- Fri Sep 18, 2009 8:03 am
#81319
I've got a friend who works with a Romer arm and a Sharkscan laser scanner. Nice equipment. Made me think if I could duplicate part of it.
I know camera systems are fairly simple these days, but I don't know if they can achieve the same accuracy.
What I was thinking of doing was using the mirror and maybe motor from an old laser printer. I have one with a 6 side polygon mirror. I'd like an encoder on it. I have a capacitive encoder that can do 2048 PPR.
The idea here is to AC couple a 1D PSD to the system. Get the mirror up to speed. We don't have to close loop control it, but we need fairly stable speed control from it. Now we replace the old IR laser diode with a red laser diode and the PSD. Lower gain but we can now see it. So, as we spin the mirror, we use one channel of the encoder to trigger the laser. We could use some logic to double the frequency by XORing the two channels together. On the rising edge, the laser turns on, on the falling edge, the laser turns off and the ADC sample and hold and conversion is triggered. We'd not have to rectify the voltage, and with stable speed control and bandpass filters on the input, we'd use this Nyquist trick to downsample directly to DC and filter out all other signals. The trigger being done by an encoder garauntees that it happens at the same place every time, and by using the index signal, we'd be able to garauntee that we have an accurate signal. For instance, with my six sided mirror, I'd have a 1/3 bit shift for each mirror. This would reduce my scanning frequency by 3, but it would triple my effective resolution too.
Got this one going in my head when I started investigating the new PSOC 3's and 5's. The onboard PLD components make things quite a bit easier.
I know camera systems are fairly simple these days, but I don't know if they can achieve the same accuracy.
What I was thinking of doing was using the mirror and maybe motor from an old laser printer. I have one with a 6 side polygon mirror. I'd like an encoder on it. I have a capacitive encoder that can do 2048 PPR.
The idea here is to AC couple a 1D PSD to the system. Get the mirror up to speed. We don't have to close loop control it, but we need fairly stable speed control from it. Now we replace the old IR laser diode with a red laser diode and the PSD. Lower gain but we can now see it. So, as we spin the mirror, we use one channel of the encoder to trigger the laser. We could use some logic to double the frequency by XORing the two channels together. On the rising edge, the laser turns on, on the falling edge, the laser turns off and the ADC sample and hold and conversion is triggered. We'd not have to rectify the voltage, and with stable speed control and bandpass filters on the input, we'd use this Nyquist trick to downsample directly to DC and filter out all other signals. The trigger being done by an encoder garauntees that it happens at the same place every time, and by using the index signal, we'd be able to garauntee that we have an accurate signal. For instance, with my six sided mirror, I'd have a 1/3 bit shift for each mirror. This would reduce my scanning frequency by 3, but it would triple my effective resolution too.
Got this one going in my head when I started investigating the new PSOC 3's and 5's. The onboard PLD components make things quite a bit easier.
Technical Alchemy
Strange ideas, rambling technical thoughts
http://technicalchemy.blogspot.com/
(yes, I'm working on a real website eventually)
Strange ideas, rambling technical thoughts
http://technicalchemy.blogspot.com/
(yes, I'm working on a real website eventually)