| *** tpb has joined #yosys | 00:00 | |
| *** roh has quit IRC | 00:56 | |
| *** seldridge has quit IRC | 01:30 | |
| *** roh has joined #yosys | 01:36 | |
| *** digshadow has joined #yosys | 02:44 | |
| *** emeb has quit IRC | 02:50 | |
| *** emeb_mac has joined #yosys | 02:54 | |
| *** proteusguy has quit IRC | 04:21 | |
| *** xrexeon_ has quit IRC | 04:35 | |
| *** AlexDaniel has quit IRC | 04:39 | |
| *** proteusguy has joined #yosys | 04:52 | |
| *** digshadow has quit IRC | 05:06 | |
| *** seldridge has joined #yosys | 05:07 | |
| *** proteusguy has quit IRC | 05:08 | |
| *** dys has quit IRC | 05:14 | |
| *** proteusguy has joined #yosys | 05:20 | |
| *** sklv has quit IRC | 05:33 | |
| *** sklv has joined #yosys | 05:33 | |
| *** AlexDaniel has joined #yosys | 05:40 | |
| *** proteusguy has quit IRC | 05:45 | |
| *** shivam has joined #yosys | 06:03 | |
| *** shivam has quit IRC | 06:09 | |
| *** emeb_mac has quit IRC | 06:26 | |
| *** seldridge has quit IRC | 06:28 | |
| *** GuzTech has joined #yosys | 06:48 | |
| *** xerpi has joined #yosys | 06:50 | |
| *** digshadow has joined #yosys | 07:24 | |
| *** leviathan has joined #yosys | 07:40 | |
| *** leviathan has quit IRC | 07:53 | |
| *** leviathan has joined #yosys | 07:53 | |
| mattvenn | In my efforts to make a sliding DFT in Verilog | 08:04 |
|---|---|---|
| mattvenn | https://github.com/mattvenn/fpga-fft | 08:04 |
| tpb | Title: GitHub - mattvenn/fpga-fft: sliding DFT for FPGA, targetting Lattice ICE40 1k (at github.com) | 08:04 |
| mattvenn | I'm a bit stuck on how to handle the integer maths | 08:04 |
| mattvenn | the 'twiddle factors' are scaled to make the most use of the precision available | 08:05 |
| mattvenn | but then when I'm doing the multiplication, things quickly overflow | 08:05 |
| mattvenn | in my python implementation, the frequency bins stay low, because all the twiddle factors are <= 1 | 08:05 |
| mattvenn | how do I handle keeping repeated multiplications within the bounds of my registers? | 08:06 |
| sorear | your repo is misnamed, that is not a fft | 08:08 |
| mattvenn | I started off doing an FFT and then changed to SDFT | 08:08 |
| mattvenn | but what is the difference? | 08:08 |
| sorear | do a bit shift? (x * 85) >> 8 ~= x * (1/3) | 08:09 |
| sorear | hmm, finite precision is a problem, disappointed that the paper doesn't cover numerical stability at all | 08:15 |
| mattvenn | this one touches on it: https://www.dsprelated.com/showarticle/776.php | 08:18 |
| tpb | Title: Understanding and Implementing the Sliding DFT - Eric Jacobsen (at www.dsprelated.com) | 08:18 |
| mattvenn | but just mentions that you can tweak the coefficients down a little if it overflows | 08:19 |
| mattvenn | however, I definitely some scaling happening somewhere before I consider that | 08:19 |
| *** AlexDaniel has quit IRC | 08:48 | |
| *** digshadow has quit IRC | 09:17 | |
| *** proteusguy has joined #yosys | 09:23 | |
| *** zkrx has quit IRC | 09:24 | |
| *** proteus-guy has quit IRC | 09:26 | |
| *** zkrx has joined #yosys | 09:29 | |
| *** leviathan has quit IRC | 10:35 | |
| *** AlexDaniel has joined #yosys | 10:54 | |
| *** xerpi has quit IRC | 11:07 | |
| *** develonepi3 has joined #yosys | 11:09 | |
| *** sklv has quit IRC | 11:30 | |
| *** sklv has joined #yosys | 11:31 | |
| *** AlexDaniel has quit IRC | 11:45 | |
| *** AlexDaniel has joined #yosys | 12:57 | |
| *** massi has quit IRC | 12:59 | |
| *** jwhitmore has joined #yosys | 12:59 | |
| keesj | there was an other amp hour interview that was ICE/FPGA related | 13:01 |
| keesj | https://theamphour.com/395-an-interview-with-luke-valenty/ | 13:01 |
| *** massi has joined #yosys | 13:01 | |
| keesj | I wonder how stable the tinyFPGA clock is and how the bootloader performs. | 13:02 |
| keesj | it was not 100% clear to me what functionality is present when using Yosys | 13:03 |
| *** jwhitmore has quit IRC | 13:10 | |
| *** AlexDani` has joined #yosys | 13:29 | |
| *** AlexDaniel has quit IRC | 13:31 | |
| *** ccole has quit IRC | 13:39 | |
| mattvenn | it all works with icestudio, which depends on apio, which depends on icestorm/yosys | 13:52 |
| *** xerpi has joined #yosys | 14:29 | |
| *** captain_morgan has quit IRC | 14:37 | |
| *** captain_morgan has joined #yosys | 14:38 | |
| *** promach_ has joined #yosys | 14:39 | |
| *** emeb has joined #yosys | 14:59 | |
| *** roh has quit IRC | 15:00 | |
| *** digshadow has joined #yosys | 15:20 | |
| *** roh has joined #yosys | 15:47 | |
| tinyfpga | keesj: the TinyFPGA BX uses a MEMS oscillator from Microchip | 15:59 |
| tinyfpga | keesj: http://www.microchip.com/promo/dsc6000-family-mems-oscillators | 15:59 |
| tpb | Title: DSC6000 Family MEMS Oscillators | Microchip Technology (at www.microchip.com) | 15:59 |
| tinyfpga | keesj: its accurate within 25ppm and has a jitter on the order of 10s of ps | 16:00 |
| tinyfpga | keesj: like mattvenn mentioned, it’s 100% compatible with icestorm/yosys. | 16:01 |
| mattvenn | hey tinyfpga ! | 16:20 |
| tinyfpga | hey mattvenn :) | 16:20 |
| tinyfpga | I lurk around here once and a while...get push notifications to my phone when my handle is mentioned | 16:20 |
| *** digshadow has quit IRC | 16:23 | |
| *** digshadow has joined #yosys | 16:28 | |
| mattvenn | cool | 16:32 |
| mattvenn | good to see your recent boards worked out well | 16:32 |
| *** dys has joined #yosys | 16:35 | |
| *** seldridge has joined #yosys | 16:37 | |
| *** digshadow has quit IRC | 16:53 | |
| *** promach_ has quit IRC | 17:07 | |
| *** dys has quit IRC | 17:16 | |
| *** dys has joined #yosys | 17:17 | |
| *** digshadow has joined #yosys | 17:30 | |
| *** xerpi has quit IRC | 17:36 | |
| *** GuzTech has quit IRC | 17:48 | |
| mithro | tinyfpga: Hey! | 18:18 |
| tinyfpga | mithro: hey, what’s up? | 18:19 |
| mithro | tinyfpga: How goes things? | 18:19 |
| tinyfpga | mithro: very busy...was lucky to have time to assemble the samples last night | 18:20 |
| tinyfpga | mithro: need to update the website now for the BX so the reviewers have instructions to use Atom IDE with APIO/icestorm | 18:21 |
| tinyfpga | mithro: and fill out some paper work for arrow to start an order of FPGAs for the next TinyFPGA BX production run | 18:22 |
| mithro | tinyfpga: That would be awesome - interested to see how atom + apio work | 18:22 |
| *** m_t has joined #yosys | 18:42 | |
| *** digshadow has quit IRC | 18:49 | |
| *** jwhitmore has joined #yosys | 18:52 | |
| *** AlexDani` is now known as AlexDaniel | 19:11 | |
| *** xerpi has joined #yosys | 20:22 | |
| *** pie_ has quit IRC | 20:46 | |
| *** jwhitmore has quit IRC | 21:15 | |
| *** AlexDaniel has quit IRC | 21:16 | |
| *** AlexDaniel has joined #yosys | 21:17 | |
| *** jwhitmore has joined #yosys | 21:28 | |
| *** pie_ has joined #yosys | 21:32 | |
| *** jwhitmore has quit IRC | 21:56 | |
| *** develonepi3 has quit IRC | 22:03 | |
| *** seldridge has quit IRC | 22:12 | |
| *** jwhitmore has joined #yosys | 22:17 | |
| *** develonepi3 has joined #yosys | 23:05 | |
| develonepi3 | Are most on this irc using RaspBian to work with IcoBoard? | 23:07 |
| mithro | daveshah: Do you know if with the ".param LUT_INIT 0110" can I just pad it with zeros to get the 16b version? | 23:13 |
| *** jwhitmore has quit IRC | 23:15 | |
| *** xerpi has quit IRC | 23:16 | |
| *** m_t has quit IRC | 23:20 | |
| *** pie_ has quit IRC | 23:23 | |
| *** AlexDaniel has quit IRC | 23:26 | |
| *** AlexDaniel has joined #yosys | 23:26 | |
| *** dxld has quit IRC | 23:38 | |
| *** dxld has joined #yosys | 23:44 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!