nanoDLP

Hi,

I searched the forum but did not find a thread on this therefore I throw in my idea... (I did google for it as well, but to no avail)
It is well possible, that I am overengineering this quite a bit.

I am coming from microelectronics manufacturing and there you will always find a UV sensor in the light path to get exposure time correction to adjust for uv-source variance/aging to keep mJ/cm^2 constant.
Of course, the resists used in wafer lithography are much thinner than in 3D printing and over/underexposure is far more critical.
But anyways, I think, it would be nice feature to correct exposure profiles linearly with uv-power.
Also, this would reduce calibration effort after lamp exchange and it also makes testing for the LCD ageing possible. Correlating LCD temperature and uv leaking level is also interessting.
It might cost some build volume depending on the placement of the sensor but this could even be avoided if placed on the bed edge for example.

By quickly checking on AliExpress I found at least three different UV-A sensor modules as break out boards with i2c being 3.3V compatible with the RPi.
Vishay VEML6070 - not good for sensing 405nm - https://www.vishay.com/docs/84277/veml6070.pdf
Silicon Labs SI 1145 - well suited for 405nm - https://www.silabs.com/documents/public … -46-47.pdf
CJMCU-6760 - no clue as I found no datasheet yet I figured out the UV sensor on this board by image comparison. It is a Vishay VEML6070 as well and not suited for 405nm.

One catch will be the calibration of the sensor.
I assume, LCD-on and LCD-off could work but a better defined common and cheap 10-30% UV-dampening "filter" material would be much better.
The sensors have a wavelength-dependent non-linearity and 2 point calibration with flull scale to zero-offset might not suited for controlling uv levels in the full scale range.
AFAIK, paper is quite well controlled in thickness and 10g paper could be a reference but I have no clue about the UV related properties of different paper manufacturers (at least, I would not try recycled paper).
--> SI1145 is factory calibrated for UV-index calculation - therefore, at least these sensors are comparabale to each other.

I will play around with the UV sensors after they arrive somewhere in April. If this bears insights, I will share the data here.
Or somebody else already has measured this with a UV-meter and it is already clear that this is a waste of time.

So, the actual feature request:

1. compensate exposure time linearly against sensor signal
full-scale defined for LCD-bright, zero-offset for LCD-dark, eventually a cheap defined UV-dampening material to get a FS-20% point as 2 point calibration depends on sensor linearity or known non-linearity.

2. measure LCD-bright level as lamp wearout indication

3. measure LCD-dark level as LCD wearout indication

Cheers

Last edited by Quas7 (2018-04-28 22:44:48)