2003 Mercedes SL500 vario-roof fault diagnosis
aegea traced a stuck roof through stored codes, pump swaps and latch-switch checks. Operation returned after replacing the RVC roof module and tightening a loose latch connection, with the effect of each change left unresolved.
Select a photo to view its source post. The source may require sign-in.
A roof that stopped after pivot lubrication
aegea’s 2003 SL500 initially opened its roof normally. With the roof partly closed, he sprayed Dri-Film lubricant on the pivots; afterward the roof stopped responding to the console switch and the rollbar buttons did not move the rollbar. The pump still whirred. The next day, a closing command lifted only the left side of the trunk lid by two or three inches before stopping. This timing was the owner’s starting observation, not proof that the lubricant caused the failure.
Stored codes did not identify a confirmed failed part
The owner read stored B1648, B1197 and B1262-001 faults involving the rollbar deployment solenoid and roof limit-switch states. He cleared them and tried the roof again: it stayed stuck, while those codes did not return. He noted that he had recently disconnected the right rollbar assembly, which could explain the stored open-circuit code. The account therefore separates stored fault text from a confirmed component diagnosis.
The pump swap did not settle the diagnosis
aegea initially suspected insufficient hydraulic pressure and swapped in a freshly rebuilt pump. After that swap, the pump no longer made a sound and the roof still would not move. He checked fuses and exchanged relays without restoring the pump noise. The original pump, suspected of low pressure, subsequently worked in his separate 2004 SL500. He also reported that lubricant had reached microswitch casings and that later electronics cleaning had not restored roof movement. None of these observations establishes a single failed pump or switch.
A solenoid question and unresolved switch states
Later stored B1659-002 and B1661-002 faults led the owner to ask whether a pictured gold-colored part was the relevant solenoid. In the same post he corrected his reading: Y55/1y6 referred to the load-assist solenoid valve. His later live-data list still showed invalid or unrecognized roof-switch states. The photograph accompanied a component-identification question; it is not confirmation that the pictured part failed or needed replacement.
Pump noise returned, but the roof still stopped
The owner reported that Top Hydraulics had checked the valve body on another car. He then cleaned the circuit board, connectors and solenoids, and reattached the left windshield latch containing switches S91/2 and S91/3. The pump whirred again, yet a closing attempt still raised the left side of the trunk lid only a few inches. The return of pump noise therefore preceded the eventual recovery and did not by itself resolve the roof fault.
Recovery after an RVC replacement and connector work
Weeks later, aegea reconnected the windshield-latch wiring instead of using bypass latches. The roof eventually closed, but the display did not recognize it as closed and the windows remained partly down. He replaced the roof module/RVC computer and tightened a loose windshield-latch connection accessible in the trunk. Afterward, the console switch closed the roof with the display correctly reporting roof closed, and he said the roof worked properly. Both the module replacement and connection work belong in this reported sequence.
The final cause remained uncertain
JettaRed asked whether the loose connection alone explained the recovery or whether too many adjustments had been made to know. The thread supplies no isolated test separating the effect of the RVC replacement from the connector work. The result remains aegea’s report of restored operation after several changes, rather than a proven loose-connector-only fix.
A separate owner still had a non-running pump
EHiser described a separate 2004 SL500 whose roof and rollbar still did not work after a pump rebuild and a roof-ECU replacement. This was another owner’s unresolved problem, not a continuation or confirmation of aegea’s result.
Source reply ↗ · EHiser