On Sun, 04 May 2014 21:14:43 -0700, John Larkin
wrote:

On Sun, 04 May 2014 18:36:25 -0500, John Fields
wrote:

On Sun, 04 May 2014 15:31:10 -0700, John Larkin
wrote:

On Thu, 01 May 2014 14:35:25 -0700, Jim Thompson
wrote:

On Thu, 01 May 2014 14:25:49 -0700, John Larkin
wrote:

On Thu, 01 May 2014 15:12:23 -0500, John Fields
wrote:


Download all the files into a single folder and then run ZCD7.asc.

John Fields

I recall that one. It has some logic-race bugs.

May well be. Why don't you enumerate them? Be precise. This is
supposed to be what this group is all about.

I did, back when he first posted it.

---
That was a completely different circuit, so why don't you just go
ahead and post the errors you think you see in this one?

John Fields

Looks similar to the other one. It sure has the same hazards, multiple sloppy
async inputs into a clocked state machine.

---
In this one there are no races being run, so your "critique" is
bogus.
---

But the complexity alone makes it absurd as a zcd.

---
You might think so because your sow's ears can never become silk
purses, but the price for precision is often complexity.

For example, you can't use a sundial to check an atomic clock's
accuracy.
---

It needs a floating 5 volt
power supply, too.

---
You're grasping at straws since even a simple 5VDC out wall-wart
connected to the mains through a transformer will satisfy that
requirement.
---

By the time you do that, and the clock, you'll be up to 30
parts maybe. Still not isolated!

---
Doesn't need to be, since the output pulse can be.
---

I real life, that bridge rectifier, loaded by 100K, will not give clean zero
indications.

---
When the AC cycle goes through zero volts, where's the output of the
bridge going to be, then?
---

Try something clean and simple now and then.

https://dl.dropboxusercontent.com/u/...uits/ZCD_4.JPG

---
How close can you get to the 20kHz zero crossings the OP asked for
with that?

Or _any_ of your circuits?
---

But Jim loves yours, and that's all that matters.

---
Huh???