Tuesday, 2018-06-26

*** tpb has joined #photonsdi00:00
*** se6astian|away is now known as se6astian06:36
*** se6astian is now known as se6astian|away06:39
*** se6astian|away is now known as se6astian07:54
se6astianGood morning felix_, how is the routing going?08:29
*** futarisIRCcloud has quit IRC08:52
*** Bertl_zZ is now known as Bertl10:40
*** RexOrCine|away is now known as RexOrCine10:50
felix_i have more or less routed everything that can be routed before the placing is finalized. still working on the global placement, to get the board smaller. the routing is the easy part; the placing the tricky one ;)14:31
felix_can the eeprom also store the plugin module hardware revision and feature set populated or should i rather add some locations for pull-up/down resistors on some spare fpga pins?14:37
felix_or is the plan only to produce fully populated boards and no partially populated boards that e.g. only have one sdi output and no sdi input or legacy synchronization input?14:42
felix_ah, i need to improve the axiom connector; that's currently the default pcie x1 edge connector from the kicad library, which isn't nice14:58
*** Bertl is now known as Bertl_oO15:12
*** se6astian is now known as se6astian|away17:43
*** se6astian|away is now known as se6astian18:48
se6astianthe eeprom is envisioned to store a lot of information about the plugin moldule not just name and version/revision19:25
se6astianthe actual structure is not defined yet19:25
se6astianbut the eeprom has plenty of space19:26
felix_ok19:26
se6astianpartial population does make sense I guess19:26
felix_the module will be about 7cm long (including the pcie-style connector)19:29
felix_won't get much smaller that this19:29
se6astiancan you share a preliminary layout of the whol module ?19:30
se6astianscreenshots19:30
felix_i ripped apart some parts yesterday to optimize some stuff, but sure i can share a screenshot later19:32
se6astiangreat21:10
felix_http://i.imgur.com/RCjbOGw.png21:14
felix_connectors are from to sdi in, 2x sdi out, legacy sync in21:14
felix_the footprints of the crystal and crystal oscillator are still missing and the footprints of the voltage regulators and the pcie-style connector still need some work21:16
felix_the line at the top is 7cm long21:17
felix_but yeah, the pcb will not get much smaller and the general position of things won't drastically change any more21:21
felix_well, some millimeters at the top and botten still can be shoved off, so that it will fit the form factor21:22
KjetilIs there any actual benifit to placing the DC-blocking caps that close to the connector?21:32
KjetilI guess U1101 is the sync separator. Could probably save a bit by using a cheaper connector for it?21:34
felix_yeah, the sync seperator doesn't need some super high end connector21:41
KjetilI'm guessing a lot more thermal vias are recommended on the EGP on the jitter cleaner21:42
Kjetil"Use no fewer than 25 vias from the center pad to a ground plane under the device"21:43
Kjetil(Details at this stage)21:43
KjetilRemember that you don't really need to care about polarity on SDI so you can swap around the high speed serial lines to ease routing21:49
felix_yep21:54
KjetilAre there missing some rubber bands between the sync sep. and the FPGA?21:56
felix_i ripped apart a lot of the fpga connections yesterday and haven't fully reconnected them, so yes, a lot of the fpga connections are missing at the moment21:58
Kjetilhave you connected hsout to both the jitter cleaner and the fpga?21:59
felix_hsync from the sync seperator?22:00
felix_there should also be a connection to the fpga22:00
Kjetilyes22:00
*** se6astian is now known as se6astian|away22:01
*** RexOrCine is now known as RexOrCine|away22:07
*** futarisIRCcloud has joined #photonsdi23:49

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!