On Thursday, July 3, 2014, Jerry Davis <[email protected]> wrote: > I really haven't done any serious real-time since my DEC days. > I was a software specialist for DEC in Dallas, who specialized in > real-time apps for RT-11 and RSX. > At that time I was handling in excess of 3000 interrupts per second in > assembler on a downloadable RSX-11S image. > This was on old hardware by today's standards, I would think. > > For real-time what is expected interrupt rate on an ARM? BBB specifically? >
If you are talking Linux events and not interrupts, then you are probably (without me bothering to check or look things up) in the hundreds of microseconds or a millisecond. If you write a kernel driver to handle the interrupt as you should on a Linux system, then you are talking likely a few microseconds as the A8 pipeline is a dozen clocks deep and the context save/restore is fairly burdensome. If you are talking PRUs, it only polls interrupt sources and depends entirely on the inner loop. http://beaglelogic.net is sampling 14 pins and saving over 1GBytes of data at 100MHz, so the PRUs can be dedicated to functions and handle interfaces with extremely low latency typically in the hundreds of nanoseconds based on your event loop. > > > > On Wed, Jul 2, 2014 at 9:56 PM, William Hermans <[email protected] > <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: > >> The term real time is subjective in this context anyhow. There is always >> going to be latency, it is just a matter of how much you can put up with. >> After all we're not talking about some medical device, or Automobile >> control system. UAV ? >> >> But looking at this from an automobile perspective, you have a main >> computer system with many MCU's Performing sometimes critical tasks, and >> communicating back and forth with this main system via CANBus. >> >> The OP really needs to define "Robot" more clearly. I mean we're not >> talking about an AT-AT walker model with stepper motors, while also having >> the ability to walk around are we ? . . . >> >> >> >> On Wed, Jul 2, 2014 at 5:30 PM, Robert Nelson <[email protected] >> <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: >> >>> On Wed, Jul 2, 2014 at 7:28 PM, Jerry Davis <[email protected] >>> <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: >>> > I have used BBB, RPi, and arduino. >>> > >>> > arduino is closer to real-time if you need that. However, it does only >>> what >>> > is programmed in C in a loop fashion. There is a setup() function and >>> loop() >>> > function. >>> > just about everything is done in the loop. If you need tight >>> integration >>> > (real-time) with some piece of hardware, then arduino is the way to go. >>> >>> Unless you use the "pru" on the bbb.. ;) >>> >>> Regards, >>> >>> -- >>> Robert Nelson >>> http://www.rcn-ee.com/ >>> >>> -- >>> For more options, visit http://beagleboard.org/discuss >>> --- >>> You received this message because you are subscribed to the Google >>> Groups "BeagleBoard" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected] >>> <javascript:_e(%7B%7D,'cvml','beagleboard%[email protected]');> >>> . >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- >> For more options, visit http://beagleboard.org/discuss >> --- >> You received this message because you are subscribed to the Google Groups >> "BeagleBoard" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] >> <javascript:_e(%7B%7D,'cvml','beagleboard%[email protected]');> >> . >> For more options, visit https://groups.google.com/d/optout. >> > > > > -- > Extra Ham Operator: K7AZJ > Registered Linux User: 275424 > Raspberry Pi and Arduino developer > > There are 10 kinds of people in the world: > Those who can read binary and those who can't. > > -- > For more options, visit http://beagleboard.org/discuss > --- > You received this message because you are subscribed to the Google Groups > "BeagleBoard" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected] > <javascript:_e(%7B%7D,'cvml','beagleboard%[email protected]');> > . > For more options, visit https://groups.google.com/d/optout. > -- For more options, visit http://beagleboard.org/discuss --- You received this message because you are subscribed to the Google Groups "BeagleBoard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
