Suppression: Interpretation

I'm using AMOS. I have 3 predictors and one dependent. they are all positively correlated with each other. However one predictor has a negative regression weight, which is somewhat illogical. Removing that predictor causes the other IV regression weights to drop, so it seems that it is having a suppression effect.

The means of my dependent variable are pretty high. So if it is having a negative effect, it isn't very strong, despite the fairly high negative regression weight (-0.6). The positive regression weights of the other two IVs are 0.5 and 0.9 respectively. I'm trying to figure out how to interpret this result. Is it possible that the positive IVs are simply overpowering the negative IV? I'm an SEM newbie.

Just as an experiment, I removed the covariance lines between my IVs. The two positive IVs saw a light regression weight drop, but the negative regression weight dropped to -0.13. That seems significant but I have no idea what it means or even if that is a valid thing to do. I know I can't leave them like that but I just wanted to see what would happen.

Any advice is appreciated.