help with led sequencer

[Mazikowski]

I'm trying to make a sequencer to control LED's (or other items via relays or transistors) that will stack from 1-10.

I need it to work something like this (1=lit, 0=off)
1000000000
1100000000
1110000000
1111000000
1111100000
1111110000
1111111000
1111111100
1111111110
1111111111

and all of that needs to happen in the course of about 2 seconds then repeat as long as the circuit has power.

I figured I could use either a 4011 NAND gate or 555 as my oscillator and a 4017 counter, but that would only give me one output active at a time. I want one on step 1, two on step 2, ... , ten on step 10, then back to 1.

Any suggestions as to what I should do?

 

[audioguru]

Instead of a CD4017 you need to use a shift register.
Or a microcontroller can be programmed to do it.

 

[Mazikowski]

What kind of shift register setup would I need to use. I'm not very familiar with them

I'm trying to avoid microcontrollers if at all possible, as I have NO experience with them and for what I'm doing, they seem like overkill.

 

[ericgibbs]

What kind of shift register setup would I need to use. I'm not very familiar with them

I'm trying to avoid microcontrollers if at all possible, as I have NO experience with them and for what I'm doing, they seem like overkill.

You require serial input with parallel output S/R's type, Dual 4bit S/R HEF4015


EDIT: Corrrected the S/R type.

 

[on1aag]

Hi Mazikowski,

How many light chasers do you want ?

https://www.electro-tech-online.com/attachments/light-chaser-gif.12478/

https://www.electro-tech-online.com/attachments/light-chaser-984-gif.13410/

on1aag.

 

[bennettdan]

Hello Mazikowski,
I basically used the circuit you describe and added diodes to feed the prior LEDs as it pulses along and the 555 should give you the 2 second ramp you are asking for. If you want to change the speed use 1.44/2RC .

 

[on1aag]

It won't work, when the first transistor drives the led it will light.
When the second transistor is driven, the second led will light
but the first one will be off . . . the same goes for all other
transistors.

on1aag.

 

[Mazikowski]

Wouldn't the diodes allow current to flow back to the other LEDs? So when the transistor for 1 was active 1 came on, but when the transistor for 2 was active, power would go to 2, but also in one direction via the diode to 1?

 

[on1aag]

Hi Mazikowski,

The diodes have also a voltage drop, it's possible that the
previous led will light a bit and the driven led will light a bit
less but if the number of diodes increases . . .
And even if you would use "ideal" diodes the current through
the leds would be divided evenly and the more leds light, the
dimmer they will be.

on1aag.

 

[Mazikowski]

Ah, I forgot about that... Well, would that circuit work if I had it driving relays or other transistors? This is going into a car, by the way. It will be driving an undetermined number of LEDs per channel as well, so I will have to change things around. Each channel would have it's own resistor(s) according to how many LEDs are present.

I primarily need a circuit to trigger the banks of LEDs.

 

[matk95]

A pic micro would be the best option for this becuse it would be simple. if not your going to end up with a complicated circuits

 

[geko]

The attached code used with the 16F628A version of the circuit shown on the link below will do what you want.

http://picprojects.org.uk/projects/ledchaser/index.htm

Of course you can do it purely in hardware as others have shown, but the PIC results in a lower component count and gives you a lot more flexibility to change things

 

[on1aag]

Hi Mazikowski,

I know it's tempting to use the circuit with the CD4017
because it does it all in one box but if you want a led
bar it's getting rather complicated and you're gonna get
more trouble than you can handle.
Stick to the circuit with the shift register, if you add
a second CD4015 you'll get 16 outputs.
And if you add transistors you can drive as many leds
as you want.

https://www.electro-tech-online.com/attachments/light-chaser-984-gif.13410/

on1aag.

 

[ericgibbs]

Hi Mazikowski,

I know it's tempting to use the circuit with the CD4017
because it does it all in one box but if you want a led
bar it's getting rather complicated and you're gonna get
more trouble than you can handle.
Stick to the circuit with the shift register, if you add
a second CD4015 you'll get 16 outputs.
And if you add transistors you can drive as many leds
as you want.



on1aag.

hi on1aag

Your link is not linked.

Regards

 

[OutToLunch]

I know everyone loves digital, but this can also be accomplished with a ramp oscillator and a number of comparators that turn the LEDs on. All you need is a stack of resistors and tap off of each point to get the next comparator level (place these points on the non-inverting input) and those DC levels are compared to the ramp (placed on the inverting input). Transistors to turn the LEDs on are probably not necessary as the comparators can sink the LED current, so all that is needed is a limiting resistor on each LED. A ramp oscillator can be created with the 555 timer (see here: ).

So basically, a 555 timer, two quad comparators and a single comparator (one LED is always on), two transistors (or more preferably a matched pair), and some Rs and Cs should do it.

 

[Ubergeek63]

I know everyone loves digital, but this can also be accomplished with a ramp oscillator and a number of comparators that turn the LEDs on. All you need is a stack of resistors and tap off of each point to get the next comparator level (place these points on the non-inverting input) and those DC levels are compared to the ramp (placed on the inverting input). Transistors to turn the LEDs on are probably not necessary as the comparators can sink the LED current, so all that is needed is a limiting resistor on each LED. A ramp oscillator can be created with the 555 timer (see here: ).

So basically, a 555 timer, two quad comparators and a single comparator (one LED is always on), two transistors (or more preferably a matched pair), and some Rs and Cs should do it.

True but that is clumsy compared to a couple 74HC164 shift registers

 

[on1aag]

So basically, a 555 timer, two quad comparators and a single comparator (one LED is always on),
two transistors (or more preferably a matched pair), and some Rs
and Cs should do it.

So basically 9 comparators, a resistive divider and a voltage
reference could do the trick. Ehhhh . . .like a LM3914 ?

You shouldn't skip your lunch !

on1aag.

 

[OutToLunch]

cram your ego - there are a ton of different ICs out there and i'm not about to keep track of every one of them - even if it is a simple bar graph driver (clue to you - i've never used one).

i was offering a different option which is made even simpler with the part you brought up.

 

[Ubergeek63]

cram your ego - there are a ton of different ICs out there and i'm not about to keep track of every one of them - even if it is a simple bar graph driver (clue to you - i've never used one).

i was offering a different option which is made even simpler with the part you brought up.

Ego Gracie?

you are adding a dozen resistors which may or may not be well matched in addition to the RC curve introduced by the 555 "ramp". admittedly a nit picky point.

two simple shift register chips and your choice of oscillator gives you any length up to 16 LEDs as opposed to your "simpler" solution that would require 3 quad comparator packages and a hell of a lot of wiring.

 

[OutToLunch]

the response was to on1aag. a differing opinion on whether a circuit is viable or not is fine. being arrogant because you know of the existence of a particular ic is ridiculous.