MC4024

Well? Is someone going to analyse the schematic and tell me how it works? Accurate Algebraic solutions only, please! No hand waving :)

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona Voice:(480)460-2350 | | | E-mail Address at Website Fax:(480)460-2142 | Brass Rat | |

Hi Jim,

when I was a boy of 12 or so, I spent several years trying to understand the MC4044 datasheet. Alas I knew no engineers, but instead tried schoolteachers etc. No-one could tell me how to pick Wn or delta (I was attempting to design a floppy disk controller and needed a PLL. Knowing what PLL meant would have helped :)

Years later I learned it is quite possible to have a BE, ME or even PhD and STILL not know how to choose Wn and delta. And of course that you cant be a "real" engineer until you have mastered 2nd order systems.

I love old datasheets - they used to have information on them. I hate microcontroller datasheets - many manufacturers give little or no electrical specs.

Cheers Terry

Jim, I went through that exercise years ago, and came up with a modification that worked with random data. But, as I recall, there are two flip-flops. One gets toggled by a clock transition, and cleared by a reference transition. The other gets toggled by a reference transition, and cleared by a clock transition. So, one of the FFs tells you how much the oscillator is leading the reference, the other FF tells you how much the oscillator is lagging the reference. These control the U/D charge pumps.

How's my memory?

Tam

Lousy ;-) You're reciting Phase/Frequency Detector jargon... the MC4024 is a VCM (Voltage-Controlled-Multivibrator), with better than a

...Jim Thompson

that

transition.

leading

the

Isn't he thinking of the 4044?

Graham Holloway

that

transition.

leading

the

So you win some, lose some. I actually went to your web site, but must have clicked on the wrong thing. Came up with a phase detector. Never used a

Tam

Probably, except it can't handle "random" data, i.e. with missing transitions... it'll run off into la la land. But the "phase jerk" methods that I have posted will.

...Jim Thompson

"the ECL one that works with an LC", that would be another of my designs, the MC1648 (or, IIRC, now revitalized on a faster process as the MC12048).

...Jim Thompson

-- snip --

I learned how to do digital circuit design as a boy, just from reading datasheets. I was quite competent until I went to engineering school :>.

Are you thinking of the MC12148? Sadly discontinued long ago.

Thanks, - Win

(email: use hill_at_rowland-dot-org for now)

:)

modification

One

clock

lagging

That was the change. It did handle random data. Missing data transitions caused equal positive and negative error pulses.

Tam

See Lansdale....

and, for my analog/linear designs....

...Jim Thompson

Faster process?? I went Googling for PLLs, previously under the impression that a PLL was really cooking if it was running at 100 MHz. Man, was I in for a surprise! Of course, some of what I found was research papers on experimental devices, but...

The MC1648 would do about 300MHz (designed around 1968). The MC12148 probably will do close to 1GHz.

I've done custom chips as high as 3.8GHz.

I'm working a chip design right now (actually writing up the final design review as we speak) that runs at 2112MHz.

...Jim Thompson

And that one, "Introduction to Single Chip Microwave PLLs," is over 10 years old.

Thanks, - Win

(email: use hill_at_rowland-dot-org for now)