SEM Gloat

[Jerry]

A little over a month ago a friend of mine found a SEM on craiglist. Only $300, seemed too good to be true. He tried emailing the seller with no response so I gave it a try and about a week later we were loading it up. And it was $300, plus another $200 for the sputter coater.

It's an old Hitachi S-450, made around 1980 or so. Came with all sorts of goodies which even fit the darn thing didn't work (Which they said it did not) it would be worth it. Came with a Kevex X-ray spectrometer (8" Floppies!), a 50l dewar for LN2, Edwards #12 direct drive pump, and a whole box of spare parts, stubs, etc. They told us they stopped using it about 10 years ago when the board in the X-Ray spec went bad. One of the other guys also mentioned something about a bad tube.

A couple weeks after we got it back to my friends lab we put it back together. Eventually it pumped down and tried to get an image, nothing. I machined a piece of acrylic to make a window for one of the accessory ports on the chamber and coated a stub with phosphor from a old fluorescent lamp. From this we were able to tell that we were getting an e-beam and you could see the raster on the phosphor. After that I got a HV probe from a friend and checked the other high voltage points. We were getting about +500v to the SED detector screen but when I checked the PMT power supply in the cabinet I found it was completely dead. The transformer was warm on it though so I figured it was the output diodes.

Remembering that I had noticed the cable had been spliced going to PMT head I decided to see what that was all about. I pulled the electrical tape off and found the problem. It seems some genius decided it would be too much trouble to remove the lid and disconnect the cable from the console when they went to move it so they just lopped the cable about 2' from the pmt. And then it appears they hired monkeys to solder it back together. The HV line is coax and the splice eventually shorted out and carbon tracked killing the PMT power supply. I ended up just removing the short length and redoing the connections. I hooked up a variable HV supply to the PMT and started to see something before I gave it too much voltage and killed one of the resistors in the PMT divider circuit. I also killed the coax so I replaced that as well. I tried again with a smaller Ultravolt power supply but that did not provide enough current as they used rather low resistance resistors in the PMT divider circuit. Since that was a bust I decided to just wait till the new diodes for the old power supply came in.

Today I installed the diodes and it works. buttoned everything up. Next is to try and figure out what is going on with the X-ray unit. Also need to figure out how to align the aperture. We have been running with it on the retracted position which does kill the DOF.

Here are some pics. The blue one was taken with a tripod in front of the screen and using a slow scan.


IMG_1340 by macona, on Flickr


IMG_1353 by macona, on Flickr


IMG_1350 by macona, on Flickr


546660_10150789638465614_708640613_9649901_1137363641_n by macona, on Flickr

[Doug Coulter]

Man, you have to talk to BillF more and tell him how you find this super nice stuff (so he can find more of it for me!).

[Bill Fain]

Hi, Yeah Jerry seems to have the knack for finding some cool stuff. I found a SEM about 15 years ago at a surplus sale. About the size of one of those rooms of furniture IKEA sells. If I had bought it, I was undoubtedly be divorced now. It's extra tough finding stuff around here in Podunk. I can find you all the copper worms and whiskey still parts you need, but, High Tech, that's a different story. -bill

[Jerry]

Like I mentioned, found it on CL. My friend thought it was a typo. It was not.

Finally got the manuals for it from Hitachi. They wanted a Dunn and Bradstreet number just to get the darn thing. Good thing too, the fast scan went kaput, we;; need the schematics.

I looked at the Kevex X-ray box. +12v is dead and +24 is at about 30. A friend at PSU gave me a newer control box for a Kevex and it looks like it will work with the sensor head. Thermo is sending us the manuals and stuff to get it hooked up. The software as well, which runs under Win 98. The new control is significantly more advanced. It will give you a map of the item being scanned and tell you what the different materials are. For example an IC die might show aluminum traces and silicon substrate. It shows a color map for the different materials.

[johnf]

Jerry
Very Good
Nice to see that you are trying to get the EDAX working.

I too have made a SEM gotcha
I Bought a Leo 440 (2001)for 1NZ$ (=US$0.76) for work and I hope I get the Philips PSEM-500 (1974) that we still use for nix

I did get the Oxford x-ray detector for the leo for my own use --it was in their rubbish pile not working with a US$ 30,000 for a fix quote attached

We'll see ---bet the internal preamp has had a hernia


post pics soon of both

[Jerry]

From what I was told it is often the Si(Li) detector. The guy that I got the box from said thats the one thing he has not figured out how to work around.

Going to get LN2 thursday. I was surprised how cheap it is. About $70 to fill the 50L dewar.

[Jerry]

Well, the SEM is now mine. My friend bought a Jeol 6300F and a 8600 microprobe off a guy in Cali. Trying to get those working now, the 6300F was out of Santovac 5 in one of the diff pumps, Jeol uses two diff pumps in series to lower back streaming.

The hitachi had burned out the last filament it came with so with the help of a friend at Portland State I rebuilt the two bases I have and then a guy on the internet send me 43 more bases I can rebuild. I got the filament wire off eBay, .089mm dia.

Next thing I am going to do is clean the apertures and then remove the diff pump and install a turbo in it's place so I don't need water to cool the machine. I have a Varian V301 to stick in it's place.

Then I need to see if I can get the EDX running.

The SEM that replaced mine:

Untitled by macona, on Flickr

[Jerry]

Well, I have the SEM up and running again plus a few mods. I ditched the diff pump and installed a Varian 301 turbo pump in it's place and am now backing that with an Edwards XDS10 oilless scroll pump. Since I dont have to worry about cooling or oil vapor I move it into my living room.

The scan issue did pop again and is now fixed permanently. The problem was a bad STK070 audio amplifier module that they used as the scan coil drive for the CRT. It went bad and the image went wonky. Lucky for me there was a spare module with the parts that came with it.

I am working on getting the EDX unit going now. I got the software for the MCA running on XP and talking to the VME chassis and have modified cables to work with the older preamp on the sensor. The EDX unit can also control the beam of the microscope and make digital images from the results. That would be real nice. Need to get the LN2 dewar filled next.

I wonder if I can use the EDX unit as a gamma spectrometer. I know the same kind of sensors can be used (Si(Li)) and I read some calibration info and they used a Co sample source for calibration. It works pretty much the same way. E-beam hits the sample and x-rays are bounced out to the sensor. The sensor spits out proportional pulses out to the VME frame to a pulse processor and then to a MCA and the results are displayed as a histogram.

Untitled by macona, on Flickr

[Donovan Ready]

I think I hate you.

[Doug Coulter]

I'll settle for just being jealous. I wish I had space for one for one thing. Maybe we can get Jerry to take us some pictures of samples so we can see flaws in what we make? After all, in private hands, it's gotta be a case of mostly looking for a use.