Documentation:

• Agilent 9000A
  
• programmers reference guide (.pdf)
  
• users manual
• Rhode & Schwartz rto
  
• Users Manual (.pdf)
• tektronix mdo4000 oscilloscope
• programmers manual (.pdf)
• users manual (.pdf)
  

Agilent MSo9104A mixed signal oscilloscope

    Borrowed from agilent by jonathan to test for lidar usage.

I/O measurements

• Setup:
• 20ghz sampling, interpolation off
• segmented memory. Segment length  20000.
  
• 2 byte word data, LSB FORMAT used.
  
• steaming mode on
• use 1.2 millisecond testpoint output square wave for trigger.
• after trigger we acquire 20000/20e9 = 1usec of data. so we should be able to trigger every 1.2 millisecs
• I didn't figure out how  to turn off the data display during aquistion. this might speed things up.
  

• timing measurements
• The table  below shows how long it took to acquire data for different number of segments (each 20000 points). The first column contains the number of segments acquired.The 2nd column is the number of millisecs we had to wait before data became available.This is after the scope has triggered once for each segment. The last column is how long it took to read in the data.
• 
  

• Number of segments (each 20000 points)	wait for data millisecs	read data millisecs
  1	110.7	29.188
  2	114.9	14.578
  5	118.0	19.5
  10	126.8	25.7
  20	139	37.5
  50	180.7	70.0
  100	245	133.4
  500	775	626.
  1000	1432.3	1244.0
  2000	2750	2544
  3000	4071	3784
  4000	5388.6	5138.8

• Page 1 top:
• Time to wait for data vs number of segments. The red line is a linear fit to the data
• the fit is: Tm= 112.8 + 1.319*Nsegments
• 1.319 is about the time for a trigger (1.2336 milliseconds)
• There is a 112 ms offset. This must be how long it takes to get going.
• Page 1 bottom:
• error in fit for each number of segments. This error may be the startup time fluctuating.
  
• Page 2: i/o rate. MB/sec i/o rate versus number of bytes input. At around 2MegaBytes the i/o rate saturates at 32MB/sec.

<- page up
home_~phil