Sign up Calendar Latest Topics
 
 
 


Reply
  Author   Comment  
frankdopamine

Junior Member
Registered:
Posts: 9
Reply with quote  #1 

Hello everyone,

I work as PhD student in the Aldo Genovesio’s Lab at La Sapienza in Rome, Italy. We are using Monkey logic (ML) for our behavioral experiments for some time. We would like to use ML also for our recording sessions. 

We connected ML to our TDT system composed of the RZ2 BioAmp Processor and the RS4 Data Streamer. We are using a National Instrument PCIe-6323 data acquisition board plus two NI CB-68LPR screw terminal (one for sending our behavioral codes to TDT system and the other one for the others input/output sources (es: reward system)).
To send our behavioral codes as digital output to RZ2 BioAmp Processor via a serial DB25 cable, we followed the schematics available on National Institute of Mental Health site (https://www.nimh.nih.gov/research/research-conducted-at-nimh/research-areas/clinics-and-labs/ln/shn/interface-for-monkeylogic.shtml) (please, see the attached file "Schematics", page six, we are using the “16-bit option”).
So, we assigned in ML input/output submenu the Port 0, line 8 to 23 as Behavioral Codes and Port 0 line 24 as Strobe Bit.
For the strobe duration, we left the default setting (T1 125 us and T2 125 us) in the Strobe timing specification submenu, because I read that these values are pretty safe for both Plexon and TDT in another post (http://forums.monkeylogic.org/post/how-to-set-behavioral-code-and-strobe-word-9646000).
RZ2 receive the word (bit 0-15, our behavioral codes) in the addressable bytes of the Port A and B, plus a single strobe bit (bit 16) in the addressable bit in the Port C (C0) via the serial cable as shown in NIMH schematics.

We did a test, and we saw that TDT received the Behavioral Codes and stored them in its own data file format (Data Tank file) with the timestamps.

We would like now doing some controls to be sure that there is no excessive delay between machines in the timestamps of our codes in order to start a real recording session after that.

So, my question is: there is some way/ strategy to quantify the delay of the timestamps in my recording system compared to the same timestamps in ML or more generally a way/ strategy to be sure there is no excessive delay between machines?

If there is someone who has already connected ML with TDT recording system following the schematics of NIMH and would like to share the result of his/her tests and/or how he/she did them; we would appreciate that.

Many thanks in advance!

 
Attached Files
pdf Schematics NIMH.pdf (37.53 KB, 5 views)

0
Jaewon

Administrator
Registered:
Posts: 873
Reply with quote  #2 
I don't think you need to worry about the timestamp delay, when you use a well-proven system, like TDT. You can make the TDT system send back a signal to ML and measure a turnaround time, but setting up things to do so is a hassle (although I don't discourage you if you want to do it).

I once have made a loopback connection from digital output of NI to analog input (running at 40 kHz) and measured the intervals between the timestamps of eventmarkers and voltage changes. The numbers were a little different, depending on the MATLAB version, but the median was about 0.2 ms. So, if the TDT system checks the digital input at about 40 kHz as well, you may assume a similar delay.




0
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.