktop toy, which I could play with later on at university.

That's also the reason why those old goodies are still expensive, retrofit digital image acquisition and those things will outperform new budget instruments.

[tmbinc]

I have two Zeiss SEMs from the '80s; a DSM950 and a DSM940. Both Tungsten, 30kV max., SE+BSE detectors. I've paid ~4500 EUR for the first one, and much less for the second one (missing some parts though). Resolution is spec'ed (and was initially demonstrated) at 5nm but the stars (and apertures) need to align right for that. The first one I bought at an auction (I was the only bidder), second one from a private seller.

Getting the device shipped is a pain (the column is ~500kg - doesn't sound like a lot, but believe me - it is!, and on top of that there is the console with another 300kg), but once you have it on a pallet, it's all doable.

In terms of operation, they are surprisingly robust. Full schematics are available, and they are in a sweet spot of being new enough to not use analog circuits with unobtainium parts all over the place (and hand-drawn schematics!), but also not new enough to use highly integrated, potentially custom, parts. Signal processing and control is all mostly TTLs with a few Z80s sprinkled in (for which the firmware can be easily reversed). I've retrofitted digital image acquisition (with an OpenVizsla and custom software), and I'm in the process of modernizing the complete electronics to get rid of the console (and the '80s technology) completely.

It's not necessarily an expensive hobby (though to have fun, you need at least a sputter, otherwise what you can image is severely limited), but it can be a very time consuming hobby. I'm still fighting with being able to get a good resolution (better than, say, 50nm), but it's the combination of a well-tuned and -maintained device (which I don't have), lot of experience (which I'm only slowly gaining), and well prepared samples (which I'm still learning).

[MindBender]

Basically the same with all used equipment that is more expensive and rather heavy.
Go there, inspect, pay someone with a truck or rent a trailer, get it home and get it back into working order.

Clear. But what are good sources to find them? There're the usual instrument brokers, but they double the price, lack the knowledge to help out getting it installed and operational, and they don't offer any guarantees.

An who am I competing with? Do those 50k instruments advertised on eBay really sell? That seems a bit much for an as-is item, even if it was decommissioned in working condition.

Getting the device shipped is a pain (the column is ~500kg - doesn't sound like a lot, but believe me - it is!, and on top of that there is the console with another 300kg), but once you have it on a pallet, it's all doable.

That's a good point: ~500kg sounds like a lot, but still feasible to me. Practically though, it may be a lot harder indeed.

In terms of operation, they are surprisingly robust. Full schematics are available, and they are in a sweet spot of being new enough to not use analog circuits with unobtainium parts all over the place (and hand-drawn schematics!), but also not new enough to use highly integrated, potentially custom, parts. Signal processing and control is all mostly TTLs with a few Z80s sprinkled in (for which the firmware can be easily reversed).

The sweet-spot between hand-drawn circuits and custom silicon: That is a good point and very good advice, thanks!

It's not necessarily an expensive hobby (though to have fun, you need at least a sputter, otherwise what you can image is severely limited), but it can be a very time consuming hobby. I'm still fighting with being able to get a good resolution (better than, say, 50nm), but it's the combination of a well-tuned and -maintained device (which I don't have), lot of experience (which I'm only slowly gaining), and well prepared samples (which I'm still learning).

I was already surprised it is possible to get a decent image without coating it all, but sputter-coaters aren't that hard to find, or even to make.

I still have doubts about the real-estate problem. When I still was a bachelor, I would have proudly installed it in my living room, but I don't think I need to explain why that isn't an option anymore, right? I have a pretty spacious lab in the basement, but for that it's still rather large, and there's no way I can get that 500kg column down the stairs. What's left it the garage, where also my X-ray machine lives. I had planned to build a small mechanical workshop there, but I don't think I'd use that a lot. I would have to remodel it a bit, but it is very feasible. Does it need to be heated? I does get cold there in winter, but not (much) below freezing.

And finally: Do you guys know about any SEM microscopist workshops I could take, to see how much fun it is, before I take on this endeavour?

[DonIdai]

Hi tmbinc,

I just got miself a DSM 960. Just the column box as the whole controlling unit was lost years ago. I have been looking for service manuals, column drawings and circuit schematics for a while now with no luck. I found some pdf files from a german forum but they were mostly on the control unit, which I dont have...
Do you happen to have any manuals on this SEM or perhaps you now where I could get them?

Thanks,
Toni