Sweep with ADC control

Hi Skfir,

Your trigger generator is fine, the only change to make is its repetition rate. As with the VCO and similar blocks, frequency runs from zero to the Nyquist rate (fs/2) with a 0 - 1.0 input. Thus at a sample rate of 192KHz, the pulse frequency is f = 96000 * input. At 0.5 input, your pulse frequency is 48KHz (or 21uS between pulses). The sweep itself takes much longer than this to complete, so presumably you'd want the repetition rate to be much slower. For example, to make the sweep repeat twice per second, set the top DC source to 0.000021

Problem is, even with this change the triggered sweep doesn't output anything. I tried substituting every sort of imaginable input short of hitting it with a hammer, and simply could not get it to trigger. Also frustrating is the lack of documentation on this block, both from built-in help and on the wiki. Sorry but that's as far as I got.

Bob

Hello Bob! Thank you for the reply, I have already figured out my mistake with the frequency. What I found out is that a sweep is triggered fine if I use an ordinary (not externally controlled) pulse generator, but with a "voltage controlled" one there is no triggering. So might it be some bug in either the sweep, or VC pulse blocks?

Should I report about this in Technical Support, or AD monitors this forum?

Sorry Bob one more additional question, I found out that the linear sweep block "eats" the same number of instructions as the logarithmic one. Should it be this way, or it is also a bug? Because, I may be wrong, but if I am correct, logarithmic function is more difficult to calculate, then linear, so linear is supposed to "eat" less power, or am I wrong?

Hello Skfir,

ADI support engineers, including those involved with SigmaStudio programming, do monitor this forum -- but it may help to get their attention if you report this apparent bug in a separate post, using the format at:

https://ez.analog.com/docs/DOC-2043

If I'm not mistaken, all SigmaDSPs implement a hardware base-2 log function, which would explain why the log and the linear sweeps use a similar number of instructions. The new ADAU1452 even makes log and exponential functions available in the toolbox, and also uses them to perform division. By contrast, the ADAU1701's division algorithm uses Newton's iterative method, which takes way more instructions.

Bob

Thank you Bob!

I will report at once!

Thanks again.