UV-meter design
Introduction
This is a simple design for a handheld wideband UV-meter.
It uses as much as possible ready-made components and will need a
minimum amount of soldering.
It fully complies to the requirements, but will only need a sticker over
the rightmost digit to show in mMED/hr.
Components
Box
-
Serpac M6 - Design choice
- 105.41 mm x 60.96 mm x 21.59 mm
- Integrated 9V battery compartment with contacts
- Can be customised with full colour printing
- Arrow (black)
- Mouser (all) (black)
-
- BOS 400 with compartment for 9V battery
- 100 mm x 65 mm x 24 mm
- Alternatively BOS 500
series
(120 mm x 60 x 25 mm)
- RS BOS502 (attention, 502 has area for display, should use 503).
Design
- Cutouts for display, pushbutton, sensor, hole for calibration
- Printed customisation by factory or with sticker
Display/voltmeter
- Murata 3½ Digit, LCD Display Digital Panel Voltmeter DMS-20lcd-1-5-c
- DMS Application note 3: Analog Common and Reference in-out
- DMS Application note 12: Signals with Zero Offsets
UV-sensor
- sglux SiC UV TOCONs / Broadband UV
sensor
- TOCON_ABC5
- Gives 0-5V output
- "Scatter" (tolerance) of ±20%. Should be able to calibrate this
out.
- See sglux FAQ: What about the TOCON scatter?
- Lower and higher sensitive sensors exist
Calculation of required sensitivity and sensor
- We want 0.2 MED/hr at 2V output = 0.5 MED/hr at full scale 5V output.
- Scaling factor is 0.24 MED/hr/W/m² so we want 0.5/0.24 = 2.08 W/m²
full scale
- Correction factor of 0.24 for a wideband TOCON ABC (that is more sensitive for UVA) to correct so that under natural sunlight gives same reading in med/hr as a perfect erythema meter.
- See Comparison of different UV sensors for use in an XP radiometer, A. Butterworth, December 16, 2018 (paragraph above Table 1).
- 2.08 W/m² = 2.08e-4 W/cm² = 208 µW/cm²
- The closest is ABC5 with 180 µW/cm² full scale.
- 0.2 MED/hr would then give an output of 0.2/0.24 * 1e-4/180e-6 * 5 = 2.3 V.
- The tolerance of the TOCONs is ±20%, some of them may not give us the required output: 2.3V*0.8= 1.84V instead of 2.000V. Should take this in account that 1.84V can be calibrated to read as 2.00V. 2.3V*1.2=2.76V. So calibration range must be 1.8V - 2.8V.
- For comparison 0.2 MED/hr should give 0.79 V on the TOCON E1. Calibration range 0.6V - 1V.
- Assuming the powers that are quoted by sglux are in W/cm² are normalised to the sensitivity spectrum of the sensor.
Likely not needed: digital calibration could be made with a TOCON Adjustment board (needs specific box to make setting).
Printed Circuit Board
- Holds
- TOCON
- Needs to be mounted at 90 degrees as on front-side of box.
- Needs an additional board?
- Ask sglux if TOCON can come with wires or a low-cost connector allowing assembly at 90 degrees
- Needs to be mounted at 90 degrees as on front-side of box.
- Push button (height should be right)
- Wires up to 9V holder
- Display/voltmeter
- Needs careful adjustment in height, or with socket that allows movement? Or wire the display up to the PCB?
- Allow for hole in PCB to reach calibration screw
- Resistors for the full-scale adaptation from 0-5V to 0-2V (1%
metal-film). Can be done with Vref.
- Tolerance of TOCON ±20%. Should be able to handle this. Can
be done with Vref. 1.8V should be able to read as 2.0V.
- Use high-precision potentiometer?
- Use selectable resistors?
- Use a digital potentiometer or sglux setting board/schematics (requires programming with special box and connector)
- Final fine (±50 counts or ±2.5%) calibration on Display/Voltmeter
- Tolerance of TOCON ±20%. Should be able to handle this. Can
be done with Vref. 1.8V should be able to read as 2.0V.
- TOCON
Various
- 9V battery
- Assembly: mask off the rightmost digit with tape or a dedicated sticker. Or possibly not even needed as no UV will read as 000 anyway.
Erik van der Bij - 4 April 2019