Troubleshooting
Upload a PanelView 600 Program When DF1 AutoConfigure Fails
Short answer
When RSLinx AutoConfigure fails on a PanelView 600, configure the DF1 driver manually at the terminal's power-up defaults - 9600 baud, CRC error checking, no parity, 8 data bits, 1 stop bit, no handshaking - then upload with PanelBuilder32. The unreadable display does not block serial recovery, and the memory card is not a guaranteed source copy.
On this page
AutoConfigure failing in RSLinx does not stop the upload. Configure the RS-232 DF1 driver manually at the terminal's power-up values - 9600 baud, CRC error checking, no parity, 8 data bits, 1 stop bit, no handshaking - browse the driver, and run the transfer from PanelBuilder32 once the terminal appears and stays. The unreadable display costs you the terminal's own configuration menus, not the recovery; every value you need is entered PC-side. The memory card in the slot is not a source copy until a file from it opens as an application, so serial stays the primary path.
What the failure actually is #
PanelBuilder32 does not own the serial link. RSLinx builds the DF1 path; PanelBuilder32 rides it for the transfer. When AutoConfigure fails, the driver never had a DF1 conversation, so retrying the upload inside PanelBuilder32 fixes nothing.
The RS-232 connector on this terminal family carries shared duty. The PanelView Standard user manual assigns that port to printer and file-transfer work, with moving applications between a computer and the terminal as its file-transfer job. A port left assigned to a printer will not answer DF1 traffic even though the connector and its electrical interface are healthy. Normally you would read the port's assignment off the terminal's configuration screen; the unreadable display takes that check away, so you work from the PC side and treat printer mode as the last configuration suspect, not the first.
What you see, and where to look first:
- PanelBuilder32 cannot see the terminal: no DF1 path in RSLinx yet, the wrong COM port, or another program holding the port open.
- AutoConfigure fails on a proven port: the terminal is not speaking DF1-compatible framing, or the cable pinout is wrong.
- The manual defaults fail too: a COM-port or cable fault, or the port is on printer duty.
- The terminal appears once, then vanishes: connector seating or a broken conductor in the path.
- The display is unreadable: a diagnostic loss only, not evidence the communication port failed.
- A memory card is installed: proves nothing about a recoverable application file.
The six DF1 defaults to enter by hand #
| RSLinx driver setting | Value |
|---|---|
| Baud rate | 9600 |
| Error checking | CRC |
| Parity | none |
| Data bits | 8 |
| Stop bits | 1 |
| Handshaking | none |
These are the terminal's documented power-up values, which is why they are also what AutoConfigure expects to find. CRC is the default error checking for DF1; BCC is the alternative the protocol allows, and a mismatch on that one setting alone will sink the link.
The framing follows EIA-232 asynchronous characters: one start bit, the data bits, a parity bit only when parity is enabled, then the stop bits. Handshaking is not part of the frame at all - flow control runs on the port's control lines, which is why EIA-232 handshaking is not a baud fix and why "none" is the value to select here.
Worked example: a 2711-B6C10x at FRN 4.48 on a native serial port #
Inputs:
- Terminal 2711-B6C10x, firmware FRN 4.48, read from the unit being recovered.
- A Windows XP computer with a native serial connector, so no USB-to-serial adapter sits in the path.
- An RSLinx RS-232 DF1 driver added on the COM number that connector carries.
- The six values from the table above.
Rule in words: an EIA-232 character frame is the start bit plus the data bits plus the parity bit when parity is on plus the stop bits. Divide the baud rate by the frame length for characters per second; the reciprocal is the line time of one character.
Arithmetic, written out:
- Frame length: 1 start + 8 data + 0 parity + 1 stop = 10 bits per character
- Character rate: 9600 bit/s ÷ 10 bit/character = 960 characters per second
- Character time: 10 bit ÷ 9600 bit/s = 0.00104 s, about 1.04 ms per character
Result: the port puts one character on the line about every millisecond, so a working poll-and-reply exchange crosses in milliseconds.
Decision: browse the driver and hold it to a one-second standard. A terminal that appears and stays visible goes straight to the PanelBuilder32 transfer. A full second of silence with these values applied means the values are not the problem - the fault is the COM selection, the cable pinout, or a port left on printer duty, in that order of checking.
Restore the DF1 path step by step #
- Record the identity - 2711-B6C10x, FRN 4.48 - and plan a read-only recovery. Leave the terminal powered and in place.
- Connect the PC's native serial connector to the terminal's RS-232 port. Leave the DeviceNet side alone; the upload never crosses it on this model.
- Confirm which COM number Windows gives that connector, and close terminal emulators, programming packages and any background service able to open a COM port.
- In RSLinx, select RS-232 DF1 devices from the available drivers, add a new one on that COM port, and run AutoConfigure.
- If AutoConfigure fails, type the six values from the table, apply them, and browse again.
- Judge the browse against the one-second standard from the worked example: steady visibility proceeds to step 9, silence goes to step 7.
- Substitute hardware in order. Prove the PC COM port against any other serial device, then prove the cable. A cable sold as straight RS-232 says nothing about its pinout, and these terminals fail exactly there, on how DTE-versus-DCE wiring defeats a PanelView 600 serial connection. With no known-good cable on the bench, uploading an HMI program without the right cable pinout builds one from the documented pins.
- Only after the port and cable pass does printer assignment become the leading suspect - and with an unreadable display it is the one fault you cannot clear from the PC side.
- With the terminal visible steadily, open PanelBuilder32, select the DF1 driver as the transfer path, and upload to a new file on the PC.
- Stop there. Save the untouched copy and verify it before any migration work touches it.
What the memory card can and cannot give you #
The card may hold a usable application file; its presence proves nothing about what is on it or whether it opens. Serial upload stays the primary recovery path. A card file counts as a backup only when it opens as an application in PanelBuilder32 - the same acceptance standard the uploaded file has to meet. Until then, leave the card in place and keep the whole recovery read-only.
Prove the upload before the migration #
- The transfer completes with no communication error. A discovered node is not a recovered project.
- The saved file reopens in PanelBuilder32 and shows its screens and tags; why an uploaded HMI copy is not an editable project is the failure mode this reopen test closes out.
- One copy stays untouched permanently; conversion, retagging and cleanup happen on a second copy.
- The recovery record carries the COM number, the six values, the cable that worked, 2711-B6C10x and FRN 4.48.
Before the terminal comes out of the door, read the old HMI program ahead of panel replacement so the PanelView Plus project keeps the screens and tags the operator already uses. And when the PanelView 600 finally retires, the station can be rebuilt around industrial panel PCs instead of another dedicated terminal.
Dead ends that waste an afternoon #
- Rerunning AutoConfigure against an open circuit, a wrong pinout, or a COM port another program owns. It cannot correct any of the three.
- Troubleshooting DeviceNet while physically connected to RS-232.
- Debugging a USB-to-serial adapter that is not in the path; that question exists only when the path uses one.
- Treating the memory card as the source copy.
- Mixing recovery with migration edits in one file.
- Downloading anything to the terminal to prove the link. The application on that terminal is the only copy until the upload succeeds, so prove the path read-only and leave the write functions alone.
Questions asked
Do I upload over DeviceNet or over the RS-232 port?
Over RS-232. On the 2711-B6C10x, DeviceNet is the terminal's link to its controller and plays no part in an upload. The RS-232 connector is the printer and file-transfer port, which is exactly why a printer assignment there can shut the door on DF1 traffic. Leave the DeviceNet network alone while you recover the application.
Can I use a USB-to-serial adapter instead of the native port?
You can, but it adds a variable the native port does not have: USB-to-serial adapters vary in how they handle DF1 traffic, and not all of them work. Prove the path on the native serial port first. If the native port communicates and the adapter does not, the adapter is the fault and the terminal settings are not.
Does the unreadable display stop the recovery?
No. Every value the recovery needs is entered on the PC side in RSLinx and PanelBuilder32, and the upload never asks you to read the terminal's screen. The display matters in one case only: if the port proves to be assigned to printer duty, reassigning it means changing the terminal's own configuration, which needs the display. That is the one path where fixing the display comes first.
Is the memory card a copy of the application I can fall back on?
Not until it is proven. A card in the slot may hold an application file, but its presence says nothing about what is on it or whether the file is usable. Keep the serial upload as the primary path, and count the card as a recovery copy only if a file taken from it opens as an application in PanelBuilder32.
Does the FRN 4.48 revision matter for the upload?
Not to the serial settings. The DF1 values depend on how the port is configured, not on the firmware revision. Record FRN 4.48 with the catalog number anyway: the recovery notes, and any PanelView Plus conversion that follows, need to identify exactly which terminal the application came from.