Hi,
I essentially have the same issue as a previous post
http://stauby.de/sarah_userforum/viewtopic.php?t=221
in which I cannot obtain a low scale BLSSM spectrum.

If I take the example BLSSM GUT parameter input file --> generate spectrum with SPheno --> take output parameters and fill entries of Leshouches.in.BLSSM_low then I obtain negative mass squared errors,
"Mass spectrum converged, but negative mass squared present.
Note: The following floating-point exceptions are signalling: IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
STOP Subroutine TerminateProgram".

I attach SPheno_out_BLSSM.txt, which is the output file of the example Leshouches.in.BLSSM file created with MakeSPheno[] in SARAH. I have copied the relevant information as required by the skeletal LesHouches.in.BLSSM file, and commented out parts not mentioned in this file. I have also added MINPAR 8, the MZp (mass of Z') otherwise this throws an error concerning the tadpole equations for BMuP; and also tan beta'. This can be found in the attached LesHouches_in_BLSSM_low.txt file.

What is the mistake here, or perhaps more easily: is there an example Leshouches.in.BLSSM_low which does not throw an error?

Many thanks!
Simon King

When you get this type of error messages it is usually a good idea to run some analytical checks. You can see whether all scalar squared mass matrices have positive eigenvalues by just using the analytical output of SARAH and replacing your input numerical values.