|
Needs to label the pins on the chips. Or what the chips are. Or something. I'm sure the article describes what IC1 and IC2 are, but it makes it hard to read the schematic if they don't say what each pin does. I realize it's for implementation (the target audience doesn't understand what each pin does anyways) but with that addition it will be a lot nicer I think.
|
#
?
Jun 5, 2011 23:21
|
|
|
|
| # ? May 8, 2024 23:33 |
|
|
It looks like some fisher price baby's first circuit thing.
|
|
#
?
Jun 5, 2011 23:52
|
|
|
Delta-Wye posted:Needs to label the pins on the chips. Or what the chips are. Or something. I'm sure the article describes what IC1 and IC2 are, but it makes it hard to read the schematic if they don't say what each pin does. I realize it's for implementation (the target audience doesn't understand what each pin does anyways) but with that addition it will be a lot nicer I think. Is it sad that I could tell that they are 555 timers from the way they're hooked up? Positive on pin 8, negative on pin 1, a capacitor connected between negative and pin 5, what appears to be an output on 3 and IC2 seems to be hooked up in an astable configuration connected to IC1 in some bizarre way I've never seen before. Looks like some kind of tone generator but they left the reset pin disconnected which is no good because it doesn't have an internal pullup. Edit: I still hate it because it's less clear than something done up in say Eagle. BattleMaster fucked around with this message at 00:14 on Jun 6, 2011 |
|
#
?
Jun 6, 2011 00:11
|
|
|
Delta-Wye posted:Needs to label the pins on the chips. Or what the chips are. Or something. I'm sure the article describes what IC1 and IC2 are, but it makes it hard to read the schematic if they don't say what each pin does. I realize it's for implementation (the target audience doesn't understand what each pin does anyways) but with that addition it will be a lot nicer I think. Disagree. I've never seen a circuit that looked good with all the pins labeled, unless it's within a datasheet (and even then, only on the ones with a <10 pins). Labels like R1 and IC1 are much more important than putting values and IC names directly into the schematic. Either a box on the side or a list at the end is sufficient.
|
|
#
?
Jun 6, 2011 03:10
|
|
|
Slanderer posted:Disagree. I've never seen a circuit that looked good with all the pins labeled, unless it's within a datasheet (and even then, only on the ones with a <10 pins). Really? You've never seen a circuit look good with all the pins and values listed? It really is common in the industry. Standard, even. A list at the end would be loving useless when you have hundreds of components. Or when you have a 196 pin DSP. Having to cross reference pin names every time you look at the schematic would make it completely worthless.
|
|
#
?
Jun 6, 2011 12:30
|
|
|
Yeah, I design schematics all the time, usually with large pin count parts, and every pin is labeled. Bigger parts also typically have the part name on them, though not necessarily the long name you would use to order them.
|
|
#
?
Jun 6, 2011 13:03
|
|
|
Wow, I'm an utter idiot who hasn't designed anything in a CAD program for way too long. I just opened my past couple projects, and noticed that every single part in the schematic has every single pin labeled (and I should know, since I had to manually build the footprint and schematic symbol for each and every one of the dozen different ICs in my last one). However, I realize why I got confused--it was another weird element of that schematic! The ICs are drawn out as DIP packages, instead of as generic block symbols. With the exception of nice simple DIP-8 packaged ICs (like generic EEPROMs), I've been moving more towards having my schematic symbols be generic. I settled upon that style after realizing how stupid and annoying it was to try to make the design of the schematic symbol for weird packages that resembled the pin config of the package.  That's a nice schematic, Forrest Mims. I still use the little camel hump things (not sure if there is an actual term) for wire crossings when I draw stuff by hand half the time, since I occasionally leave out node dots. I've been taught from basically the beginning that this has been deprecated, but I still like it for hand drawn stuff, since it makes things clearer at first glance (probably because my stuff is messy as hell)
|
|
#
?
Jun 6, 2011 14:18
|
|
|
Slanderer posted:
I do it too, despite having gotten into arguments about it with people with more seniority. Why would you deliberately make your schematics less clear and more prone to error?
|
|
#
?
Jun 6, 2011 15:47
|
|
|
sixide posted:
Well, so it turns out that once it's working the signal is actually pretty low and I'd really like to boost it. I'm not sure I understand how to run the signal twice though--what's the method behind that? The blog I was following doesn't make it clear to me. But I dig the idea of the tape loop/record head/playback head... if I can rustle up another cassette player I'm gonna try some things. Took me forever to locate this one though.
|
|
#
?
Jun 6, 2011 15:47
|
|
|
ante posted:I do it too, despite having gotten into arguments about it with people with more seniority. Why would you deliberately make your schematics less clear and more prone to error? My argument is that if I leave out a node dot by accident, it's equivalent to a completely different circuit. But if I use node dots and the little wire-crossing things, than if I forget to use either one I can immediately tell I made a mistake and that I need to clarify it.
|
|
#
?
Jun 6, 2011 15:54
|
|
|
We actually have a design standard at work where schematics should not be drawn with 4 way intersections because it's too easy to confuse them with one line passing over another on printouts. Here's examples of a part and nets on a typical schematic of mine:
|
|
#
?
Jun 6, 2011 18:01
|
|
|
What do you guys think of this as a power supply for an EE student looking to start a small hobby workshop? I was mainly looking for something that was at least dual output, and bipolar so I could hook up op amps and whatnot. Is it overkill? Any recommendations?
|
|
#
?
Jun 6, 2011 22:31
|
|
|
alwayslost posted:What do you guys think of this as a power supply for an EE student looking to start a small hobby workshop? I was mainly looking for something that was at least dual output, and bipolar so I could hook up op amps and whatnot. Is it overkill? Any recommendations? That one looks fine. As an less expensive alternative, you might like http://www.mpja.com/prodinfo.asp?number=9333+PS I have not used this particular supply, but I have a couple others from mpja and they work as advertised.
|
|
#
?
Jun 6, 2011 22:44
|
|
|
Hillridge posted:We actually have a design standard at work where schematics should not be drawn with 4 way intersections because it's too easy to confuse them with one line passing over another on printouts. This is also what I do, as taught to me by my father, who was an electronics engineer for many years (and still technically is, except his work mostly do software now), and is what was done at many places he worked. Given that I always found the Forrest Mims style of a little bump from the Radio Shack/Tandy books my dad gave me really weird, although a little amusing, looking. I do love those books though.
|
|
#
?
Jun 7, 2011 17:14
|
|
|
jacteh posted:Pg = (rho)*g*h You just need a differential pressure transmitter with a high and low leg to compensate for any changes in water level. It would be difficult to find a transmitter capable of measuring the tiny change in something the size of a fishtank cheaply though.
|
|
#
?
Jun 8, 2011 01:35
|
|
|
I went and bought a Tektronix 11801B oscilloscope off ebay. This thing must have been awesome back in the late '80s/early '90s when it first came out. It's still pretty cool, being able to go down to 5ps/div and all.  The menu says it's been in use for a total of 82,000+ hours.  Oscilloscope Display
|
|
#
?
Jun 9, 2011 00:42
|
|
|
Krenzo posted:I went and bought a Tektronix 11801B oscilloscope off ebay. This thing must have been awesome back in the late '80s/early '90s when it first came out. It's still pretty cool, being able to go down to 5ps/div and all. 50Ghz analog bandwidth is certainly impressive, but the sampling rate is like... 200Ksps, how does that work? I understand equivalent time sampling but even so. Then they also say it as a 10 femtosecond 'sampling interval' whatever that is.
|
|
#
?
Jun 9, 2011 01:30
|
|
|
Rescue Toaster posted:50Ghz analog bandwidth is certainly impressive, but the sampling rate is like... 200Ksps, how does that work? I understand equivalent time sampling but even so. Then they also say it as a 10 femtosecond 'sampling interval' whatever that is. It receives a trigger and then samples a point relative to the trigger. On the next trigger, it samples 10 fs after trigger. On the next, it samples 20 fs after trigger, and then 30 fs and so on. It does this until it's built up many samples that are some multiple of 10 fs after the trigger. You can choose to have 512, 1024, ... up to 5,000 samples to display at a time. So at the best resolution, you could have a waveform made up of 5,000 samples that are 10 fs apart. If you want to see a longer span of time with less resolution, it would use a larger sample interval of picosecond, nanosecond, microsecond, etc. Since it only does 200Ksps, say if your trigger is a 1 MHz signal, it will catch one trigger pulse and then ignore any triggers until it's done sampling the first trigger and ready for another. Krenzo fucked around with this message at 02:46 on Jun 9, 2011 |
|
#
?
Jun 9, 2011 02:44
|
|
|
Hello! I hope this is an appropriate place to posit my question. I recently came across this : http://www.etsy.com/listing/74059377/working-arcade-light-switch And I decided that by golly, I wanted one! Except. You know, $35. For $3 of plastic. This will not stand, nope, so I want to make one. The plate is like $.50 or free or whatever, the buttons are like $2, easy enough. Except the buttons are momentary switches and do not normally function as toggle switches, of course. This is where I lose track because I only know how to work on arcade cabs a bit and nothing about interior wiring. Fine fine. The solution that was posited to me was that I should be using a latching relay rated for at least 110v/15a for a standard US light switch. Is there anything MISSING in this solution or can you brilliant people suggest a more elegant solution? If it is beyond my scope I will probably make my boyfriend do it, but I want to make sure I'm at least headed in the right direction. $35! 
|
|
#
?
Jun 9, 2011 23:50
|
|
|
Krenzo posted:It receives a trigger and then samples a point relative to the trigger. On the next trigger, it samples 10 fs after trigger. On the next, it samples 20 fs after trigger, and then 30 fs and so on. I don't know a whole lot about the blood and guts of an oscilloscope, but 10 femtoseconds? Light travels something like 3 microns over that timescale, and in fact reliable event timing at the fs level won a physics Nobel as recently as 2005. The Tektronix website is pretty clear that they do in fact mean 10 femtoseconds, but I'm having a hard time understanding how they might possibly accomplish that.
|
|
#
?
Jun 9, 2011 23:52
|
|
|
Captain Stinkybutt posted:Hello! I hope this is an appropriate place to posit my question. No, that's fine. The "on" button should be a "normally open" button that powers the latching relay, and the off button should be a "normally closed" button that interrupts power to everything and makes the relay switch off. Just google these terms if you need more detail.
|
|
#
?
Jun 10, 2011 03:36
|
|
|
Captain von Trapp posted:I don't know a whole lot about the blood and guts of an oscilloscope, but 10 femtoseconds? Light travels something like 3 microns over that timescale, and in fact reliable event timing at the fs level won a physics Nobel as recently as 2005. The Tektronix website is pretty clear that they do in fact mean 10 femtoseconds, but I'm having a hard time understanding how they might possibly accomplish that. I suspect there is some marketing BS in there. You would be hard pressed to find a real world clock that has phase noise low enough to have <10 fs of jitter. If I had to guess, they use a TDC (time to digital) converter to measure how long it took from the trigger condition to the edge of the sample clock. A common way to build this is to start and stop a current source that charges a capacitor, and then measure the voltage on that cap to determine how much charge was transferred. Once they know this, they can determine where each sample was in relation to the trigger event. This is called random interleaved sampling, since each acquisition begins at a random time from the trigger event. It might also be possible to phase shift the clock as the acquisition is occurring by using a DAC in the clock PLL loop, however I have no idea if this would actually work. I think the problem with the idea of "walking" (i.e. each time you sample, you shift by some small delta in time) across the phase of the signal you want to acquire is that it requires your sample clock to be phase locked with the data signal. Otherwise you're back in the random interleaved case because your trigger is asynchronous to the sample clock and you don't know where it occurred (other than between two clock edges).
|
|
#
?
Jun 10, 2011 04:23
|
|
|
If you wanted the most stable DC voltage possible from line power, would there be any appreciable [or hell, even measurable] advantage to using a zener in combination with a linear regulator IC? This is after a step down transformer and a 4 diode bridge + capacitor. In theory the less ripple goes into the regulator the less you'll see on the output, but i'm not sure if the resistor to keep the zener from exploding won't cause more problems then it solves. I suppose in the least it could be used to split the heat dissipation between the two elements.
|
|
#
?
Jun 10, 2011 16:28
|
|
|
A resistor/zener combination won't do you much good at all. A simple transistor/zener regulator has pretty poor load regulation but the ripple rejection can be pretty good. If you're trying to annihilate ripple, take a look at capacitive multipliers and depletion MOS preregulators.
|
|
#
?
Jun 10, 2011 16:42
|
|
|
Pizer posted:If you wanted the most stable DC voltage possible from line power, would there be any appreciable [or hell, even measurable] advantage to using a zener in combination with a linear regulator IC? This is after a step down transformer and a 4 diode bridge + capacitor. Good questions to answer would be: How stable do I need the voltage? How accurate do I need the voltage? How much current will I need to draw? I would just throw a big RLC on the output of your rectifier to get it quiet, and then feed that into a good LDO (make sure the input is at least 1V above the output). What are you doing with this voltage that requires it to be so stable?
|
|
#
?
Jun 10, 2011 17:18
|
|
|
sixide posted:A resistor/zener combination won't do you much good at all. A simple transistor/zener regulator has pretty poor load regulation but the ripple rejection can be pretty good. quote:What are you doing with this voltage that requires it to be so stable? Was thinking of using a filter with some inductors but wasn't entirely sure how they would interact with the "load" since it's an IC and thus active and the concept of input/output impedance with respect to active devices still confuses me :\ Seems like you can keep adding stages of capacitors to ground and inductors in series ad infinitum to get better output at the cost of spikes on the startup current Pizer fucked around with this message at 19:06 on Jun 10, 2011 |
|
#
?
Jun 10, 2011 18:56
|
|
|
While we're on the topic of power, I've got a low power (<100uA) microcontroller device that I'm working on powering off a battery and I came across something that I'd be interested in a more detailed explanation. I was seeing an order of magnitude higher power usage than I was expecting and it came down to the fact that I had an MCP1702 (3.3V) also on the board which I used when powering the microcontroller from a 5V power supply. When running normally, the board produces its own power from a AA battery and a boost converter, so the Vout pin of the MCP1702 gets 3.3V applied to it. Vin is floating, and ground is connected to ground. In this configuration, the MCP1702 is pulling ~1.4mA just doing nothing. My question is where does this come from? I can see in the datasheet that there's a voltage divider for the feedback, which should pull some power but I assume they'd use higher value resistors than ~2.4k combined. I'm also seeing the diode on the gate there, which would be conducting in this configuration, which is my more likely culprit. So being a guy who knows nothing about things, I ask, is this one of those no-brainer things that everyone knows a voltage regulator will pull a decent amount of power even if not used? Is there a quick solution like grounding Vin or pulling it up to Vout?
|
|
#
?
Jun 12, 2011 13:22
|
|
|
The MCP1702 actually has a much lower quiescent current than a lot of linear regulators you'll run into. I don't know much about how the internals work in a linear regulator, but it's not as if they draw that current for fun. Microchip put a conscious effort into keeping it low for that particular part. If you're going to be powering your application via battery you might want to use a 3.0V CR2032 or a pair of AAs or something instead of regulating down a higher voltage which will be inefficient for what you're doing no matter what. BattleMaster fucked around with this message at 16:47 on Jun 12, 2011 |
|
#
?
Jun 12, 2011 16:45
|
|
|
CapnBry posted:So being a guy who knows nothing about things, I ask, is this one of those no-brainer things that everyone knows a voltage regulator will pull a decent amount of power even if not used? Is there a quick solution like grounding Vin or pulling it up to Vout? If you're using a barrel connector for +5V, could you disconnect the regulator's common terminal when it's unplugged with a switched jack? I think you figured correctly that the clamping diode is what causes the higher current with the input floating.
|
|
#
?
Jun 12, 2011 17:02
|
|
|
Anyone hosed around with the LeafLabs Maple? Considering picking one up in the near future (taking circuits now over the first half of the summer, taking signals & systems over the second half, will be nice to finally have an understanding of how the analog side of the house works beyond v=ir so I figured it might be a good time to step up to a decent uc for prototyping 
|
|
#
?
Jun 13, 2011 02:17
|
|
|
Otto Skorzeny posted:Anyone hosed around with the LeafLabs Maple? Considering picking one up in the near future (taking circuits now over the first half of the summer, taking signals & systems over the second half, will be nice to finally have an understanding of how the analog side of the house works beyond v=ir so I figured it might be a good time to step up to a decent uc for prototyping Pretty beefy, would this be your first uC? A $4.30 LaunchPad would serve equally as well to learn the art and proper usage of C for embedded environments, not to mention giving you experience with a uC family that's ridiculously easy to breadboard and prototype, since you can get it in DIPs.
|
|
#
?
Jun 13, 2011 16:04
|
|
|
movax posted:Pretty beefy, would this be your first uC? A $4.30 LaunchPad would serve equally as well to learn the art and proper usage of C for embedded environments, not to mention giving you experience with a uC family that's ridiculously easy to breadboard and prototype, since you can get it in DIPs. I've taken an embedded control class that used 8051s and I've used an Arduino for a semester project (solar tracker, read voltages from solar panels and drove a servo and a stepper motor) and for personal fuckaround stuff (kegerator display), so I'm pretty confident with the C and assembly side of things e: I guess part of the reason a convenient ARM package interests me besides the nice bump in power is that my arch class last semester got me Pretty Enthused with mips asm and its simple pipelined datapath and convenient instruction set, which ARM is relatively similar to afaict? Blotto Skorzany fucked around with this message at 18:34 on Jun 13, 2011 |
|
#
?
Jun 13, 2011 18:31
|
|
|
Otto Skorzeny posted:I've taken an embedded control class that used 8051s and I've used an Arduino for a semester project (solar tracker, read voltages from solar panels and drove a servo and a stepper motor) and for personal fuckaround stuff (kegerator display), so I'm pretty confident with the C and assembly side of things I had a quick look at it and it seems interesting. The only main problem is it's an amateur board by a small company. There doesn't seem to be a massive community either. It's quite a neat idea having it's own IDE and toolchain but they don't look that mature. There is nothing more annoying trying to work out what a bug is when you can't trust the development tools you are using. (I haven't used that board so I may be completely wrong). An alternatives is to get a STM32 discovery board http://www.st.com/internet/evalboard/product/250863.jsp . They used to give them out free, so you may want to look around for that. (If you pay for postage from the UK I'll send you one). The only downside is you are going to having to use a demo IDE. Usually 32kb size limited or have annoying pop-ups.
|
|
#
?
Jun 13, 2011 21:46
|
|
|
Otto Skorzeny posted:e: I guess part of the reason a convenient ARM package interests me besides the nice bump in power is that my arch class last semester got me Pretty Enthused with mips asm and its simple pipelined datapath and convenient instruction set, which ARM is relatively similar to afaict? Oh, and it's only $27
|
|
#
?
Jun 14, 2011 03:12
|
|
|
Yeah, one of the PIC32 Starter kits from Microchip may not be a bad idea. 3 tiers of kit, with the highest end one giving you Ethernet and USB and all the jazz. Off hand I think they all sport the PIC32MX795, so you get every peripheral, and you can downscale accordingly for your final "product".
|
|
#
?
Jun 15, 2011 19:07
|
|
|
Anyone ever actually built things around the MSP430? I pulled the dev boards that TI made for them a while back out of storage, and I'm considering making something with one that specifically uses it's hilariously low power consumption. Specifically, I'm hoping to make a programmable digital timer, both to improve my coffee brewing skills, and to give me something to do. It would be a pretty simple timer set in minutes + seconds, with my purposes requiring only 3 digits (ie, <10 minutes total). Time would be set using a rotary encoder (ideally one with detents to give some tactile feedback)], with a single button to start or pause the timer, and a piezo buzzer to signal the alarm. The most important aspect would be the auto-shutoff and the (invisible to the user) power saving stuff. The device + display would "shut off" after N seconds of no activity (when the timer isn't counting, obviously), and could wake up on a pin change interrupt from either the button or the encoder. If I wanted to be even crazier. Just as important as this, though, would be what is happening while the timer is counting down. Since the display only needs to change once a second, I can potentially sleep during most of that, and only wakeup to shift digits. This, of course, relies on a built in RTC (or just a timer counter connected to a watch crystal), and a non-crappy uC that can keep outputs set while in sleep mode (unless I use a fancy e-ink/e-paper display, or one of the other persistent display techs). The MSP430 is great in that it keeps its auxiliary 32khz clock operating in the 2nd to lowest power sleep mode, while still consuming practically no power, and is able to wake up and start the regular clocks within like 6us. With properly written code, I can sleep the device most of the time while it's operating, and should be able to run it off a watch battery for a long rear end time. So, for anyone who has worked with MSP430's, are there any good resources, tutorials, and example projects you'd recommend? While I always do end up reading this datasheets, it's so much easier to start with other demonstration stuff and work from there (especially since it acquaints me with the peculiarities of the particular c compiler being used).. Also, does anyone have any good suggestions for resources, books, or anything else on programming for low-power devices? It's rare that I see a hobbyist project give any thought to power consumption--it's a terrifying world where everything has a wall wart and an antique linear regulator (not even LDO), sleep modes are terrifying magic that breaks projects, and where the streets are paved with superfluous quantities of flash memory.
|
|
#
?
Jun 15, 2011 19:29
|
|
|
Slanderer posted:where the streets are paved with superfluous quantities of flash memory. I've done most of my uni course material with MSP430s. They are crazy little processors; beware that all of the low-power options might make configuring peripherals a little harder than you are used to, but it's not that bad. Check out the 4xx series for ones with built in LCD controllers. I've used IAR a bit (which is free) but I like crossstudio better (not free, as I recall). I can pm/email you one of my homework assignments from last year that uses a timer to generate an interrupt every 2 ms, has a code listing and a real quick study of power usage (actually, awake time vs. sleep time, but that's the idea). Let me know if you have any specific questions.
|
|
#
?
Jun 15, 2011 22:12
|
|
|
Slanderer posted:Also, does anyone have any good suggestions for resources, books, or anything else on programming for low-power devices? It's rare that I see a hobbyist project give any thought to power consumption--it's a terrifying world where everything has a wall wart and an antique linear regulator (not even LDO), sleep modes are terrifying magic that breaks projects, and where the streets are paved with superfluous quantities of flash memory. You're right that there is a serious lack of low power design in the hobbyist world. Unfortunately, I don't know of any specific low power tutorials - every one seems to assume that power supplies are black magic, power is infinite, and that all devices will operate at room temperature. Hence you get people using a 3 terminal regulator from the 1970's to go from 12V -> 3.3V. poo poo, even at work I've seen designs where someone was using a resistor/zener combination to regulate a high end DAC (fortunately it was a customer design, and not something someone internal produced). For power supplies, I would begin looking at app notes from Linear Tech and National/TI. You're going to want to use a power optimized converter, and I know Linear Tech has some that have extremely low quiescent current (LTC3388). I'm sure that National/TI also have something similar. The downside is that the main driver for low power is the portable device market, so the parts tend to be fairly small. You're going to want to run all parts at the absolute lowest voltage they can go, since voltage is directly proportional to power consumption. You're also going to want to run your main system clock as slow as you can to reduce switching losses. Don't forget that you still have to keep time, so your clock needs to be somewhat accurate. Not sure how good the 32KHz RTC on the MSP430 is, but if it's stable enough I would use a divided version of it at 1KHz (or less) if possible. For absolute lowest power, everything in the design will have to be considered - even pull up/pull down resistors can contribute a sizable amount of power loss (driving 3.3V in to a 10K is 1mW!). edit: Apparently the MSP430 can only do divide by 1/2/4/8 on clocks, so with a 32kHz crystal, the best you could get is a 4kHz clock. That's probably slow enough - according to the datasheet I found, that should be on the order of 5uA@2.2V (assuming no outputs are drawing current). SnoPuppy fucked around with this message at 22:40 on Jun 15, 2011 |
|
#
?
Jun 15, 2011 22:25
|
|
|
Delta-Wye posted:I've used IAR a bit (which is free) but I like crossstudio better (not free, as I recall). I can pm/email you one of my homework assignments from last year that uses a timer to generate an interrupt every 2 ms, has a code listing and a real quick study of power usage (actually, awake time vs. sleep time, but that's the idea). I'd be interested in that. I picked up a couple of Launchpads in order to teach myself uC programming but I'm a bit stuck on the "what the hell do I do with this" part. Looking at more projects might kick my butt in that direction
|
|
#
?
Jun 15, 2011 22:35
|
|
|
|
| # ? May 8, 2024 23:33 |
|
|
SnoPuppy posted:You're right that there is a serious lack of low power design in the hobbyist world. Unfortunately, I don't know of any specific low power tutorials - every one seems to assume that power supplies are black magic, power is infinite, and that all devices will operate at room temperature. Hence you get people using a 3 terminal regulator from the 1970's to go from 12V -> 3.3V. MSP430s make it easy. As long as the MCLK is pretty low, you can run them down to absurd voltage levels. They should run comfortably with the VCC pin at 1.8V or so at 4 MHz (depending on the part) according to the datasheet, so you might be able to sneak it a bit lower. They are tolerant up to 3.6V according to the datasheet; I've overclocked them and ran them at higher VCCs no problem. Easy power solution is two 1.5V cells in series. Assuming your LCD would be happy with that, you don't even need the regulator.  Set up ACLK to run off of the 32KHz, so you should get 32768 ACLK pulses per second. You can configure a timer to count ACLK pulses up to a certain number and then generate an interrupt (32k, obviously). Inside the interrupt, you simply do your timekeeping and update the display as required. The MCU is pretty much off except for when it's in the interrupt, when it can easily run off of the DCO at the default speed. EDIT: What is the easiest way to share a PDF? Delta-Wye fucked around with this message at 23:44 on Jun 15, 2011 |
|
#
?
Jun 15, 2011 23:41
|
|