For anyone who is interested, I am trying to resolve the blurry image cold weather issueI've had the drone since Christmas. My first impression of the recorded video was that it was a bit soft. This is a screenshot of early video after a few minutes of flying. Outside temp 40 deg. Drone started out at 68 degrees.
My first indication of a real problem was when I left the drone in my car, and then took it for a flight Drone and outside temp approx 40 degrees:
Did a bunch of online research (including here) and discovered that this is a common problem. Now this was never a problem with my two BB1 drones, and the cameras shouldn't be that different. So I decided to try the "Anthony Fix". He was able to refocus his lens such that the image would be sharp over a wide temperature range.
Before tearing it apart, I took this video as a baseline. This is with the drone and outside temperature at 30 degrees:
As a reasonable goal, I decided that I want good video from summer temps down to 40 degrees.
I took the lens apart and discovered the following:
The sensor chip appears to be glued to the camera barrel.
As Anthony pointed out, the rear of the sensor is missing heat sink compound.
The lens barrel includes a fine thread, probably intended for focus adjustment, either in an early design or in a different application.
There is no physical contact of the plastic parts. The body of the lens barrel measures .460" and the bore in the camera barrel measures .490". This leads me to believe that the camera is adjusted and permanently fixed using sophistcated optical methods of controlling the relative positions through fixturing, with X-Y-Z movement, using the video image as feedback. The adhesive might be a UV-cure variety.
Having picked all the adhesive off of both flanges, I now have two reasonably clean reference surfaces which I can use to control the lens position along the optical axis. I used a couple of Post-It pads as adjustable shim, each page being .004" thick. The goal is to evaluate focus as a function of flange separation, and then to find a separation that will produce a sharp image throughout the temperature range.
I quickly discovered that the .004" spacial resolution of the Post-It pad is not fine enough. In order to get close to sharp focus I need to add layers of magic tape, .002" to the mix. Even at the spacial resolution of .002", I'm not confident that I'm hitting the sweet spot. I'll be using real feeler gauges(+/-.001") when I revisit. So, what does this all mean?: Lens distance is super-critical, as is lens tilt. Finding a spot that accommodates a wide temperature range is hard to imagine. But I'm not there yet.
To be continued...
-Lars
My first indication of a real problem was when I left the drone in my car, and then took it for a flight Drone and outside temp approx 40 degrees:
Did a bunch of online research (including here) and discovered that this is a common problem. Now this was never a problem with my two BB1 drones, and the cameras shouldn't be that different. So I decided to try the "Anthony Fix". He was able to refocus his lens such that the image would be sharp over a wide temperature range.
Before tearing it apart, I took this video as a baseline. This is with the drone and outside temperature at 30 degrees:
As a reasonable goal, I decided that I want good video from summer temps down to 40 degrees.
I took the lens apart and discovered the following:
The sensor chip appears to be glued to the camera barrel.
As Anthony pointed out, the rear of the sensor is missing heat sink compound.
The lens barrel includes a fine thread, probably intended for focus adjustment, either in an early design or in a different application.
There is no physical contact of the plastic parts. The body of the lens barrel measures .460" and the bore in the camera barrel measures .490". This leads me to believe that the camera is adjusted and permanently fixed using sophistcated optical methods of controlling the relative positions through fixturing, with X-Y-Z movement, using the video image as feedback. The adhesive might be a UV-cure variety.
Having picked all the adhesive off of both flanges, I now have two reasonably clean reference surfaces which I can use to control the lens position along the optical axis. I used a couple of Post-It pads as adjustable shim, each page being .004" thick. The goal is to evaluate focus as a function of flange separation, and then to find a separation that will produce a sharp image throughout the temperature range.
I quickly discovered that the .004" spacial resolution of the Post-It pad is not fine enough. In order to get close to sharp focus I need to add layers of magic tape, .002" to the mix. Even at the spacial resolution of .002", I'm not confident that I'm hitting the sweet spot. I'll be using real feeler gauges(+/-.001") when I revisit. So, what does this all mean?: Lens distance is super-critical, as is lens tilt. Finding a spot that accommodates a wide temperature range is hard to imagine. But I'm not there yet.
To be continued...
-Lars
