Solution Finder Get Tech Support

PI EXpert and PIXls report different values

Posted by: mk27 on

Attached is the same design done using PIXls Designer 9 and PI Expert 9.
A number of the items of reported data are different.
Can you please tell me which is correct.

 

First, I entered identical input parameters in both programs.
The transformer design it created was different.  (This is problem #1.)

 

I then made the following changes to the inputs, so that they would generate the identical transformer:
a) Efficiency in PIXls: 75%.
Efficiency in Designer: 77.3% (a 3% difference in value)

b) VD in PIXls: .94 volts
VD in Designer: 1.00 volts (when added to 24V output, a difference of 0.2%)

 

Now, with the identical transformer, I find the following differences in reported values:
1) KP .63 vs. .58 (difference of 8.6%)
2) KP_Transient: .31 vs. .36 (a difference of 14%)
3) IRMS: .41 vs. .37 (a difference of 10%)
4) BM: 2441 Gauss vs. 2270 Gauss (a difference of 7.5%)
5) BAC: 773 Gauss vs. 571 Gauss (a difference of 35%)
6) ISRMS: 1.512 A vs. 1.35 A (a difference of  12%)
7) IRIPPLE: 1.28 A vs. 1.09 A (a difference of 17%)

 

Can you please tell me which set of values is correct.

Files

Comments

Submitted by PI-Spock on 04/17/2013

Both sets of values are correct.

There are many solutions dependig on the set of inputs and constraints. Clearly the input values (such as efficiency) are not the same it is not surprising that the designs are different.

Regardig which ddesign is the better one - I would use the optimized design from PI Expert as it uses better input and constraints and gets a better overall design.. 

Submitted by mk27 on 04/17/2013

You recommend using the "optimized" design from PI Expert.  However, if I try running the optimizer on this design, while keeping the transformer selection fixed, PI Expert says that it cannot optimize.

To review what I did here - I tweeked two input values by a very small amount so that I got the identical design in both programs.  So, the question I asked  is not about which design to use, but rather: for an identical design, which reported values are correct.  As noted, some of the value differences are order of magnitude larger than the very small difference in input values.

Perhaps the more fundamental question is why, when I use identical input parameters, and identical transformer, the two programs give me different designs (e.g.: different inductance specification)?

Submitted by PI-Spock on 04/23/2013

The 2 designs are not exactly identical. For example the effieicny between the 2 designs is different. The calculations in both applications are accurate (a few differences may exist in the amount of derating used by both applications for instance) but the results should be very close.