[3dem] Current measurement with viewing screen

Thu Nov 11 02:38:19 PST 2010

Dear Fred,

We have characterized our 3 in-house detectors 
( see http://journals.iucr.org/d/issues/2010/01/00/ic5061/index.html and  http://www.diplib.org/add-ons ) 
and initially used  screen current measurements to calculate the conversion factors of the cameras. Fortunately, we had an opportunity to validate them using a Faraday’s cup (Gatan, model 646) and to compare the results. The relative errors between screen current and Faraday’s cup measurements for 2 cameras were +3.5% and -2%, but for the third camera it was 35% (@ 120kV) and 30% (@200kV).  

The conversion factor (from the screen current measurement) of another (4th) camera was compared, this time, with another approach – (GIF drift tube measurement) - and the difference was around 1%.  

Additionally, the effective gain (plot variance in the image vs. intensity in the image [in counts] ) gave us some insight about statistical properties of the cameras , although it differed from the conversion factor about one order of magnitude.
One of the reasons for this difference is quite large point spread function of the scintillator-based EM cameras. If we rebin the images the influence of PSF is smaller.
One could get a rough estimate of the conversion factor in the following way:
- Rebin your large uniformly illuminated image (with factor  16x16).
- Calculate the effective gain (slope of the plot variance-intensity in the image) 
(For this purpose you can also use the function ‘ef_gain.m’ from our toolbox available on the website mentioned above)

For the cameras that we characterized this rebinning (16x16) reduced the difference between the effective gain and the conversion factor (measured by Faraday’s cup) to less than 10%.

Hope this can help
Best Regards,
--------------------------------------------------
Dipl.Ing. Milos Vulovic
PhD candidate

Delft University of Technology
Faculty of Applied Sciences / Quantitative Imaging Group
Lorentzweg 1  / 2628CJ  Delft
The Netherlands
E-mail: m.vulovic at tudelft.nl
Tel: (+31)15-278-31-91
Fax: (+31)15-278-6740

Leiden University Medical Center
Department of Molecular Cell Biology / Section Electron Microscopy
Postal Zone S1-P  P.O.Box 9600 / 2300RC Leiden
The Netherlands
E-mail: vulovic at lumc.nl
Tel: (+31)71 526 9449
Fax: (+31)71 526 8270

-----Original Message-----
From: 3dem-bounces at ncmir.ucsd.edu on behalf of William Rice
Sent: Wed 11/10/2010 6:16 PM
To: 3dem at ncmir.ucsd.edu
Subject: Re: [3dem] Current measurement with viewing screen

Hi,
A similar way to do it is to use film. We have the following table, 
provided by John Berriman when he was at our lab:

Using Kodak SO 163 film, exposure with 1 e/um^2 (16 pA/cm^2 for 1 sec) 
then developing 12 min in full strength D19 (complete development) 
should give the following OD's:

Voltage       OD
80 kV         2.2
120 kV      2.0
200 kV      1.4
300 kV       1.0

Bill
Liang Jin wrote:
> Hi, Fred
>
> In our camera, we have an integrated Faraday plate to measure electron beam current. Therefore, I got a chance to calibrate the screen readings of many different microscopes including FEI and JEOL ones. Unfortunately what I have found out is that the calibration factor is different for different scopes, ranging from 0.6 to 0.9. The only accurate way is like you mentioned to use a Faraday cup and calibrate your instruments. Maybe you can simply request FEI to do a calibration? 
>
> Another easy way of calibrating the beam current is to use your camera. If you know the exact count per electron, you would be able to take an image and use the mean count to estimate the actual electron dose. This method relies on an accurate measure of the sensitivity of your camera in terms of count/electron. The conversion factor for fiber-coupled cameras is determined by the scintillator and the CCD paramters. Given that every scintillator is unique, you will have to check with your camera manufacture to figure out if there is a number measured on your camera.
>
> Best regards, 
>
> -Liang
>
> Dr. Liang Jin
> Direct Electron, LP (http://www.directelectron.com/)
> 13240 Evening Creek Drive S., Ste. 311
> San Diego, CA 92128
> Tel: (858) 384-0291  x104
> Email: ljin at directelectron.com
>
> -----Original Message-----
> From: "Bob Grassucci" [rg2502 at columbia.edu]
> Date: 11/09/2010 09:16 AM
> To: "Fred Sigworth" <fred.sigworth at yale.edu>, 3dem at ncmir.ucsd.edu
> Subject: Re: [3dem] Current measurement with viewing screen
>
> Hi Fred,
> We have been using a BS factor of 0.85at 200 kV and 0.75 for 300 kV.  It 
> can be set so that it is used in low dose  if you are comfortable 
> editing the registry.by the following settings.
> regedt32
> Hkeylocalmachine
>      software
>          FEI
>              Lowdose
>                  BS factor    200 0.85
>                                      300 0.75
> After you do this the low dose options tab should have the correct dose 
> readings,  I am told that this is routinely done with the newer scopes 
> but was not early on.  Of course a faraday cup is the most accurate way 
> to go but not everyone has access to one.  I hope this helps.
> Bob
>
>
> On 11/9/2010 12:01 PM, Fred Sigworth wrote:
>   
>> To calibrate our CCD camera we have been using the viewing screen to measure the beam current in our Tecnai F20.  Of course this measurement is in error due to backscattering, secondary electrons, etc. and the right way is to use a Faraday cup.  But, has anyone compared the value from a Faraday cup with the screen current?  If the conversion factor is known, even approximately, that would help us a lot.
>>
>> Fred Sigworth
>>
>> Fred J. Sigworth
>> Department of Cellular and Molecular Physiology
>> Department of Biomedical Engineering
>> Yale University
>> New Haven, CT  06520
>> tel.  203 785 5773
>> fax 203 785 4951
>> fred.sigworth at yale.edu
>>
>>
>>
>> _______________________________________________
>> 3dem mailing list
>> 3dem at ncmir.ucsd.edu
>> https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem
>>
>>
>>     
>
>
>   

-- 
William J. Rice, Ph.D.
Manager for cryo-EM Computing
New York Structural Biology Center
89 Convent Avenue, NY, NY 10027

_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu
https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mail.ncmir.ucsd.edu/mailman/private/3dem/attachments/20101111/49052913/attachment-0001.html