Post by thread:[PIC]: Operation of PICs from low voltage

Still coming to grips with this new fangled system - gotta put : after PIC they tell me to ensure all the pure PIC people see it .... mumble, mumble ... try again. QUESTION Based on actual experience. - What is the lowest Vdd at which a 16F84 will operate (including occasional EEROM read/write) reliably using an RC clock at about 1 MHz? Yes. I know what the data sheets say. Yes, I know the answer is try it and see. Yes, I know the results will vary. No - I don't (yet) know what other people have tried and found works. BACKGROUND: Based on random aged 16F84 data sheet: - The std 16F84 is specified to operate at a minimum Vdd of 4.5 V - The 16LF84 is specified down to 2 volts I am building a small 16F84 device where portability and battery life is important (wearable aid for a disabled person), Essential aim is to drive a beeper subject to certain constraints - irrelevant to following. Device will be carried in top pocket / worn on cord around neck / clipped to clothing so user but hopefully few others can hear the occasional beep. Another variant may drive a pager vibrator instead. I am running a 16F84-10 (10M parts happened to be in parts drawer) . This runs happily with physically small 3.6V 80 mAH NiMH battery and RC clock (33K, 18pF) at 1MHz osc = 25o KHz PIC basic instruction rate. This is very adequate for my task. At 3.8V it draws 250 uA At 2 volts it is still running and clock frequency has not moved markedly and draws 100uA At 1 volt it is STILL running but clock is MUCH slower (~ factor of 10?) and draws under 10uA (!) I don't need to run at 1 V but ability to run at 2V is interesting. I haven't tried eerom read/writes yet at 2V. What have other people found about low voltage operation of "standard" parts. Is this perhaps another example of Microchip 'select on test"ing parts and many std parts are in fact capable of low voltage operation? Yep - I could (try to) buy some 16LF84's but ... TIA Russell McMahon

Only experience I've had is running a PIC16F84 on two alkaline C cells at 4 MHz. Ran fine down to 2.6 or 2.7 volts, whereupon it had trouble starting the oscillator, but I didn't spend any time tweaking the oscillator, because by that time the batteries were low enough that the rest of the circuit wanted fresh batteries(not pic-related). This was crystal controlled, likely RC oscillator will run at lower voltages with no tweaking just like your experience showed. Also it was not the low-voltage part, don't know if there's any real difference in silicon. For those interested, it was a radio-controlled remote "talkback" switch that generated MIDI for a Yamaha O2R mixer. I used a wireless doorbell and disabled the boinger. Put a 16F84 in the receiver to generate MIDI note-on and note-off(the Yamaha mixer accepted all sorts of commands through MIDI and talkback was one of them). Then the "producer" got to hold the doorbell button and could push it whenever he wanted to talk to the "talent", saving him from having to ask the engineer to push the talkback button on the mixer. heh. Birthday gift for my nephew the engineer. -Bob

At the old company, we were running PIC's @ 2V and below before anything was spec'd that low. The xtal oscillators were very finnicky(?), but the RC would always get up and fly (pretty much it has too by design). We did see RAM problems when the processor was cold (<32F). Basically, bits flipped on their own accord. Processors we were using were the C54, L84, and the LF84. Had success with all of these (>85% yield). LV parts are what is standardized on now, 100% yield. LP oscillator leaves alot to be desired at 5 volts in term of startup, wouldn't even try it at 2 volts!!! Most parts listed above would fly down to 1.8V, but I would be nervous about the bit flipping issue.

One surprising experience I had was to find that the 16F84 would run happily with a supply voltage of zero volts!, yes thats right zero volts, on further inspection of the circuit I found that the 16F84 was leaching a supply from the I/O pins (the external circuit was independently supplied) Even though I have had years of experience with other microcontrollers the PIC chip still surprises me with it's flexibility, robustness and other unexpected qualities (I am amazed to hear that it will run down at 1V). On Wed, 7 Jun 2000, Bob Blick wrote: {Quote hidden}> Only experience I've had is running a PIC16F84 on two alkaline C cells at > 4 MHz. Ran fine down to 2.6 or 2.7 volts, whereupon it had trouble > starting the oscillator, but I didn't spend any time tweaking the > oscillator, because by that time the batteries were low enough that the > rest of the circuit wanted fresh batteries(not pic-related). > > This was crystal controlled, likely RC oscillator will run at lower > voltages with no tweaking just like your experience showed. > > Also it was not the low-voltage part, don't know if there's any real > difference in silicon. > > For those interested, it was a radio-controlled remote "talkback" switch > that generated MIDI for a Yamaha O2R mixer. I used a wireless doorbell and > disabled the boinger. Put a 16F84 in the receiver to generate MIDI note-on > and note-off(the Yamaha mixer accepted all sorts of commands through MIDI > and talkback was one of them). Then the "producer" got to hold the > doorbell button and could push it whenever he wanted to talk to the > "talent", saving him from having to ask the engineer to push the talkback > button on the mixer. heh. Birthday gift for my nephew the engineer. > > -Bob > ----- John P. Brown spam_OUTJ.P.BrownTakeThisOuTbradford.ac.uk ---- \ --- Witty remark goes here --- / --------------------------------------------------

Of course it's not the IO input circiut leaking - it's the pair of static protection diodes. One conducts from 0V to IO pin, the other from IO pin to Vcc. These diodes can be used for other things like putting mains directly into a PIC simply by current limiting. You can tie 240VAC (UK) into a pic that's battery powered. Tie Neutral to 0V and LIVE to IO pin via a 10M resistor. This can be used to synchronise the PIC with the phase of the AC and to derive an accurate 50Hz time base (very accurate over a period, less so over short times). The 10M resistor limits the current and the static prtection diodes leak the small current through to the battery. First time you rig it up take a deep breath before switching on! | -----Original Message----- | From: JP.BROWN [.....J.P.BrownKILLspam@spam@BRADFORD.AC.UK] | Sent: 08 June 2000 08:48 | To: PICLISTKILLspamMITVMA.MIT.EDU | Subject: Re: [PIC]: Operation of PICs from low voltage | on further inspection of the circuit I found that the 16F84 | was leaching a | supply from the I/O pins (the external circuit was | independently supplied)

Yup I had one running with no supply as well. I was watching what happened to the output on a scope when I removed the battery supply. The signal just kept on going! (way off freq and greatly attenuated) but still measurable. MikeW >>> .....J.P.BrownKILLspam.....BRADFORD.AC.UK 06/08/00 09:47AM >>> One surprising experience I had was to find that the 16F84 would run happily with a supply voltage of zero volts!, yes thats right zero volts, on further inspection of the circuit I found that the 16F84 was leaching a supply from the I/O pins (the external circuit was independently supplied) Even though I have had years of experience with other microcontrollers the PIC chip still surprises me with it's flexibility, robustness and other unexpected qualities (I am amazed to hear that it will run down at 1V).

Any relation of this one with your previous note about blowing up the JW part? Andy Jilles Oldenbeuving <EraseMEjillesspam_OUTTakeThisOuTRENDO.DEKOOI.NL> on 06/08/2000 01:12:25 PM Please respond to pic microcontroller discussion list <PICLISTspam_OUTMITVMA.MIT.EDU> To: @spam@PICLISTKILLspamMITVMA.MIT.EDU cc: (bcc: Andrew Kunz/TDI_NOTES) Subject: Re: [PIC]: Operation of PICs from low voltage >First time >you rig it up take a deep breath before switching on! Where i work they let pic's run off 220V AC, trough a couple of resistors and an X2-type capacitor, or capacitive power supply. Jilles