2015-11-10, hazen
Recipe to run AMC13
To start the software tool
- Log on to PC (daqlabamc13.cern.ch) with user 'daq' password 'M......e' (see Dominique)
-
cd work/amc13
-
source setup.sh
-
AMC13Tool2.exe -c 192.168.1.130
(for this AMC13 S/N 166)
Initialize to send fake data
Enable DAQ link
-
> daq 2
(two links active)
-
> fed 0 100
(set link 0 FED number to 100)
-
> fed 1 101
(set link 1 FED number to 101)
Send triggers
Reconfigure firmware (to change from e.g. 10G to 5G firmware)
- Download mcs file as instructed from Mr WU's directory.
- e.g. latest version is
AMC13T1v0x023d_7k325t.mcs
-
AMC13Tool2.exe -c 192.168.1.130
(for this AMC13 S/N 166)
-
>pk
(to program Kintex chip)
Read status
Other Resources
Testing log for 2015-11-10
At CERN, trying to test 10G link. Found an AMC13 S/N 166 which I can talk to at 192.168.1.130.
Firmware is 0x236/0x2d. Updating to 0x23d/0x2e.
NAT-MCH IP address is 192.168.1.41 (same as BU NAT).
Note: eth1 was set to mask 255.255.255.0, changed by hand to 255.255.0.0.
Don't know why this is. Updated AMC13 software in ~ehazen/work/amc13 to trunk version. Reprogram / reboot AMC13. Versions now AOK.
Program v0x0002 (test version with 1 10Gb link) and it partially works! Dominique can read the status from the other end of the link. No event data transfer due to a known bug.
2012-01-22, hazen
Arrive CERN, install AMC13 in MCH1 slot (upper) in DAQ lab. Move short Ethernet cable from eth1 to upper right RJ-45 on MCH (
GbE).
Now we can talk to AMC13! See
AMC13spec.txt for some minimal documentation.
AMC13 occupies two I/P addresses at:
- Spartan-6: 254 - 2 * serial_no
- Virtex-6: 255 - 2 * serial_no
So, for S/N 6 installed, I/P addresses are 192.168.1.242 and 192.168.1.243.
Install duplex LC fiber between upper two SFP+ and run DAQ loop-back test:
(log in as daq to pc)
$ cd amc13_python/src_amc13
$ python ipbus.py 243
...
> w 0 4
> w 1 1
> w 0 2
> r 0 16
0x0000: 0x00006558
0x0001: 0x00000001
0x0002: 0x00000111
0x0003: 0x008965e1
0x0004: 0x00000000
0x0005: 0x019c32ed
0x0006: 0x00000000
0x0007: 0x019c3433
0x0008: 0x00000000
0x0009: 0x00000000
0x000a: 0x06000000
0x000b: 0x12341234
0x000c: 0x00036cc0
0x000d: 0x00000000
0x000e: 0x07f9c6f7
0x000f: 0x0ff70241
Address 3 should be counting, not address 4. This is what we observe!
--
EricHazen - 22-Jan-2012