On Saturday 08 April 2017 01:46:45 Erik Christiansen wrote: > On 07.04.17 13:16, Gene Heskett wrote: > > I do not know what the current clock speed is, could be as low as > > 700 or as high as 1200 MHz. > > What I usually do is no more than: > > $ dmesg | grep MHz > [ 0.000000] Detected 1499.981 MHz processor. > [ 0.062078] CPU0: Centaur VIA C7 Processor 1500MHz stepping 00 > [ 1.992110] Refined TSC clocksource calibration: 1499.999 MHz. > > Erik > (Who has filed away all the good oil on this thread, like a squirrel > with cheeks full of nuts.) > On the pi 3b, the closest I can come is: dmesg |grep timer [ 0.000000] Architected cp15 timer(s) running at 19.20MHz (phys). [ 0.000018] Switching to timer-based delay loop, resolution 52ns [ 0.002230] Calibrating delay loop (skipped), value calculated using timer frequency.. 38.40 BogoMIPS (lpj=19200)
1.752415] bcm2835-cpufreq: min=600000 max=1200000 Which is not a heck of a lot of help. 153 bogomips is later claimed for all 4 cores. About 2 hours of latency-test watching servo-thread at 1000000, showed latency at around 220000, or about 1/4 of the 1ms servo-thread. I found in the hostmot2 manpage a stepgen.timer-num reference, which was actually timer-number to a halmeter, which operates on all stepgens enabled and perportedly is good for following errors, so I looked at the first dpll timer at -100 and enabled that. I have serious doubts it will be any help as I think some sort of a noise pulse is triggering a framing error in an spi packet, everything gets bad data, and shuts down everything. The totally unfunny part about that is that while the gui shows both joints errored, it is not latched! Anyplace. Hell of a way to run a train. OTOH, I did sign up for this job. But I didn't know I ws going to be the only one swinging a corn knife in an r-pi jungle. But I do hope the trail I am cutting will be useful for others. I put a couple pieces of 16 gage into the 7i90 power connector, makeing a local single point feed out of it, and then fed it 5 volts, going on from there to the pi, without the ferrite choke in its power cord. I think that helped, the error halts went from maybe 3 minutes to an hour or more, but I've not tried any hf bypassing on that point, the 5/8" long bits sticking out of the back of the 7i90's pwr plug are about full. I'd made up some surge absorbers out of 5 1n914's per wire from the jog wheels, and added a ceramic cap labeled 500M to ground on each of those lines, cleaning up 95% of the noise there and preventing the inputs to the 7i90's 3.3 volt circuitry from exceeding 3.75 volts, which made that ugly, noisy signal nice and clean, and its still perfectly square at any speed I can spin the wheels. I am tempted to do the same with the spindle encoder inputs too but will need to make more mounting provisions for more perfboard. 1n914's I have bunches of. I don't seem to ever get to the last shoe drop with this thing... Back to bed & see if I can get enough sleep for a change. Thanks Erik. 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 Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
