Hi everyone,

I have to make a circuit that will output a 9V signal (or something to that effect) only once for a few minutes, with a delay configurable between one and seven days.

That's it... it's to trigger a burn wire on an ocean anchor, so the accuracy of the timing is not at all important - in fact, this is going to be the backup of the backup(!).

I've looked into it quite a bit and have concluded that a 4060 wired into a 4017, wired into another 4060 (to keep the ON signal high for a few minutes), will be sufficient. It's basically a slightly modified version of this circuit (the first one on the page, also many thanks to its designer). The 4060 sets the 24 hour signal, and lets you delay it up to 10 times.

My main question is: what are folk's thoughts on this? Is there a better way? Would a 9V battery last anywhere near long enough to power this for up to 7 days (or more)?

The key is simplicity and reliability. I'm already using a fancy microcontroller for the first signal and a Teensy for the backup signal. Hoping to come up with something even more basic for the backup-backup. Any input would be greatly appreciated!

 

The battery should last for far longer than 7 days.
The 4060 and 4017 only take static leakage current, plus a small dynamic current, depending upon the clock frequency, the total probably being in the low tens of microamps.

What is the highest frequency in the divider chain?

 

I am curious, what is the function of a burn wire on an ocean anchor?

 

Certainly a small processor like a ATTINY85 could manage this, and guarantee
no false triggers during power up. Use its A/D converter to read a pot or R divider
network for delay. Put it into alternate run / sleep mode to conserve power.

Program it with Arduino or one of the block GUI langauges like Ardublock, mBlock,
or Snap4Arduino.

Regards, Dana.

 

The circuit looks pretty good. The only place I suspect you will have difficulty is in setting the frequency of the RC oscillator. Would it be any problem if you are off by 5% of the total time? If so maybe it would be better to use a crystal reference.

If you do become tempted to do this with a microcontroller, be assured that even very simple microcontrollers are up to the task.

Seeing as this is the backup for a backup your best solution might be to use a circuit similar to the one you've shown but with a crystal oscillator using (perhaps) a 32768 Hz crystal. Yes, it will take more divider stages but the circuitry will be simple and not hard on the battery.

 

I'm generally a fan of discrete component circuits, but if this is a mission-critical project I think the number of components involved needs to be kept to a minimum, for fewer points of possible failure. For this reason I'd go with the microprocessor approach.
I've read that the oscillator section of the 4060 can be a tad touchy. Are you sure it will work, and work reliably, with a 22uF timing cap and 10meg?

 

The ATTINY shipped from factory is speced at 10% timing accuracy.

With a xtal that would be << 1%. But challenge there is xtal ruggedness and
ability to startup in cold temps.

3'rd alternative is a simple internal function at manufacturing test that
does a cal routine to generate fuse changes for its OSC. That will get you
to a couple of % at room T and V.

All timing components, R and C, internal, not subject to cap failure or humidity
in hi R timing implementations.


Regards, Dana.

 

The ATTiny would be give a more robust and accurate design than the 4060. Instead of a crystal I would use a ceramic resonator. The accuracy is ±0.5% and temperature stability is ±0.3% over the range of -20 to +80°C. They are available over a wide range of frequency. Refer to the attached datasheet.

 

Thanks for the replies everyone. To answer the questions:

What is the highest frequency in the divider chain?
- Not sure.

What is the function of a burn wire on an ocean anchor?
- The burn wire is what releases the anchor, allowing whatever unit that's attached to it to float to the surface.

Would it be any problem if you are off by 5% of the total time?
- Not at all. As I said this is the backup-backup, so it's simply there in the event that the first two microcontrollers fail. Precise timing is insignificant.

Are you sure it will work, and work reliably, with a 22uF timing cap and 10meg?
- Not at all. I've never used these chips and am in no way proficient in anything electronic... yet!

Now for my own question...

I've read that the oscillator section of the 4060 can be a tad touchy
- Touchy in terms of keeping time or in terms of breaking down and not functioning altogether?

I should have mentioned that this is going deep in the water, so it's going to be quite cold, though no lower than, say -2 degrees C.

I'll definitely look into the ATTiny.

 

Touchy in terms of keeping time or in terms of breaking down and not functioning altogether?

Touchy as in being fussy over what components are attached to the oscillator section in order for the oscillator to start up.

 

If the oscillator section of the 4060 is a concern, then you could use a LMC555 timer (CMOS version) configured as an astable oscillator to drive the divider chain, which would be very reliable.

 

Right on, thanks for the advice! Just waiting on the chips to show up now. I'll post updates as soon as I figure out what I'm going do.