Battery power for ATtiny85

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

I am considering using an LM1117T 3.3 regulator to supply an ATtiny85.
The only micro function is to drive a 12 VDC output (via a N channel MOSFET).
The project power supply is a 12 VDC battery to supply both the Tiny85 as well as
the 12 VDC output.

The current project is using a 7805 to supply the Tiny85 with 5 VDC VCC.
It works fine but I wish to extend the battery life if I can.

Question: Will using the 3.3 vs. the 5 VDC regulator offer any power savings ?

Note: I am already using a sleep mode and a wakeup on WDT interrupt.
I am thinking that perhaps driving the Tiny85 with 3.3 volts will increase some battery life ?

Thanks .

I'll believe corporations
are people when Texas executes one.

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

Quote:

Will using the 3.3 vs. the 5 VDC regulator offer any power savings ?

P=IV so I'd have said yes.

(but what do I know - I'm just a bit mangler!)

(having said that dropping 12V->3.3V rather than 12V->5V presumably involves more heat loss but then again the other side of that is that the LM chip is likely more efficient than the old 7805)

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

The micro will draw less current @ 3.3V than at 5V, it's in the datasheet. Make sure you obey the maximum frequency vs voltage limits.

Will your mosfet turn on fully with a 3.3V gate signal?

Check regulator specs for their internal current draw, since they'll be drawing that current all the time.

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

Quote:

I am thinking that perhaps driving the Tiny85 with 3.3 volts will increase some battery life ?

1) What does the data sheet say about power consumption at the different Vcc levels?
2) Is the (mostly sleeping) AVR's power consumption over a period of time even significant compared to the usage when the 12V device is actually firing?
3) Is the (mostly sleeping) AVR's power consumption over a period of time even significant compared to the regulator's [in]efficiency? Especially when dropping 12V to a much lower level? (Aren't there some nice regulators for this that have a nice "sleeping" feature when there is low draw?)
4) Have you done your "power budget"? I made a spreadsheet to help, see this thread https://www.avrfreaks.net/index.p... and a description with worked example in https://www.avrfreaks.net/index.p...

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

I did wonder about that. Over a period of a day or a week, how many mAh will the 12V load draw, vs how many will the regulator-AVR draw?

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

kk6gm:

Quote:
Will your mosfet turn on fully with a 3.3V gate signal?

I will check the data sheet for the device. I used a pulldown on the MOSFET input and I am thinking I can make that smaller.

theusch:

Quote:
(Aren't there some nice regulators for this that have a nice "sleeping" feature when there is low draw?)
Do you have any recommendations?
Also, thanks for the spreadsheet. I will have to digest that later. I did find the referenced posting on using the WDT interesting. I also use the WDT interrupt (4 sec.) to wake up from the sleep mode. I think the code for the T85 is ok.

But, thanks to all who replied and I will continue to trash this out.

I'll believe corporations
are people when Texas executes one.

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

My favourite linear reg. is MCP1702 series http://ww1.microchip.com/downloa...

2uA (1uA typical) quiescent max, 13.2V i/p.

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

A 7805 uses almost just as much current as the AVR just to feed itself; about 7mA. It's totally worthless for battery operation.

The LM1117 doesn't seem to be any better in this regard (5-10mA).

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

I like the TPS76333 (3.3V output). Admittedly only 10 v max input.

Ross McKenzie ValuSoft Melbourne Australia

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

chartman, thanks for the reference, I will take a look at that device. I think it is only surface mount? That would be a problem for me. But I will look again at the datasheet and make sure about the mounting for the device.

I'll believe corporations
are people when Texas executes one.

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

Now I am beginning to think that it could be a solution (?)
to just power the micro with a 3 Volt battery, no regulator and associated current lost with the regulator.I think that must be a problem in driving the MOSFET which controls the 12 Volt output.

The current status doesn't look too bad for battery life. This is just a personal project and I can charge the battery (Gel cell 12 Volt 1.2 AH) every so often and be ok with that. This will never be a commercial product that I can see. The curiosity does make me want to investigate and implement a best design.
Thanks for the inputs from all who posted. The best way to learn :-)

I'll believe corporations
are people when Texas executes one.

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

@ tubecut:- it's available as TO92 as well from Farnell UK and probably Digikey USA

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

chartman: OK, great information. I will check out Digikey here in the US. That could well be the best approach.
I did not look at all of the data sheet yet (I will now), will it handle the MOSFET (MDP4050L) drive?

edit update: Looked at Digikey, this looks like the best possible device for 3.3 V regulator for battery operation.
I will order some today and modify my board. If I understand correctly, the MOSFET is a voltage driven device and I am thinking that the 3.3 (or near that) volt output of the Tiny85 could control the MOSFET. The curren project has a pulldown at the mosfet input and I think it
may only require changing the pulldown value (decrease).

I'll believe corporations
are people when Texas executes one.

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

Looks like your mosfet is marginal at 3.3V (if I found the correct datasheet - the designation wasn't exactly the same). This one is a favorite of mine, and cheap:

http://www.mouser.com/ProductDet...

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

kk6gm: I looked at the data sheet. I think it is showing
the turn on voltage as 1 volt? Is that correct. I will have to order a couple of those along with the regulator chartman suggested.

I'll believe corporations
are people when Texas executes one.

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

tubecut wrote:
kk6gm: I looked at the data sheet. I think it is showing
the turn on voltage as 1 volt? Is that correct. I will have to order a couple of those along with the regulator chartman suggested.

More like 2V, but that's a nebulous figure. If you look at Fig. 1 you can see that it turns on quite strongly @ 3.2V, at least for loads of 10A and less. What is your load current through the device?

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

kk6gm :

Quote:
What is your load current through the device?

The load is small, more or less 35mA. It is nothing but two LED's and what looks like at least one resistor. It is a sealed unit, so it can't be dis-assembled.
I selected the mosfet because the lamp module is designed for 12 volt operation.
edit: Also I will add that other than supplying the
uP (tiny85) power, the 12 volts is not applied to the lamp module except for a short burst of time (50% duty cycle) for a period of 2.5 seconds. At the end of that 2.5 seconds, then the tiny85 is placed in the sleep mode with a WDT interrupt set for 4 seconds. The WDT interrupt wakes up the tiny85 and the next 50 % duty cycle repeats. That is all this does. In actuality this is a LED blinker for the bicycle when I have to exercise. I wanted to use the 12 volt module because it is much brighter in day time operation and the 12 Volt 1.2AH battery allows for a longer lasting battery operation.
The 2.5 second 50% part is the blinking of the lamp module. BTW, the module is built with 2 each LED's.

The width of, say, a positive going pulse is 48 ms and of course the negative pulse is the same.
That is everything that is done in this simple project.
Thanks for all of the assist.

I'll believe corporations
are people when Texas executes one.

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

Quote:

I wanted to use the 12 volt module because it is much brighter in day time operation

You sparkies will need to educate me--why would an LED be brighter if a 12V supply is used versus e.g. 6V or 3V?

(and why did I need to do a captcha for the above? Is "sparky" a swear word?)

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

Yes, I understand that LED's are current driven. However, the module is designed to operate at 12 volts, therefore I drive the module at 12 volts. In testing I did indeed use a 5 VDC and I could see a difference. Probably because of the
built in resistor(s)? Of course that may invalidate my comment that was quoted. As soon as I entered that comment I knew I would get some feedback on that. I did measure the current draw of the module (supplying 12 volts) and it was 35 mA.

The captcha to me is a pain in the butt. I didn't see where I had entered any words that might be a problem. It looks like to me it is required to enter a post.

I'll believe corporations
are people when Texas executes one.

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

I don't know of a small thru-hole mosfet that will work well at 3.3V, but you could certainly use an NPN transistor - even a "digital" one with built-in base resistor if you wanted to get fancy.

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

Well I was thinking that the mosfet would require less current (yes, I know it is voltage controlled) and can consider that if I change to a 3.3 volt operation. I do like the MCP1702 pointed out by chartman. I am not familar with a

Quote:
digital one with built-in base resistor
. Can you point me to info one that one?
Thanks.

I'll believe corporations
are people when Texas executes one.

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

It's true a mosfet would require less current, but you're still only talking maybe 1/2mA of base current.

Here's a typical digital transistor:
http://www.mouser.com/ProductDet...