Description.
In case of making QR Code with Japanese Strings. Actobat DC makes wrong codes. This bug is the same as InDesign CC's behavior.
However, InDesign fixed this problem since ver.14.

Step to reproduce...
1. Make a form field with QR Code.
2. Type Japanese characters in the form field and Acrobat will make QR Code that is incorrect.

Here is more detail and prease refer attached image.

Original Strings 「Text1<CR>問題は日本語を正常に扱えるかどうか」
Decode Stirngs 「Text1<CR>問題は日本語を正常に扱縺医ｋ縺九←縺�縺�」

42 75 46 57 87 43 10 de 59 58 fe 9a 18 ce 38 1a
fe 69 7a 5e 69 ca ce 8a a9 ee 38 29 2e 6a da 3e
5b 8b 8e 38 1a be 68 9b 18 09 ce 09 78 c2 17 9c
14 11 81 a7 38 24 63 02 58 00 ec 11 ec 11 ec 11

0100 0010 0111
8bit byte mode　4
Characters Length　0x27(39)
T　54
e　65
x　78
t　74
1　31
<CR>　0d //Carrage Return
問　e5 95 8f
題　e9 a1 8c
は　e3 81 af
日　e6 97 a5
本　e6 9c ac
語　e8 aa 9e
を　e3 82 92
正　e6 ad a3

常　e5 b8 b8
に　e3 81 ab
扱　e6 89 b1
To Kanji mode　8 //Here is an wrong behavior.
Characters Length　09
....
ce 09 78 c2 17 9c

Text1<CR>

At start, Actobat selects 8bit byte mode and its correct behavior. However, after adding 39 bytes, Acrobat changes encoding mode. Why change mode here?
The String that I provided is the same code range and doesn't need change mode. Additionally, Your Kanji mode has crucial bug.
The Kanji mode needs Shift JIS Character encoding. But Acrobat doesn't change encoding. Do not use Kanji Mode with UTF-8.

I recommend discard Kanji mode and forced to set 8bit binary mode with UTF-8 for all characters the same as InDesign.
Please refer below report in InDesign uservoice.
https://indesign.uservoice.com/forums/601180-adobe-indesign-bugs/suggestions/37181575-indesign-makes-incorrect-qr-code-words-and-we-can

Adobe Community Professional
Ten A.