Kas,
I think you have a problem with the data or the model.
But first, let me address the question you asked. Because you had two different cluster datatypes, the Variant to Data function in the Model VI could not interpret the data put into the variant called "data" by the Main VI. So the values coming out of the Unbundle by Name in the model function were always zeros. This resulted in some very small values fro the rest of the function. In some cases it may have been small enough to be interpreted as zero.
With the variant problem fixed I started looking at the reasons for the Maximum iterations exceeded errors. I put probes several places within the model VI and plotted the occurrences of the error in the main VI. The error occurred more than 20 times. I also noted that NaN and Inf values were showing up on the Probes. When these values occur, some of the intermediate values are in the 1E80 to 1E100 range. When you try to take the exponential of a value that large, you get values out of the range of a double precision number. So I tried scaling the frequency (which is where the big numbers occur) by 1E-12. This eliminates the NaN and Inf values but the Maximum Iteration Errors still occur.
I also changed the stopping criteria to tolerances of 0.01 and max iterations = 10000. This slows the program significantly but does not eliminate the errors. Even at 100000 iterations and 500000 function calls the error occurs over 30 times. For the points where it converges the number of fucntion evaluation is rather randomly distributed between 15 and 300. Thius is what makes me think you have a data or model problem.
Lynn
