RPM Meter
in [Design]
|
Prev: Bad Omen?
Next: Posts with threaded holes
From: markp on 30 Jul 2010 15:58 "John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message > [1] I *still* need a good, small, low-drop current limiter for use as > a self-resetting fuse. The ideal part would be a SOT-23 whose current > limit is set by one resistor, thermal limits, and retries at full > current once in a while. How come nobody makes one, at least past 5 > volts? Polyfuses sort of suck. Here's an EDN article about low dropout constant current source (figure 2 in the article): http://electronicdesign.com/article/components/low-dropout-current-regulator-improves-led-driver-.aspx I'm wondering whether you could add another resistor between R2 and ground, and in-between them put an N channel MOSFET. You could detect current limiting condition with an ADC input from the voltage at the top of that second resistor and turn off the FET, thereby isolating the supply from the output. You could program it to retry once in a while once it had detected current limiting. Not exactly a 1 chip solution though.... Mark.
From: Joe on 30 Jul 2010 17:08 On Fri, 30 Jul 2010 09:14:49 -0700, John Larkin wrote: [snip] > Circuit topologies are fun all on their own, and are appropriate as a > starting point for discussion of an underspecified problem, like the > ones we usually get here. Circuits are sort of like improv comedy, or > jazz, or slam poetry. The one we had on current limiters was fun [snip] > I don't use Spice much, so I start with circuits without values and > "simulate" them in my head. I visualize the way I'd want the circuit > to work with correct but in fact unknown values, and if it looks > interesting, real values could be plugged in later (one at a time > even!) to force the dynamics to align with the image. People who > "visualize" a circuit's operation in Spice have no choice but to use > real component values and then fiddle for results. > > I do admit to limited expertise with rusty patio furniture. [snip] > I like to drift "off topic" by injecting technical asides into the > henpecky personal rants. But that makes some of the hens even madder. [snip] Asides, if they address technical issues and are not too far out of some range of reasonable s.e.d topics, probably offend nobody. But snarky asides [like the "rusty patio furniture" one above] intended as putdowns accomplish nothing positive and make you look petty, stupid and nonprofessional. If you have something useful, accurate, interesting, new, creative, inspiring or educational to say, then say it. If you want to say something petty, untrue, ad hominem, incomplete, obamic, vague or snarky, then don't.
From: David Eather on 30 Jul 2010 18:10 On 30/07/2010 1:38 PM, Grant wrote: > On Fri, 30 Jul 2010 10:14:56 +1000, David Eather<eather(a)tpg.com.au> wrote: > >> 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 > > Look at your ftp or firewall settings, works fine here on Win7 with > Google Chrome browser, and on Slackware with wget. > > Check for passive FTP enabled? > > Grant. Thanks for the help, but no joy. I'm giving up on it.
From: Jim Thompson on 30 Jul 2010 18:50 On Sat, 31 Jul 2010 08:10:15 +1000, David Eather <eather(a)tpg.com.au> wrote: [snip] > >Thanks for the help, but no joy. I'm giving up on it. How about the RPM meter? IIRC you wanted pulses to count. Trivial to do. Contact me if you need help. Likewise if you want a DC voltage versus RPM. Let me know. ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | Spice is like a sports car... Performance only as good as the person behind the wheel.
From: Grant on 30 Jul 2010 19:00
On Fri, 30 Jul 2010 09:14:49 -0700, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >On Fri, 30 Jul 2010 15:15:11 +1000, Grant <omg(a)grrr.id.au> wrote: > >>On Thu, 29 Jul 2010 20:06:55 -0700, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Thu, 29 Jul 2010 18:30:04 -0700, Rich the Cynic <cynic(a)example.net> >>>wrote: >>> >>>>On Wed, 28 Jul 2010 16:36:00 +0100, markp wrote: >>>>> >>>>>> John "The Bloviator" Larkin's post is a slap-it-on-paper-quick without >>>>>> thinking "solution". Use at your peril. He'll never label it with >>>>>> values. If he did, his "solution" could be checked. >>>>> >>>>> I rest my case. >>>> >>>>I guess he has to go after John since he's plonked practically everyone >>>>else on the group, presumably to "punish" us. (Ha, ha, I can't hear you, >>>>nanner nanner nanner!) ;-) >>>> >>> >>>Right. He only distributes technical excellence now in private emails, >>>because we village idiots don't deserve such wisdom. Darn, I was >>>hoping for further adventures in rusty patio furniture. >> >>Thing is, being reminded of a topology that may suit a particular >>case or problem is a valuable reply, doesn't take much time or >>effort to post, one gets from paper to camera to web site fairly >>easily these days. Even I can do it ;) > >Circuit topologies are fun all on their own, and are appropriate as a >starting point for discussion of an underspecified problem, like the >ones we usually get here. Circuits are sort of like improv comedy, or >jazz, or slam poetry. The one we had on current limiters was fun [1] > >I don't use Spice much, so I start with circuits without values and >"simulate" them in my head. I visualize the way I'd want the circuit >to work with correct but in fact unknown values, and if it looks >interesting, real values could be plugged in later (one at a time >even!) to force the dynamics to align with the image. That's where I'm at too, I found spice useful in modeling changes to a commercial (poorly designed) PWM motor controller, even though I had to play with defining the load so it matching CRO waveform, but in that situation it saved much spider's nest soldering to improve matters. And when I did change the board to new component values, the spice'd effort was good match to the real. But not for design from scratch, only sections where I'm not sure of the outcome. But then, I've not used LTSpice enough to use it effectively. I don't think it helps me define component values for say a calibration network, where I'm prepared to put a trimpot and two E24 1% resistors north and south to get wanted range. Spice will let me check the result, but will it find the component values (with 1 or 2 1% E24 each side of trimpot) if I tell it the trimpot value and reference voltage? > [2] People who >"visualize" a circuit's operation in Spice have no choice but to use >real component values and then fiddle for results. Which is okay in some situations. > >I do admit to limited expertise with rusty patio furniture. This must've been before I joined this group? Or in a thread I killed. I kill many threads by subject because they're way out of my area or about USA issues I know nothing about. > > >>Whatever fun there is in a flameful reply and/or the occasional >>flamefest is lost if it drags on for weeks, or starts becoming >>personal attack. Ideas can be silly, doesn't mean the person is >>(well, not all the time :) > >I like to drift "off topic" by injecting technical asides into the >henpecky personal rants. But that makes some of the hens even madder. Just shows up the intolerance some have? A little madness helps :) > >John > >[1] I *still* need a good, small, low-drop current limiter for use as >a self-resetting fuse. The ideal part would be a SOT-23 whose current >limit is set by one resistor, thermal limits, and retries at full >current once in a while. How come nobody makes one, at least past 5 >volts? Polyfuses sort of suck. The sad thing here is that Jim T. could probably design one, but unless a zillion others want it too, it'd never make it to silicon. Your spec sounds like a low power, low dropout regulator that self protects and can be made to oscillate via thermal lag or feedback. My first polyfuse 'hands on' experience is the 1.85A one in series with a hobby drill, with a bright red LED across it. It's showing me much about the beastie's behaviour on overload, and near overload while recovering. They're quite slow, take seconds to recover. So they're very good for protecting stuff like a motor from losing its magic smoke, or a speaker driver, actuators without limit switches. Not so good for electronic protection except as another place to share the heat part of coping with the overload? > >[2] Suppose you had a Spice model that worked just right, but the >values were hidden. Based on the waveforms, you could determine the >actual parts values, one at a time, in any order. That's sort of the >idea: visualize the circuit and the waveforms, then start poking in >values. Sort of the anti-Spice. But in order to build the spice model, one first visualises the circuit and guess/estimate the component starter values. It'd be fine if the spice schematic could be imported to the CAD for PCB layout and product documentation, BoM, etc. Grant. |