low cost spectrometer ish ALTA II tear down

I wanted to share a quick teardown and some photos of the parts inside the ALTA II reflectance spectrometer.

The ALTA II uses eleven LEDs covering different wavelengths from visible light into near infrared. Each LED provides one measurement band, and the photodetector measures the reflected intensity at that wavelength.

To use it, press the ON/OFF button and then one of the numbered wavelength buttons. The color of each wavelength is printed below its number. The selected LED turns on, and the display shows how much of that light is reflected from the object back to the photocell.

For example, a blue object should produce a high reading when illuminated by the blue LED because it reflects much of the blue light. The same object should produce a much lower reading under a wavelength it absorbs.

A white reference can represent approximately 100 percent reflection and a dark reading can represent 0 percent. The sample reading is then compared with those two values to estimate its reflectance at each wavelength.

Inside are the eleven LEDs, the photodetector, the display and the control electronics. Each LED also has its own calibration potentiometer. These were most likely adjusted at the factory using a known white reflectance reference so that every wavelength could be calibrated separately.

More information about the ALTA II:

https://www.lpi.usra.edu/education/products/spectrometer/

My two cents is that it would be cheaper to build one yourself, and for many experiments it would probably work just as well.

Teardown photos below.

Photos and videos

neato speaker hack (stereo + bluetooth)

I opened my Neato D85 to repair the side brush after its O ring dried out and failed.

While waiting for the replacement O ring, I noticed that Neato had fitted a real speaker instead of using a much cheaper piezo buzzer. Since the speaker was already there, I thought, why not use it to play music?

I used parts that I already had. The Bluetooth receiver is a VHM 314 module that cost about $1, and the amplifier is a PAM8403 module that cost about $0.25. I also added a second speaker for the other stereo channel.

Both speakers should have the same impedance. The original Neato D85 speaker is approximately 25 mm, 8 ohms and rated at 1.5 W. A matching replacement should cost about $0.40.

The tiny device beside the original speaker appears to be marked AJT 436. The package is only about 1 mm by 1 mm, so I was not certain whether it might be a MAX436 family part.

The Bluetooth and amplifier electronics consumed about 190 mW while idle and around 850 mW while playing at full volume. If the extra load prevented the robot from completing a full cleaning cycle, a small separate battery could be added for the music section.

The added speaker was mounted on the side facing the middle of the room. The opposite side carries the side brush and normally faces walls and corners.

It sounded much better and, most importantly, was louder than the boring vacuuming sound. There was still some annoying distortion at high volume, so stronger neodymium speakers would be the next improvement.

Videos:

https://www.youtube.com/watch?v=b6EKfvYa0yk

https://www.youtube.com/watch?v=T5zsCcu4-6s

More photos and videos below.

Photos and videos

srs sr620 universal time interval counter

Another eBay find, a Stanford Research Systems SR620 universal time interval counter.

I bought it thinking that it would be an interesting repair challenge, but it turned out to be such a quick and easy repair that I completely forgot to document it. It was working within an hour and passed every built in test without a single error.

The SR620 measures frequency, period, pulse width, phase and the time interval between two events. It can also collect repeated measurements and calculate statistics and Allan variance.

According to SRS, it has a single shot time resolution of 25 ps and can measure frequencies up to 1.3 GHz.

Internal photos below.

Photos and videos