On Tuesday 21 February 2017 01:56:45 Erik Christiansen wrote: > On 21.02.17 01:13, Gene Heskett wrote: > > If perchance this magic new device has some fast gpio, and a spi > > driver could be written for it, the 7i90HD with spi firmware might > > be usable? > > The on-board Intel Curie microcontroller, which provides the Arduino > 101 compatible extra stuff, has "SPI Flash" listed as an "Other > Interface": See: http://www.udoo.org/udoo-x86/ > > Perhaps more guaranteed to be really hard real-time; the quad-core > main cpu has "up to 20" GPIO. That'll depend on what else we want to > use pins for, clearly. It ought to be feasible to ferret out some > Linux SPI foo out there in the googleverse. (Some time to faff with it > is then item 2 on the agenda. (Well, 3 until the board arrives and I > push some Linux into its brain.) > > > That spi bus protocol is a 32 bit, 4 byte packet going each way, > > with a 32 megabaud transfer rate in and out of a Raspberry pi 3b. > > Thats no slouch in the ability to do realtime control. Said another > > way, if the cpu power was there, one could run the servo-thread much > > faster than 1 kilohertz. Without the hal calculations required, an > > update rate of 4 megahertz for every bit of the 7i90's 72 pins of > > i/o could be achieved. > > > > LCNC is sitting idle out there, and the pi doesn't isolate the > > isolcpus=3 from being monitored by htop like it does on the x86 > > hardware, and the idling rtapi task is using 13.2% of cpu-3. So I > > can add quite a bit of processing overhead in my .hal file's yet > > before it actually gets "busy". Or run a servo-thread at 5 > > kilohertz. > > Well, it's X86. So we have to hobble 3 of the 4 cores, but not on ARM? > > > Yes, you'll use a pile of ferrite snapon chokes, and pay real > > attention to a single bolt grounding system, but once thats > > understood it seems to Just Work(tm). And that card is only a tad > > over a $60 bill on your front deck here in the USA. And its > > interface versatile, offering the same i/o features at the slower > > EPP parport rate if its a true 3.3 volt EPP port. But on the pi, the > > spi is only 4 signal wires not counting the 8 commons, and still > > faster than the parport version is. I have no clue if a true EPP > > parport driver has even been written for the pi's. It doesn't have > > one natively that I'm aware of. > > The Udoo X86 isn't as cheap as a Pi, and I added a good sized SSD to > my order, which didn't help. Their promotional video claims it's a > _heck_ of a lot faster than the Pi, but they may be focussing on > graphical performance. > > Erik
That can and is a minor problem on the pi, gfx isn't yet drm optimized. Its fast enough, but not "instant". 20 FPS maybe. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
