After playing with the software part and getting PS3Eye to capture under Windows, I decided to see how difficult is to remove the IR blocking filter from the lens of this camera.
The pics are included below, but since last update of my blog, I cannot get the text back.
Please check my blog for more and up to date information on this camera.
For information and pictures of the newer version of the PS3Eye camera go here.
Both versions of PS3Eye camera feature 24C64 (8K) EEPROM chip that contains USB descriptor table as well as the camera’s firmware.
I checked the status of the WP pin and it is tied to the GND (EEPROM is not write protected!).
This means that camera’s firmware is software upgradeable.
Excellent tutorial, with a nice tour of the PCB to boot!
Although I think i’d class this as a medium for difficulty, since you had to cut around the lens to get it out. Would you mind if this goes on the new wiki?
Excellent tutorial, with a nice tour of the PCB to boot!
Although I think i’d class this as a medium for difficulty, since you had to cut around the lens to get it out. Would you mind if this goes on the new wiki?
Keep up the good work!
The plastic around the IR filter is very thin. With a sharp took it will take seconds to get it lose. I removed mine with a small flat screwdriver, and this is why it took me a bit longer and why it doesn’t look as clean.
Hadn’t kept up for a while so I spent a few hours doing the same thing without the benefit of your great disassemble post.
Couple of gotchas for those attempting this. My Ps3 cam had some glue sealing the tongue on front half of the base-sphere to the rear half. (this is the tongue that traps the USB cord). Finally had to dremel through the seams to release it.
Have done a complete disassembly of the lens trying to figure out the best way to pop out the IR filter. I have not done this yet as the rear most element - that does indeed appear to be the IR filter - also seems to have some curvature to it. In my mind this means removing it will affect the focus(which might not be a bad thing - see below). Any thoughts?
Normally you can just break the spot of glue on the focus locking ring to reset the focus in these kinds of situations, but again, when they were making this one there must have had some extra epoxy on the employee snack cart because there is so much glue locking the lens in place I’m afraid I will shatter the lens before I break the seal. It’s beyond a gentle knifing as the glue has seeped into the threads. For use as a web cam placed a foot or so from the user, the focus is sharp, but for use at any distance it’s very soft and desparately needs a refocus. I’m amazed the glue hasn’t prevented the wide/narrow ring from rotating.
I might just get another to see if they are all like this or kit-bash the two to get one really good one.
Now armed with Alex’s pictures and confirmation that the rear element is indeed the IR filter, I’ve had at it.
Mine had a metal collar sealing the filter in. Cutting didn’t really work but a gentle prying back of the collar all the way around released the filter no problem. In the lens tube below the filter is a spacer ring, and then below that the first of three lens elements. These appear to have been simply held in place by the presence of the IR filter and spacer rings creating a lens sandwich. In popping out the IR filter both lens element and spacer have moved back up the lens tube slightly taking up a bit of the space that the IR filter took. So I pushed them back down the tube and tacked the spacer ring with a spot of glue to restore the sandwich. So this should have put all the lens elements back where they belong.
Upon reassembly of the camera I have discovered two things.
1)There is a huge focus issue now. This could be caused by either this loose element not being in the right position, or the IR filter was needed to complete the light path. I’m not sure how the floating element could be out of place - there is just no other place it could live with the IR filter in position, and if the lens was meant to be up the tube more (assuming I pushed it down too far and the IR filter is not as thick as I think it is) then the focus would have changed every time the camera got knocked about as the lens would have had room to slide around with nothing to lock it in place. I’ve checked the other elements and they are in the original spots as well.
The second thing that I discovered upon firing it up post surgery is that the rear most element was indeed an IR filter. The camera picks up more IR than it did with its filter in place (although out of the box the camera had no trouble seeing to at least 900 nm before, albeit with lower sensitivity than my real IR 850-tuned camera.)
But now, even with an 800 nm and above passing filter in front of the lens to block all visible light, I’ll bet I can spot neutrinos!. It’s so bright as to be unusable. Who needs an IR source - everything seems to be emitting it. Any light from a projector hitting my touch screen (even with the projector having an IR blocker over the lens) blooms the camera (which, remember, still has the that 800 and above passing filter taped in front of the lens.)
If I hold the original IR blocking filter in front of the lens instead, it brings the levels back to what I would expect for a normal camera.
Even with the projector off, and barely enough light in the room to see; through the PS3 cam, it’s like daylight.
Going to try an 850 or 900 nm and above passing filter instead of my 800.
Also going to try to reinstate the IR filter to check focus. Now that it’s out the IR filter does look flat so it shouldn’t have anything to do with the focal distance.
Now armed with Alex’s pictures and confirmation that the rear element is indeed the IR filter, I’ve had at it.
Mine had a metal collar sealing the filter in. Cutting didn’t really work but a gentle prying back of the collar all the way around released the filter no problem. In the lens tube below the filter is a spacer ring, and then below that the first of three lens elements. These appear to have been simply held in place by the presence of the IR filter and spacer rings creating a lens sandwich. In popping out the IR filter both lens element and spacer have moved back up the lens tube slightly taking up a bit of the space that the IR filter took. So I pushed them back down the tube and tacked the spacer ring with a spot of glue to restore the sandwich. So this should have put all the lens elements back where they belong.
Upon reassembly of the camera I have discovered two things.
1)There is a huge focus issue now. This could be caused by either this loose element not being in the right position, or the IR filter was needed to complete the light path. I’m not sure how the floating element could be out of place - there is just no other place it could live with the IR filter in position, and if the lens was meant to be up the tube more (assuming I pushed it down too far and the IR filter is not as thick as I think it is) then the focus would have changed every time the camera got knocked about as the lens would have had room to slide around with nothing to lock it in place. I’ve checked the other elements and they are in the original spots as well.
The second thing that I discovered upon firing it up post surgery is that the rear most element was indeed an IR filter. The camera picks up more IR than it did with its filter in place (although out of the box the camera had no trouble seeing to at least 900 nm before, albeit with lower sensitivity than my real IR 850-tuned camera.)
But now, even with an 800 nm and above passing filter in front of the lens to block all visible light, I’ll bet I can spot neutrinos!. It’s so bright as to be unusable. Who needs an IR source - everything seems to be emitting it. Any light from a projector hitting my touch screen (even with the projector having an IR blocker over the lens) blooms the camera (which, remember, still has the that 800 and above passing filter taped in front of the lens.)
If I hold the original IR blocking filter in front of the lens instead, it brings the levels back to what I would expect for a normal camera.
Even with the projector off, and barely enough light in the room to see; through the PS3 cam, it’s like daylight.
Going to try an 850 or 900 nm and above passing filter instead of my 800.
Also going to try to reinstate the IR filter to check focus. Now that it’s out the IR filter does look flat so it shouldn’t have anything to do with the focal distance.
Indeed, the camera is really sensitive to the IR light. This is a good thing. The reason why everything is so bright even with the IR bandpass filter on, is that we have enabled auto brightness control by default. In the future versions of PS3EyeTest app I will add a full manual control of these parameters. Stay tuned…
Regarding the focus issue, we may be able to find some disolvent for the epxoy that they used to glue the lens casing and loosen it up enough so it can be turned.
Another route might be the use of a standard mini lens instead.
if you look closely at the second picture you can see two places around the edge where small black plastic loops have to be broken to get the plastic cover off. I had to use a small screw driver to pry under the edge to see what was stopping it from coming off. Then I saw those little clasps that must be broken. My suggestion is pull a little bit harder, but be prepared for them to give way. Thanks Alex I had no problems removing the lens hopefully I did not accidentally mess anything up, though it appears I have not yet.
I have it open, removed the IR filter, and now the focus is off, replaced the IR filter and it’s sharply focused again.
(Same trouble SteveB has)
So I have some questions:
Does anyone have an idea what kind of lens you would need to replace the standard one? (type/name?)
Is there a name for the mount (the square piece with the two pins at opposing corners and the holes for the screws 1/2 a side away clockwise of those corners)? Or is this not a standard mount and am I just going to have to ducktape/glue a new lens on the board?
Also, any ideas on the upper band limit of the CMOS? I’d like to use 980nm lasers for my LLP setup, and general info I’ve found on cmos/ccd chips gives me the impression I should be alright (most cmos/ccd go up to 1000+), but if someone could give me the definite proof on this one, I’d be much obliged!
Thanx for your tutorial. Very detailed. Step 2 was a bit harder then described, I had to use quite some force to get it of
After putting everything back I had some dust on the ccd. So ppl check before you put everything together again…
BTW: If you need to mount a 35mm IR bandpass filter. I have used the hood of a spraycan. It was an exact match. I drilled a hole in the topside that fits around the lans shaft.
