Look the note of metalink: ID 452924.1
You can create a diskgroup with one member, but is not the recomendation,
for external redundancy:
5) Maximize the number of disks in a disk group for maximum data distribution and higher I/O bandwidth.
6) Create LUNs using the outside half of disk drives for higher performance. If possible, use small disks with the highest RPM.
7) Create large LUNs to reduce LUN management overhead.
8) Minimize I/O contention between ASM disks and other applications by dedicating disks to ASM disk groups for those disks that are not shared with other applications.
sorry by my bad english,
Normal - Two Way Mirroring. One failure group.
High - Three Way Mirroring. Two failure group.
That's incorrect. You need at least two failure groups for normal and three failure groups for high redundancy.
By default every disk or LUN is a failure group, but you can also add several disks to the same failure group to setup striping. The setup depends on your hardware layout, disk space requirements and redundancy needs. ASM provides redundancy only between failure groups.
Keep in mind that ASM is not disk mirroring. It's not RAID. ASM mirrors based on file level and allocation units. It knows about Oracle database files and therefore can use variable and optimal stripe sizes based on file templates. You can add and remove disks to reconfigure your disk grouips online without having to re-initialize your storage, which you cannot do with RAID or external redundancy.