Hi! I have some problems with the **conversion** **value** from what I read in the scheme, and the Hex data that it sould be wrote in the AD1701...

For example, with a Tone module, if I put a

- 1 Hz oscillation, I receive the 0x015E data value (I read it in the output window)

- 2 Hz oscillation, I receive the 0x02BB data value

- 3 Hz oscillation, I receive the 0x0419 data value....

So, it seems that if I pass from 1Hz to 2Hz, I have to add 0x015D, but if i pass from 2Hz and 3 Hz I have to add 0x15E (which is the 1 Hz value...)

With some calculations, I assume that it is an approximation problem. In fact, if we want 1 Hz, we have to

349.5 * 1Hz = 349.5 ==> so we assume the decimal value 350 that is in fact 15E...if we want 2 Hz, we do

349.5 * 2Hz = 699 ==> 2BB, that is the value for 2 Hz...

In this way, I can get my desired Frequency, considering the approximation, but I can't calculate any value under 1. With the ToneGenerator, I have a minimum resolution of 0.01 Hz...so

- 0.01Hz = 0x3

- 0.02Hz = 0x7

- 0.03Hz = 0xA...

So here we have to add 3 or 4, so 3.5... but for now I don't need a so little resolution...

A similar problem is that in the "State Var Filter" (the one with numeric Q) the

-1Hz = 0x44A (the minimum resolution)

-2Hz = 0x894

-3Hz = 0xCDE

It seems that it is more linear, because I have to add 0x44A, from 1 to 2, and for 2 to 3...so HEX=(frequency)*0x44A = frequency* 1098...but

if I calculate it for 10 Hz, I expect for 10980=2AE4...but I see 2AE5(10981)!!

I need the RIGHT formulas to get the RIGHT values from ANY blocks in SigmaDSPs....

For now, it is very difficult to implement in a uC the right functions to know what number we have to put in the DSPs...

OR am I in WRONG with something!?!? if so, I apologize...but I really need your help!! Thank You All!!

It looks like both of your calculations for the different parameters in SigmaStudio are correct. The difference between 0x2AE4 and 0x2AE5 is just one lsb at a level of 2e-23, which could simply be due to rounding the calculation at each step. I believe that you do have the right formulas to an acceptable precision that you need for your application; these differences shouldn't affect the performance of your calculations.