Mainframe to linux file transfer issue for special character

When I transfer mainframe file into linux environment using FTP.
special character { is getting translate into space.

File contents getting managed if we transfer file from mainframe to windows if we use below
codepage conversion it works for windows.
quote site sbdatacon=(ibm-237,iso8859-1)

Can anyone help me how it can be done for linux.

Did you try setting the transfer mode to binary and not text?

After using binary mode, I can not even read single character.

Given the fact that a binary transfer is a byte-to-byte (exact) file transfer, what does that mean on your mainframe side?

Just in case you're dealing with the EBCDIC from IBM... after the binary ftp transfer, try to convert the transferred file to ASCII:

dd if=your-ebcdic-file conv=ascii 

Step 1 > Transfer ebcidic file from mainframe to linux as binary

Step 2 > Converting binary file to ascii (Please share command)

Converting back from the mainframe to linux you can look at this construct:

iconv -f IBM037 -t utf-8 input_ebcdic.txt -o output.txt

Haven't we already "shared" the command?
There're plenty of discussions on the subject that Google comes with - one of which is this - you might not even need dd/iconv as a separate step - it seems that some FTP servers on the mainframe side provide the conversation ability OOTB.