This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LightCrafter Start Up Troubleshooting

Other Parts Discussed in Thread: TPS63020

Hi,

We recently purchased a lightcrafter kit with the intent of replacing the LEDS with a white one to do structured light imaging.

Following some of the posts in this forum:

http://e2e.ti.com/support/dlp__mems_micro-electro-mechanical_systems/f/850/t/162664.aspx#592028
http://e2e.ti.com/support/dlp__mems_micro-electro-mechanical_systems/f/850/t/159950.aspx

First thing we did was removed the ribbon cable that supplies power to the 3 LEDs.

Since we wanted just white light, we inserted our own externally powered white LED into where the blue LED used to be.  We also removed the one dichroic mirror in front of the blue LED to pass the full spectrum of white light.

After we put everything back together, the lightcrafter wasn't able to display anything.  We think there's something wrong with the booting process.

Our question is, is there some kind of synchronization with the LEDs when the system boots up?  What is the purpose of the photo sensor on the smaller flex cable?

We tried putting everything back together (i.e. the old LEDs back in).  The only thing missing is the dichroic mirror.  We notice that when the system boots up, all three LEDs flashed a few times.  Then every ~10s or so, the lightcrafter would make the three LEDs flash again.  But the lightcrafter remains unable to display anything.

Does anyone know what the problem might be?

Thank you for your time.

-Mike

  • The Light Crafter runs fine without the LEDs and the sensor

    Probably etendue:

    http://e2e.ti.com/blogs_/b/dlp_mems_blog/archive/2009/12/12/who-is-ed-tondue-and-why-due-i-care.aspx

    A more informative article is here

    http://catalog.osram-os.com/catalogue/catalogue.do;jsessionid=53F098CB7D0B85DCBFC957903A395B1F?act=downloadFile&favOid=020000020003fea8000200b6

    Injecting light into the light engine is quite tricky LEDs are Lambertian (emit in all directions) while the DLP only accepts light over a 22 degree angle

    As a result you have to use very fast aspheric lenses to "collimate" (to the extent an extended incoherent light source can be collimated) the light into the homogenizer window.

    I'm still perfecting my technique but here are the best practices:

    • Get the brightest white 1mm (no larger) die you can get, use a an "in air" LED not one cast in an acrylic lens (etendue is proportional to n^2 where n is the index of refraction of the media enclosing the emitter) an acrylic casting costs you a 2.25 factor over air mounted.
    • Bust out the dichroic combiner from the light engine but leave the rest of the optics in.
    • Gently pry the lens assembly off the Blue LED which came with the unit, an X-Acto knife under the dark grey holder will pop the assembly right off.
    • Now use a dental pick to gently pry the curved lens off of the assembly.
    • Now knock the remaining clear non lens portion from the assembly, it will shatter but no loss.
    • Look at the assembly under a microscope and note the shelf on which the lens rests
    • If your White led fits in the assembly (lucky you) mount the assembly back in the light engine and micro-position the LED in place watching the brightness of the checkerboard image to guide you once close use a photodiode to do final optimization
    • If it doesn't fit cast the assembly LED side down into Paraffin wax and very gently polish the assembly down to the lens shelf us very fine grit sandpaper resting on a sheet of glass. 3m ultrafine wet dry finishing sandpaper is pretty good.
    • Follow by micropositioning the LED in place. Position is very critical!

    This should work for white light, I'm doing near ir.

  • Mike,

    Stephen points to a great app note from OSRAM that talks about etendue matching. Thanks Stephen! This is certainly an issue when replacing LEDs. Some time ago, we did some calculations for optimal etendue matching of a 0.3" DMD with a f/2.4. The calculations showed that the optimal LED area was about 1.7sqmm. With an f/1.5, the optimal LED area was about 3.2 sqmm. This gives you an idea on the range of etendue and LED areas possible. On LightCrafter, the Red and Blue LEDs are 0.81 sq mm while the Green LED is 1.06 sqmm. 

    Now, LightCrafter works with the LEDs disconnected. If you see the schematics, on page 5 at the lower right, there is an overvoltage protection circuit that will power down the TPS63020. This DC-DC regulator powers the LightCrafter's LEDs and will shut-down if the LEDs are disconnected. The rest of the system will work correctly, except no light is emitted. You can look into the DMD to see that it is still operating. 

    IF you powering the LED externally, the LightCrafter should boot normally. The LightCrafter uses a sequence that triggers the LED enables and the PWM to the LED so that the LEDs are on when the mirrors are set to the desired image (you do not want to turn on the LED when mirrors are flipping). There are test points for these signals (TP9-TP14) where you can wire cables to your external LED circuitry. The LED enable is controlled by the selected LED in pattern sequence mode. The PWM is controlled by the bit-weight of any grey scale pattern.

    The photo sensor is not currently being used. The LEDs are too bright, so the photo sensor is saturated. The original intent of this sensor was to measure white point and perform white point correction. However, the light is so bright that it is saturated and we could not come up with a simple and cheap way of controlling the amount of light incident on it.

    There is also a thermistor on the green LED board. This thermistor is being measured by the MSP430 to shut down the system if the LED heats up.If the thermistor has been damaged or this line is shorted, the MSP430 will shut the system down and the driver board's Red LED will turn on (next to power connector). This thermistor is checked every 65ms.

  • That's good to know I'd love to use some of the larger OSTAR units for my app. What is the f/# on the existing optics?

  • Stephen,

    I am not sure of the exact f-number for the LightCrafter's optics. I will update you by Monday.

    Divya

  • Hi Stephen,

    I have forwarded your query to the Young Optics who is the Original Design Manufacturer (ODM) of the LightCrafter's optical engine. I will update you once I receive the response from the Young Optics team. Thanks for your patience.

    Regards,

    Divya

  • Stephen,

    Young Optics has indicate that it is f/ 2.2.

    Thanks for your patience while we found this answer.

  • Hi Stephen,

    Would you mind letting me know what NIR LED's you have tried to integrate into the lightcrafter and had some success with?

    I was thinking about trying to use the SFH 4235 from OSRAM.

    (http://ca.mouser.com/ProductDetail/OSRAM-Opto-Semiconductors/SFH-4235/?qs=5ndlnLNLvJXVq3vqAS%2fhd9PFPJtuus05GTIWL%252bl4a8U%3d)

    Thanks again,

    Adrian

  • I want to know more about your case