Four plots with different settings:

combined_8723_200.pdf  neg polarity, gain 0, shaper 0
combined_8723_280.pdf  pos polarity, gain 0, shaper 0
combined_8723_281.pdf  pos polarity, gain 0, shaper 0, preampmode
combined_8723_287.pdf  pos polarity, gain 3, shaper 0, preampmode
combined_8723_3ff.pdf  shutdown oc PCA16

Mail from Fan 20140721
======================


Attached are our test results with different PasaCtrl settings. None of 
them get the result (around 100) as you expected.
(We let X24 and X25 unconnected)

Followed is an example about how the attached file is named:
combined_8723_200.pdf  : "8723 is CPLD firmware version"; "0x200 is 
9-bit SaltroCtrl setting"
PasaCtrl = SaltroCtrl[7:0]
Bias_saltro = SaltroCtrl[8]
saltro_global_reset =  SaltroCtrl[9]

Could you please give us some advice on the test methods?


ANSWER 20140722
===============

Dear Fan & Yifan,

thank you for the tests!

You seems to have solved the data format problem. Good!

Though the results are difficult to interpret.

Channel 1 seems lower than the others.
Perhaps the input has a short to something (neighbour?)
Can you measure that?   

The reference voltages might be wrong?
Could you measure them and send me the result?

Some channels seems sometimes to jump by 256(?), a bit error?
Timing problem (clock ...)?
See plots for 0x281.

The setting 0x200 is the only settings that seems to be what one would expect.

The shift in pedestals with gain/shaper values do not correspond to what I
thought should happen when changing gain and shaping time, though you did not
change gain/shaping with the same preampmode.

Could you run with:

nonpreampmode:

0x200 (gain 0, shaper 0)
0x202 (gain 1, shaper 0)
0x204 (gain 2, shaper 0)
0x206 (gain 3, shaper 0)
0x208 (gain 0, shaper 1)
0x218 (gain 0, shaper 3)
0x238 (gain 0, shaper 7)
 
preampmode:

0x201 (gain 0, shaper 0)
0x203 (gain 1, shaper 0)
0x205 (gain 2, shaper 0)
0x207 (gain 3, shaper 0)
0x209 (gain 0, shaper 1)
0x219 (gain 0, shaper 3)
0x239 (gain 0, shaper 7)

and the other polarity (0x28x).

to see if one can find what bits are doing what.

Could you send me the data files, then I can plot them myself in more detail.

There seems to be very low noise, all time simples are mostly the same,
within the plot resolution.

In the script I sent you can play with the x,y axis through the lines:
set xrange [0:1000]   
set yrange [0:1024]

What settings do you have for the ALTRO?
(pedstal, subtraction etc...)?

Can you run with more time samples?

It would also be interesting to get a signal in on the lemo connector, to see
if there is anything seen from the SALTRO.