divide-by-N

If I have a clock, 10 MHz maybe, and I'd like to divide by N, with an

8-pole dip switch setting N, what's a cheap standard chip to use? Output duty cycle doesn't matter.

John

Reply to
John Larkin
Loading thread data ...

google is your friend

You might also want to try cracking a textbook.

Reply to
AZ Nomad

any good ideas in this?

formatting link

something with a presetable counter

-Lasse

Reply to
langwadt

Not very helpful in this situation.

Textbook? I don't need theory, I need parts.

Somewhere I remember an 8-bit divide-by-N with built-in switch pullups, but I can't find it. Or maybe I'm thinking of an address comparator with pullups, which would lead to a 2-chip solution.

Some old CD-series stuff is nice but too slow.

I guess two LS/ALS or F-series 4-bit presettable counters would work, cascaded, no pullups needed.

74F579 is sort of a creaky old beast but might work.

There is always the old dipswitch-feeding-bipolar NAND gate trick.

John

Reply to
John Larkin

CD40102 (binary) or 103 (BCD) perhaps?

-- Silvar Beitel

Reply to
Silvar Beitel

HC4059

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

Gack! Think modern... 'HC292 family or similar. ...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Modern? I am surprised nobody yet suggested a PIC.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

I was going to say, find some spare pins on your FPGA. He puts the things in everything else he makes...

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
Reply to
Tim Williams

Actually, the 74HC40103 is the binary divider,and the (less popular)

40102 the BCD part.

Farnell stocks the NXP 74HC401013 in DIP and SOIC packages; it has a maximum clock frequency opf 35MHz versus the TI equivalent that falls over at 25MHz.

Farnell does also carry (but with no stock at the moment) the CD74HCT40103 which is even slower, falling over at 14MHz, but even that would be good enough for John Larkin's application.

I've been using the part for at least twenty years now - it has always struck me as a chip designed exactly for that job.

Programmable parts give you more options, but they rarely sell for a dollar apiece in small quantities.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

Might be difficult to do 10MHz with a PIG^hC.

ngs

Perhaps he is using it for FPGA pre-scaler?

Reply to
linnix

I've seen a tricky solution in the last Elektor (a German paper magazine). The input signal is delayed with some gates of a 74HC04 and then the delayed and original signal is XOR'ed with a 74HC86. With this trick you'll get a pulse for each edge. The pulse is fed into a 74HC40103. Finally a

74HC73 makes a symmetrical output. This provides divides from 1 to 255 with 8 dip switches.

Of course, I wouldn't use such a circuit. Delay lines with gates usually are bad design pratice and you'll need 4 ICs for it. This is a solution with one chip:

formatting link

The webpage is a bit outdated, meanwhile I've fixed some problems with the SPI communication and now even the gate level simulation works. The latest project files, with updated testbench:

formatting link

With 8 dip switch inputs and without all the other stuff I've implemented, like the prescaler and configurable duty cycle, it should fit perfectly into a XC9536, which you can get for a buck:

formatting link

I have some XC9572 here and can program and test it with real hardware, instead of just the simulator, if you like and send you a programmed chip by snail mail, but I assume you might have already some CPLDs lying around.

If you don't need symmetrical output, a 74HC40103 should be fine. But it costs nearly the same as the much more powerful XC9536.

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de
Reply to
Frank Buss

Actually, the HC40103 looks great. SO16, cheap, doesn't need to be programmed. The switch inputs will need pullups, but that's to be expected.

I tend to forget about NXP.

Thanks,

John

Reply to
John Larkin

I've got a customer who wants us to do a pulse generator board that's so simple it's embarassing. DIPswitch settable rate, pot settable pulse width... and that's all! A few HC chips will do the whole thing.

The HC40103 looks ideal, just a SO16. The 4059 suggestion is appreciated but it's huge and very, very weird.

John

Reply to
John Larkin

I'd think maybe a PIC, but there might be a programming-less solution like a counter/divider chip that's less trouble.

-- Les Cargill

Reply to
Les Cargill

On a sunny day (Tue, 25 Aug 2009 21:24:01 -0500) it happened "Tim Williams" wrote in :

Yes, FPGA, PIC, 8 pin DIP switch ????? These days we have EEPROM and using RS232 you only need _one_ pin to set it. Very few DIP switches in modern equipment.

In the old days sound cards, and I/O cards especially, were full of DIP switches. Some mobos have jumpers.... do it right and you need no pullups. But avoid mechanical contacts if you can. FLASH will be OK, as it will lose value when everything else FLASH based does...

Reply to
Jan Panteltje

Right. You'd think I'd remember that, since I designed the parts at RCA back in 1974! :-)

It was.

-- Silvar Beitel

Reply to
Silvar Beitel

I WOULD TALK HIM INTO THUMB WHEEL SWITCHES. WHY HAVE THAT POOR GUY DO BINARY????

THEN YOU CAN USE A PIC TO DO THE BCD TO BIN AND TIMING, AND USE POT ON THE ADC FOR PULSE WIDTH,.

I HAVE DONE THAT THUMB WHEEL THING ALMOST HALF A CENTURY AGO, AND CUSTOMER WAS VERY HAPPY WITH IT.

TODAY I JUST MIGHT CONSIDER A NUMERIC FOIL KEYPAD.

I AM NOT READY FOR TOUCH SCREEN DISPLAY YET FOR THIS..

DONE IN UPPER CASE TO SAVE BITS AND REDUCE GLOBAL WARMING.

Reply to
Jan Panteltje

ings

I've used the 4059. Once you figure it out is is very handy. It is just the ticket for a thumb switch controlled divider. The funny

2/5/10 divide on the end means that you can often handle the case where the number the user sees needs to be other than a simple integer.
Reply to
MooseFET

A CD4040 that gets reset when a giant diode AND says to could work too.

Reply to
MooseFET

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.