The real experiment did not start yet because a few problems occured and a I had to find a new test couple. This gave me the opportunity to test one of the displays myself. The real experiment/pilot-test will start tomorrow evening.
All posts tagged display
Final test experiment
finished building the displays
More pictures and a video after the break
Read on… →
Light leakage in ‘objective’ display
Looking pretty good without frontplate (video)…
… but vague with front plate. So we will have to patch this one too.
Problem with white perspex
I bought the white perspex with a thickness of 2mm that, according to the added specifications, did not let any light through. Well this is not true and causes me a lot of problems. The company that sold me this, has agreed with me and gave me back my money, but this doesn’t solve the problem…
back of the display
light leaks at the front of the display (with plastic ‘diffusing’ foil)
Lighting up the words ‘Jij’ and ‘Vandaag’.
But the translucency of the material isn’t the only problem. We tried to make the display as thin as possible in order to increase the light intensity in each box. This assumption was correct. But we forgot that by reducing the scale, the influence of the imperfections would increase. So now we had 3 light leaks:
section through the display panel
The solution is now to use aluminum foil and filler to close the gaps and reflect the light. Also the diffuser would be excluded.
building the subjective display
The circuit board with 5 SiPo’s (74HC595) to control 40 outputs.
View of the back of the circuit board.
Putting the laser-cut parts together to form an early structure. The material used is white and black 2mm thick perspex.
The front plate with all the letters on it. Due to the high density of characters, the temperature probably got to high and parts started to melt. The plate is curved and I’ll have to think of a different approach to create the plate.
zoomed in on the front plate
Inserting the logic in the structure
adding the LEDs and resistors + connecting them with the circuit board.
Switching on the power to light up the LED’s in ‘groups’.
video:
problems during the assembly:
- I forgot to cut out some holes to put the LEDs through. This was solved by using a construction hole as a LED hole. The downside of this will be a small light leak to neighboring words… Reason of this mistake was that those holes were on a non-printable layer in Adobe Illustrator.
- The ‘ground line’ of the transistors was broken on 3 places. This was first discovered after assembling two rows so it was still doable to detach the whole circuit board and solve the problem. The strength of the bonding was tested while doing this and found to be more than good. After assembling some more rows, I noticed a similar problem. It appeared that the ground line was now broken on two extra places. I tried to repair it but had to remove a transistor in order to reach the parts. Luckily, there were some spare outputs left. The ground line wasn’t interrupted before assembly (was tested and can be seen on photo) so it must have happened during assembly. Probably by pressing to hard on the board while making a glue connection.
- The perspex plates were cut out with the ‘plastic 3mm’ setting. The white plates cut out nicely but the black ones were not completely cut through. I could not find any specifications about the material I used, nor if black perspex is denser than white perspex. For the front plate, I decided to change the settings to ‘plastic 4mm’ to be sure the characters are completely cut out. It appeared that the laser now passes two times along the same path, increasing the temperature. The edges of the characters are melted and the faces was bent afterwards… Next time I will try the 3mm setting again but remove the protection foils in advance.
Display prototype two
The previous prototype was insufficient to conclude anything because the front plate was too thick to let enough light come through. This meant I had to redo some tests to conclude how many LED’s are needed per word.
I used a 2mm thick white cardboard that reflects more than the wood I used before. The depth of the lightboxes was reduced from 20mm to only 6mm which is just enough to fit a 5mm LED. The lights are controlled using an Arduino Uno with an adjusted version of my Charlieplexed wiring and coding. The ‘words’ are between 3 and 7 characters long and a white LED was placed on both sides. Read on… →
