Interesting solutions being offered here, however I must admit that I'm not able to follow the one with the LM2902 
But I really appreciate the support and posts made on my topic! A lot of knowledge over here.
I've changed indeed the post a little by adding the status led, but well I'm developing and ideas pop up.
The solution Mooly gave by using NAND looks easy to make, but can a NAND directly drive a LED? Isn't there also the same current problem?
But I really appreciate the support and posts made on my topic! A lot of knowledge over here.
I've changed indeed the post a little by adding the status led, but well I'm developing and ideas pop up.
The solution Mooly gave by using NAND looks easy to make, but can a NAND directly drive a LED? Isn't there also the same current problem?
There are many families of logic. 7400, 74L, 74S, 74AS, 74AC/ACT, 74HC/HCT...
All have different drive capabilities. You find these in the datasheet. Many will drive an LED direct, most I think. Some LEDs work with a tiny current. Pick the right relay and you can drive it direct.
Things are getting over complicated.
Think 8080 uProcessors. They have a clock and counter.
Too complicated!
Use a 555 timer to drive a 2N5000 FET switch for the relay. If you want you can easily move the 2/3 point for triggering and therefore time delay.
The fewer the components, the cheaper and more reliable the product gets.
Sorry, not 2N5000 that is an npn transistor but would do. As soon as my brain is in gear I will advise the correct number.
Got it, 2N7000
Think 8080 uProcessors. They have a clock and counter.
Too complicated!
Use a 555 timer to drive a 2N5000 FET switch for the relay. If you want you can easily move the 2/3 point for triggering and therefore time delay.
The fewer the components, the cheaper and more reliable the product gets.
Sorry, not 2N5000 that is an npn transistor but would do. As soon as my brain is in gear I will advise the correct number.
Got it, 2N7000
Last edited:
Interesting solutions being offered here, however I must admit that I'm not able to follow the one with the LM2902
The opamps CC showed are all configured without feedback. That means they obey the simple rule of "if the + input is higher than the - input then the output will swing toward the rail". So take the caps out of the circuit and you can work out the DC conditions. Then imagine the caps back in place and see how they slow things down.
Yes, any of CMOS logic family will drive an LED but when it comes to LED's a lot depends on the efficiency of your chosen device. I would suggest you try yours with a current of say 1.5ma (so that would be a 4.7k and a 9 volt battery as a test). If the brightness is OK then all is well.
If you need more current then we could replace the NAND gates (which are all used as invertors) with a 4049 HEX invertor. That has 6 invertors of "high current ability". The word "high" is relative but it is high for CMOS. Each invertor would be used in place of one NAND gate. You then have another option (if needed) of paralleling invertors to drive the LED's as there are six in the package.
Note the positive supply is on pin 1 for this IC. Don't leave any unused inputs floating.
Attachments
I have had a little project book for logic circguits sitting on the shelf.
Just recently I finally decided to dip my toe into logic circuits and bought a few of all the different 14 pin dip chips (mostly cd4000 series) needed to complete all the projects in the book.
I used CPC and the average cost came to >40p incl VAT.
It would be cheaper to use opamps and/or comparators, of which I have loads.
Some day I might start to understand logic circuits !
You've come up with a totally different approach to anything so far...
One thing I would say... and I'm sticking with the CMOS over this... is that you have to look at how the circuit behaves during the transitions. I like to see a 'clean' change in logic level. The opamp solution of CC requires low voltage rail to rail devices and the circuit doesn't seem to have any hysteresis mechanism.
I suspect the same 'clean switching' issue may happen with the discrete version too as the devices slowly come into conduction (but I see you simulate a clean transision
). It all depends on the ultimate goals of the design. A Babani book ?
I used to play with logic a lot many years ago building an Elecktor 'Digiscope' and a fully featured TV pattern generator. My first ever logic project was a Practical Wireless traffic light controller that simulated the UK sequence correctly. All TTL.
A 4017 and slow running clock (so no hf generated) is all you need to make a cracking 'push button selectable' input selector.
I must add using logic to the LTSpice Tutorial although only basic gates and so on are available as a standard.
I'm a radio amateur.
There's a long tradition in ham radio of employing logic chips for purposes for which they were not originally designed. 1 gate + 1 crystal oscillators. Little CW transmitters with six gates all yammering away in unison.
Logic chip families differ widely in the voltages they employ and the currents they are capable of.
Old 4000 CMOS typically have a fanout of 30+ gates, but this is because of the high input impedances. They operate at voltages well above TTL (5V), but output is often less than 1mA depending on voltage.
This is why I was a bit surprised to see CMOS in the circuit.
OTOH modern logic families 74HC, 74HCT, still sometimes referred to as CMOS, have voltage and current levels expected of TTL (7400 series). 74HCT will happily put out (or sink) 25mA.
You just have to be aware that some NAND gates are more equal than others.
Of course you have to be careful with your choice of opamps too, but most members deal with them a lot and have them on hand, while still regarding a 555 as outré.
Obviously there are a whole range of options available if we are concerned with how clean the transitions are, starting with using comparators instead of opamps, but this is hardly a demanding application.
I still recommend that anyone requiring this, or any other simple logic function learn to use PIC micros, they're so versatile you can sub one in for any gate, they're not fussy about voltage, and they'd slaughter all these solutions for component count.
Say 5mA each for the LEDs, you can put 75mA out those 3 in parallel as long as you switch them simultaneously. 'Course you need to program it.
There's a long tradition in ham radio of employing logic chips for purposes for which they were not originally designed. 1 gate + 1 crystal oscillators. Little CW transmitters with six gates all yammering away in unison.
Logic chip families differ widely in the voltages they employ and the currents they are capable of.
Old 4000 CMOS typically have a fanout of 30+ gates, but this is because of the high input impedances. They operate at voltages well above TTL (5V), but output is often less than 1mA depending on voltage.
This is why I was a bit surprised to see CMOS in the circuit.
OTOH modern logic families 74HC, 74HCT, still sometimes referred to as CMOS, have voltage and current levels expected of TTL (7400 series). 74HCT will happily put out (or sink) 25mA.
You just have to be aware that some NAND gates are more equal than others.
Of course you have to be careful with your choice of opamps too, but most members deal with them a lot and have them on hand, while still regarding a 555 as outré.
Obviously there are a whole range of options available if we are concerned with how clean the transitions are, starting with using comparators instead of opamps, but this is hardly a demanding application.
I still recommend that anyone requiring this, or any other simple logic function learn to use PIC micros, they're so versatile you can sub one in for any gate, they're not fussy about voltage, and they'd slaughter all these solutions for component count.
An externally hosted image should be here but it was not working when we last tested it.
Say 5mA each for the LEDs, you can put 75mA out those 3 in parallel as long as you switch them simultaneously. 'Course you need to program it.
Last edited:
Yep, you could do some neat things with a handful of CMOS or TTL back in the day.The stores guy at college was heavily into amateur radio (very late 1970's) but that didn't capture my imagination as much as audio. I have a book at my side now... and its been there for nigh on twelve months, "PIC in practice" by D W Smith. Lol, I've got a PICKIT2, its just writing the code that's the hard bit.
hoihen... sorry, we seem to have hijacked your thread a little bit
No problem at all
The post already enabled me to make the basic version work and showed some other improvements that are possible and which I will test.
For me this is also a learning experience. And a lot of knowledge is being shared.
I do not have a background in electronics, but I'm very much interested and recently started to build some circuit with help of this forum, Googling and logic thinking
The post already enabled me to make the basic version work and showed some other improvements that are possible and which I will test.
For me this is also a learning experience. And a lot of knowledge is being shared.
I do not have a background in electronics, but I'm very much interested and recently started to build some circuit with help of this forum, Googling and logic thinking
Just tried the LED I will be using with the 9 V battery.
With a 4.7 K brightness is no problem at all, I plan to have even less brightness. Made a try with a 5.7 K and I can even go further up in resistance.
So then the current drawn by the LED should not be a problem.
Then to find the correct NAND type.
With a 4.7 K brightness is no problem at all, I plan to have even less brightness. Made a try with a 5.7 K and I can even go further up in resistance.
So then the current drawn by the LED should not be a problem.
Then to find the correct NAND type.
That sounds promising. You are around the 1.5ma figure with a 4k7 and 9 volts and so the standard 4011 NAND gate should be fine.
Yes, logic should always be decoupled... I'd taken it as a given but thanks for pointing it out.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.