Don't ad-block us - support your favorite websites. We have safe, unobstrusive, robotics related ads that you actually want to see - see here for more.
0 Members and 1 Guest are viewing this topic.
ARM9 microcontroller with DSPMicrocontroller has a 600 MHz 16/32-bit dual-MAC DSP core and a 300 MHz ARM9 RISC core; includes 16-Mbit DRAM, eFPGA block, reconfigurable interfaces, plus a wide range of peripherals such as A/D and D/A; targets wireless infrastructure equipment.STW22000ea/large qty, $30
[...] chips that clocks 300 mhz or faster? I believe I may need a fast system clock for a long distance laser range finder and vision systems.
Hi,Quote from: ib1yysguy on February 10, 2009, 02:37:08 AM[...] chips that clocks 300 mhz or faster? I believe I may need a fast system clock for a long distance laser range finder and vision systems. Speed of light = 299,792,458 m/sFor a resolution of eg. 100 mm, each counter step should be 334 ps and if you want, say 5% tolerance max, you need jitter and other timing errors to be under 16.7 ps.Even if you meant 300 MHz and if each oscillatorcycle was worth a system clock cycle, your clock would be 3,3 ns (and the tolerance??), so even with single clock instructions it won't cut it the way you are planning.There is a reason that T.O.F. LASER rangers cost a soul (when on offer) and I can guarantee you that you need quite a bit of (discreete) high speed circuitry under your west before you even think of beginning a T.O.F. design.Go slow and get a triangulating LASER to work for starters, that should keep you from boredom for quite a while, if you want to design it from the ground up.
Quote from: Soeren on February 10, 2009, 04:37:03 PMHi,Quote from: ib1yysguy on February 10, 2009, 02:37:08 AM[...] chips that clocks 300 mhz or faster? I believe I may need a fast system clock for a long distance laser range finder and vision systems. Speed of light = 299,792,458 m/sFor a resolution of eg. 100 mm, each counter step should be 334 ps and if you want, say 5% tolerance max, you need jitter and other timing errors to be under 16.7 ps.Even if you meant 300 MHz and if each oscillatorcycle was worth a system clock cycle, your clock would be 3,3 ns (and the tolerance??), so even with single clock instructions it won't cut it the way you are planning.There is a reason that T.O.F. LASER rangers cost a soul (when on offer) and I can guarantee you that you need quite a bit of (discreete) high speed circuitry under your west before you even think of beginning a T.O.F. design.Go slow and get a triangulating LASER to work for starters, that should keep you from boredom for quite a while, if you want to design it from the ground up.Thats not entirely true... I built a successful TOF laser rangefinder circuit a couple months ago able to determine distances up to 15mThis was with an Atmega8Even though distance wasnt too much, it all depends on the sensors and lasers you usedFor example, if I increased the voltage to the photo transistors I used, this would increase the range of detectionif I increased the laser's intensity (plan on going from 5nw up) then I could increase the rangeand the minimum processing speed I could use it on (switching chips while switching wiring and altering code a bit) I found that 8MHZ is the minimum for accuracy. Beyond that, its like an asymptope if I were to graph it, it would slowly and slowly get closer to a fixed "distance" but never approach it
Thats not entirely true... I built a successful TOF laser rangefinder circuit a couple months ago able to determine distances up to 15mThis was with an Atmega8Even though distance wasnt too much, it all depends on the sensors and lasers you usedFor example, if I increased the voltage to the photo transistors I used, this would increase the range of detectionif I increased the laser's intensity (plan on going from 5nw up) then I could increase the rangeand the minimum processing speed I could use it on (switching chips while switching wiring and altering code a bit) I found that 8MHZ is the minimum for accuracy. Beyond that, its like an asymptope if I were to graph it, it would slowly and slowly get closer to a fixed "distance" but never approach it
#picaxe 14m scanmotor: pwmout 2,249,0 'turn off motor in preparation for direction change let w2=w1/b6 'store distance result in w2 debug b6=0 'sample counter used for averaging w1=0 'samples are added together and stored here. pause 10 'gives scan motor time to stop before changing direction readadc10 4,w0 'read position of laser pin1=bit9 'bit9=512 (about halfway) if position >=512 then relay is on pwmout 2,249,140 'turn on motor at 14% to start up pause 40 'wait 40mS for motor to start up pwmout 2,249,80 'drop motor speed to 8% (slow scan) b7=1 'this flag indicates that the scan motor is out of scan rangescanloop: readadc10 4,w0 'read position of laser if pin3=0 then 'has laser been detected by sensor? let w1=w1+w0 'add its position for averaging inc b6 'inc number of samples end if if w0>20 and w0<1000 then 'is laser within scan range? b7=0 'reset out of range flag and continue scan goto scanloop end if if b7=1 then goto scanloop 'laser out of range and flag not reset (has not come into range yet) goto scanmotor 'laser out of range and flag has been reset (just gone out of range)
It appears that you aren't calculating time of flight but are doing triangulation instead.