Fullscreen HTML5Med Res (??MB)HTML4

MORE FUN WITH THE PICAXE – PART 8 To sleep, perchance to dream (or nap) . . . (and PICAXE datalogging too!) Quite aside from all their other benefits, the inbuilt micro-power standby features of Picaxes – send­ ing them to sleep – can greatly enhance battery life. A new use for apparently “dead” batteries, perhaps? by Stan Swan G iven the prevalence of portable devices (refer June “SILICON C HIP ” editorial!) this alone should win favour for projects such as torches and data loggers used away from a mains supply. There are four distinct commands, and several code techniques, to organise this. PAUSE – introduces a program delay in milliseconds. Example: pause 100 = 100ms    Maximum delay is 65535ms (a little over one minute) with about 1ms overhead. Of course, we’ve used pause a lot in earlier articles for LED flashing etc. No power saving. WAIT – equivalent to pause, but with larger units Example: wait 10 = 10 seconds    Up to 65.5 seconds wait possible, but again no power saving activated NAP – Enters a low power short period mode. Example: nap 3 yields a 144ms delay.    The eight nap period values (060  Silicon Chip 7), yield a duration given by the formula: Delay (in ms) = 2^time value x 18.    Hence Nap 0= 18ms, while nap 1= 36ms, nap 2= 72 ms , nap 3 = 144ms etc. Nap 7 = 2304ms (2.3 seconds). SLEEP – Also enters a low power long period mode (2.3 seconds units). Example: Sleep 10 = ~23 seconds. Although of high accuracy, resolution is lower and overall times can amount to ±1% deviation (perhaps up to 30 seconds drift in an hour.) The maximum sleep value, of 65535 (being 256 x 256) extends to days, but could be temperature dependent. Hence don’t set your alarm clock by this… Other delays Unusual delays, not catered for above, can perhaps be organised by “do nothing” nested loops that may run to minutes. Hence a 100 x 100 = 10,000 count, could involve code looping inside loops and be set up perhaps   for b0=1 to 100: for b1= 1 to 100: next    b1: next b0 A further alternative may be to exploit the serin command, which patiently awaits the arrival of serial data bytes – perhaps from a linked PC timer. Rev. Ed’s AXE033 LCD display in fact includes a DS1307 real time clock chip offering this feature, allowing precise intervals to be set. OK – you know all about pause, so let’s first put nap to work in yet another LED flashing circuit. Another one? As we’ve mentioned before in this series, pulsed or flashing LEDs attract attention, help identify the source and also save battery power. And the PICAXE is perfect for doing it! The classic 1970s LM3909 IC enjoyed decades of use in just this field but that was – well – the 20th century. If the flash rate is fast enough (>20Hz or so) human persistence of vision comes into play as well and www.siliconchip.com.au the pulsing light “looks” to be just a steady source. Hence it’s “win win” – appearing to be on but saving significant power –and with today’s ultrabright white LEDS, battery life can be hugely extended – an important issue in emergencies or even lesser developed countries rural lighting. Take note however – flash rates around 7-10Hz are particularly irritating (they’ve even been used for riot control) and in extreme cases may bring on epileptic fits. Power down Rather than just have a short pause between such LED flashes, it’s maybe better to use the brief power down features of nap. When devices are being worked hard, such as the “over driven” white LED here, this brief cool-down spell may help to ease the thermal stress on the LED (normally limited to ~30mA but capable of withstanding 100mA pulses). Recall that readadc command from the “Door Minder” article (SILICON CHIP March 2003)? This has been further extended here so that the presented resistance from a 500kΩ pot selects the flash sequences. Such resistances could have been set with a multi-position switch and assorted resistors, or even maybe a stepped push switch but the pot simplifies things nicely. Note the generous program comments! (Above): the circuit diagram and protoboard layout for the first part of this month’s PICAXE series – using the various “slumber” commands to save battery power in the LEDs. This is used in conjunction with the “LEDNAP” program overleaf. As usual, the photo is just slightly different from the protoboard layout above (we’ve moved some components for clarity in the drawing). The pulsed white 5000mCd LED (from Jaycar) used here stands out like a lighthouse at night and is easily visible several kilometres away. www.siliconchip.com.au September 2003  61 In contrast to catnaps, sleep is intended for some serious PICAXE downtime resting! During such a sleep, power demands were found to drop to the 100µA range, (although brief wake up surges have been reported), hence almost offer a new use for otherwise “dead” batteries. The 2.3 second unit (being the upper value of nap7) implies a super nap is invoked. The maximum sleep LEDNAP.BAS value (65535), extends to some days and sequential sleeps could stretch to (maybe) months. For portable work, battery life may ultimately limit the program duration but a 1000mAh NiMH should last 10,000 hours (over a year) which may compare with its shelf life! Or a small photo-voltaic panel could keep a rechargeable pack trickle-charged. These aspects are naturally (Also downloadable from: www.picaxe.orconhosting.net.nz/lednap.bas) ‘LEDNAP.BAS program for September “Silicon Chip” Picaxe-08 article. Ver 1.0 12/7/03 ‘Potentiometer acts as multi position switch! Stan. SWAN = s.t.swan<at>massey.ac.nz ‘Values as set here allow 7 different flash types to be dialled up via the 500k pot ‘Much tweaking still possible-better reading b0 ranges,alter pot(Log?),R value etc ‘White LED (10mm ?) via BC547 with 470 Ohms from its base to Picaxe pin 2 via “NAP”. ‘Actual nap value depends on application - steady light,standby, rescue beacon etc ‘Even nap 0 has distinctive ~20 Hz flicker = ideal bike front for attention getting? ‘NB ~10Hz flicker is most irritating to many observers - may cause epileptic fit?! ‘On 3AA 4.5V supply & 470 base R,typical DSE Lux meter values (pulsed of course) ~ ‘nap 0=35mA 340 Lux, nap 1=22mA 216 Lux, nap 2= 13mA 120 Lux, nap 3= 8mA 70 Lux ‘nap duration = 2^period x ~18ms , with period values 0-7 ( rolled over if beyond ) ‘nap 0 ~18ms, nap 1 ~36ms, nap 2 ~72ms, nap 3 ~144ms, nap 7 ~ 2secs ‘Inbuilt loop o’head of course distorts M/S ratio. Red LED direct driven from pin 4 ‘Program download from = www.picaxe.orcon.net.nz/lednap.bas Comments (‘) optional ‘———————————————————————————————————————— ledflash: ‘main adc read to set nap time or dim with pwm etc readadc 1,b0 ‘ADC read pin1 -via 500k pot(Log?) & 47k V.divider b1= b0/22 'divide returned “nap” value so fits in 0-7 range if b1=0 then redflash '“08” readadc values are in 16 blocks 0-160,11 wide if b1=1 then redpwm 'Could use as is,but division OK for just 7 values if b1=7 then beacon '1/2 Hz beacon mode nap 7 forced via b2 variable '———————————————————————————————————————— whiteflash: 'routine for normal adjustable white flash rates b2=b1-2 'tweak returned b1 rates,since too short as divided    high 2:pause 10:low 2 'LED pulse - somewhat overdriving via transistor nap b2 'brief snooze to allow LED/BC547 cool down ! goto ledflash 'return to main pot. reading procedure '———————————————————————————————————————— redflash: 'red led at pin 4 flash - rear bike light style high 4:pause 10:low 4:nap 7 'pulsed ~ every 2 seconds - adjust to suit needs goto ledflash 'return to main pot. reading procedure '———————————————————————————————————————— redpwm: 'red led attractive pulsing effect for b3=0 to 255 step 2 'loop so red LED has pleasing brightness increase    pwm 4,b3,1 :next b3 'PWM pin 4 LED one cycle at increasing pulse width     for b3=255 to 0 step -2 'loop to fade led out    pwm 4,b3,1 'PWM pin 2 led one cycle at decreasing pulse width next b3:pause 300 'led displays a pleasing “heartbeat” effect ! goto ledflash 'return to main pot. reading procedure '———————————————————————————————————————— beacon: 'battery life prolonging (weeks ?) beacon flash     high 2:pause 10:low 2 'brief led pulse - via transistor.Approx 1mA draw nap 7 '~ 2 sec delay between pulses goto ledflash 'return to main pot. reading procedure 62  Silicon Chip an issue for portable applications such as our data logger. What – a Picaxe data logger? Yes – although just a baby, the “08” has a 64 bytes non volatile memory (EEPROM). Any data (of values 0 –255) can be stashed away here (in ascending locations 0,1,2. – 64), although programs (stored downward 128,127 etc) share the same RAM and can be overwritten if care is not taken! Key commands used are write and read, which store or retrieve during a program run, much as you’d keep a pencilled scratchpad of (say) items during a stocktake. The EEPROM (Electrically Erasable Programmable Read Only Memory) command is intended to “pre store” values, either ASCII or data, so they are available once the program gets to work. No battery backup is needed to hold this data, since the memory is Flash RAM based. Here’s a simple code snippet example – EEPROM (13,7,19,69) ‘ pre stores bytes 13,7,19,69 at locations 0,1,2,3 if free write b1 ‘ stores b1 value at memory location staring from next available (4 here) read b0 ‘ read b0 retrieves this value for program use Due to the “Von Neumann architecture” nature of the 08, its memory can’t be easily extended, sadly ruling out use of cheap I2C RAM chips. Even though 64 bytes may seem too trivial to exploit, it’s proven ideal for recording voltage divider network values via the “08” readadc command. Don’t get too excited - only low resolution is possible, and an upper limit value of 160 applies, but some 16 unique values may be detected and stored. Data loggers of course allow valuable monitoring of such “real world” values as temperature, earthquakes, wind speeds, voltages, traffic, pH etc - maybe too tedious or hazardous for human recorders. Educators may quibble but humans are arguably better employed than just watching dials and writing down numbers and additionally we’re often devious, lazy or deceitful – perhaps “snoozing” when a key value arrives! Direct computer analysis is eased with machine gathering too. OK, you’re convinced – but aren’t such data loggers costly? Read on! The deceptively simple Picaxe-08 www.siliconchip.com.au Part two for this month: data logging with the PICAXE-08. Here’s the circuit diagram and protoboard layout. circuit (at right) and program (overleaf) uses a 100kΩ/25oC NTC thermistor as a temperature sensor in simple automated application, with a sleep period initially set to take readings every minute for an hour. Since only 64 program bytes were available (the other 64 being used for data of course) refinements were limited, and the readout technique a compromise. But as the Excel graph shows, very distinctive thermal environments were easily logged, and applications abound even as it stands! Mmm – how about checking the heating/cooling rate of a spa pool as a guide to its insulation performance – waterproof the sensor of course. Monitoring an air conditioner - the period to bring rooms to a desired temperature perhaps? Replace the NTC with an LDR and note illumination changes – room lighting /security etc. Check actual voltages of discharging battery packs for relative performance? Incidentally, we’ll be extending this design later with a Picaxe 18A (and 18X – due late 2003), offering higher resolution, multi-channel and more memory, so stay tuned. To overcome the “08” lo-res limitation, an elegant ramping solution has just been suggested. This sets up the PWM command to feed incremental pulsed PWM bursts into a capacitor OK, this one is really different because it is built on a mini protoboard, whereas the diagram above is based on the standard protoboard we have used throughout this series. Either is perfectly suitable (especially if you already have one or other!). That’s a 4V 20mA PV panel at top – rescued from a budget garden lamp. An Excel spreadsheet plot of the data obtained from our PICAXE logger, in this case being used as a temperature logger. It shows the readings over one hour in various locations – just to give you some idea of what the PICAXE logger can be put to! www.siliconchip.com.au September 2003  63 DATALOG8.BAS (Also downloadable from: www.picaxe.orconhosting.net.nz/datalog8.bas) and NTC/LDR until the preset digital threshold is reached, at which point values are recorded and “serout” passed to be read (or graphed) on an attached screen. This naturally feeds live values rather than logged ones, but implies up to three channels of hi-res data (values 0 – 255) could be simultaneously gathered and PC stored. Check www.picaxe. orconhosting.net.nz/datagath .bas for the full program. ‘DATALOG program for September “Silicon Chip” Picaxe-08 article. Ver 1.02 24/7/03 ‘Use with attached 100k thermistor etc pin 1. Via Stan.SWAN => s.t.swan<at>massey.ac.nz ‘When “08” powered up,any prior stored EEPROM values sent as pin 4 serial port data ‘Display this gathered data via any terminal program -LCD,BananaCom,F8,StampPlot etc. ‘If saved via a terminal program,the “.csv” data can of course be Excel graphed too ‘NB-Gives you 30secs to turn unit OFF before fresh storage begins & thus wiping old! ‘**** CARE - BE PROMPT ! REPROGRAMMING/RELOADING “08” TOTALLY WIPES DATA TOO ****. ‘As set up logs temp in 0-30 C range every min for 1 hr.WAIT more accurate than SLEEP? ‘Tweaking V divider network Rs may allow narrower temp range.Alter 47k to 100k maybe ‘Picaxe 18A should run this OK too, but give high res readings & store 256 values.Yah! The circuit ‘———————————————————————————————————————— Son of a Picnik box! ‘Picaxe data storage value range 0-255,although only to 160 via “08” readadc of course. Until now, all our circuits have ‘EEPROM builds up from location 0,but program builds down from 128. Just 61 bytes used been based on a full-sized solderless protoboard. And that’s what the dia‘Basic EEPROM syntax is ex. EEPROM (13,7,19,69 ) where bytes 13,7,19,69 EEPROM stored gram overleaf shows. (We figure that ‘at locations 0,1,2,3. Code use READ & WRITE to access this data at these spots maybe by now anyone experimenting with ‘To retain program simplicity & maximise number of readings, data can not be viewed as PICAXEs would have one of these ‘gathered.In practise this should not be an issue, since data logger likely to be used protoboards). ‘stand alone/ outdoors etc, then retrieved to display values back at an indoor PC etc. But as you can see, there is an awful ‘N.B.SLEEP not 1:1,as unit=2.3 secs.Elapsed times ~x2 expected.By trial SLEEP 25=1 min lot of unused real estate on that board. ‘Stored values are non volatile - thus no need battery backup connection once gathered Let’s look at a couple of alternatives. ‘PIC makers (Microchip) say data retained in EEPROM >40 years unless overwritten.Bravo A one-third size breadboard (sold ‘Typical (baby Wish board!) hardware setup pix=> www.picaxe.orcon.net.nz/datalog8.jpg here in NZ as the Global EXP-325) just ‘With sample Excel graph resulting (1 hour run)=> www.picaxe.orcon.net.nz/datalog8.gif fits everything, (including the 3.6V ‘ ****** Download this program via => www.picaxe.orcon.net.nz/datalog8.bas <= ****** NiCd battery – rescued from an old PC ‘———————————————————————————————————————— motherboard) we’ve become familiar ‘ASCII art schematic Typical temp. readings/readadc values with. It’ll snap shut in a sealed food ‘ - - - - - - -ve rail 0 Celsius 11 container for outdoor or submersed applications too, and of course a sol‘ Piezo _____ Pin | | 4 21 dered version can be easily made too. ‘ Pin 0 ___ | 2 LED 47k 8 32 ‘ ——— | |_______| | 12 43 Something more permanent? ‘ |Picaxe| |_______________| 16 53 So you want a permanent version? ‘ | 08 | Pin 1 ADC | 20 64 Time to transfer your circuit to, say, a ‘ ——— 100k 24 75 PC board? ‘ |______Pin 4 NTC 28 86 etc During recent time with the “08”, ‘ ||| serout | 3-5 V ‘Usual 3 wire + + + + + + +ve supply NOTE - Temp values approximate ‘prog.input & may need better calibration ‘———————————————————————————————————————— ‘READ/PLAYBACK ROUTINE serout 4,n2400,(12,”Datalog “) ‘ASCII values 10=CR, 12=FF(=cls),13=LF, 44=comma for b0= 0 to 63 ‘stored data values readout to terminal or LCD read b0,b1 ‘polls & reads out stored eeprom values ( .csv) serout 4,n2400,(44,#b1) ‘comma,then value <at> pin 4. LED to show data too? next b0 ‘read next stored EEPROM value out wait 30 ’30 secs “reading” delay -modify if too short etc ‘———————————————————————————————————————— ‘WRITE/DATA LOGGING ROUTINE for b0= 0 to 63 ‘begin 64 data readings at time set by SLEEP sound 0,(75,10) ‘Beep to alert data logging commencing pulsout 2,500 ‘brief flash from pin 2 LED indicates datalogging readadc 1,b2 ‘b2 has 16 blocks 11 wide (range 0-160),so 21 etc write b0,b2 ‘sequentially write values to EEPROM locations sleep 25 ’25x2.3secs ~1 min delay (+/- 1% )-alter to suit next b0 ‘Ex. Sleep 782 yields 64 x 1/2 hr =32 hrs data! Here’s a stylised version of how a ‘Data gathering stops when 64 readings taken permanent hookup on Veroboard or ‘———————————————————————————————————————— similar might work. 64  Silicon Chip www.siliconchip.com.au References and parts suppliers . . . (also refer to previous months articles) 1. Suitable NTC thermistor (R1895 100kΩ/25C), and PC prototyping board (H5605), assorted resistors, transistor & 500kΩ pot: Dick Smith Electronics www.dse.com.au 2.White LEDs, battery (3.6V 70mAh SB-1609 ~ A$9) and small piezo sounder: Jaycar Electronics www. jaycar.com.au 3. Humans perceive pulsing lights under “flicker fusion frequency” laws (Ferry-Porter etc) http://webvision. med.utah.edu/temporal.html 4. Mini EXP-325 protoboard (~A$5): www.globalspecialities.com 5. “Dataloggerama” insights are gained at www.rogerfrost.com 6. Picaxe supplies via MicroZed www.picaxe.com.au, with program listings and diverse links www.picaxe. orconhosting.net.nz 7. Thanks to Eltham Tech Centre and Andrew Hornblow (Taranaki, NZ) for DIY ideas. and building on feedback from students, older hobbyists and emails, it’s remained apparent that soldering should be the last thing you do when investigating such electronic circuits. Maybe your design is unstable, draws too much current, has wrong colour LEDs or is even (duh!) too big for the box! Once soldered up these aspects may be a nightmare to alter without tedious redesign or desolder-ing, risking ESD or heat damage. Solderless “breadboard” proto-boards remain ideal for rapid PICAXE development (I’ll use them again with the 18A), even though some of the tweaking can be of course via software. Naturally, once your design is finalised, if you want to keep it then it’s standard practice to produce a soldered version. If full PC board design costs and time are not justified, consider several rapid cost-effective approaches instead. 1. Many suppliers sell a pre-drilled 1/10th inch grid board, with copper tracks matching a normal breadboard. It’s almost “paint by numbers” to move items from one to the other and solder as you go, with the final result looking convincingly professional. 2. VeroBoard (and all its variations). The hobbyist’s standby of course and is easily scored and cracked to size. Considerable lateral thinking by NZ “BrightSparks”, with minds refined by crosswords, it transpires, have produced soldered Vero versions to suit many “08” circuits. In the most general form (which even includes a programming socket) only a tight block of track needs removal but a grander version even provides for driver transistor placement. IC header strip can be soldered in for flying leads too. Recommended ! 3. Copper clad “spider board”. Deep cuts made quickly and neatly across the copper produce ten sections (yeah, we know spiders have eight legs, not ten!) that suit top surface mount soldering of components and an IC socket, or even edge clipping of leads. It perhaps best suits junior users who SC need WYSIWYG insights. NEXT MONTH: All new: The Picaxe 18A Similarly, here’s a “Spider Board” which you can make yourself with a sharp knife and scrap of PC board. www.siliconchip.com.au Our baby Picaxe zero eight, Will now take a rest, so its mates Can show off their bytes, With more circuit delights, And greater PIC magic create! September 2003  65