I read the information on the DDS that uses a 32 bit word command to generate a frequency from 0 to 30 mhz. I also got info that the prop chip can do it but I don't know assembly or Spin.  I have my circuit working perfectly Thanks to Bean with a freqout up top 20Khz with 1 hz accuracy of 1 hz.  The um-fpu chip is 14.00 dollars just to write 32 bit words. Then another 14 for the DDS chip.. I also was wondering what is the highest freqout possible (it does not have to be accurate above 20khz.) It just needs to send a high freqout to pulse a neon transformer and light up a tube. I was testing one and I set my Function generator to 271khz to light the tube I was using. I did my PCB layout to include both chips and a fram chip too incase I need to go that route... I am trying to figure out which way to go on this?

Also the EEPROMs are 2.40 @ I rather buy another sxchip 28 or 48 for 2.51 and get 50 million times better chip.

:eyes:

The SX uses an 8-bit core.  Internally all math and logic operations are done using 8-bit numbers.  Larger numbers can be "processed" through the use of additional instructions and logic.  SX/B adds this additional processing whenever Word variables are used instead of Byte variables.  So the answer to your question is that the SX can generate a 32-bit number if you explicitely program it to do this.  However it can not natively do this with a built-in instruction.

  You may want to look at the 32-bit math routines found on SxList.com  If you want to learn to do 32-bit math on the SX you can save $14.00 - $2.51 = $11.49.  If you think it will take you more than an hour or two to set that up and want to save time. $14 might not seem so steep!

  I can not speak with authority on the maximum frequency limitation of the SX.  However, I expect it will be in the MHz range when coupled with an appropriate oscillator.  It should be more than enough to drive your neon transformer.

I hope this helps.

Thanks Sparks..... I thought it could do it which will save me on getting the CO-Chip. I think I will switch to the SX48 which should have room to do  everything I need it to do and I won't have to add another SX 28. I will still use the FRAM chip to store the values that I need to run in automode.
BTW Sparks you have a great website with a lot of fun stuff. Thanks for all your help.

The SX48 should be able to process your 32-bit numbers fine if you are willing to take the time to figure out how to do it.

  As for the website, I assume you are referring to SparkFun.com.  I agree.  It is a great site!  Unfortunately I had no part in that and am not affiliated with the site.  I have noticed, however, that a good formula for success in the hobby world is to give information away for free and charge reasonable prices for your products.  Parallax and SparkFun both do a great job of this!

  Let us know if you get 32-bit math working.  I am sure there will be others interested in doing 32-bit math at some point.

If you want to do 32bit math easily, check out C4SX:
http://forums.parallax.com/forums/default.aspx?f=7&m=203602
The cc1b compiler supports 32bit integer, fixed point and floating point math.
It uses operator overloading so you can use the normal expression symbols

I looked at the link but I like using SX/B for now. I will try to do it in SX/B if possible I don't know how I will do it yet. It looks like I will have to eventually learn assembly, or C or another language to do more complicated programming. There should be a way to do it with SX/B I was hoping someone pointed me in the right direction......Thanks for the info Peter.

Carlos,
 Here is a routine you can add to the program I already wrote for you that will set the frequency to 8 times the value given. It is a modified version of the original subroutine. You can have both in your program.

 This SetFreqX8 works up to 388KHz (by specifying a value of 48500).

Bean.

Wow Bean thanks again. You make it seems so easy. I would of tried to change the entire Sub. It just needed your touch to a few lines and there it is.  This chip and Parallax just keep amazing me. When I think I have to go to a bigger chip or another chip people like yourself have solutions for newbies and others on this forum.

The code for 32 bit
DO
     IF freqWanted < 12208 THEN
       freqWanted = freqWanted << 1 ' * 2
       INC optReg ' Adjust prescaler
     ENDIF
   LOOP UNTIL freqWanted >= 12208
Original code
DO
     IF freqWanted > 24414 THEN
       freqWanted = freqWanted >> 1 ' / 2
       DEC optReg ' Adjust prescaler
     ENDIF
   LOOP UNTIL freqWanted <= 24414
 
With just a few lines of code you can do it just like that..  I was reading all the other stuff and did not have a clue where to begin. I am out of Ram but you just changed a few lines of the original code and it now generates 32 bit.  I don't think I can find the words to thank you bean and the other great members on this site.. I am not sure enough to explain what you did but it looks like you multiplied the freqwanted by 2 and inc the adjust prescale .
new code
IF freqWanted > 70 THEN
   optReg = 1 ' Assume prescaler is 1:1
   ' Adjust prescaler to keep frequency above 12207 original code
IF freqWanted >= 382 THEN
   optReg = 4 ' Assume prescaler is 1:8
I posted what I see and hope that maybe this will help other members that may have the same question..