**Language:** [Deutsch](6.-Ulbrichtkugel) | English

# Integrating Sphere

![alt text](images/2.-Bauanleitung/Ulbricht-1.jpg)
*The base is a white ping‑pong ball. With a utility knife I cut two 11 mm holes: the entrance pupil for the LED and the exit pupil for the gate. The holes should be roughly 90° to each other. Between the two holes goes a slot. Why?*

![alt text](images/2.-Bauanleitung/Ulbricht-2.jpg)
*To prevent the LED from shining directly into the exit pupil. For that you build a small light barrier. Fittingly, I cut a piece from a lab spool; it’s nicely light‑tight.*

![alt text](images/2.-Bauanleitung/Ulbricht-4.jpg)
*The cut barrier is slid into the slot and glued.*

![alt text](images/2.-Bauanleitung/Ulbricht-6.jpg)
*A control view from one hole through the other shows: we see only black. That’s how it should be.*

![alt text](images/2.-Bauanleitung/Ulbricht-7.jpg)
*For the ping‑pong ball to scatter and reflect perfectly and neutrally, it should be coated with barium sulfate. The light barrier should be coated as well. Here I mix a suitable liquid in a magnetic stirrer; it takes a while until the barium sulfate has no more clumps.*

A mix that works quite well:
- 100 g barium sulfate
- 65 ml ethanol
- 35 ml distilled water
- 20 ml polyvinyl alcohol (PVA), clear

Use a pipette to put the emulsion into the ball and swirl it so the coating spreads everywhere without pooling. After about 60 minutes at room temperature, it’s dry enough for a second coat. Keep it moving during drying!

*Note: This mixture developed slight cracks after a few weeks and seems a bit too glossy. So far that hasn’t been a practical issue, but I’ve since found another recipe that still needs testing.*

- 20 g BaSO₄
- 10 ml 5% PVA solution (1 part diluted with 4 parts 90° distilled water)
- 10–15 ml distilled water (should be “a little thicker than cream”)
- 1 drop IPA

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-08.jpg)
*For space reasons I placed the integrating sphere inside the lamp‑house cover. That’s not ideal because you have to power the lamp there... see below. In any case, the black heat‑shield must be removed.*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-09.jpg)
*The top vent grille is removed and the housing widened to make room for the fan (needed with the YujiLED).*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-10.jpg)
*A lot of complicated measuring and calculating followed to place everything correctly... after three attempts I had it.* :)

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-11.jpg)
*Mark the center of the LED COB on the heatsink...*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-12.jpg)
*LED attached with thermal adhesive...*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-13.jpg)
*...and an old PC fan screwed onto the heatsink in blow mode.*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-14.jpg)
*Fits pretty well! The original latch mechanism holds everything together.*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-15.jpg)
*Meanwhile I feed the LED via a spring‑contact strip...*

![alt text](images/2.-Bauanleitung/Ulbrichtkugel-16.jpg)
*...whose counterpart is mounted on the projector. Not pretty, but it works. This is just the first version of a 5‑year tinkering project.* :)

The exit pupil of the integrating sphere should sit as close to the gate as possible—then the effect of completely unidirectional light is strongest.

Have fun experimenting!


---
<p align="center">&laquo; <a href="5.-Circuit-Description-EN">Circuit Description</a>     |     <a href="Home">Home</a> &raquo;</p>