Finding And Removing Records That Sum To Zero Based On Two Other Columns?

I have a dataset of about 70 columns but would like to remove any two rows that sum to zero (based on two columns) if the column "Units" are the same for the same "Bar code" thus to say "for the same 'Units' (for any two lines), check to see if their respective 'Bar Codes’ are the same and if so remove if their amounts sums to zero".

I have the following code which seems to be working but for relatively large data ( about 5k), it does not select all the data that meets the criteria:

Code:
SELECT *
FROM datatable
WHERE ((([datatable].[Unit] & [datatable].[Bar Code] & Abs([datatable].[Amount]))
In (SELECT datatable.Unit&datatable.[Bar Code]&abs(datatable.Amount)
FROM datatable
GROUP BY datatable.Unit&datatable.[Bar Code]&abs(datatable.Amount)
HAVING count(*)>= 2 and sum(datatable.Amount)=0)))
ORDER BY [datatable].Unit & [datatable].[Bar Code] & Abs([datatable].Amount);

identifying the bug in this code that makes it not pull all the data meeting the criteria.

See sample data below:

# Unit bar Code Amount
1. AAB Mac1 2.75
2. AAB Mac1 -2.75
3. AAB Mac1 1.24
4. AAC Mac3 35.00
5. AAC Mac3 -20.00
6. AAC Mac3 20.00
7. AAD Mac3 16.11
8. AAC Mac2 11.00
9. AAC Mac2 -11.00
10 AAC Mac3 12.05
11 AAF Mac3 -12.05
12 AAD Mac3 -16.11
13 AAC Mac4 35.00

I want to select records # 1, 2, 5, 6,7,8,9 and 12 since they are same units having the same bar code and their sums total zero.