Is the ATtiny13's operating current *that* low?

Go To Last Post
17 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Someone pinch me if I am dreaming, but is the ATtiny13 datasheet
realistic and does it really draw so little current when run off
its internal, 128khz clock? I am working on a little controller, just
for fun, and am not sure I should bother with a voltage regulator. It
appears a 'tiny13 could probably draw all the power it needs through
the controller's inputs.

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Current consumption is pretty well linearly related to clock speed. If in doubt, make some measurements yourself! Just because the cpu is low power, doesn't mean it will tolerate poor power supply. Depending on your circuit configuration, a regulator (be it an ic or zener diode)may still be necessary.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I should have phrased the bit about the regulator a little differently. The ATtiny13's power consumption appears to be so low that there doesn't seem to be any need for the usual power connector pin and three-pin regulator. I assume everyone considers how to arrange and protect the supply and inputs.

I am just curious before I build and measure, if anyone has confirmed the datasheet's typical current values. The last time I got to apply an 8-bit processor consuming micro- rather than milliamps at useable clock rates, there were only two possibilities, and one of them emulated a PDP-8!

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm still unclear. You propose to run an app with the Vcc pin of the Tiny13 unconnected and floating? If not, what do you intend to hook up there?

I cannot speak for the Tiny13 directly. All values for power draw that we checked for the Mega48 were very close to those stated in the datasheet.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
I'm still unclear. You propose to run an app with the Vcc pin of the Tiny13 unconnected and floating? If not, what do you intend to hook up there?

I plan to connect the Vcc pin, although perhaps there's some neat advantage to floating it I have never considered. :)

I was just working through the details for including a micropower regulator and realized that it might actually not do anything other than consume current since there would appear to be enough leaking in through the board's inputs to power the ATtiny13. I was already planning to power the '13 through its load -- an incandescent lamp with one side tied to +12V and a MOSFET grounding the other -- but I may want to do things a little differently if the datasheet's consumption values are realistic.

Thanks for confirming they are for the Mega48!

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I was already planning to power the '13 through its load -- an incandescent lamp with one side tied to +12V and a MOSFET grounding the other --

If you were not planning to float Vcc, then what >>were<< you going to connect it to? ??

Anyway, a floating Vcc would violate chip maximums if it floats too high. I'm also very confused about your concern for microamps when the app includes a 12V incandescent light.

True, a regulator isn't needed if you have a steady power source within specs, like a battery.

Does your app have any timing requirements? If so, a stable clock source is needed, and the internal oscillators may vary widely with changes in supply voltage.

I'll leave "load powering" to those that know more, but it doesn't sound like a good idea to me.

You might want to consider something like charging a supercap when your light is on, then live off that when you are idle and there is no juice to power you, if that is a summary of your app.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can trust the datasheet's info.

---
Formerly Torby. Stitch626 just seemed a more descriptive nicname.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
If you were not planning to float Vcc, then what >>were<< you going to connect it to? ??

Not to worry. I am pretty orthodox about power pins.

Originally, I was going to connect Vcc to a voltage regulator, the regulator to a transient suppression stage and the suppression stage to an instrument cluster indicator lamp. (The controller is going into a car.) However, since an ATtiny13 will apparently draw so little current I will delete the regulator, hook the 'tiny13 to what would have been the pre-regulator transient suppressor and lower the suppressor's clamping voltage to yield 5V.

Quote:
Does your app have any timing requirements? If so, a stable clock source is needed, and the internal oscillators may vary widely with changes in supply voltage.

I am pretty sure I can tolerate at least +/-30% timing variations which will more likely be caused by temperature rather than voltage changes even after I make the above changes. The supply may droop quite a bit when the engine is cranking but the side effects of that should be tolerable if not desirable.

A quick scan of the ATtiny13 datasheet suggests the stability of the 128khz watchdog oscillator under temperature and voltage variations is similar to the internally calibrated oscillator, only the factory doesn't calibrate it. I will, with a frequency counter and an edited flash constant, as this will be a one-time project. I think.

Quote:
You might want to consider something like charging a supercap when your light is on, then live off that when you are idle and there is no juice to power you, if that is a summary of your app.

I plan to do something like that, only more quickly and constantly so no one notices the switching and so the controller can use a smaller storage capacitor.

It should be a fun project. I owe a friend some content for his website so I will write up and post everything when I am done for the amusement of any who might be interested.

Thanks for the comments and suggestions!

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But wouldn't that cause the suppressor to draw lots of current?

---
Formerly Torby. Stitch626 just seemed a more descriptive nicname.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It might, especially if the car's electrical system voltage increases because of a fault in the charging system. However, this particular suppressor stage only needs to protect the ATtiny13. The stage can use a fairly large series input resistor since the '13 will draw less than 150µA at 5V and (in the V-version) operate down to almost 1.8V.

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

In that case, look for a 70k or so resistor and a zener diode.

---
Formerly Torby. Stitch626 just seemed a more descriptive nicname.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The resistance can't be that high as the voltage source is intermittent and I don't want the controller to reset when you start the car in winter. After totalling up all the voltage drops, a 3.3K resistor and 1/2W zener should work. That will give a modest, 2-3 ma quiescent current when the ignition is on (in normal to extreme conditions) which won't heat the diode and will be negligible compared to the current being drawn by other, higher power consumers.

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would continue to power the device through its VCC pin. Expecting reliable operation when relying on power from the esd diodes on I/O pins is not a good idea. Typically such diodes tend to have an effective forward resistance of 50 ohm or so, and I would expect that the internal VCC would experience higher levels of noise this way.

Some secondary effects I would assume MAY happen:

1) internal oscillator instability as they tend to be supply sensitive
3) Reduced drive strength and associated changes in I/O edge rates
3) program execution instability when a bunch of I/O all go high at the sajme time

etc.

The LAST thing you want to experiment with on a uC is POWER.

PDP8? I remember they had one of those at McGill U playing SRTTRK and Adventure! Been out of the game for ahile have you? :)

Dana
Dana

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As I wrote, I am pretty orthodox in how I connect a chip's power pins. I have always connected them, I was planning to connect them, and I will probably continue to connect them. But now I am beginning to wonder if this is a habit I should unlearn. Enough have now suggested floating the Vcc pin, there must be something to the idea. :D

I will supply the ATtiny13 all the current it wants through its Vcc pin, and bypass it. However, there may be 50 µA leakage currents into three input pins that forward bias the protection diodes. People have been feeding that much current -- and more -- into CMOS inputs for ages to simplify designs, without ill effect. Will such currents cause problems into an AVR input?

Atmel's response was, essentially: we don't guarantee anything outside what's in the datasheet.

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Enough have now suggested floating the Vcc pin, there must be something to the idea.

I must have missed something in this thread. I know of nothing said other than your initial implication about "pwoering from the inputs". I don't remember anyone suggesting the idea.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Originially, I wrote:

Quote:
I am working on a little controller ... and am not sure I should bother with a voltage regulator. It appears a 'tiny13 could probably draw all the power it needs through the controller's inputs.

By that I meant the ATtiny13 could draw the very modest power it needs from the controller [board]'s input signals and that the controller board wouldn't need the usual dedicated power connector pin and voltage regulator. I didn't state how I would take power from the board's inputs or how I would route it to the ATtiny13 to power it.

Somehow that got interrupted as floating the ATtiny13 microcontroller's Vcc pin and powering it off its inputs, through its ESD diodes.

I hadn't thought of that, but am amused by the idea. (Although, again, I am too orthodox about chip power connections to do it.)

- John

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

danafraymond wrote:
PDP8? I remember they had one of those at McGill U playing SRTTRK and Adventure! Been out of the game for ahile have you? :)

Yes, in one big way I have been, although the first PDP8 I saw, with its little, spinning reels of DECtape, struck me as quaintly old-fashioned.

I have been a CMOS fan since the days when people thought 74LS TTL was pretty neat because of how little power it used and heat it generated compared to its predecessors. One thing I always appreciated about CMOS was how easily you could design a system that demanded so little of its power that the supply could dispense with a heatsink and often an IC voltage regulator.

I have since managed to completely miss out on using chips suited for micropower systems. They certainly existed and exist, but the parts and systems I got involved with always seemed a bit power hungry. Sure, there were tricks you could use to make them less so, but deep down, all seemed better suited to using milliamperes or amperes to solve problems.

Rightly or wrongly, I have thought of the AVR microcontrollers I have used (8515, 90LS2343, ATtiny26) as milliampere parts. (I should mention that I have had to run them at 5V.) Imagine my surprise to stumble upon a chip in the same family, the ATtiny13, that can easily run off only tens or hundreds of microamperes -- without tricks or extra parts.

I still remember some of how we designed things in the days of the original CMOS. I now find myself in the happy situation of having a nice, low-power part and being able to ask: How can I now do things? How should I now do things? I have a fun project ahead of me.

- John