First Stupid Question of the Year.

[captainkirksdog]

Okay, the wheels have come off this thing. I'm not asking for advice, I'm asking for design help. Yeah, I concur that there's a billion and nine things to do that are easier, but not any fun. I like to build stuff, period. For those that came late--I do not want a water valve. I've got plenty of those. As succinct as I can say it: I want to build a selectable interval (5, 10, 15, 20, 25, & 30 minute) timer with a four-digit LED countdown display. Read the rest of the particulars above.
Here's the why of this endeavor: I was looking for a specific LED a while back and mom walked by and saw these really dinky little red seven-segment displays (1/4") I have. She asked me what they were and me being the brain trust that I am, bread-boarded one up to show her. Welllll, one thing lead to another blah, blah, blah. You know the story. And now I'm stuck. She's kinda hoping for more magic. And I am plumb out.
Hey, I want to say thanks to all you guys for responding to my problem. I really do appreciate it. Don't give up on me just yet. Some of your ideas are pretty good. <ckd>

p.s.: Atferrari: I got it. We're cool.

 

[spec]

Okay, the wheels have come off this thing. I'm not asking for advice, I'm asking for design help. Yeah, I concur that there's a billion and nine things to do that are easier, but not any fun. I like to build stuff, period. For those that came late--I do not want a water valve. I've got plenty of those. As succinct as I can say it: I want to build a selectable interval (5, 10, 15, 20, 25, & 30 minute) timer with a four-digit LED countdown display. Read the rest of the particulars above.
Here's the why of this endeavor: I was looking for a specific LED a while back and mom walked by and saw these really dinky little red seven-segment displays (1/4") I have. She asked me what they were and me being the brain trust that I am, bread-boarded one up to show her. Welllll, one thing lead to another blah, blah, blah. You know the story. And now I'm stuck. She's kinda hoping for more magic. And I am plumb out.
Hey, I want to say thanks to all you guys for responding to my problem. I really do appreciate it. Don't give up on me just yet. Some of your ideas are pretty good. <ckd>

p.s.: Atferrari: I got it. We're cool.

Nice post CKD- now we know just what you really want.

spec

 

[spec]

OK, here are some design defining questions CKD:

(1) Do you just want to use parts from your spares box and, if so, can you supply a list of relevant parts that we can base a design on, or can you order from a distributor?
(2) Can you say if you want your counter to be battery or mains powered?
(3) If you can order from a distributor can you tell us where you are (US, I would imagine) so that we know the parts you can buy and the characteristics of your mains supply?

spec

 

[tcmtech]

I'm not asking for advice, I'm asking for design help.

Same thing in most peoples books. hence the replies you got.

Help me buy a car but I don't need help buying a vehicle. That's the equivalent of what I got out of the above quote.

 

[throbscottle]

Going back to Les - post #5.
This is just out of my head, it may be complete rubbish.
So, you would have 4 74LS192 (or equivalent which should save you some batteries), but you make the 10's of seconds counter reset to 6 every time it reaches 0. Each one drives a 74LS47 (or equivalent)
Since you are only dealing with a maximum time of 30 minutes you don't need to introduce any more 6's.
You cascade the counters so the sequence of digits is taken care of.
To get the 6's you catch the borrow output and use it to pre-load 6 into the counter. If you connected the pre-load inputs P1 and P2 together they could share a pull-up resistor, and the borrow output could pull the ones you want (P0 and P3) down directly, as well as the PL input. So nothing fancy there assuming it works. I don't know if you might have an issue with it changing to 6 when you reset the chip however. We can work that out if you do.
So thinking about the values you want to pre-load. So for each one you set the seconds to zero - you can use the reset input for that.
For the minute units you are setting only 5 or 0, so you set the preload inputs as 0101 or 0000, which can be wired direct to the setting switch.
For the minute tens you are setting some other number - this is where a diode matrix comes in handy, which gets it's input from your rotary switch. It might be adequate to use the very crude method of series capacitors to activate the pre-load inputs when the switch is set to a new position, so long as you use a break-before-make switch. They will need a discharge path. If not, a 74LS121 (or equivalent) will give you the pulse you need.
The 555 timer now just needs a 1 second interval, I think you should be able to re-set it together with the counters so you don't get a partial second at the start.
Doesn't look so daunting now I hope!

 

[cowboybob]

Here's a dual 555 timers rig that meets some of your design criteria:

Minimum time is 5 minutes with 5 minute increments selectable up to a max of 30 minutes. The DIP switch in the schematic is serving as your 6 position rotary (which my SIM lacks). As with ANY 555 timer, "spot on timing" is not possible (RC timer arrangements are notorious for variability) but, given the task, accurate enough. As a reminder, for this circuit, timing is derived from the formula: 1.1(R)(C)=seconds.

Note that I've only included 2 LEDs for timer condition, i.e., a GREEN ON for the selected watering time duration. at the end of which it turns off and immediately followed by a RED LED turning on for 5 minutes (approx) to let you (or your Mom) know it's time to turn the water off.

Your desire to have a digital timer readout component vastly complicates the issue. There's no real way to incorporate that feature without going digital in one form or another.

You might, however, consider adding to the RED LED circuit a small piezo buzzer.

The 555 should be able to supply power for both the LED and the buzzer I noted above without the need of a driver transistor (it could be easily added, however, if need be).

The whole point of this arrangement is to give you a circuit that will be relatively accurate, use components you have on hand and will count down for a selectable time period that, when reached, will display an alarm indicating the need to turn off the water.

 

[throbscottle]

Err, cowboybob? You missed the part about his mum really liking the 7 seg displays...
(Edit) Well I say "his". You never know with these people from the gas giants...

 

[cowboybob]

Err, cowboybob? You missed the part about his mum really liking the 7 seg displays...
(Edit) Well I say "his". You never know with these people from the gas giants...

Didn't miss it throb, just ignored it under the circumstances. I know how Moms can be. I figured a little change in the "readout" (and adding a buzzer) might satisfy her "whiz-bang" need. Otherwise the project gets overly complex. In my book, anyway.

Could be wrong, of course. I can't tell, for sure, how much time/effort CKD wants to put into this rig. Not that I haven't over-engineered my share of projects... .

 

[Les Jones]

Hi ts,
What you say in post #25 is basically what I was thinking. You would need to preset the tens of seconds count to 5 not 6. (The seconds count would go count would go 03 02 01 00 59 58 ) Your method of using the "borrow output" (Called terminal count down on the data sheet.) is better than the way I was thinking of doing it. I was thinking of decoding the count of 9 (Which it would go to on the clock pulse when it was in state zero.) and using this to do the parallel load. This would had a potential race hazard problem that may need an additional monostable to solve. When I posted post #5 the OP had not specified the value of his preset times so had assumed that the seconds counter may have to be set to a value other than zero. If this had been the case it would have required additional logic around the parallel load inputs of the tens of seconds counter. (As it would have had to be loaded with the initial value at the start and with a value of 5 after each time the tens of seconds reached zero.) I don't like drawing schematics and I think this would have taken me quite a long time. (If I was building it I would just have started building it on a piece of veroboard and maybe draw the schematic at the end.) Other comments. I do not see the need for a seconds display on a water timer. I did think it could be quicker to write the code to put on a PIC or 8 bit AVR chip and post a programmed chip to the OP but I don't know of a postal service to other planets.

Les.

 

[AnalogKid]

Back in the DIP days there were various alarm clock, watch, stopwatch, and timer chips by Mostek, Intersil, and National. Probably some available on ebay. Single chip, presettable, count up, count down, multiplexed 6 digit drive, relay signal output - all the feature combinations in 24 to 40 pins. They were the core of products like darkroom timers. I've still got some but I forget which ones. Where are you located?

MK50250
http://www.decodesystems.com/mk5017.html

ak

 

[throbscottle]

I do not see the need for a seconds display on a water timer.
Les.

Me neither but in post #1 CKD did say a 4 digit display.
I am reminded of some funky little 7 seg displays I had from the 70's, they were in gangs of 4, fitted 0.3" dip spacing, tiny displays which had a lensed casing so they looked bigger. Didn't have any use for them at the time (my electronics knowledge at that tender age being considerably less than my enthusiasm). I think they went the way of all Cool Stuff, unfortunately.

 

[captainkirksdog]

Throb, the units you're referring to are the Hewlett-Packard QDSP6064. Same size digits as mine except mine are single digit units. Just FYI.

I don't like drawing schematics and I think this would have taken me quite a long time. (If I was building it I would just have started building it on a piece of veroboard and maybe draw the schematic at the end.) Other comments. I do not see the need for a seconds display on a water timer. I did think it could be quicker to write the code to put on a PIC or 8 bit AVR chip and post a programmed chip to the OP but I don't know of a postal service to other planets.

Les, I DO like drawing schematics. If fact, they're essential to me. I'm not an engineer and it sounds like most of you guys are. I hear "clock pulse" and my brain goes wonky. And I don't see the need for any of this; but I do see the potential for having a good time building it.
TCM, advice is "don't build it yourself, buy some POS ready-made unit." Help is "do I use a 7490 into a 7447 or something different." At least that's my definition of it.
I came to electronics late in life. I was a woodworker by trade with a focus on building kosher coffins (yes, even coffins follow kosher law). I have obviously failed to learn enough about electronics and especially digital electronics. My hobby has been radio, mostly crystal radio. I know it's beginner stuff but don't bail just yet. I do know a lot of other stuff. <ckd>

 

[throbscottle]

QDSP6064 very similar but I don't think the same as what I had, I think mine had the pins sticking straight out of the bottom. They were in a bargain bulk pack of 7 seg's from my local branch of Tandy when we used to have them in this country, and they actually had a fair (if expensive) selection of components.
I might have to get some of those just 'cuz they're cool.
Interesting point about the size. I think everyone here has probably been imagining you have bigger displays, even though you did say 1/4". It's good to have a reference picture.
I'm fully with you on the building stuff because of journey. I speak as a hobbyist, not a pro. Actually sitting down and attempting to design stuff is a huge learning experience, maybe the best way so long as you have access to information, easy now we have the internet instead of text-books which always seemed to be over-simple or over-mathematical, beyond me.
So you are definitely doing the Right Thing by my book.
The folks here on ETO have been a gigantic help to fill the gap between my largely forgotten TV repair training from the '80's, which I never went on to actually use, and a level needed to design anything above a very rudimentary level.
Digital is pretty easy, at least at a basic level (ie, mine!). If you can get your head round a truth table and a timing diagram, you can do quite a lot. It's quite nice knowing how to create out of gates etc, when you see young people who learnt how to program in school who see a micro as the only solution (even if you have 5 times the parts count...)
Interesting about the coffins. I didn't even know kosher was a law, I though it was just something that applied to food.
So, do we need to draft schematics for you or do you have all the info you need to be going on with?

 

[throbscottle]

Might have been this (second picture):
http://www.decadecounter.com/vta/articleview.php?item=788

 

[Les Jones]

Hi ts,
I thought I'd try your preset idea today. I couldn't find any 74192s but I found a 74193 (Count of 16 instead of 10) I think it is similar enough for testing the idea. It did not work as expected. When I set the preload value to 5 the counter only seemed to divide by 4. I changed it to 6 and then it did divide by 5. When I looked at the outputs with my 4 channel scope I found it was being preset to a count of 6 but the count went straight from 1 to 6 instead of zero to 6 as we expected. (It counted like this 6, 5, 4, 3, 2, 1, 6 and so on.) We may be able to invert the terminal count down signal and use it's trailing edge via a capacitor to generate a very short pulse to do the preload.

Les.

 

[throbscottle]

It's because the zero condition is being used to do the pre-load. It never gets a chance to be seen at the output.
I wonder if the borrow output could be gated with the clock to get the desired result, so the zero has a chance to display?
I think I might have to get some of these chips and try the experiment myself!

 

[Les Jones]

Hi ts,
The timing diagram on the original data sheet that I used was not very clear. The waveforms did not seemed to be lined up very well. I assumed that the transistion took place at the end of the borrow pulse. (Rising edge.) I have found a data sheet with better timing diagrams. It shows the borrow pulse starting half way through the zero state. If that was true I would have expected to see the zero state for half a clock cycle. Going back to my initial idea of decoding the 9 state for a 74192 (or the 15 state for the 74193) I realised that a full decode would not be needed as the only time Q3 would be set counting 0 to 5 would be when the count rolled over from zero to 9 (Or 15 for the 74193) so just Q3 inverted could drive the preload signal. The fact that the display would show some illegal value for a few hunndred nS would not be noticed. I will try thes methods tomorrow if I get time.

Les.

 

[Les Jones]

Hi ts,
Q3 inverted connected to the preload input works. I did not try the method I suggested in post #35.

Les.

 

[throbscottle]

Yeah I was thinking a bit wonky when I wrote that. Best ignored
At least you found a working circuit. Now we just need the OP to show us his funky stuff

 

[Les Jones]

Hi ts,
I thought that your method would work when I first looked at the timing diagrams so we both made a mistake. We need the OP to tell us what logic ICs that he would like to use. If he wants to use the original 74 series the timer will eat batteries. (The 74193 took almost 50 mA.) We also need to know how he wants to start the timer. Does he want to set the time selection switch then switch the unit on and it starts timing straight away or does he want a button to start the timer running. Also at the end does he want it to sound a buzzer for a set time and shut down or does he want the buzzer to continue until the unit is switched off.

Les.