ATTiny13 voltage and current ratings

It should be just fine.

Eh ?

It means you can't exceed that current on those pins *including any load currents*.

Graham

NiMH batteries will routinely show 1.25V/cell when they are fully charged, which puts you right at 6V. 1.3V/cell is not unheard of, and that puts you over. I'd recommend a low drop out* 5V regulator. Even if you could just go straight off the battery you want to have some healthy capacitance on the power supply line, as batteries have significant internal resistance that can cause supply-line droop, which messes up the processor.

The DC current input restriction is there because something in the chip will burn up if you exceed it. The chip by itself shouldn't consume much power (the amount will be in the data sheet), but you will be driving things from it's output pins -- the maximum current directive is to you, the circuit designer, to make sure you don't hang too many things off of the chip.

• The 78L05 needs more than 1.5V of overhead, depending on whose you buy and how much current you pull -- I figure that 8V into the regulator is the minimum safe value, and I prefer 9V.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

text -

Thanks for the useful information. I'll look more into the specs on my 78L05.

Say... you wrote "Applied Control Theory for Embedded Systems", eh? Are Ziegler-Nichols and Cohen-Coon covered in your text?

Thanks,

Michael

text -

No, but if I ever get a chance to do some addenda, Z-N is on my list of things to add, and now I'm going to have to familiarize myself with Cohen-Coon.

Ziegler-Nichols tuning tends to result in an underdamped system; Astrom and Hagglund improved on this in a way that uses the Z-N measurements followed by different calculations.

If you can get your hands on the system for long enough, it's much better to do a more formal system identification, followed by a controller design. This covers a lot more bases than Z-N tuning does.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

text -

Forget the 78L05 or any other regulator. You don't need it, and it won't work, in any event.

You have 4 cells. The maximum voltage per cell is 1.43. So the maximum total voltage is 5.72, and your device can handle up to 6.00. The cells will drop to nominal very quickly, so you'll have 4.8 volts. No 5 volt regulator will work with 4.8 V input.

If you *must* have 5.0 volts exactly, you'll either need to add more cells and a regulator, or use a dc-dc converter, which

*can* produce 5V regulated output with 4.8V input.

Ed

text -

I could use a 9V batt. Don't like to, but I could.

I suppose I could even go to Goodwill, get some wall warts for $1, and use a regulator on that.

I just have a house full of 1.2V NiMHs.

On Jul 3, 6:44 pm, Tim Wescott wrote: ...

I took a process control class in college; didn't learn much from it, I'm afraid, and I didn't do so great in the class first time around (wink) - but those terms came up during class (PID controllers, etc.)

Personally, during class, I wondered why a simple on-off thermostat couldn't work. I just couldn't visualize the need for such esoteric mathematical exercises for control.

If Z-N and Cohen-Coon have been superseded by something even better, I'm all ears...

Michael

If you're going to charge the battery while the controller is connected, or if you want to protect it against misuse then the regulator would be needed, but otherwise no.

if you put large loads on the output pins those currents flow through VCC or GND

Bye. Jasen

What is it with arithmetic this week? everyone seems to be messing it up! anything less than 1.5v per cell is OK.

--

Bye.
   Jasen

Often a simpler approach _is_ all that's needed -- witness the millions of homes with on-off thermostats that keep the temperature nice and comfortable.

But then, often that simpler approach has its limitations -- witness the millions of homes that use continuously-variable heat pumps with fancy "thermostats", that keep the temperature nice and comfortable _and_ save a considerable amount of power over a heat pump with an on-off thermostat.

On-off control also forces the process variable to oscillate around some value (hopefully the right one). If this isn't tolerable, then you need continuous control.

And if you want to know ahead of time if on-off control is adequate, you need to do at least some of that fancy math.

Entire books have been written on self-tuning PID controllers.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

Oh gawd. (4)(1.25) = 5. Not six. Ooh I hate it when I carry twice.

Ehsjr is right -- you don't need a regulator. Just put a honkin' big cap on the output, and have fun.

Make sure the processor will still be happy when you get down to

0.9V/cell, though -- anything more and the battery still has lots of useful charge. You should also consider shutting your system off automatically when the voltage drops to that level, to save the battery from damage.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html