Challenge: Figure Out How This Logic-based Chaser Works

[Andrea] built this LED chaser using one logic chip. It illuminates all but one of the six LEDs, with the dim bit moving back and forth along the row in a chase sequence. This is something like an inverse Larson Scanner without the fading tail. But doing it with a logic chip instead of a microcontroller is a fun challenge.

Which brings us to the point of this feature. [Andrea] didn’t really post an explanation of how the circuit works. Usually missing details mean that we archive the tip and move on to the next one, but we think this provides a fun activity. Can you figure out how the circuit works? We already know that it’s using a CD4017 decade counter/divider chip. This gets its clock signal from a 555 timer circuit. [Andrea’s] schematic is a bit hard to read, but grab a copy, blow it up a bit (or use your browser zoom) and study the CD4017 datasheet (PDF) if you need to.

Want proof that it does actually work? It’s embedded after the break.

36 thoughts on “Challenge: Figure Out How This Logic-based Chaser Works

  1. quite easy really, it Should count to higher than the leds shown, the other half is folded back on itself to simulate a count Down, drive half with resistor, the other half with diodes that will overide them on the count back.
    I did something similar with 74154 driven by a binary counter that inverted its output when it hit 16 and trigered a flip-flop (zero reset the flip-flop).
    but yeah drive One way with a resistor, and fold over with a diode to drive the rest backwards.
    nice little circuit by the way ;)

  2. Uh? This is very basic. NE555 as timer, 4017 as counter, and diodes as “OR gates”. Oh and transistors as inverters. It’s the 2nd schematic I made… ever, first one being a 2 led blinker.

    Kuddos for the extreme flying resistor soldering job.

  3. The resistors, diodes, and PNP transistors form DTL NOR gates, translating the 10 counter states into the LED outputs. State 0 has the left LED off, state 1 blinks off the next one, state 2 the next and so on until state 5 blinks off the right one. Then, state 6 blinks off the next one and so on until it’s back around to state 0.

  4. The decade counter has 10 outputs, and we only see six. The four LEDs in the middle are tied to two pins each (5 & 7, 4 & 8, 3 & 9, and 2 & 10). When the pins go high, they turn off the connected LED. Shore looks purty.

  5. Digital logic 101 stuff here. and “one chip” is a lie as a 555 timer is a logic chip. So there are TWO chips there.

    I could build this with NO chips and just Transistors.

    He Could have done it without the 555 if his mad haxxor skills were high enough. He needs to hack more until he levels up.

    1. Sounds like a challange!
      -Transistors only
      -No Chips
      -Must be doccumented (Pictures & Video)
      -Bread board or Perf board accepted, No PCB
      -Points will be deducted for non-transistor components.
      -6 Leds and one Power Supply will be allowed with out a deduction.

      prove your level.

  6. This was called ‘diode transistor logic’ in the distant past before it got marginalized by ‘transistor transistor logic’ – TTL. Still a useful solution when you need it.

    1. I noticed as well and looked at the pictures of the PCB to see if the resistor values were all the same. They look like they are. The BC327 has a fairly high gain so the different base resistors won’t matter. The 330 ohm resistors are effectively limiting the current to the LEDs. It doesn’t look like the LED is pointing at the camera, so maybe the LED itself has a higher millicandela rating. *shrugs*

    2. The first and the last led are polaryzed by different resistances at the central pin of the 327, but I think this is not the reason for which the first led is brighter.
      If the reason was that, those leds should be both brighter than the middle leds.

    1. That’s a nice simplified solution, but if you built it for real it wouldn’t work. The Falstad simulator is not taking into account the output current available from the 4017, which is very low at high levels. This is why you need the transistors in there to amplify things. The current is so low that it will not be enough to overdrive those resistors to a high level, even if they were 330 ohms.

      Here’s a Falstad circuit to explain what I’m talking about:

      1. Yes it matters. If you want to turn the LED off in your circuit, you have to get more than 3V developed across the resistor to forward bias your diode to less than 2V (ideally 0.7V), which becomes the new path for current… stealing it away from the LED until the LED can no longer be forward biased (i.e. OFF). Check out my Falstad simulation to see this effect… play around ;-)

      2. That’s ok, I see your reply! Hope it was helpful. Best way to learn about electronics is to build as many circuits as you can, then you will start to “encounter” all of those little things in the specifications that are pretty much a huge puzzle piece waiting to be misaligned.

  7. Sorry, you call that a challenge? It’s a 4017 counting, with discrete inverters (the PNP transistors) driving the LEDs; the transistors to the four center LEDs are ORed using two diodes (remember DTL logic?) to two counter outputs.

    Given the way he describes the circuit, he’s just cargoculting; he doesn’t understand half of how the components used are supposed to work.

    1. Hahaha, yeah right. His English is not very good, but he knows what he’s talking about. Don’t forget, he built the circuit and took a video of it… not you. Maybe if you knew how the circuit worked you’d understand HIM. ;-) No offense meant!! really… just trying to make you think twice before mouthing off.

      1. So why do you think I don’t know how the circuit works? This is just Digital Logic 101: a 4017 decimal counter, driven by a 555 oscillator, and all but outputs 0 and 9 ORed in pairs (1-8, 2-7, 3-6 and 4-5) to create the scanner effect.

        The 470 ohm resistors are superfluous; when you drive the base of the BC327 high there’s no current running into it, and the only current drained from the 4017 output is the current going into the 47k resistor, so less than 0.2mA at Vcc=9V. That’s a 0.09V drop over the 470, totally negligible, especially since there’s a series diode dropping 0.6 .. 0.7V. And running those 47k resistors back to the 4017 is nonsense as well, just tie one 27k resistor between the base of each transistor and ground.

      2. I don’t think you know what you are talking about because you are being rude like anyone cares to back you up… (fictitious troller backing you up) “Yeah man, that guy doesn’t know what he’s talking about! He’s just copying and pasting a bunch of circuits from Forrest Mims. Somebody should revoke his right to represent electronics knowledge on Instructables. We’ve gotta enact SOPA on his @ss and shut down Instructables so we can all be safe from his lies and deception! I’m calling my senator right now!!! Damnit… I can’t even believe he put those superfluoride 470 ohm resistors in there as well!!! He’s robbing my wallet man… this is just unacceptable!!!”

        So because of that, what you say just goes in one ear and comes right out the other. I focus on your negative comments and feel bad for the guy that took time to post a circuit, and now he has to listen to you complain and whine about his electronics skills. You can’t just offer up advice in a polite way, you take pride in forcing it down someones throat.

        I suppose though if you had just posted how you thought it worked, and your suggestions… your comment would go unnoticed giving you no attention, which you obviously want… because you signed up to get replies.

        Have a nice day.

      3. I was under the impression that to qualify for a Hackaday entry you had to create something remarkable. Well, maybe it is, but not in a good way. An otherwise unremarkable circuit, weirdly implemented (several people have commented on superfluous resistors, and how straightforward it is to understand the logic if you look past his description) and built in a way to make anyone who knows how to hold a soldering iron weep. It would not surprise me at all when the left LED being brighter was caused by shoddy soldering.

        But apparently I should simply appreciate the effort.

  8. As other people say, this is (at least was) really simple circuit a while ago. Very popular in kit form.

    Still, it’s useful for people learning, especially if they haven’t had much experience of digital logic.

    Somethings for people to think about:

    How would you expand this to a longer chain of LEDs?
    How would you attach something that draws more current?
    What’s the fastest you can go?
    etc etc.

    1. And it doesn’t even contain an Arduino!!! Dang you are right!!

      Seriously though… It’s not your typical way of building a Larson scanner… and the 4017 doesn’t give you a rolling LOW level… it gives you a HIGH one, so there is… ahh forget it.

      It’s on Instructables.. isn’t that good enough for you???

  9. You can reduce the parts count by 25% by reducing the number of resistors for the NOR-logic.

    Connect one resistor between base and ground.
    Connect diodes between outputs and base.

  10. not reading all the other replys i say this:have the 555 paused for a delay, then delay on the next led and do this to pwm all but one , then a second later switch the led u dont pause on to the next and there ya go, use capacitor timers everywhere i bet theres a bunch of SMD resistors and capacitors hidden on the back. diodes (form a gate and) change the delay-delay depending on which led/output it is paused on.

    more analog adjustments to do then a tv tuner! me thinks.

    i’ll come back later and examine the schemastic

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.