RPM Meter
in [Design]
|
Prev: Bad Omen?
Next: Posts with threaded holes
From: langwadt on 29 Jul 2010 13:58 On 29 Jul., 19:22, John Larkin <jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote: > On Thu, 29 Jul 2010 09:42:42 -0700 (PDT), "langw...(a)fonz.dk" > > > > <langw...(a)fonz.dk> wrote: > >On 28 Jul., 05:07, John Larkin > ><jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote: > >> On Tue, 27 Jul 2010 18:30:29 -0700 (PDT), Dookie > > >> <abstract.dissona...(a)gmail.com> wrote: > >> >I have a contact/photo rpm meter but it is acting funky and would like > >> >to create a simple meter for some testing. > > >> >I've created the pic code and seem to have gotten it to work but need > >> >some sensor to debug it. The pic code basically counts the time > >> >between pulses. I do not use the ADC but rather the digital inputs and > >> >monitor when the pin goes high. The code does have built in dead time > >> >to prevent multiple countings. > > >> >My idea was to stick a magnet on the device I want to measure the > >> >rpms(flywheel) and use a coil of wire as the sensor. > > >> >The problem is that I'm getting much lower voltage than I expected. > >> >Faraday's law says the voltage on the coil is the number of turns > >> >times the change in flux. The change in flux depends on the rpm's but > >> >I'm simply using my hand to move the magnet across the coil. I get > >> >something like a sinusoid damped(or alternatively the derivative of > >> >the erf function). I have about 200-300 turns and the coil is getting > >> >unwieldy. > > >> >This is fine but it most I get about 30mV(surely to be much larger > >> >with the flywheel at full speed). What I'm thinking of doing is simply > >> >using an opamp or bjt to act as a switch/comparator but the only > >> >problem I see is that the noise may add extra pulses. > > >> >Is there a simple way I can increase the noise floor, say, by > >> >offseting the ground of the op amp and effectively setting making dc > >> >negative relative to the opamp. I'm thinking that I could add a > >> >resistor to the -V of the op amp. I only have a + supply. > > >> >Or is there any better way that doesn't involve to much work? It > >> >doesn't have to be completely perfect as I am taking an average for > >> >the rpm's and can through out outliers. The main thing is to get the > >> >input to the pic to act digitally. > > >> >I know there are a god awefull number of ways to do this but something > >> >simple with minimum and common parts is what I'm looking for. > > >> Your pickup is going to behave like a classic variable-reluctance > >> speed sensor. The signal amplitude and frequency will go up together > >> as the speed increases. The classic signal conditioner is an > >> integrator of sorts... > > >>ftp://jjlarkin.lmi.net/VR_integrator.JPG > > >> which flattens out the speed:voltage curve. > > >> This can be made to work nicely over a 1000:1 speed range. I use that > >> circuit in my VME tach modules. > > >> John > > >you might not like it but I've seen this works with no changes for the > >crank sensor in pretty much all cars. > > > hct14 > >VR_in >--10u---+--22K--+-|>o----> out > > | 1n > > | _|_ > > | > > 22K +--10K--+ > > | | | > > +---1K--+-o<|---+ > > | hct14 | > > 47u 10n > > _|_ _|_ > > That is a very cool way to set the threshold bias for the upper > schmitt. And add a little time dithering, if you ever need it! > > The 22K+1n is only a 22us time constant. How narrow do the VR pulses > get at high revs? I guess it's mostly a deglitcher and doesn't > actually integrate much. > > John the standard bosch crank sensor is 60 minus 2 teeth, so frequency in Hz ~= rpm -Lasse
From: David Eather on 29 Jul 2010 19:14 On 29/07/2010 9:55 PM, Jan Panteltje wrote: > On a sunny day (Thu, 29 Jul 2010 09:48:04 +1000) it happened David Eather > <eather(a)tpg.com.au> wrote in<OoydnecLgt0rXM3RnZ2dnUVZ_gKdnZ2d(a)giganews.com>: > >> On 28/07/2010 9:09 PM, Jan Panteltje wrote: >>> On a sunny day (Wed, 28 Jul 2010 12:08:10 +1000) it happened David Eather >>> <eather(a)tpg.com.au> wrote in<RuidnQS0OouWDNLRnZ2dnUVZ_t2dnZ2d(a)giganews.com>: >>> >>> >>>> PIC? Shouldn't you use something like a 555 and an analogue meter? >>> NO >> >> But a 555 is the highest example of the chip designer's art. > > It sure is a clever chip, and you can make stable oscillators with it > with a frequency that does not depend on the supply voltage much. > However, in this case, if you generate a fast duration pulse to drive > an analog meter, you still need a voltage stabiliser, as you need a constant amplitude. > > The PIC driving a LCD however, has an internal oscillator that has an accuracy of > +-2% from 0�C to 85�C, and +-5% from -40�C to +125�C, this no longer needs the regulator, > other then a simple zener perhaps to make 5 or 3.3 V, and no calibration needed. > Better display possibilities, added calculation possibility for other things, > like long time average, possibility of setting alarms, possibility of loop control. > All in one PIC chip mind you. > > > > > > But the cost of the time to program it outweighs the cost of a voltage regulator. I am not actually a big fan of 555's - I just like how some "features" (like the switching generated power spikes) annoy some peoples.
From: David Eather on 29 Jul 2010 19:27 On 30/07/2010 9:14 AM, David Eather wrote: > On 29/07/2010 9:55 PM, Jan Panteltje wrote: >> On a sunny day (Thu, 29 Jul 2010 09:48:04 +1000) it happened David Eather >> <eather(a)tpg.com.au> wrote >> in<OoydnecLgt0rXM3RnZ2dnUVZ_gKdnZ2d(a)giganews.com>: >> >>> On 28/07/2010 9:09 PM, Jan Panteltje wrote: >>>> On a sunny day (Wed, 28 Jul 2010 12:08:10 +1000) it happened David >>>> Eather >>>> <eather(a)tpg.com.au> wrote >>>> in<RuidnQS0OouWDNLRnZ2dnUVZ_t2dnZ2d(a)giganews.com>: >>>> >>>> >>>>> PIC? Shouldn't you use something like a 555 and an analogue meter? >>>> NO >>> >>> But a 555 is the highest example of the chip designer's art. >> >> It sure is a clever chip, and you can make stable oscillators with it >> with a frequency that does not depend on the supply voltage much. >> However, in this case, if you generate a fast duration pulse to drive >> an analog meter, you still need a voltage stabiliser, as you need a >> constant amplitude. >> >> The PIC driving a LCD however, has an internal oscillator that has an >> accuracy of >> +-2% from 0�C to 85�C, and +-5% from -40�C to +125�C, this no longer >> needs the regulator, >> other then a simple zener perhaps to make 5 or 3.3 V, and no >> calibration needed. >> Better display possibilities, added calculation possibility for other >> things, >> like long time average, possibility of setting alarms, possibility of >> loop control. >> All in one PIC chip mind you. >> >> >> >> >> >> > > But the cost of the time to program it outweighs the cost of a voltage > regulator. > > I am not actually a big fan of 555's - I just like how some "features" > (like the switching generated power spikes) annoy some peoples. But seriously for a simple and temporary RPM meter without a chance to act "funky" I would have put some form of one-shot and an analogue meter as the ideal solution (fast, simple, cheap, little circuit/firmware debugging). It's not without possible problems, but whatever they are are very unlikely to be related to the problems in the original device.
From: John Larkin on 29 Jul 2010 19:59 On Fri, 30 Jul 2010 09:27:40 +1000, David Eather <eather(a)tpg.com.au> wrote: >On 30/07/2010 9:14 AM, David Eather wrote: >> On 29/07/2010 9:55 PM, Jan Panteltje wrote: >>> On a sunny day (Thu, 29 Jul 2010 09:48:04 +1000) it happened David Eather >>> <eather(a)tpg.com.au> wrote >>> in<OoydnecLgt0rXM3RnZ2dnUVZ_gKdnZ2d(a)giganews.com>: >>> >>>> On 28/07/2010 9:09 PM, Jan Panteltje wrote: >>>>> On a sunny day (Wed, 28 Jul 2010 12:08:10 +1000) it happened David >>>>> Eather >>>>> <eather(a)tpg.com.au> wrote >>>>> in<RuidnQS0OouWDNLRnZ2dnUVZ_t2dnZ2d(a)giganews.com>: >>>>> >>>>> >>>>>> PIC? Shouldn't you use something like a 555 and an analogue meter? >>>>> NO >>>> >>>> But a 555 is the highest example of the chip designer's art. >>> >>> It sure is a clever chip, and you can make stable oscillators with it >>> with a frequency that does not depend on the supply voltage much. >>> However, in this case, if you generate a fast duration pulse to drive >>> an analog meter, you still need a voltage stabiliser, as you need a >>> constant amplitude. >>> >>> The PIC driving a LCD however, has an internal oscillator that has an >>> accuracy of >>> +-2% from 0�C to 85�C, and +-5% from -40�C to +125�C, this no longer >>> needs the regulator, >>> other then a simple zener perhaps to make 5 or 3.3 V, and no >>> calibration needed. >>> Better display possibilities, added calculation possibility for other >>> things, >>> like long time average, possibility of setting alarms, possibility of >>> loop control. >>> All in one PIC chip mind you. >>> >>> >>> >>> >>> >>> >> >> But the cost of the time to program it outweighs the cost of a voltage >> regulator. >> >> I am not actually a big fan of 555's - I just like how some "features" >> (like the switching generated power spikes) annoy some peoples. > >But seriously for a simple and temporary RPM meter without a chance to >act "funky" I would have put some form of one-shot and an analogue meter >as the ideal solution (fast, simple, cheap, little circuit/firmware >debugging). It's not without possible problems, but whatever they are >are very unlikely to be related to the problems in the original device. You could use my world-famous double-tach circuit... ftp://jjlarkin.lmi.net/DoubleTach.jpg John
From: David Eather on 29 Jul 2010 20:14
On 30/07/2010 9:59 AM, John Larkin wrote: > On Fri, 30 Jul 2010 09:27:40 +1000, David Eather<eather(a)tpg.com.au> > wrote: > >> On 30/07/2010 9:14 AM, David Eather wrote: >>> On 29/07/2010 9:55 PM, Jan Panteltje wrote: >>>> On a sunny day (Thu, 29 Jul 2010 09:48:04 +1000) it happened David Eather >>>> <eather(a)tpg.com.au> wrote >>>> in<OoydnecLgt0rXM3RnZ2dnUVZ_gKdnZ2d(a)giganews.com>: >>>> >>>>> On 28/07/2010 9:09 PM, Jan Panteltje wrote: >>>>>> On a sunny day (Wed, 28 Jul 2010 12:08:10 +1000) it happened David >>>>>> Eather >>>>>> <eather(a)tpg.com.au> wrote >>>>>> in<RuidnQS0OouWDNLRnZ2dnUVZ_t2dnZ2d(a)giganews.com>: >>>>>> >>>>>> >>>>>>> PIC? Shouldn't you use something like a 555 and an analogue meter? >>>>>> NO >>>>> >>>>> But a 555 is the highest example of the chip designer's art. >>>> >>>> It sure is a clever chip, and you can make stable oscillators with it >>>> with a frequency that does not depend on the supply voltage much. >>>> However, in this case, if you generate a fast duration pulse to drive >>>> an analog meter, you still need a voltage stabiliser, as you need a >>>> constant amplitude. >>>> >>>> The PIC driving a LCD however, has an internal oscillator that has an >>>> accuracy of >>>> +-2% from 0�C to 85�C, and +-5% from -40�C to +125�C, this no longer >>>> needs the regulator, >>>> other then a simple zener perhaps to make 5 or 3.3 V, and no >>>> calibration needed. >>>> Better display possibilities, added calculation possibility for other >>>> things, >>>> like long time average, possibility of setting alarms, possibility of >>>> loop control. >>>> All in one PIC chip mind you. >>>> >>>> >>>> >>>> >>>> >>>> >>> >>> But the cost of the time to program it outweighs the cost of a voltage >>> regulator. >>> >>> I am not actually a big fan of 555's - I just like how some "features" >>> (like the switching generated power spikes) annoy some peoples. >> >> But seriously for a simple and temporary RPM meter without a chance to >> act "funky" I would have put some form of one-shot and an analogue meter >> as the ideal solution (fast, simple, cheap, little circuit/firmware >> debugging). It's not without possible problems, but whatever they are >> are very unlikely to be related to the problems in the original device. > > You could use my world-famous double-tach circuit... > > ftp://jjlarkin.lmi.net/DoubleTach.jpg > > John > > Actually I can't. I have never been able to view or download anything posted at ftp://jjlarkin.lmi.net |