At times it makes me think they started this to see if theirs a solution to it. Because I have never seen anyone use a ULN2003 or a ULN2803 with 4000 gate logic (commercially). Its always been 74xx000 series logic. Because the low on a 4000 logic gate will bias them on. But to verify this, they need to make a voltage measurement to see what volts their low is. Even if the ULN2803 has a 2.7K on its base, I still wonder if another 1K in series, or a 1K to gnd would change the low logic behavior,

Using a 4017 with transistor bank works just fine. Try it! (see AnalogKids prior post)

 

Since the Arduino is clocking the 4017, it can keep track of what state the 4017 it is in.

It can then put it in any other state, simply by supplying very fast clock pulses, so fast they could not be perceived by the eye.

I would use the freed up Arduino pin as an input connected to, say Q0 or whatever is the reset state so you can re-sync just in case it gets out if sync.

This gives you the ability to set any LED as the active one, or all off any time you like.

Does this help?

I appreciate the input, and I do see how this could work in theory. Again, using MaxMSP and not coding within the Arduino IDE I cannot send those pulses fast or accurately enough. Also, I like how the sequencer looks pseudo random when LEDs are strobed with a separate pin.

 

I have never seen anyone use a ULN2003 or a ULN2803 with 4000 gate logic (commercially).

Really? See post #35.

Your statement about using ULN drivers with CMOS 4000 logic parts is woefully misinformed, or just naive. I've probably posted 50 such basic circuits, plus many more variations. Crutschow has way more than that. There are hundreds of such circuits from many of the long-time posters here, and we also post on other sites.

Commercial circuits from me:
Broadcast audio and video
Automotive
University research labs
Industrial controls
Telecom equipment
MRI scanner signal processing
19 years of assorted military projects

Because the low on a 4000 logic gate will bias them on. But to verify this, they need to make a voltage measurement to see what volts their low is. Even if the ULN2803 has a 2.7K on its base, I still wonder if another 1K in series, or a 1K to gnd would change the low logic behavior,

Your statement is strange. If you mean that the driver will be biased on unintentionally, then - No. A 4000-series gate will not turn on a ULN driver when the gate output is low. Among other reasons, a ULN driver has (as noted above) internal pull-down resistors. Thus, even if left completely floating (or connected to a tri-state output), the driver will not come on unintentionally.

Loading a CMOS output stage with a 1 K resistor to GND is a problem, especially at higher Vdd values, because standard 4000-series parts are not rated to source 5 mA or 12 mA output current. For such situations, the obvious thing to do is use a ULN2004 / 2804 that has an input stage designed specifically *not* to overload a PMOS or CMOS output stage.

ak

 

OK, its not often I see 4000 series into a TTL chip.

I wonder if you have seen the internal schematic of a ULN series driver. Have a look at post #37. As you can see, the ULN2803 is not a TTL circuit. In fact, it is the exact opposite of a TTL circuit. A standard TTL gate input stage actually is an output stage. You have to sink current *out* of the input to GND to cause the input to see a "low". This is the opposite of a ULN driver, which is a version of a common-emitter amplifier; it requires the driving circuit to source current *into* the input stage. Also, a ULN driver has built-in pull-down resistors, which makes me wonder about post #9.

Some of your posts sound like they've come through a translator. Where are you located?

ak

 

the plot thickens . . .