Searching \ for 'SquareWave' in subject line. ()
Help us get a faster server
FAQ page: techref.massmind.org/techref/index.htm?key=squarewave
Search entire site for: 'SquareWave'.

Truncated match.
'SquareWave'
1998\04\23@195218 by

Hi All,

I am facing a big challenge and before I will go ahead I would like to
ask anybody who has a good idea.
The solution not need to be cost effective.
I need to generate square wave with close to 50% duty cycle,
but quartz precision.
The frequencies are :   400Hz      -   4KHz   10Hz steps
4KHz       --   40KHz  100Hz steps
40KHz     --   400KHz 1000Hz steps
400KHz     --   4MHz 10000Hz steps

I really appreciate any idea.

Thank You,

Peter     ( schupetdvp.com )

At 04:31 PM 23/04/98 -0700, you wrote:
>Hi All,
>
>I am facing a big challenge and before I will go ahead I would like to
>ask anybody who has a good idea.
>The solution not need to be cost effective.
>I need to generate square wave with close to 50% duty cycle,
>but quartz precision.
>The frequencies are :   400Hz      -   4KHz   10Hz steps
>                                        4KHz       --   40KHz  100Hz steps
>                                        40KHz     --   400KHz 1000Hz steps
>                        400KHz     --   4MHz 10000Hz steps
>
>I really appreciate any idea.
>
>Thank You,
>
>Peter     ( schupetdvp.com )
>
>

How close to 50%? A simple "D" type F/F will "Square" up the incoming
waveform. As for the rest a simple Johnson decade counter will provide the
/10 (And other programmable divide values below 10). As for the base
generator/divider, perhaps something like a 74HC4060 will do, as this will
take a highish ref frequency, of say 16 or 32MHz. If I remember correctly,
the first output on this will be 4MHz "Square"

Other than that, some form of PLL could be used say a 74HC4046, followed by
a programmable / by n counter. With this you will easy get greater
resolution. But may introduce more problems.

Other factors to consider including phase noise, jitter, long term accuracy/
stability etc (The list goes on).

Just remember that there is more than one or two possible solutions, you
will have to find the one that best suites your needs.

Dennis

In message <000001bd6f0f\$e7a22e00\$362aa8c0jupiter.dvp.com>, Peter
Schultz <schupetDVP.COM> writes
>Hi All,
>
>I am facing a big challenge and before I will go ahead I would like to
>ask anybody who has a good idea.
>The solution not need to be cost effective.
>I need to generate square wave with close to 50% duty cycle,
>but quartz precision.
>The frequencies are :   400Hz      -   4KHz   10Hz steps
>                                        4KHz       --   40KHz  100Hz steps
>                                        40KHz     --   400KHz 1000Hz steps
>                        400KHz     --   4MHz 10000Hz steps
>
>I really appreciate any idea.

This is trivial if you use a DDS chip like the AD9850, I have a web page
describing a simple AD9850 system. The device actually generates a sine
wave, but has on-chip circuitry to convert the output to a square wave.
You'll find data on the ADI web site:

http://www.analog.com

You can use a PIC to control it.

Leon
--
Leon Heller: leonlfheller.demon.co.uk http://www.lfheller.demon.co.uk
Amateur Radio Callsign G1HSM    Tel: +44 (0) 118 947 1424
See http://www.lfheller.demon.co.uk/dds.htm for details of my AD9850
DDS system. See " "/diy_dsp.htm for a simple DIY DSP ADSP-2104 system.

Peter Schultz <schupetDVP.COM> write:
> I am facing a big challenge and before I will go ahead I would like to
> ask anybody who has a good idea.
> The solution not need to be cost effective.
> I need to generate square wave with close to 50% duty cycle,
> but quartz precision.
> The frequencies are :    400Hz    -   4KHz    10Hz steps
>                           4KHz  --   40KHz   100Hz steps
>                          40KHz  --  400KHz  1000Hz steps
>                         400KHz  --    4MHz 10000Hz steps

Peter,

You might want to take a look at the specifications for the Cypress
ICD2053B Programmable Clock Generator chip that was mentioned here a few
days ago.  Two-wire serial interface, tristate output, 8-pin package.

http://www.cypress.com/products/timi/icd2053b.html

The "down side" is the it will only cover the last part of your desired
range, since it is limited to [391KHz - 100MHz].  However, Cypress (or
another manufacturer) may have a similar chip designed for lower-frequency
operation.

Hope this helps get thigns started...

Frank McKenney            / OS/2 Advisor (OS2BBS)
McKenney Associates       / Richmond, Virginia / (804) 320-4887

if you have a chip that generates 391KHz to 100MHz
you could multiplex between the direct osc. output
and the output of a div-by-10 counter and a div-by-100 counter,
with multiplexing control done by the pic.

the multiplexing / clock dividing probably
would fit into a 16v8 or 22v10 PAL (from Lattice, AMD, etc.),
if space is a concern.

Jason Sachs
Electrical Engineer

Deka Research & Development
340 Commercial Street
Manchester, NH 03101
(603) 669 5139 x327

> {Original Message removed}
On Fri, 24 Apr 1998 04:26:45 +0100 Leon Heller
<leonLFHELLER.DEMON.CO.UK> writes:

>This is trivial if you use a DDS chip like the AD9850, I have a web
>page
>describing a simple AD9850 system. The device actually generates a
>sine
>wave, but has on-chip circuitry to convert the output to a square
>wave.

You need only the "counter" part of a DDS system (leave off the sine wave
table and DAC) to generate approximately square waves.  The "on-chip
converter" is likely just the MSB of the phase accumulator brought out to
a pin.  (Somewhat better results could be had by analog filtering the
sine wave and comparing it to zero.  This would help interpolate the
edges of the square wave between sample points)

At lower frequencies (up to maybe 50 KHz) DDS is possible with PIC
software (Ob PIC).

_____________________________________________________________________
Get completely free e-mail from Juno at http://www.juno.com
Or call Juno at (800) 654-JUNO [654-5866]

Hi everyone,

My name is James and I have been following this thread and maybe can help, or
at least make a suggestion.
On the Maxim web page you can check out a chip called the MAX038 High-frequency
Waveform Generator, its features are:
.1Hz to 20MHz
Triangle, sawtooth, sine, square, and pulse waveforms
15% to 85% Variable Duty Cycle

You can even order samples of this and other chips - they are free and come
with full documentation.  I have ordered several chips from them and have

Hope this helps and have a good day.

--
James E. Merritt N0SRB                  Iowa State University
jemiastate.edu                         Center for Nondestructive Evaluation
Electrical Engineering student          x-ray tech/modeller/programmer

FYI - The URL is http://www.maxim-ic.com/ and you can find info on the chip
at http://www.maxim-ic.com/ProParts.htm

{Original Message removed}
> I am facing a big challenge and before I will go ahead I would like to
> ask anybody who has a good idea.
> The solution not need to be cost effective.
> I need to generate square wave with close to 50% duty cycle,
> but quartz precision.
> The frequencies are :    400Hz    -   4KHz    10Hz steps
>                           4KHz  --   40KHz   100Hz steps
>                          40KHz  --  400KHz  1000Hz steps

Hi Frank Mckenney

I dont know if you have got your answer for the 40KHz to 400KHz in 1Khz sq.
wave problem, but here's a possible solution.

Getting a range of freq. on a linear scale is a problem, bcos the parameter
that you can ACTUALLY control is the time-period of each oscillation. This
creates a non-linearity and makes the clock division a precision problem.
eg. lets say the base clock period is 4MHz, to get 400KHz I div. by 10, OK.
But to get 399KHz I need to divide by 4000/399= 10.025. If I req. a +/-
0.25KHz accuracy I need the division to be +/-.007. For the next freq. I
req. a division of  10.050. And so on. (these steps arent linear). To get
over this, I'm suggesting a direct method. Measure the freq. and control it.
The signal path would be
PIC--> 16 bit DAC --> VCO --> triple decade counter --> PIC counter
To measure the freq. setup a 1 Hz interrupt (since this uses the PIC crystal
clock it would be quite accurate) which reads the freq. from the counter and
resets it for the next msrmt. (might have to software extend to 16/24 bit
counting). Use the measured freq. to close the control loop. Basically a
digital PLL.

I know this will not work as-it-is, but might seed some ideas.

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...