Difference between revisions of "Logic Sniffer"

From YobiWiki
Jump to navigation Jump to search
Line 42: Line 42:
 
** Problem of never ending if no changes occur
 
** Problem of never ending if no changes occur
 
** Better to ground the unused probes of one group to not loose space recording noise
 
** Better to ground the unused probes of one group to not loose space recording noise
  +
===Test on Smartcard <> Terminal communication===
  +
  +
Trying to sniff a T=0 communication:
  +
* channel 0: CLK (actually at the sampling rate given below we're largely undersampling the clock which is about 1MHz)
  +
* channel 1: RST
  +
* channel 2: I/O
  +
  +
All signals are low before card is inserted
  +
<br>I/O is high when idle
  +
<br>We sample @ 200kHz and trig on the RST signal
  +
* 3/97
  +
* complex
  +
* Stage1
  +
** Immediately, //
  +
** Mask vv.
  +
** Value v.. (RST down, IO up idle)
  +
* Stage2
  +
** On level1, //
  +
** Mask vv.
  +
** Value vv. (RST up, IO up idle)
  +
** Action: start capture
  +
We have nicely all the ATR in one window
  +
<br>Trying Tools / UART analyzer
  +
* Set Rx on channel 2
  +
* 8E1, invert
  +
=> all got wrong :-(
  +
<br>It seems it cannot understand idle high
  +
  +
I also tried serial trigger but I don't understand how it considers baudrate and of course it doesn't work

Revision as of 17:33, 24 November 2010

Description

I'm talking about this beast

Installation

My udev rules

I added a /dev/OpenLogicSniffer symlink by following this page:

#File /etc/udev/rules.d/77-ols.rules
#Rules for Openbench Logix Snifferslogic. Creates a nice link to the ols
ATTRS{idVendor}=="04d8" ATTRS{idProduct}=="fc92" MODE="0666" SYMLINK+="OpenLogicSniffer"

If you've problems with modem-manager trying to access the device (as seen in /var/log/syslog), you can add the following rule:

ATTRS{idVendor}=="04d8" ATTRS{idProduct}=="fc92" ENV{ID_MM_DEVICE_IGNORE}="1"

Firmware upgrade

Software

I tried this fork: http://www.lxtreme.nl/ols/ which is the official client for this board
and changed a bit the run.sh script to handle it through symlink :

if [ -h $0 ]; then
    DIR=$(dirname "$(readlink $0)")
    cd "$DIR"
    BASEDIR=.
else
    BASEDIR=$(dirname -- "${0}")
fi
  • Make sure that the Device is set to "OpenBench Logic Sniffer".
  • Click the capture button and type in your OLS port path in the "Analyzer port" field (i.e. /dev/OpenLogicSniffer if you installed it as I did). Leave the other settings at their default value for now and click "Capture".

Usage

See 101 and 102 tutos

  • Number scheme inside/outside in SUMP client refers to number scheme printed on PCB
    • Default: inside: pin0 is at opposite of RESET button, on buffered probe header (up to 5V)
  • External clock source in SUMP: through the CKI header (see the 4 holes in the middle of the PCB)
  • Limits:
    • 5V on buffered probe header / 3.3V on unbuffered wing header
    • 200MHz with 2 groups, no noise filter
    • 1 group: 24K / 2 groups: 12K / 4 groups: 6K
    • RLE (compression): 4 groups only (v2.12) / all (v2.11).
  • Trigger before/after ration: samples before/after the trigger
  • RLE
    • Problem of never ending if no changes occur
    • Better to ground the unused probes of one group to not loose space recording noise

Test on Smartcard <> Terminal communication

Trying to sniff a T=0 communication:

  • channel 0: CLK (actually at the sampling rate given below we're largely undersampling the clock which is about 1MHz)
  • channel 1: RST
  • channel 2: I/O

All signals are low before card is inserted
I/O is high when idle
We sample @ 200kHz and trig on the RST signal

  • 3/97
  • complex
  • Stage1
    • Immediately, //
    • Mask vv.
    • Value v.. (RST down, IO up idle)
  • Stage2
    • On level1, //
    • Mask vv.
    • Value vv. (RST up, IO up idle)
    • Action: start capture

We have nicely all the ATR in one window
Trying Tools / UART analyzer

  • Set Rx on channel 2
  • 8E1, invert

=> all got wrong :-(
It seems it cannot understand idle high

I also tried serial trigger but I don't understand how it considers baudrate and of course it doesn't work