Message-Id: <4.1.0.67.19981009073725.021f06a0@mail1.21stcentury.net> Date: Fri, 09 Oct 1998 07:48:30 -0500 To: bikecurrent@cyclery.com From: "Donald Piven" Subject: Re: 100mph cycle speedo? Quoth Josh Putnam: >Jobst Brandt has confirmed in rec.bicycles.* postings that the >Avocet 45 will operate accurately above 100mph. This may have to >do with the difference in pickup designs between most >speedometers with a single magnet and reed switch and Avocet's >multi-pole magnet ring. Back in March, this sort of question came up, and I conducted a mad-scientist experiment. The results indicated that on conventional bikes (with 26"/700c wheels), all Avocets will easily and accurately indicate speeds >100mph. However, the Avocet 45TT might not be suited for high speeds on smaller wheels, due to a change in how the 45TT processes the input pulse train. ---------------------------------------------------------------------------- Date: Wed, 25 Mar 1998 23:07:55 -0600 To: bikecurrent@cyclery.com From: Don Piven Subject: The Almost-Definitive Avocet Speed Test Quoth Simon Lee: >My avocet manual says the maximum speed is 206km/h or 100000 revs/hour >or around 30Hz . That means 1000 CPU clocks per 'sense' which sounds >pretty reasonable to me, remember the CPU must do all those trip, min, >max, stopwatch calcs etc etc.. Although my math suggests a lot less time per "sense" for Avocets: a 700c wheel with a circumference of 82" performs just under 773 revolutions over a mile. At 60 mph, this is just under 773 rpm. At 400 mph, that's 5100 rpm plus or minus. Since the Avocet magnet ring has 20 poles, you get 20 pulses per revolution, which results in a frequency of 850 Hz being sent to the computer. (The actual frequency is half the pulse rate; remember that the magnet ring is alternating N/S poles, so the sensor generates AC.) So I gave it a real-life test, kinda. I wired up an old Avocet mount to the line output of a Roland synthesizer, picked a sinewave preset, and adjusted the volume to put about 0.3 VRMS across the computer. I have a number of Avocets in my collection; I picked a 45 which I'd had for a couple of years, and a new 45TT I got for Christmas. I started with the (non-TT) 45. The computer was calibrated to a wheel circumference of 82.01", and I started my way up the scale. A pitch of A-flat two octaves up from middle C (830 Hz + a fraction) results in a stable reading of 387 mph, varying between 387.1 and 387.6 over twenty seconds or so. One semitone higher (A-natural, 880 Hz), and the speed reading plummets to 3.2 (three point two) mph; higher pitches result in the indicated speed coming back up, but erratically. (At 388 mph, the ride odometer clicks away about 0.1 mile per second.) I recalibrated the 45 to 40.00" wheel circumference. I was able to feed it double the frequency (Ab three octaves up, 1660 Hz) for a 380 mph reading; going higher again resulted in a reduction of indicated speed and erratic display. I then changed the calibration to its maximum of 130.99"; again I hit the wall at ~380 indicated (C an octave up, 523 Hz). Now for the new Avocet 45TT. The manual for the 45TT claims that this thing will register up to 400 mph. The starting calibration on this one was 85.24". Plugged it in, went up the scale . . . and not even close -- the highest speed the 45TT would record was around 128 mph when fed middle C (261.6 Hz); one semitone up and the display went right down to zero and wouldn't budge. I recalibrated to its minimum of 30.24" and -- surprise! -- the brick wall in the case of the 45TT is the pulse rate. Again it wouldn't accept any signal much over 260 Hz, and maxed out at an indicated 45 mph with this wheel size. The maximum wheel size of 130.99" resulted in an indicated 194.4 mph with the same 261.6 Hz signal. I'd thought that the differences between the old 45 and the new 45TT were incremental (the addition of a "time trial" distance/time countdown mode in the TT), but it looks like Avocet made some fundamental changes in how the thing operates -- the older 45s apparently were limited only by math overflow, while the new 45TT apparently has a brick-wall filter in its input that ignores any input higher than 250 Hz or so, which severely limits the maximum speed it can register, especially with small wheel sizes. Keep in mind that this test did not address the issue of how good a signal the inductive pickup can emit at very high wheel speeds, which would affect the short-term accuracy of the attached computer. Ain't mad science wonderful? ---------------------------------------------------------------------------- +-----------------------+---------------------------------------------------- |Don Piven - Chicago IL | Suburbs are places where they cut down the trees, |dpiven@21stcentury.net | then name streets after them. |.. or .. djp@suba.com | | '97 Waterford 1900 | +-----------------------+ [Alfred E. Neuman] +=====================================================+ Subscribe to your favorite bicycling magazines through Cyber Cyclery at the lowest possible prices. Find out more at http://cyclery.com/home/subscriptions.html +=====================================================+