ALTA II Reflectance Photometer Teardown

The ALTA II is sold as a classroom reflectance spectrometer. Inside, it is a compact multiband photometer built from eleven light emitting diodes, a silicon photodiode, a display board and simple control electronics.

What it measures

Each button turns on one LED. The light reaches a sample through the circular opening on the back, and the central photodiode measures the reflected light. The eleven bands cover visible light and near infrared from 470 to 940 nm.

It is called a spectrometer for teaching, but it does not scan a continuous spectrum. The Lunar and Planetary Institute describes it more precisely as a multiband photometer because it measures eleven separate wavelength ranges.

Inside

The ring of LEDs makes the optical layout easy to understand. The sensor sits near the center so the source and detector view nearly the same area. The main board contains the LED drive parts and the photodiode amplifier. A ribbon cable joins it to the display and keypad board.

This simple architecture is a good starting point for a home built instrument. The difficult part is not turning LEDs on. The difficult part is making the measurement repeatable.

How to get useful reflectance data

1. Keep the sample flat and cover the complete opening.
2. Block room light from entering around the edge.
3. Record a dark reading for every wavelength.
4. Record a white reference under the same geometry.
5. Measure the sample several times and average the results.
6. Calculate relative reflectance as sample minus dark, divided by white minus dark.

The LEDs have broad emission bands, and the photodiode sensitivity changes with wavelength. A homemade version needs individual current control and calibration for each band. It can produce useful relative curves, but it should not be described as equal to a laboratory scanning spectrometer without comparison data.

The Lunar and Planetary Institute ALTA page gives the wavelengths, optical principle and classroom measurement method.

Project photographs

Adding Bluetooth Stereo to a Neato D85

I opened a Neato D85 robot vacuum to replace a failed side brush O ring. The original speaker suggested a second project, so I added a small Bluetooth receiver, a stereo amplifier and another speaker.

Parts used

• VHM 314 Bluetooth audio receiver
• PAM8403 stereo class D amplifier module
• Two matching 8 ohm speakers
• Insulated wire, mounts and power filtering

The original speaker is marked 8 ohms and 1.5 W. The second speaker should have the same impedance and similar sensitivity. A lower resistance is not automatically better. It draws more current and can overheat the amplifier or supply.

Amplifier details

The PAM8403 is a bridge output class D amplifier. Neither speaker terminal is ground. Each speaker must connect only between its own positive and negative amplifier outputs. Joining either output to the robot ground can damage the module.

Diodes Incorporated rates the device for up to 3 W per channel into 4 ohms at 5 V with 10 percent distortion. An 8 ohm speaker produces less power and is a sensible load for this small installation.

Power and noise

I measured about 190 mW while the added audio system was idle and about 850 mW while it played loudly. These values are small beside the drive and vacuum motors, but an audio module that stays awake can still shorten standby time.

Motor wiring is noisy. Keep the audio input away from motor leads, twist speaker pairs and place supply bypass capacitors close to the amplifier. If distortion remains, reduce the input level before replacing speakers. Clipping, supply noise and a loose enclosure can all sound like a bad speaker.

Installation

Secure every board and wire so nothing can enter the wheels, brush, fan or lidar mechanism. Keep airflow paths clear. A separate protected battery is possible, but it adds another charger and should not be hidden where damage cannot be seen.

Amplifier limits and wiring are described on the Diodes Incorporated PAM8403 page.

Hackaday featured this modification in Modded Robot Vacuum Can Whistle While It Works.

Project photographs

Project videos

SRS SR620 Time Interval Counter Repair

I bought this Stanford Research Systems SR620 as a repair project. It returned to operation in less than an hour and completed its built in self test. The photographs show the instrument, the internal electronics and the passing display.

What the SR620 does

The SR620 measures frequency, period, phase, pulse width and the time between two events. Stanford Research Systems specifies 25 ps single shot time resolution and a frequency range up to 1.3 GHz. It can also calculate statistics and Allan variance from repeated measurements.

The inside is a useful example of precision 1990s test equipment. Fast input and timing circuits occupy the analog side of the main board. Digital control, display and interface circuits fill the remaining area. The design is serviceable, with a large board, labeled sections and accessible test points.

Return to service checks

1. Inspect the mains input, fuse, earth connection and power supply before applying power.
2. Check every supply rail for the correct voltage and ripple.
3. Run the built in self test.
4. Allow the timebase to warm up.
5. Measure the front panel reference output through each relevant input path.
6. Compare frequency and time interval results with a traceable reference.

A passing self test confirms many internal functions. Calibration with traceable equipment is the next step when specified accuracy matters. Picosecond resolution is also not the same as picosecond absolute accuracy. Trigger noise, cable delay, channel mismatch and timebase error all contribute to a real measurement.

The official SR620 page includes specifications, manuals and performance test information.

Project photographs