Technicow
08-06-2008, 11:00 PM
I've had my carvewright for 2 weeks now and personally think it is a tremendous piece of robotics. Like any high end device, there will be never ending refining of various parts.
I got hit by the "Clean Sensor" error this past weekend. After checking the forums here I checked the options-7, and saw a never ending zero return from the board sensor. Again, after reading posts on this it became apparent that emitters and/or detector had separated from the PCB in the sensor unit.
I have a replacement on it's way (I have only positive things to say about any calls or emails with carvewright, their support for their customers is top notch. It reminds me of how companies used to treat folks.).
I chose to pay the $5 core and keep the busted unit to see if I can determine a solution. My day job is making devices for the military which goes out in the field. Generally the stuff needs to be pretty rugged and everything-proof.
Through observation, I've seen what I suspect as the achilles heal on this component. The PCB is glued to the plastic enclosure. The emitters are on the "other side" of the rigid plastic (those who have looked know what I mean here). The issue is that vibration loosens and breaks the glue holding the PCB in place. Once that glue is not holding, the PCB gets shaken and it "worries" the leads for the emmiters & detector. They wind up giving way.
The solution I've devised is to fix the breaks in the connections. Then stiffening the emitters & detectors to the PCB. Afterwards, I have the PCB assembly operate as though on a "shock absorber" within the plastic enclosure. The "wall" between emitters / detectors is removed (dremel with a cutting wheel).
So far, I've disassembled the unit and resoldered all broken connections. I've put a light pipe onto the detector, and have encased the detector & light pipe in shrink tubing. The shrink tubing is about 1/32 to 1/16 inch below the emitters (to avoid overspill of light and triggering false positives). I then placed this assembly in the C & C machine and checked light levels with wood / paper / nothing etc... Everything checks out value wise (A/D is reporting valid values).
I've now encased the PCB assembly in Silicone and put it into the modified plastic housing. It is drying tonight. Tommorow (or Friday at th latest) I'll install it and test it. If all goes well, the emitters & detector will move with the PCB and not hit anything thus negating the erosion issue (in otherwords it may be a fix). If not...then I'm out 5 bucks but gave it a shot!
If it works, I'll post up a how to if anyone is interested and will gladly share this with carvewright if it looks like an avenue which may help in this part.
Other then all of that, great machine. Really enjoying it.
Andy.
I got hit by the "Clean Sensor" error this past weekend. After checking the forums here I checked the options-7, and saw a never ending zero return from the board sensor. Again, after reading posts on this it became apparent that emitters and/or detector had separated from the PCB in the sensor unit.
I have a replacement on it's way (I have only positive things to say about any calls or emails with carvewright, their support for their customers is top notch. It reminds me of how companies used to treat folks.).
I chose to pay the $5 core and keep the busted unit to see if I can determine a solution. My day job is making devices for the military which goes out in the field. Generally the stuff needs to be pretty rugged and everything-proof.
Through observation, I've seen what I suspect as the achilles heal on this component. The PCB is glued to the plastic enclosure. The emitters are on the "other side" of the rigid plastic (those who have looked know what I mean here). The issue is that vibration loosens and breaks the glue holding the PCB in place. Once that glue is not holding, the PCB gets shaken and it "worries" the leads for the emmiters & detector. They wind up giving way.
The solution I've devised is to fix the breaks in the connections. Then stiffening the emitters & detectors to the PCB. Afterwards, I have the PCB assembly operate as though on a "shock absorber" within the plastic enclosure. The "wall" between emitters / detectors is removed (dremel with a cutting wheel).
So far, I've disassembled the unit and resoldered all broken connections. I've put a light pipe onto the detector, and have encased the detector & light pipe in shrink tubing. The shrink tubing is about 1/32 to 1/16 inch below the emitters (to avoid overspill of light and triggering false positives). I then placed this assembly in the C & C machine and checked light levels with wood / paper / nothing etc... Everything checks out value wise (A/D is reporting valid values).
I've now encased the PCB assembly in Silicone and put it into the modified plastic housing. It is drying tonight. Tommorow (or Friday at th latest) I'll install it and test it. If all goes well, the emitters & detector will move with the PCB and not hit anything thus negating the erosion issue (in otherwords it may be a fix). If not...then I'm out 5 bucks but gave it a shot!
If it works, I'll post up a how to if anyone is interested and will gladly share this with carvewright if it looks like an avenue which may help in this part.
Other then all of that, great machine. Really enjoying it.
Andy.