| larsc | the downside of this method for the fifo counter is that there is quite a bit of combinatorial dealy for the gray counter output | 00:26 |
|---|---|---|
| larsc | it's still the bottleneck in my design | 00:27 |
| larsc | if you do the comparison in binary you can pipeline the conversion to gray | 00:29 |
| ysionneau | oh, I didn't know you could make an (sigma delta) ADC in a FPGA by using LVDS IO ports | 09:47 |
| larsc | because you can use the lvds input as a comperator? | 10:00 |
| ysionneau | yes I guess | 10:02 |
| ysionneau | sounds logical since differential IOs are basically comparing the two lines | 10:03 |
| larsc | found something from lattice http://www.latticesemi.com/~/media/Documents/ReferenceDesigns/SimpleSigmaDeltaADCDocumentation.PDF | 10:04 |
| ysionneau | pretty cool | 10:06 |
| ysionneau | there is an I¬¬¬ paper as well about that | 10:07 |
| ysionneau | but you need to pay like $$$30 | 10:08 |
| larsc | yea, saw that as well | 10:10 |
| lekernel | ysionneau, if you want to read I¬¬¬ papers, I have free access | 11:24 |
| ysionneau | lekernel: ah cool, if you can grab this one http://goo.gl/i2UIoM please :) | 11:25 |
| ysionneau | hum, writting {A,B,C,D} in fpga RAM and reading back into {B,A,C,D} is not a fantastic idea :) | 15:02 |
| ysionneau | now lm32 unittests pass again | 15:02 |
| ysionneau | very useful those unit tests that mwalle did for the mmu :) | 15:03 |
| ysionneau | and very well written | 15:03 |
| aeris | medium indianess | 15:48 |
| ysionneau | weird indianness :p | 15:49 |
| wpwrak | military-grade encryption. (enigma) | 15:49 |
| ysionneau | so well encrypter it could not boot anything, not even unit tests :p | 15:49 |
| ysionneau | encrypted* | 15:50 |
| lekernel | xilinx does something like that for their PLL control data | 16:22 |
| lekernel | no idea why | 16:22 |
| lekernel | but bits forming e.g. the multiplier and divider values are spread all over the place | 16:23 |
| Action: kristianpaul likes mojo | 16:44 | |
| wpwrak | (test results) sounds like a great victory. useful for my monthly statistics :) how many boards were made in this run ? | 18:29 |
| lekernel | 10 | 18:33 |
| wpwrak | so the next run will be the big one ? | 18:40 |
| lekernel | hmm, maybe | 18:45 |
| lekernel | the designers insisted I move the micro-USB connector, though that should be relatively risk-free | 18:46 |
| wpwrak | do creotech also do the SMT ? or are they just taking care of testing and maybe logistics ? | 18:49 |
| wpwrak | ("do the SMT") as in "have a line" | 18:49 |
| lekernel | yes, they do the SMT | 18:50 |
| wpwrak | great. it's nice to have the ability to check PCBs right at the end of the line. | 18:55 |
| wpwrak | the only drawback i see is that they're really hard to reach, in case you have to visit them ;-) | 19:52 |
| larsc | Now I'm getting timing errors on my reset signals... | 20:09 |
| wpwrak | that's what you get for playing god and messing with the fundamental constants of the universe | 21:00 |
| larsc | no, thats what I get for using the reset signal from the wrong clock domain... | 21:01 |
| larsc | but my dma controller is finally able to run at 295 Mhz on the datapath | 21:01 |
| lekernel | wpwrak, about 6 hours of train to Warsaw + a not-that-expensive cab ride to Piaseczno... it's not that bad | 22:58 |
| lekernel | China would be worse :) | 22:58 |
| wpwrak | oh, i had found some company with that/a similar name in lodygowice, which would be really remote. didn't look very high-tech either, though apparently doing some engineering. can't find it anymore :-( | 23:17 |
| wpwrak | let's repeat this when he's back .. | 23:19 |
| davidc__ | wpwrak: its probably creotech.pl . They build stuff for cern, so probably a reputable mfg | 23:24 |
| wpwrak | yes, they look right. the other one looked more like two guys (at most) and a soldering iron. if they're actually in electronic :) | 23:27 |
| wpwrak | can't figure out what sort of url i used | 23:27 |
| wpwrak | hah, found it ! :) http://coretech.pl/ | 23:33 |
| wpwrak | that's what i get for not using copy & paste | 23:33 |
| wpwrak | and thanks, google chromium for watching (and recording) every step i take ;-) | 23:33 |
| wpwrak | so yes, they look like as if they'd do electronics. but not quite like the sort of guys who have an SMT line in the backyard. | 23:34 |
| ysionneau | creotech .. coretech ... it's almost just a typo between one and the other | 23:44 |
| wpwrak | i guess either gets the occasional strange mail :) | 23:46 |
| --- Sat Nov 30 2013 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!