Long timers (seconds)

[ADWSystems]

Along the lines of a post I issued in the past. I'm aware of the 2 chip solution to generate 1/2 and 1 second pulses. I really need to generate pulses along the lines of 1-10 seconds, and possibly more. If I wanted to go with one chip and it was programmable; could I use a PLD, which one? If I wanted to have a chip that incorporated the crystal/oscillator, which micro would you recommend?

(so I'm looking recommendations in each category 1-which PLD, 2-which micro)

 

[canadaelk]

You do not give us much to go on: Repeating? One shot? Trigger input? Output hi/lo?

The 555 can do one, the 4020 the other. No PLD or uP needed. E

 

[KeepItSimpleStupid]

Take a look at www.picaxe.com .

 

[Reloadron]

If I am doing a home hobby experiment for long time applications of various flavors I generally run with a uC chip, be it a PICAXE or whatever. However, if I am looking at a commercial application at work I just buy an off the shelf timer along these lines for a little over $100 USD and am done with it. It can be powered by 12 VDC, 24 VDC, 48 VDC, 120 VAC or 240 VAC and gives me five functions. Repeat, One Shot, Delay on Break, Delay on Make, and Interval. It gives me selectable ranges of 9.99 Seconds to 999 Min in 5 ranges. It's compact and fits a simple Din Rail 11 pin socket with a 10 amp DPDT relay. Personally I can't design and build that package for a hundred bucks.

Ron

 

[ADWSystems]

You do not give us much to go on: Repeating? One shot? Trigger input? Output hi/lo?

The 555 can do one, the 4020 the other. No PLD or uP needed. E

Free running. The 555 cannot go that slow with any type of accuracy and the 4020 implementation requires too many chips (plus I don't care about any of the times/pulses below 1/2 second which the 4020 provides access to)

 

[KeepItSimpleStupid]

In one case it depends on the volume that you need and what you already know. The PICAXE seems to have a nice well-thought out BASIC and timing seems to be really easy.

I PIC might be cheaper in the long run because you don't have to pay for the overhead of a built-in BASIC.

And I agree, the specs are not 100% there for there is delay on make, delay on break, retrigerable, then a duty cycle thing with normally off and normally on delays. What power supply does this thing have to run at. Any issues with low power or battery powered? What is the end "pulse"? A relay, turn on a FET

Resolution? How do you want to program? DIP switch? How accurate? Resolution?

Repairability? Reliability? Power source?

Development time will cut into the cost per unit.

 

[4pyros]

Have you looked at CD4536 or CD4541? Thay are oscillators and counters in one.

 

[KeepItSimpleStupid]

4pyros:

For your problem, have you considered something like your ematch signal sets off the first device. That device delays and passes the ematch signal to the next device etc. As long as the processor start-up time is lower than the timing resolution that you need, it could work. You could power with a CR2032 battery too. About $0.50 when you buy from the right supplier.

 

[4pyros]

4pyros:

For your problem, have you considered something like your ematch signal sets off the first device. That device delays and passes the ematch signal to the next device etc. As long as the processor start-up time is lower than the timing resolution that you need, it could work. You could power with a CR2032 battery too. About $0.50 when you buy from the right supplier.

Yes that is along the lines I am thinking. I need timing down to .05 sec. with better than 1% accuracy over 50 or so units.

 

[ADWSystems]

Have you looked at CD4536 or CD4541? Thay are oscillators and counters in one.

Looking at the CD4536, it appears that it could be substituted for the CD4060 design posted in other threads. If the clock was 32.768kHz, the CD4536 would provide between 1/32 and 1024 second pulses (1/32, 1/16, 1/9, 1/4, 1/2, 1, 2, 4 seconds, etc). But what capacitor and resistor value would be used. The formula shows f=1/3RC. Therefore RC=98304. What cap and resistor values would be best?

 

[KeepItSimpleStupid]

The general idea is to keep the capacitors below or around 1-4.7 uf tops. Metalized polyester caps can easily be had at +-5% tolerances. The max R should be allow more than the input bias current to flow. Leakage current specs of the capacitors would start to cause problems when electrolytic caps are used. 3uA leakage for electrolytic caps would not be unusual.

You might be able to substitute a resonator based circuit for what you need. I did not check your calcs, but f=1/t where t is in seconds. F is in farads or 1e-6F = 1 uf. R is in ohms.