How do I digitally capture the information going to a printer?

[Atomic]

Hi, I have an old (1985) Dektak 3030 surface profiliometer. It has 2 nm possible resolution (20 nm practical) and works great. I use it to profile the thickness of thin films deposited by magnetron sputtering on various substrates. The data that the profilometer returns is in the form of a physical printout from a small printer attached by a cable tothe control box of the instrument. There is no 'computer' output from this control box such as RS232, and the instrument was created long before things like USB were available. The printer is failing and is so old I cannot be fixed and it has been a pain to analyse the data by scanning in the printouts (the practical resolution is reduced to around 50-100 by the blurring of the printout by the scanner). To replace the instrument with a modern equivalent that can offer the same data precision and accuracy would cost tens of thousands of dollars or more that I do not want to / can't spend. I would like to somehow capture the data that would normally go to the small printer and use it to digitally process the data on a computer. What I need help with is somehow capturing the data that would otherwise go to the printer. I need to caputure it in a way that will allow labview to handle it (coordinates of where the ink would go on the page for instance). The cable going to the printer is 25 pin like the parallel port of a computer.When I connect it to the 25 pin parallel port of my computer and ask labview to read it in I get nonsense. When I use a 25 pin to 9 pin serial conversion adapter and plug the output into my serial port of my computer (RS232) and ask labview to read it I get nonsense. Any help is much appreciated!

[GerdW]

Hi Atomic,

even when you read in the data from the parallel port you probably get some "nonsense". This is due to that even those old "grrrrrk" needle printers used their device specific commands for printing. You have to parse that datastream to reconstruct the picture been plotted. So grab out that printer manual from the +20year old box somewhere in the storage cabin and write your "printer language interpreter" (Though it's not that complicated for a 9 needel printer, they mostly used some simple ESCAPE codes for printer control as I learned when writing a printer driver for a 8bit computer in those days...)

So first read in the data stream that is send to the printer (I doubt this is possible with a standard parallel port, I would opt for a 8bit DI card) then parse that stream to reconstruct the picture...

Or try to get an old TencorP2, which is able to write measurement data to floppies in a simple text format

[johnsold]

Atomic,

Crawl around the back of the instrument and look to see if it has some kind of analog output for a chart recorder.  Capture that with any analog input device.

Does the instrument have any kind of front panel display of the data you would like to capture?

I developed a system which captured the data coming from an old UV/vis spectrometer by grabbing the data at the front panel displays and interpreting that (pre-LabVIEW).

Lynn 

[GerdW]

Hi Atomic,

after reading Lynn's comment I darkly remember that a Dektak3030 too stood somewhere in the corner next to the TencorP2 I used in my former job. This profilometer was merely a "backup" solution when a coworker was already using the Tencor...

So for this kind of device I agree with Lynn to use a digital camera to grab the displayed image. Use a stand/tripod to get always to same dimensions to ease automatic picture analysis.

[micro1]

Did you figure out the best method for this? I am in the same boat with the DEKTAK 3030 and IIA systems (down to only one working printer) and am wondering which option you went with.