So I scored a big bag of 74LS series logic off of Ebay for $10. I'm thinking of building a Numitron digital clock with them, 1970s style to fit the date codes on some of the chips. Of course It would be easiest to implement the divide-by-6 and divide-by-10 counters needed with a CMOS 4018, but I don't have 4000 series, I have 74LS series! Some other clocks use 74LS90s and 74LS92s to do the frequency division, but I don't have any of those either unfortunately What I do have is 74LS93 binary counters, some BCD to 7 segment decoders, and lots of AND and NAND gates, etc. What I'm wondering is if it would be possible to jury-rig the 74LS93s into divide-by-n counters by using logic to decode the BCD output - for example when the BCD output goes to 0110 use an AND gate on the middle two digits to send a pulse to the next counter and trigger the reset pin on current counter. Does this seem like it will work? Will the fact that the 74LS93 is a ripple counter cause problems? I haven't done much digital design so I'd appreciate any advice on pitfalls that might be waiting for me!

 

Maybe you could say that when the output is 0101 in the ten's place the minute counters will reset on the next high to low clock transition and advance the hour counter. Going from twelve back to one on the hours will be tricky. A 74LS192 would work good there.

 

You are thinking exactly right. You can take signals off the output digits and use them to trigger the next counter. To count to ten, AND the 8's and 1's digit together and connect to next chip. When the first chip hits 9, this signal goes high, then low on 10 which triggers the count.

Resetting the counter when you get to 10 is a bit trickier. You need the reset to be low for normal counting. Then pulse it high when you want to reset to zero. Maybe NAND the 8 and 2 digits together and connect to the reset? Not sure, I'm trying to figure out similar thing. As the the poster mentioned, going from 12 back to one will be more interesting since you will need to get signals off of two counters (digits) to figure out the reset. (That's assuming you're building a clock).

I appreciate what you are doing. I've got a bunch of TTL picked up at flea markets and am determined to build a clock without buying any more chips. But I got lucky and have several 7490's. You might consider looking at the HADES simulator, there's a 74390 4 digit counter that I found instructive, it can run in your browser.

 

Thanks for the heads up about the HADES simulator, this looks great! Hopefully I will be able to simulate some of the functional elements of the circuit first without having to mess around with protoboarding until it's absolutely necessary. If I come up with something that works I'll post it here!

 

Check out Ronald Dekker's page:
http://www.dos4ever.com/flyback/flyback.html
That'll get you going with the HV supply for the Nixie/Numitron tubes.

You might be interested in looking at his Numitron clock projects - and some other interesting stuff on his main page:
http://www.dos4ever.com/

 

Hello,

For info on numitrons look here:
http://www.tuberadios.com/numitron/

This is a page from the EDUCYPEDIA on nixie and numitrons:
http://www.educypedia.be/electronics/tubenixieclock.htm

Greetings,
Bertus

 

A couple other thoughts. The "TTL Cookbook" by Don Lancaster is a great resource. Originally published in 1974, it's got lots of practical information regarding power supplies etc and tons of counting examples. It's a bit of a collectors item so prices can be high but buy used on Amazon an it will be pretty cheap.

HADES is nice (?) but don't be afraid to put some chips on a solderless breadboard and blink some LED's. I haven't managed to damage any chips yet, only burned out an LED (forgot the resister) and a relay (too high voltage). Also for circuits and support the Yahoo group "NEONIXIES" can be useful, have to join but they're a friendly bunch. One more: searching for "ic lab 7400 site:.edu" will find a bunch of experiments from various colleges that you can start with.

 

Thank you for all the suggestions, everyone! I discovered a neat "hack" while messing around with the simulator last night. I was puzzling on the problem mentioned earlier of how to get a binary counter like the 74LS93 to reset its value to "1" instead of "0" for when the hours count rolls over from "12". I wanted to get around using a different presettable counter like the 74LS192. The way to do it is: Put the clock coming in to the hours counters through an OR gate, and connect the other input of the OR gate to the output of the gate that is driving the reset of the counters. That way, when the final two counters reset from "12", the input gets an extra pulse to drive it forward to 1 instead of 0. The reset logic for the hours count uses an 3 input AND gate to reset both counters when the count rolls over to "13". I still have to figure out the time-set circuitry, but I'm off to a good start I think!

 

I am concerned, however, that since I am using ripple counters and using decoded outputs to drive further clocked logic, that there will be glitching on the output that will cause the clocks of further counters to advance prematurely. Does anyone have any experience with this? Maybe the only way to find out is to breadboard a test circuit...

 

Sounds like you're on your way. There is a HADES example of setting the time, check it out. I wouldn't worry about glitches on the breadboard, I've not had any problems. In the TTL Cookbook it recommends a 0.1uf disc cap across the vcc and ground on each chip to avoid problems. I'm going to do that when I build on perfboard.

 

Those small capacitors will be very helpful, especially the more dense the circuit gets.

 

Well I was thinking about the timing diagram of a ripple counter, and I believe that AND decoding the outputs for a clock should be OK. For example in a tens counter, even though the outputs may not switch at exactly the same time, the first time the 2s output and 8s output are both HIGH is when the counter is at 10. They can't both be HIGH before that in a ripple counter.