Java Security

I'm been tinkering with encryption and decryption using AES. I've been having problems replicating the test vector results from the AES homepage.

http://csrc.nist.gov/CryptoToolkit/aes/rijndael/rijndael-vals.zip

For example, test 0 in the file cbc_e_m.txt:
- CBC mode
- 128 bit key
KEY=00000000000000000000000000000000
IV=00000000000000000000000000000000
Plain=00000000000000000000000000000000
Cipher=8A05FC5E095AF4848A08D328D3688E3D

I haven't been able to get the same result. I tried the PKCS5 padding, and that result was twice as long was their result. I then tried NOPADDING and got:
MyResult=66E94BD4EF8A2C3B884CFA59CA342B2E.
At least this one was of the same size.

The following link, http://www-cse.ucsd.edu/~fritz/rijndael_test.html, refers to the same test vectors, and make the follwoing quote:

"Please note the difference between the "normal" (single iteration) test vectors and the Monte Carlo tests that are a part of NIST's test set before emailing me about failing test vectors. For example, if you're looking at the ciphertext for the all-zero 128 bit key and plaintext and see C34C052CC0DA8D73451AFE5F03BE297F, you're looking at the wrong (Monte Carlo) test."

I'm wondering if the text vector page is not applicable in my instance, for encryption reasons beyond my knowledge. Anybody know of any other test vector pages? Has anybody had any similar experiences?

Thanks,

John