Millenium C questions re CD4049

[Toney]

 Has anyone done a Millenium C with a CD4049 or CD4069?

These have six hex converters per package and I have seen differing approaches to making it happen.
R.G.'s article from 2004 has two sets paralleled.

http://www.geofex.com/  scroll down.

Other advice seems to suggest using just one, with all the other inputs grounded.
I am trying to make up something to use up some CD4049's I have had for ages and I'll do a Vero once I get it firing.

Questions:

1/ What's the best practice with how many inverters to use? Assuming regular (non hi-brite LEDs)
2/ With the positive ground wiring, not quite 'getting' how the CMOS can become P type - do I tie the unused inputs to 9v+ ?
3/ Can buffered CD4049 be use effectively?

Toney.

[R.G.]

Has anyone done a Millenium C with a CD4049 or CD4069?

Yep.

These have six hex converters per package and I have seen differing approaches to making it happen.
R.G.'s article from 2004 has two sets paralleled.

There are differing approaches in terms of number of gates paralleled because the actual number paralleled makes almost no difference to the input side. On the output side, you get N times the per-gate output current with N gates. Pick your poison.

Other advice seems to suggest using just one, with all the other inputs grounded.

And you listen to advice from people other than me why?   

1/ What's the best practice with how many inverters to use? Assuming regular (non hi-brite LEDs)

If you're basing that question on driving an LED, then you take the number of milliamperes of current you want to drive into the LED, look at the datasheet for the gates you'll use for their maximum output per gate and per package, do the math and parallel up that many gates to drive the LED. The CD4049 has a very robust output current, so it may well be able to do it all with one gate. If you use high currents and CD4069, you may need two or three gates to drive the LED current.

On the input side, the DC resistance of the CMOS input is so high that from one to six gates only adds capacitance to make the turn on/off slower. Probably not slower enough to notice, but slower. If you want fast action, use as many gates on the LED end as necessary to get the current, use one gate at the input to do the DC resistance sensing, and hook up the several spare gates that remain to get the logic polarity of  signal right for your taste on the LED end.

2/ With the positive ground wiring, not quite 'getting' how the CMOS can become P type - do I tie the unused inputs to 9v+ ?

CMOS is BOTH N and P type. It reacts to the input pin being pulled to the [most positive power supply] or [most negative power supply] and it can't tell which of those is ground, and would not care whether it did know which was grounded.

CMOS inputs don't necessarily have to be pulled to ground per se, but they MUST be pulled solidly to either Most Positive Logic Level or Most Negative Logic level to keep the unused gates held in one state or the other. Exactly which state this is is not material to an unused gate.

3/ Can buffered CD4049 be use effectively?

Yes.

Collecting bits and pieces of technical advice here and there on the internet is very much like sending email to the proverbial blind men to gather data for a drawing of an elephant. 

[Toney]

When it comes to these CMOS chips I AM a blind man drawing an elephant   ...first time really looking at them.
Though the reference to other advice was, someone quoting you during my searches for info. More likely paraphrasing or "interpreting" via their own comprehension or partial degree there-of.
As always, clear, lucid advice from the man that 'wrote the book'.
 Thank you