Need very sharp squarewave circuit

[[email protected]]

Could somebody please point me to a simple, inexpensive circuit that
will generate a square wave at a frequency of about 25 or 30 MHz with
*very* short rise-times. The rise- and fall-times definitely need to
be less than 5 nanoseconds, and preferably down around 1 or 2 nsec. It
would be nice if the circuit used readily-available parts: e.g. from
DigiKey or Mouser or (better yet) Radio Shack.

The exact frequency isn't very important, neither is the frequency
stability nor the symmetry - just the sharpness.

TIA, john w.

 

[Phil Allison]

Could somebody please point me to a simple, inexpensive circuit that
will generate a square wave at a frequency of about 25 or 30 MHz with
*very* short rise-times.

** Was discussed here back in August 2003.

" Fast square wave generator -- bad ringing ... "


http://groups.google.com.au/group/s...t+square+wave+"&rnum=1&hl=en#83f430680c24bb8f




......... Phil

 

[[email protected]]

Could somebody please point me to a simple, inexpensive circuit that
will generate a square wave at a frequency of about 25 or 30 MHz with
*very* short rise-times. The rise- and fall-times definitely need to
be less than 5 nanoseconds, and preferably down around 1 or 2 nsec. It
would be nice if the circuit used readily-available parts: e.g. from
DigiKey or Mouser or (better yet) Radio Shack.

The exact frequency isn't very important, neither is the frequency
stability nor the symmetry - just the sharpness.

Fairchild Semiconductor's Tiny Logic Ultra High Speed (UHS) parts might
do it for you

http://www.fairchildsemi.com/ds/NC/NC7WZ16.pdf

If you really want fast edge-speeds, have a look at Motorola's (On
Semicondictor, FreeScale) ECLinPS parts, which are much faster, but
only produce ECL-level voltage swings (0.8V). I've used ECL to drive
long-tailed pairs of BFR92 (NPN) and BFT93 (PNP) wideband transistors
to produce bigger voltage swings with (marginally) sub-nanosecond
transition times, but you've got to put them on a carefully laid-out
printed circuit board if you want to get that kid of performance.

http://www.freescale.com/files/timing_interconnect_access/doc/app_note/AN1579.pdf?srch=1

 

[[email protected]]

Thanks to everybody who offered suggestions. What I finally did was to
get a cheap old early-Pentium motherboard with a missing CPU, but a
functional clock.

The "CLK" pin of the empty CPU socket had a pretty decent SQ wave at
about 100 MHz, and the "Pci-CLK" pins of the PCI sockets had a fairly
decent 40 MHz SQ wave.

john w.

 

[Winfield Hill]

[email protected] wrote...

Thanks to everybody who offered suggestions. What I finally did was to
get a cheap old early-Pentium motherboard with a missing CPU, but a
functional clock.

The "CLK" pin of the empty CPU socket had a pretty decent SQ wave at
about 100 MHz, and the "Pci-CLK" pins of the PCI sockets had a fairly
decent 40 MHz SQ wave.

Which "early-Pentium" still had a high-speed off-chip clock generator?
Didn't Intel give that up after their 40MHz '386 with its 80MHz clock?
I thought the Pentium clocks ran at the external bus speed, e.g. 50MHz
for the original 75 - 100MHz Pentiums, and 66MHz for the 200MHz Pentium
and 333MHz Pentium-II modules, etc.

 

[[email protected]]

You may be right. I don't actually know for sure that it was a
Pentium. The guy who gave me the board said it had been a Pentium, but
the only thing that I'm actually sure of is that it is a Socket-7
board. For my purposes, it didn't matter.

john w.