Step 1. Initialization.

This is the most time consuming and important part.
After you power on the crate start the spreadsheet.
You need to run it before starting burn-in because
Saseq initialization through GUI/C++ does not work
properly for now. To start Excel spreadsheet go to
D:\Silicon TestStand on NT\Spreadsheet
and click on saseq_617_nt.xls

On the spreadsheet that shows up click on SETUP.
Select correct number of Saseqs and put the types
of devices hooked to each chain. After you press
OK spreadsheet will initialize Saseq and download
the chains. If there are no errors take several
events (press Inject button) to make sure that the
data is OK. Close the spreadsheet.
 

Step 2.

Go to c:\users\tracker\tb directory and click on
saseq.tcl. The GUI window will show up.

The upper row of SASEQ buttons is supposed to do
SASEQ initialization. These buttons do not work
correctly and it's better not to use them.

The second row of buttons (two buttons per each
Saseq) allow you to create download files and
download chips.

a) Click on "chain 1" button. Create download file pushing CREATE button with known chip IDs. Make sure that the number of chips in the download file you created is correct.
b) Click on DOWNLOAD button. The message will appear in the empty window showing you the result of downloading.
c) Press CANCEL to kill the window
d) Click on chain 2. Create download file for chain 2. Make sure that chip IDs on chain 2 are different from those on chain 1. I used to put chip ID 90 for chain 1 and F0 for chain 2.
e) Repeat b) and c) for chain 2.

The bottom row of buttons is for data taking. Press
on SASEQ1. The window for data taking will show up.
Put the number of chips on each chain, calibration
voltage (in HEX) and modify if you want the number of
events to take and output file name. Press START.

To look at the raw data start FxEdit (you'll find the
icon on your left) and open the file containing raw data
you've just taken. Find chip IDs. Make sure that you
get back chip IDs you downloaded.
Note: Chip ID in the raw data is followed by two zero
bytes: first one just separates chip ID from all other
data and the second one is the channel number. The end
of the event is marked by a combination of two bytes
0x27 0x0F.

If chip IDs in the raw data are correct you are ready
to start burn-in. But sometimes after first cycle chip
IDs are wrong (usually for chain 2). In this case download
both chains once again and take raw data. My experience
tells that two cycles are enough to get back correct chip
IDs.
 

Step 3. Burn-in

Press "burn-in test" button in the upper right corner
of the main window. Fill in the upper (pink) part of GUI.
Put 1) number of runs 2) delay between runs (in minutes)
3) type names of devices in the corresponding windows
4) pull down menu and select the device type under test.
Press "START" button.

To interrupt the test press "STOP" button and then click
on "YES" when the window asking you to confirm the request
pops up.
Note: when the program is making the burn-in run the GUI
window is updated frequently and therefore takes your
"STOP" request quickly. During the waiting time between the
runs it is updated once in a minute. Therefore you should
wait some time after pressing "STOP" button before your
request is accepted.
 

Step 4.

The data is stored in d:\burn_in\results\device_name
directory. After the first run is completed you can look
at the data if you want. Along with the results you'll
find burn_in.kumac to make plots for gains and pedestals
and sparse.kumac to make plots for sparse readout.

To make plots start PAW. Then type
exec burn_in#plots device_type device_name run_number
For example,
exec burn_in#plots L9 3953-02 1
will make plots for 9 chip HDI with serial number 3953-02
using run 1 output.
To make plots for sparse readout type
exec sparse#plot  evice_type device_name run_number

If you have further questions feel free to send me messages
on elis@fnal.gov.