Compare invoice values from 2 tables with a 5p tolerance

Experts

We are converting from our old system to a new system (rental application) - we are importing contracts from the old to the new - I have to run parallel Invoice runs on both then extract this invoice data from both systems and do a comparison

Due to differences in rounding and VAT calculations I need to compare using a 5p tolerance e.g.

£96.60 - old data
£96.63 - new data

This will be ignored in my query but:

£96.60 - old data
£96.66 - new data

This will flag up a problem in my query

I have a one - to - one exact matching query which I need to modify to cope with this tolerance:

------------------------------------------------------------------
SELECT [OldData].[Trans_date], [OldData].[Ax_acc], [OldData].[Cust_ref], [OldData].[Depot], [OldData].[Cont], [OldData].[Ax_code], [OldData].[SumOfinvl_qty], [OldData].[SumOfAmt], "OldInvoiceData" AS DSet

FROM [OldData] LEFT JOIN  [NewData] ON [OldData].[Cont] = [NewData].[Contract] and [OldData].[Ax_code] = [NewData].[Item] and [OldData].[Trans_Date] = [NewData].[InvDate]

WHERE ((([OldData].[SumOfinvl_qty])<>[NewData].[SumOfQty]))
   OR ((([OldData].[SumOfAmt])<>[NewData].[SumOfLineAmt])) ----------Line to be modified for tolerance
   OR ((([NewData].[Contract]) Is Null))

UNION ALL SELECT [NewData].[InvDate], [NewData].[CustAcct], [NewData].[CustOrder], [NewData].[Depot], [NewData].[Contract], [NewData].[Item], [NewData].[SumOfQty], [NewData].[SumOfLineAmt], "NewInvoiceData" AS DSet

FROM [NewData] LEFT JOIN [OldData] ON [NewData].[Contract] = [OldData].[Cont] and [NewData].[Item] = [OldData].[Ax_code] and [NewData].[InvDate] = [OldData].[Trans_date]

WHERE ((([NewData].[SumOfQty])<>[OldData].[SumOfinvl_qty]))
   OR ((([NewData].[SumOfLineAmt])<>[OldData].[SumOfAmt])) ----------Line to be modified for tolerance
   OR ((([OldData].[Cont]) Is Null))

ORDER BY [OldData].[Cont], DSet;
------------------------------------------------------------------

This Exact match query works perfectly well on like for like data (I manually modified the data to generate some errors to prove this)

The Lines I need to be  modified are highlighted above

I have tried using:

   OR round((([OldData].[SumOfAmt]+0.05)/100)*100)<>round((([NewData].[SumOfLineAmt]+0.05)/100)*100)
and
   OR round((([NewData].[SumOfLineAmt]+0.05)/100)*100)<>round((([OldData].[SumOfAmt]+0.05)/100)*100)

but this doesn't give me anything (this was suggested by somone to try)

Please help as this is now very urgent

If there is any more Info please ask

Thanks in advance

Dave