diff --git a/src/get_functions.jl b/src/get_functions.jl
index 070fc70a..41b08a36 100644
--- a/src/get_functions.jl
+++ b/src/get_functions.jl
@@ -1085,7 +1085,7 @@ function get_irf(𝓂::ℳ;
                 constraints_violated = any(𝓂.obc_violation_function(zeros(num_shocks*periods_per_shock), p) .> eps(Float32))
 
                 if constraints_violated
-                    solved = false
+                    # solved = false
 
                     for algo ∈ [NLopt.:LD_SLSQP, NLopt.:LN_COBYLA] 
                         # check whether auglag is more reliable here (as in gives smaller shock size)
@@ -1093,7 +1093,9 @@ function get_irf(𝓂::ℳ;
                     
                         opt.min_objective = obc_objective_optim_fun
 
-                        opt.xtol_rel = eps()
+                        opt.xtol_abs = eps(Float32)
+                        opt.ftol_abs = eps(Float32)
+                        opt.maxeval = 500
                         
                         # Adding constraints
                         # opt.upper_bounds = fill(eps(), num_shocks*periods_per_shock) 
@@ -1106,13 +1108,13 @@ function get_irf(𝓂::ℳ;
 
                         (minf,x,ret) = NLopt.optimize(opt, zeros(num_shocks*periods_per_shock))
                         
-                        solved = ret ∈ Symbol.([
-                            NLopt.SUCCESS,
-                            NLopt.STOPVAL_REACHED,
-                            NLopt.FTOL_REACHED,
-                            NLopt.XTOL_REACHED,
-                            NLopt.ROUNDOFF_LIMITED,
-                        ])
+                        # solved = ret ∈ Symbol.([
+                        #     NLopt.SUCCESS,
+                        #     NLopt.STOPVAL_REACHED,
+                        #     NLopt.FTOL_REACHED,
+                        #     NLopt.XTOL_REACHED,
+                        #     NLopt.ROUNDOFF_LIMITED,
+                        # ])
                         
                         present_shocks[contains.(string.(𝓂.timings.exo),"ᵒᵇᶜ")] .= x
 
@@ -1123,7 +1125,7 @@ function get_irf(𝓂::ℳ;
                         end
                     end
 
-                    solved = solved && !constraints_violated
+                    solved = !constraints_violated
                 else
                     solved = true
                 end
diff --git a/src/plotting.jl b/src/plotting.jl
index e0945be0..9104c3df 100644
--- a/src/plotting.jl
+++ b/src/plotting.jl
@@ -507,7 +507,9 @@ function plot_irf(𝓂::ℳ;
                     # check whether auglag is more reliable and efficient here
                     opt.min_objective = obc_objective_optim_fun
 
-                    opt.xtol_rel = eps()
+                    opt.xtol_abs = eps(Float32)
+                    opt.ftol_abs = eps(Float32)
+                    opt.maxeval = 500
                     
                     # Adding constraints
                     # opt.upper_bounds = fill(eps(), num_shocks*periods_per_shock) 
@@ -520,19 +522,19 @@ function plot_irf(𝓂::ℳ;
 
                     (minf,x,ret) = NLopt.optimize(opt, zeros(num_shocks*periods_per_shock))
                     
-                    solved = ret ∈ Symbol.([
-                        NLopt.SUCCESS,
-                        NLopt.STOPVAL_REACHED,
-                        NLopt.FTOL_REACHED,
-                        NLopt.XTOL_REACHED,
-                        NLopt.ROUNDOFF_LIMITED,
-                    ])
+                    # solved = ret ∈ Symbol.([
+                    #     NLopt.SUCCESS,
+                    #     NLopt.STOPVAL_REACHED,
+                    #     NLopt.FTOL_REACHED,
+                    #     NLopt.XTOL_REACHED,
+                    #     NLopt.ROUNDOFF_LIMITED,
+                    # ])
                     
                     present_shocks[contains.(string.(𝓂.timings.exo),"ᵒᵇᶜ")] .= x
 
                     constraints_violated = any(𝓂.obc_violation_function(x, p) .> eps(Float32))
 
-                    solved = solved && !constraints_violated
+                    solved = !constraints_violated
                 else
                     solved = true
                 end