DOS/PDOS stuck at "Maximum index for non-local projection operator"

[bkappes]

I am running graphene on Pd in a slab geometry using USPP. The system size varies from 92 to ~500 atoms. While all SCF calculations converge, subsequent DOS/PDOS do not. The calculation is stuck with the following output to the OUTCAR:

"[redacted]
initial charge density was supplied:
charge density of overlapping atoms calculated
number of electron 3152.0000000 magnetization 410.0000000
keeping initial charge density in first step


-----------------------------------------------------------------


Maximum index for non-local projection operator 331"

The INCAR used for this calculation:
"SYSTEM = USPP Graphene on Pd(111)

Startparameter for this Run
NWRITE = 2
ISTART = 0
ICHARG = 1
IRDMAX = 525 ! IRDMAX from SCF: 479

Electronic Relaxation 1
PREC = Accurate
NELM = 90
ISPIN = 2
EDIFF = 1E-05
IDIPOL = 3
LDIPOL = .TRUE.

Ionic Relaxation
NSW = 0

DOS related values
ISMEAR = 0
SIGMA = 0.05
NEDOS = 1000
LVTOT = .TRUE.
LORBIT = 2
RWIGS = 1.3 2.6

Electronic Relaxation 2
ALGO = Fast
LREAL = Auto
VOSKOWN = 1
ISYM = 0
NPAR = 1
"

I have tried several variations including:

-- explicit inclusion of IRDMAX at 1.2 & 4 x IRDMAX (from SCF) and 10,000 (as suggested from http://cms.mpi.univie.ac.at/vasp/guide/node83.html)
-- exclusion of IRDMAX
-- Monkhorst-Pack automatic k-point generation at various grid densities

Two older posts, by ynwu on 15 Jul 2009 and dnrl on 10 Jul 2008, appeared relevant, but no resolution is reported for these entries.

Thank you for your help,

-Branden

[bkappes]

After reconsidering the categorization of this question, I moved this thread to the "Bugreports" forum.
<span class='smallblacktext'>[ Edited Wed Apr 21 2010, 06:14PM ]</span>