You should also get equivalent factor scores by using the Maximum Likelihood method under EFA in SPSS. I don't know about AMOS, but the EFA program should calculate them directly for you.
This is the procedure I suggest: (1) run an EFA (exploratory factor analysis, e.g. PCA) with SPSS; this allows you to get the "factor loadings" of each scales' item. (2) next, calculate the validity of this scale (construct) through AVE, CR, Cronbach's alpha. (3) If your construct is reliable, then you can aggregate your observed variables (items) to get your latent variable; this can be easily done in SPSS by the function "compute variable" by adding your items and then calculate the mean (it's better if you weigh your items by using the factor loadings). (4) Now you have created your latent variables: just drag them into AMOS if you're interested in performing a path analysis; otherwise, draw your latent variable and associate the items in AMOS, this allows you to simultaneoualy assess the factor loadings and the regression weights of your hypothesized conceptual model.
In sum, use SPSS to conduct EFA and aggregate your weighted variables. Next, use AMOS to run your SEM analysis.