The rank conditions ISN'T verified!---Please help!!

[apple564]

Hi,
We are trying to replicate a similar model of Iacoviello and Neri (2010). We are stucked with the rank conditions. We always got this error information, 'There are 10 eigenvalue(s) larger than 1 in modulus for 9 forward-looking variable(s)'. We also searched the forum that there could be two potential problems, including parameters and the timing of the variables. We checked the parameters that lie in the reasonable range. It seems changing parameters won't help. The timing of the variables we think are standard, but not too sure about the variable 'l'. We will be much appreciated if anyone can help us to get through this. I uploaded the mod file.