Tie down the loose ends...
- Visually inspect the aperture actuator lever for damage
- Manually test the iris mechanism on the lens by observing the blades from the lens rear while flicking the actuator lever. The blade action should be snappy with NO hesitation or sticking.
- If the lens has an aperture ring, repeat the point above at all click stops on the ring
If the iris action is not snappy, the lens iris mechanism is part of the problem and is likely the sole cause. Moving forward, lets make sure the shutter is OK:
- Spot metering is a loose end in that use for general shooting is probably the largest cause of inappropriate exposure reported on this site. It probably is not a factor in the problem described on this thread, but might mimic the symptoms you are seeing. Turn it off unless taking readings with the intent to place exposure. Spot metering is inappropriate for general photography.
- Avoid testing using auto-ISO or TAv modes. Small things like minimum and maximum range for ISO are often left out of trouble descriptions such that the shutter may be firing when the settings are dead wrong and not a reflection of the metering. This sort of thing is not always obvious in the EXIF.
- For exposure issues, M mode is our best friend in that settings are settings stuff like auto-EV comp, shadow and highlight protection and such are not active
- Don't use EV comp while testing unless essential to the test
To screen shutter function as a cause for your overexposure:
- M mode
- Set aperture to f/2.8 (wide open). In this setting, the stop-down mechanism and any issues with it are controlled for in regards to overexposure.
- Set shutter speed to 1/800s, the setting at which there was overexposure in the example photo.
- Set the ISO according to the meter reading and do a single test exposure to confirm a reasonable exposure. Use the camera's matrix metering. Do not use the spot meter.
- Shoot a burst in continuous drive mode the same as before and scan for overexposed frames
If there are overexposed frames, either the light changed during the burst or the shutter timing is wrong or there is a mechanical issue with the shutter action or there is a serious issue with the image capture/processing pipeline.
There is one case that was not tested and that is a problem with aperture control by the body. It is relatively easy to check for underexposure traceable to aperture control, but somewhat difficult to test directly for inconsistent frame-to-frame stop-down. The closest we can do is to repeat the above test with the shutter at the same setting (1/800s) and the aperture set to f/10.0, the set aperture for the failed example photo. The assumption of course is that the shutter has tested good at that speed.
Well, there it is in detail. If there is no clear cause from direct examination or M mode evaluation, the answer may be a subtle issue of configuration or usage.
Steve