ewaldType="optEwald" with default parameters gives NANs for high temperature & small box size

[GoogleCodeExporter]

What steps will reproduce the problem?

run a single charged particle in a small box at high temperature
using ewaldType="optEwald"

<?xml version="1.0"?>
<Simulation>
  <SuperCell a="0.25 A" x="1" y="1" z="1"/>
  <Species name="eup" count="1" mass="1 m_e" charge="-1"/>
  <Temperature value="1000 eV" nslice="512"/>
  <Action>
    <SpringAction/>
    <CoulombAction norder="2" rmin="0." rmax="10.0" ngridPoints="10000" useEwald="true" ewaldType="optEwald" dumpFiles="false"/>
  </Action>
  <Estimators>
    <ThermalEnergyEstimator unit="eV"/>
    <VirialEnergyEstimator unit="eV" nwindow="128"/>
  </Estimators>

  <PIMC nclone="48">
    <RandomGenerator/>
    <!-- Thermalize-->
    <Loop nrepeat="1000">
      <ChooseSection nlevel="8">
         <Sample npart="1" mover="FreePBC" species="eup"/>
      </ChooseSection>
    </Loop>

    <!-- Sample data -->
    <Loop nrepeat="200">
      <Loop nrepeat="200">
        <ChooseSection nlevel="8">
          <Sample npart="1" mover="FreePBC" species="eup"/>
        </ChooseSection>
        <Measure estimator="all"/>
      </Loop>
      <Collect estimator="all"/>
      <WritePaths file="paths.out"/>
    </Loop>

  </PIMC>
</Simulation>

What is the expected output? 
sensible energies (presumably the same as one sees with ewaldType="trad")

What do you see instead?
NANs

thermo_energy (eV): nan, Av=nan +-nan
virial_energy (eV): nan, Av=nan +-nan

Original issue reported on code.google.com by jonathan.dubois@gmail.com on 12 Dec 2011 at 2:36