I've been looking for evidence of the secondary clock matter in this old book I bought years ago, Introduction to Microcomputers V I Basic Concepts, by Adam Osborne. Concerning ROMs and microprocessors, and during a discussion of the external bus system, and the 4 groups the bus lines can be separated into, on page 5-5 I read: 4)Clock, power, and ground. There will usually be more than one clock signal on a system bus. We show just one, the [clock out symbol] output from the CPU. Real microcomputer systems will frequently have two or more signals, where secondary signals have frequencies that are some multiple or fraction of the primary clock signal.

-- Robert Manny (rmanny@thegrid.net), July 13, 1999

Answers

No Robert - it is not related.

The hypothetical secondary clock is that which is programmed in the firmware of a semi-conductor such as an embedded processor. This hypothetical clock is exactly that - a clock - just like the one in your kitchen only it has no physical hands or numbers on a dial. The clock signal on the bus is a portioned hi/low signal that if slowed way down would sound like a busy signal on the phone if you could listen to it. This clock signal is used to synchronize events. Many times this clock signal is divided or multiplied in some form via AND/OR/NOR gates or special TTL circuitry which will give either a faster or slower pulse depending upon the needs of the hardware. Example - A 486-33mhz system generally used a 66mhz oscilator with a divide-by-2 circuit to bring the "clock" signal to 33mhz. This is a sort of watered down answer, but there you have it!

Yours in COBOL... Dino!

-- (COBOL_Dinosaur@yahoo.com), July 13, 1999.

A old computer book written 40 years ago about the basic chip or circuit that was later incorporated into larger systems may shed some light on the secondary clock issue. It may have some technical information that later users of the chip never did learn about. If anyone is aware of such a book, it may help the debate.

-- Moe (Moe@wow.gom), July 13, 1999.

Robert:

I have used such a clock signal to keep real time, so I know it can be done. I'm not sure if this is what Beach is talking about, but any clock signal that can be accurately monitored can be used to keep real time.

-- Flint (flintc@mindspring.com), July 13, 1999.

The best explanation of the "secondary clock" issue (apparently a misnomer) might be that by Peter de Jager, which can be found at his website www.year-2000.com The article is entitled "A Crock of Clocks." According to Mr. de Jager, the secondary "clock" is really a counter. Hey, don't ask me--I'm a stranger to such things. But at least I could follow Mr. de Jager's explanation. Well, sort of.

-- Don Florence (dflorence@zianet.com), July 13, 1999.