In the AD9914/15 Profiles mode, can anyone confirm that:
1. The POW word should be calculated according to equation given on page 19 of datasheet
2. This equation gives a 14 bit word
3. This word should be stored in the 14 LSB bits of the 16 bits Profile POW register.
As per usual with the many things in the AD9914/15 datasheets you can sort of make a guess but can't be absolutely sure because the datasheets aren't absolutely clear. For the ASF it seems a bit clearer, it's a 12 bits word from the equation in page 19 and is stored in the 12 LSB bits of the 16 bits ASF register, this is clearer because the register map says the 4 MSB bits are open. From the ASF I made the deduction for the POW but want to be 100% sure.
Please can anyone help.
We're still thinking of ways on how to verify this in bench. We'll get back if we will find any leads.
Why you are at it could you also determine what the answer is for 'Parallel Data Port mode' as well please. For this mode, page 29 says 16 bits POW word and so does figure 41 which show all the POW registers in 'direct mode' to be 16 long. Yet we know (I think) that it is 14 bits at the DDS core (figure 30). So what is it eaxctly? 16 or 14 and if it is 14 as I think it is then for 'Parallel Data Port' is it the 14 LSB or 14 MSB bits?
I suggest to visit this thread because this has something to do with your inquiry: http://ez.analog.com/thread/39519. Upon reading this,you'll know that the appropriate divisor to be used is 2^16 since the width of the POW register is 16 bits.
This follows that function mode 0100b and 0101 has no problems because it fits to the 16 bit POW register. Now for function modes 1001b and 0111b, you'll notice that POW is using an 8 bit word but with different bit placements;POW[7:0] and POW[15:8]. This means that using these modes, the DDS device still uses the 16 bit wide register. However, you can only fill in 8 bits of it. This could be from 0 to 7 or 8 to 15.
With the AD9914/AD9915 POWs should be 16-bits long. This is mentioned right in the features section on page 1 and as you found in other places as well. Figure 30 I believe is a typo. If you notice in that same figure on the left hand side it also says that same phase offset control data bus is 16-bits wide.
Thanks Loujie and Kevin
Could you then confirm the followings:
1. POW in both Profiles and Parallel Data Port are 16 bits
2. The equation on page 19 for POW (delta phase = 2PI(POW/2^14) or 360(POW/2^14)) is incorrect, 2^14 in the denominator should be 2^16 instead.
3. Figure 30 showing POW truncated from 16 to 14 is incorrect.
I couldn't find any relevant information from your thread Louijie.