Oracle 11g – spool file – Application of unlisted columns

I'm trying to spool large data sets (over 200 KB) in csv format and do not need to cite the columns. This is possible via the export settings of the SQL developer GUI, but I try to do it via a SQL script. The resulting csv file is then read by another application. To demonstrate this, I run the script below.

disable ECHO
disable FEEDBACK
define LINESIZE 60
disable RECSEP
put TRIMSPOOL on
Enable verification
set PAGESIZE 0
define SQLFORMAT csv
spool y: aa.csv
SELECT * FROM PRODUCT_COMPONENT_VERSION;
reeling

This gives the exit

"PRODUCT", "VERSION", "STATUS"
"NLSRTL", "11.2.0.4.0", "Production"
"Oracle Database 11g Enterprise Edition", "11.2.0.4.0", "64-bit Production"
"PL / SQL", "11.2.0.4.0", "Production"
"TNS for Linux:", "11.2.0.4.0", "Production"

I need the output as below (NO ROPE STRINGS)

PRODUCT, VERSION, STATUS
NLSRTL, 11.2.0.4.0, Production
Oracle Database 11g Enterprise Edition, Production 11.2.0.4.0.64bit
PL / SQL, 11.2.0.4.0, Production
TNS for Linux :, 11.2.0.4.0, Production

I would prefer not to postprocess the file (powershell, perl, etc.) after generation and before loading, because it simply adds unnecessary steps to a process.

How can I create a spool file in which I can specify the left and right speakers, even setting them to NULL.

Regards,
Tim