Here are a few things that I have seen that can casue this.
-Western farm center had thier PCC on XP PRO and more than 10 machines. over 10 machines MUST be setup on a server
-PCCharge suggested that we set the TIME OUT VALUE IN PCCHARGE to 10 SECONDS LESS than the RPRO time out value. with western farm the original time out in PCC was 99 seconds (max for the processor) and in RPRO 120 seconds. THIS WILL CASUE DUPES RPRO GIVES UP BEFORE PCCHARGE. they suggested setting PCCHARGE time out for 85 seconds and RPRO for 99 Seconds. this gives PCC 9 seconds to communicate to RPRO that it timed out. We had a double charge that literally took 100 seconds, 1 more that RPRO and PCC were configured for.
-ARCHIVE the DATABASE under UTILITIES>TRANSACTION ARCHIVE/PURGE> there is a setting for date, her database was 45MB Verifone says that the maximum size for a HEALTHY database is 20 MB for PCC. they said that over 20 MB the database starts to become unstable. We set the archive settings from the default 90 days and reduced it to 45 days. it made the database sie 21 MB.