Primer details for "In Silico PCR" give different results

Reproduced in version 38.1, but the bug may have existed before.

1. Open _common_data/pcr_primer_design/gfp.fa.
2. Open the OP "In Silico PCR" tab.
3. Fill in:
   1. Forward primer: GAACTATATTTTTCAATGGGCACAAATTTTCTGTC;
   2. Reverse primer: TACTCATTTCCTCTTCTTGAAAAGTGACCTCAACAGGGTTAAGAACAACTTAATCTACCACTACAATTACCCGTGTTTAAAAGACAGTCACCTCTCCCACTTCCACTACGTTGTATGCCTTTTGAATGGGAATTTAAATAAACGTGATGACCTTTTGATGGACAAGGTACCGGTTGTGAACAGTGATGAAAGCCAAT;
   3. 3' perfect match: 0.
4. Repeat the following steps many times (~20 or more):
   1. Click "Show primers details". The "Primer Details" dialog appears;
   2. Enlarge the dialog to see the heterodimer characteristics (Delta G and Base Pairs);
   3. Expected: the hetero-dimer and its characteristics are always the same across runs, e. g. "Delta G=-7.4, Base Pairs=5",
      Current state: hetero-dimers may differ, for example 
   4. Close the dialog (you can use Alt+F4) and repeat the steps until the bug appears.

Perhaps repetitive steps need to be done quickly.