Divide by N chip

That should have been: "Try to make sure you get a Philips 74HC4059."

Martin.

keith skrev:

yep, stay away from schematics, though I'd pick verilog its simpler and much less to type, but that a whole different discussion.

yep go look for "webpack", and the programming cable is basically a buffer on the parallelport you could even build your own, I think there's a schematic on the xilinx site.

-Lasse

Keith Williams skrev:

how about this in verilog:

module divider(clk,divider,clkout); input clk; input [7:0] divider; output clkout;

reg clkout; reg [7:0] count;

always@(posedge clk) begin if(count == 8'd0) begin clkout

but

I guess you know that spurs and phase noise go down by almost 6dB every time you divide a signal by two in frequency, so you could get pretty good phase noise if you took one of those Analog Devices ADF4360-X PLL chips with the integrated VCO, and divide down the output to 160-190kHz. If you start with a 2GHz-ish frequency and end up with 200kHz-ish, then the phase noise ought to go down to 80dB below whatever the PLL puts out, which should be pretty good and is probably limited by thermal noise etc. You do need a fixed divider which runs at a few GHz to divide down this VCO, and you could use something like an ADF4112 as the divider if you set it to mux the divider output to a pin. You will however need a PIC or something to program the two chips. You'll probably need a proper PCB to get decent performance, though I have had some success soldering 0402 components directly to the back of a LFCSP and wiring the RF up with 1mm semi-rigid coax - under a good microscope! I have not tried this with the parts I mentioned above though. Hope this helps Chris

Were these files posted to the a.b.s.e? There was nothing in the time frame that looked applicable. Thanks. -jon

My entire project involves a programmable binary divider, with programming being done by outboard switches to gate the desired combination of outputs back to the reset on the counter so that free running divide by N from 2 to 255 is all that's needed, with a 20 Mhz input to divide by N.

Do you really need 30MHz? Or is 20MHz also good as you mentioned in an earlier post. Because if 20MHz is ok, then you could do what I recently did with a PIC micro. The PWM output can be set to FOSC/N with N=[4..4096]. I used it to generate a clock/6 signal I needed in a PLL circuit. Just a few lines of code required.

Mine was an 8 pins 12F683 @ 1 Euro. But if you need more inputs for your switches, an 16F88 or similar would do I guess. If you don't really need fixed frequencies the 12F683 could read a pot with its analog input and vary N this way. Just an idea.

Joop

Thanks Martin,

There is good news. By tightening up some other parts specs, I was able to get the minimum usable clock frequency down to 12.8 Mhz. This allows a TCXO similar to those used in cell phones to be used...which makes it even easier to implement with high stability.

By the way, Phillips has sold their digital chip division to TI and the Phillips digital cmos parts are slowly disappearing from the market as the remaining stock is sold by distributors.

I did a spot check however, and the HCT and the HC parts are made by TI now, which is good news. In fact, TI is distributing the Phillips PDF's through their website....so it appears the Phillips parts will live on, at least for awhile longer.

I think I've found a source for inexpensive 12.8 Mhz TCXO's, so it looks like I'll use them with a couple of AND gates to get my divided down LF and VLF frequencies.

Regards,

These TCXO's are interesting. Where did you find them?

SioL

That could be *very* helpful in a project I'm working on, where'd you find those 12.8 Mhz TCXOs?

If you can make do with 8 stages (rather than 16) for your LO frequency, you may use the HC40103, which fits into a 16-pin package. It is more readily available and cheaper (about USD 0.50) than the HC4059 while the speed is the same (SGS-Thomson state a maximum clock rate of 40MHz typ at 5V - no temperature specified).

Note that standard AT-cut crystals in a HC-49U package have temperature coefficients of typically -0.5ppm/K around room temperature - that's only 50mHz/K at 100kHz. (Cell-phone frequencies are another matter.)

Martin.

The 74HC4059 or 74HCT4059 can go up to 40MHz at 5V. They are available from Philips and probaly other manufacturers such as Fairchild and TI as well.

Regards Anton Erasmus