Error #28, ring buffer overflow V24, occurred only when transferring data from a notebook to an 820 controller. Transfers from the controller to the notebook continued to work, and the previous computer supported both directions. The evidence points to host-side serial timing or buffering rather than a general cable or controller failure.
Identify the Direction-Specific Failure
Start by testing each transfer direction separately. In this case, controller-to-notebook communication succeeded, while the reverse direction produced error #28 immediately after transfer began. Note that the available description calls these directions “download” and “upload”; confirm the software’s terminology before troubleshooting so the failing direction is recorded unambiguously.
Confirm the V24 Serial Configuration
| Setting | Configured value | Required check |
|---|---|---|
| Baud rate | 9600 | Match both endpoints |
| Data bits | 7 | Match both endpoints |
| Stop bits | 1 | Match both endpoints |
| Parity | Even | Match both endpoints |
| Software flow control | XON/XOFF | Verify against the controller handshake mode |
| Controller handshake indication | RTS-LINE | Resolve the apparent mismatch with XON/XOFF |
The controller indicated RTS-LINE while the notebook program used XON/XOFF. Although communication worked after a timing adjustment, do not assume mixed software and hardware handshaking is correct. Select one supported flow-control method at both endpoints, or verify the required cable and controller behavior from the 820 documentation.
Apply the Confirmed Timing Correction
- Open the serial communication settings in NC-linc.
- Set
LINE DELAYto50 milliseconds. - Keep the baud rate at
9600and retry the notebook-to-controller transfer.
This setting restored operation at 9600 baud. Attempts at 4800 and 2400 baud did not provide a complete solution, so baud-rate reduction alone was not the confirmed correction.
Verify the Repair and Avoid Unsupported Settings
Repeat transfers in both directions and confirm that error #28 does not recur. Then repeat the failing direction under the same notebook workload that previously exposed the overflow. Controller locations 5019 and 5021 and baud-rate bit patterns were discussed only for an 810 controller; do not apply them to the 820 without model-specific documentation.
FAQ
How do I fix error #28 ring buffer overflow V24 on an 820 controller?
In NC-linc, set LINE DELAY to 50 milliseconds and retest at 9600 baud. This timing change restored the failing notebook-to-controller transfer in the documented case.
What serial settings were used with the 820 controller?
The recorded settings were 9600 baud, 7 data bits, 1 stop bit, even parity, and XON/XOFF in the notebook program. Both endpoints must agree, and the controller’s RTS-LINE indication requires separate verification.
Will reducing the baud rate eliminate V24 ring buffer overflow?
Not reliably in this case. Tests at 4800 and 2400 baud did not fully solve the problem; a 50-millisecond LINE DELAY allowed operation at 9600 baud.