Multi STOP TDC

[easyrider]

Hi, I want to make a Time to digital convertor (TDC, 100MHz)with multi STOP facility. I work on positron spectroscopy and the problem i have is following- I have a gamma ray (START signal) followed by emission of 2or 3 electrons . THe energy of the electrons are determined by the Time of flight of the electrons. Right now i am using a single STOP TAC, so i am loosing the information about the second electron. hence the plan to make a Multi STOp TDC. Can anyone please suggest me how to make one?

[DFGray]

More information about your hardware setup would be nice (signal levels, signal types, etc.).  Assuming you are using a scope device, do something like the following:

1. Trigger from the gamma ray start pulse (analog or digital?)
2. Scan the scope device long enough to pick up all the electron signals from the electron detector.
3. Analyze the scope trace for electron signals.  This will typically be some sort of peak detection algorithm.  Knowing the gain of your detector and energy of your electrons, you can probably figure out if you have one or more electrons in a single peak.  The peak location relative to the trigger gives you the time and hence the energy.
   

Please post more info on your signal types and levels and we can give you more help.

[easyrider]

Hi 

Thanks for the reply. I did not get the Scope part. Also our signals are analog. So right now I am using a ORTEC Time to amplitude Convertor(single stop) . I was thinking more along the line of Multi stop TAC. Also How i am supposed to Clock the event. And I was looking for the device to convert the time difference between the various pulses to convert to pulse which will be fed to Pulse height analyzer(MCA) where the information will be binned in different channels. Can you suggest me something on these lines

thank you  

[DFGray]

You can do everything you want to do with a single dual channel digital oscilloscope and some relatively straightforward programming.  The instruments you mention are hard-wired versions of the same thing, but lack the flexibility of the digital oscilloscope.  I believe the following will work for you.

1. Wire the gamma pulse into channel 1 of the oscilloscope.  This will be your trigger.  Since the gamma pulse is wired into one of the channels instead of the analog trigger, you will also get a trace of it.
2. Wire the electron detector into channel 2 of the oscilloscope.
3. Set the acquisition time of the oscilloscope long enough to capture all electron events you are interested in.  Note that if you get another gamma pulse in the same time period, you will be able to see it in the channel 1 trace.
4. The two digital signals from the oscilloscope are aligned in time and have a well determined period between each point.  This allows you to easily determine the time from the gamma pulse to an electron pulse.  Since it is a digital signal, captured on your computer, further analysis such as peak integration or peak height are also fairly easy.
5. Create an array of integers in your computer memory to serve as a multi-channel analyzer.  Increment these as additional data becomes available.  A very simple, but highly effective method of doing this is to simply add the channel 2 traces together.  This will give you an average arrival time of the electrons relative to the gamma pulse.

All of this assumes some sort of computer based data acquisition system is in place.  If you are only using standalone instruments with no computer control, you will require a specific, multi-stop TDC.  I would recommend against this, since the price of the TDC will probably be comparable to the digital scope (assuming you do not already have one) and you can reuse the scope for other things.  I implemented a similar system using an embedded PDP11/23 over twenty years ago, but modern software environments make things much easier (LabVIEW would have cut six months off my time as a graduate student).  Are you using a computer based acquisition system?  If so, what is the operating system and acquisition software you are using?

If you need more help, let us know.

NOTE:  I am a National Instruments employee, so I am biased.  However, in my years as a graduate student and industrial physicist, I have written data acquisition systems in assembly, FORTRAN, various flavors of BASIC,  C, C++, and LabVIEW.  In my opinion, LabVIEW beats them all.

[easyrider]

Hi

Thanks again. So the next question I have is the kind of scope I should go for. I have a Tektronix -TDS 210. I am thinking of buying TDS 1012B .This has a communication port to the PC. Then I would probably need a Data acquisition system(?) from NI. Can you suggest me that too, please? I have never worked with Labview but one of my friend hasexperience and he has agreed to help me out. Thanks again for the reply

saurabh

[DFGray]

The TDS-210 has an option for a communications port (TDS2CM Communications Extension Module).  From your answer, I will assume you do not have this.  The TDS-1012B looks like it would probably work for you. It is fully supported in LabVIEW.  However, before recommending anything, I would need to know what sort of timing accuracy and/or pulse widths you are measuring.  Make sure whatever you buy is fast enough and has enough time accuracy.  Another thing to watch for is that most digital oscilloscopes have 8 bit digitizers.  This is good enough for most applications, but you may require more.  For example, if you are analyzing a small peak on a large background, 8 bits (256 levels) will probably not be enough.  Finally, look at the signal length the oscilloscope can store in a single trace.  Your application may benefit from a relatively long capture time.  Many older digital oscilloscopes only have a 1024 or 4096 point buffer.  Make sure the instrument you pick can handle the data length at the time resolution you require.  Plan for the future.  This instrument will be around long after your current project is over.

Finally, I have to put on my National Instruments hat and recommend you check out our digitizers.  These digitizers plug into your computer and use the computer as the display.  They are somewhat more flexible than a traditional instrument in that you can easily use them as building blocks for another instrument type (e.g. lock-in amplifier, vector voltmeter, multi-channel analyzer, pulse detector, ...).  Data acquisition performance tends to be higher than conventional instruments.  However, you need a computer to use them and they are not as easy to just play around with (no knobs  ), so traditional instruments are better in this regard.

Let me know if you have any more questions.  My apologies about not being more specific in my first paragraph.  Our network was just upgraded and I am having difficulties connecting to parts of the Tektronix site, so could not get the specs for the 1012B.

Message Edited by DFGray on 08-25-2008 08:23 AM

Message Edited by DFGray on 08-25-2008 08:24 AM