BTC M850 Optical Mouse Teardown

Michael Stanley & EE HomePage.com
This report is licensed under a Creative Commons Attribution 3.0 Unported License.

Introduction

Computer mice are ubiquitous. Most of us have several, and we take them for granted. But under the hood, they incorporate some interesting technology. This teardown report will examine the inner workings of a BTC M850 Optical Mouse.

Figure 1: An LED illuminates the desktop surface, which is imaged by the mouse sensor.

Theory

In past times, mariners could tell how far they had travelled in a given day by sighting the stars each day, and noting the differences from one sighting to the next. An optical mouse works much the same way. Figure 1 illustrates the vision system found in any optical mouse. Components include:

An LED to illuminate the desktop surface
A lens to focus the LED illumination on a specific surface area
A second lens to focus an image of that surface area on the optical sensor
The optical sensor itself

Figure 2: A detail of our example surface.

Elaborating on this, Figure 2 shows a hypothetical surface area under the mouse. The area outlined in black shows the surface as initially observed by the mouse sensor. Moving the mouse towards the left and top of the figure, the field of focus then encompasses the area outlined in red.

The mouse sensor is essentially a camera. From the sensor's perspective, it was able to record two images, shown as "A" and "B" in Figure 3. For clarity's sake, the area has been subdivided into smaller areas, each of which is the size that can be recognized by a single pixel in the sensor array.

From these, we can tell that the mouse has traveled a distance equal to -3 "pixels" in the X direction and +2 "pixels" in the Y direction. The direction of travel by the mouse is shown by the red arrow. The size of a "pixel" is determined by the resolution of the sensor and the magnification created by the optics.

Figure 3: The world from the mouse's perspective at times A & B.

Mouse navigation is by dead reconning. Your mouse can tell how far it travelled from one instant to the next, but doesn't know exactly where in space it started or ended.

The discussion above covered mouse navigation in two-dimensions. A "wheeled" mouse adds a third, or "Z", dimension. Add a few switches for the mouse buttons, and you arrive at the block diagram shown in Figure 4.

Figure 4: Block diagram of an optical mouse.

Before proceeding, it would be a good idea to touch on the idea of "quadrature encoding". This is a scheme in which 2 logical signals can be used to encode direction and speed data for one dimension of travel. Our recent tutorial: Quadrature Encoding Explained, provides a quick, but complete description of the technique.

Figure 4 is not a detailed schematic, but it does illustrate major components within the mouse. These are:

U₃: The PS/2 Mouse Controller is responsible for decoding quadrature encoded signals from the optical mouse sensor and scroll wheel, checking the status of mouse switches, and transmitting that information back to the PC. PS/2 mice are becoming harder to find. U₃ is more likely to have a USB interface in mice being designed today. Some units will implement a wireless interface.
U₁: The Optical Mouse Sensor performs the optical navigation functions discussed above. The sensor includes not only the camera, but sophisticated circuitry to compare images and thereby compute distance and direction variables. Fortunately, the IC takes care of all of that for you. The sensor on our mouse board is an Avago Technologies HDNS-2000. In case you didn't know, Avago Technologies is a spinoff from Agilent, which itself was a spinoff from Hewlett Packard. So this chip has a long and illustrious family history.
SW₁, SW₂ and SW₃: These switches are mechanically linked to the center, right and left mouse buttons respectively.
Y₁: The ceramic resonator provides a timebase for IC2.
Q₁: The NPN transistor provides current gain for the LED control signal output by U₁.
IRLED1: D1 is the light emitting diode which illuminates the mouse surface. We're not sure why the PCB silk screen identifies this as an IR LED, as the light emitted is quite visible
R₂: Resistor R₂ limits the maximum amount of current which flows through IRLED1 and Q₁. On this PCB board, R2 has a nominal value of 110 ohms.
R₃ & L₃: Here is an example of just how cost sensitive this market is. Rather than provide a voltage regulator for the 3.3V supply for U₁ (as called for in the HDNS-2000 datasheet), BTC simply placed a series resistor and inductor in line with V_DD3 to drop the 5V supply down to an acceptable level. R₃ is also nominally 110 ohms.
L₁ & L₂: Inductors are placed in series with the PS/2 data and clock signals to control noise.
Encoder: The wheel encoder is mechanically linked to the scroll wheel, and provides quadrature encoded signals ZA and ZB back to the PS/2 controller.
Mouse Wheel: The scroll wheel is mechanically linked to the encoder and to the center mouse button. More details will be shown in the teardown details.
CON1: A standard PS/2 mouse connector.

Some notes on the optical sensor: We should note that the Avago Technologies HDNS-2000 can act as the sole IC in a non-wheeled mouse. It has a built-in PS/2 interface. However a secondary IC (U3 in this case), is required for applications requiring a different interface (i.e. USB) or addition of a scroll wheel. We cannot tell which specific IC is used for U3 in this mouse because of the "COB" assembly techniques used (more on that later). But any number of components can be used. The quadrature decoding function is quite standard, as are USB and PS/2 interfaces. The consumer market moves at a frantic pace, and Avago no longers sells this specific sensor. Avago's newest offerings are quite similar, although they appear to favor SPI and I²C interfaces in place of the integrated PS/2 port.

Teardown

As is the case in many EEHP reports, detailed versions of most of our illustrations are available simply by clicking on the figure of interest.

Figure 5: BTC M850 Optical Mouse

Figure 6: Bottom Side of the Mouse

Figure 5 shows a top view of the mouse prior to disassembly. Figure 6 shows a bottom view. The top and bottom are joined by two screws, as well as a plastic tab on the faceplate that secures top and bottom assemblies. The optical sensor is located near the center of the mouse bottom, and will be shown in more detail later.

Figure 7: Mouse Lid Assembly View 1

Figure 8: Mouse Lid Assembly View 2

In Figures 7 and 8, we've separated the mouse faceplate from the main mouse body. The faceplate provides the user-visible portion of the left and right mouse buttons. In Figure 8, you can see two tabs which project from the bottom of the faceplate, through holes in the upper mouse body. These will contact directly with SW₂ abd SW₃, which will be shown later. Note that the faceplace/button interface is a single piece of flexible plastic.

Figure 9: Underside of the Mouse Lid

Figure 9 shows the underside of the mouse lid. The faceplate projection on the left will fit into a hole on the mouse bottom, securring the upper and lower portions of the mouse body. The faceplate projections on the right will rest on the tops of SW₂ and SW₃. Click on the photo to get a clearer view of the faceplate/mouse body interface.

Figure 10: PC Board Revealed: Click on this photo to bring up a higher resolution version with an image map feature to identify key components.

The detailed view of Figure 10 has been enhanced to identify key components interactively. Identified components use the same instance identifiers used on the PCB silk screen and the block diagram presented earlier. A key item not seen so far is the black plastic clip which overlaps the LED and sensor IC. The clip holds the LED in place and contains and redirects light from the LED downward into the lens.

Figure 11: Notice the integration of the scroll wheel and center button feature.

Figure 11 shows a closeup view of the scroll wheel from above. In this figure, you can see SW₁, which was hidden in the previous view. (we've removed the PCB from the mouse body to take this photo.) The scroll wheel is designed to rest on SW₁. Depressing the wheel with your finger closes SW₁. The detent you feel when you do this actually comes from the switch, not the wheel. Simple and elegant!

Figure 12: A closeup of the thumbwheel/encoder interface

The shaft of the scroll wheel fits directly into a mechanical rotary encoder which is shown in figure 12. This encoder appears to be a Panasonic EVQVX GS Encoder. The encoder will output quadrature encoded ZA and ZB signals to the mouse controller. These convey the rate and direction at which the scroll wheel is being rotated - and are encoded in precisely the same manner as X and Y dimension information output by the optical sensor.

Figures 13 and 14 provide closeups of the mechanical encoder. The hexagonal shaft of the encoder wheel fits snugly into the opening in the encoder shown in Figure 13. Figure 14 clearly shows where that "clicking" sensation comes from when you rotate the mouse wheel.

Figure 13: Here's just the encoder itself.

Figure 14: A REALLY close view of the top of the encoder shows how that nice "detent" feeling is mechanically generated

Figure 15: The same basic switch is used for all three mouse buttons.

Before flipping the PCB over, we'll note that all three switches on the PCB appear to be identical. Those, the scroll wheel, and the mechanical encoder are the only moving parts in the mouse.

Figure 16: Here's a closeup showing the sensor lens peeking out from the botton of the still intact mouse.

Figure 16 shows a closeup of the mouse optics as viewed from the bottom of the mouse. In Figures 17 and 18, we've removed the PCB and optical lens from the mouse body and flipped them over. Figure 17 shows the lens in relation to the PCB. Figure 18 has removed the lens to expose the bottom of the image sensor.

The black blob in the center of the PCB in both Figure 17 and 18 is the epoxy used to cement U3 into place on the bottom of the PCB. This type of assembly is sometimes referred to as a "blob-top" or as "Chip-On-Board" or "COB" assembly. In this type of assembly, a bare die is mounted directly to the PCB. Bond wires run directly from the die to the PCB, and finally a blob of epoxy is used to secure the whole thing in place.

Figure 17: Underside of the PCB, with lens assembly in place

Figure 18: Underside of the PCB, with lens assembly in place pulled to the side

Figure 19: Closeups of the lens assembly

Figure 19 is an animated GIF which rotates through several views of the lens assembly . This one piece of plastic incorporates the LED lens, a prism to redirect LED light downward onto the mouse surface, and the sensor lens. The assembly is shaped to fit snugly between the PCB board and the bottom of the mouse housing. The lens projects into the PCB area, and the mouse housing has projections that in turn hold the lens assembly in place. The whole thing is designed for easy assembly and loose manufacturing tolerances. Avago sells the LED, the LED clip, sensor IC, and lens assembly - all designed to work together. Collectively, they make it very easy for anyone to design their own optical mouse. Avago also publishes Optical Mouse Designer's Kit Design Guides which take it all the way, including schematics and PCB designs.

Figure 20: LED Duty Cycle

You may have noticed, as we did, the the light emitted from a mouse LED sometimes varies. We suspected that the LED output might be pulse-width modulated, so we placed a scope probe on that circuit path to see. Sure enough, during operation, the LED is normally modulated at roughly a 125Hz rate (see Figure 11). We also noted that when the mouse is first plugged into a computer, the LED is on 100% of the time - which is why it is brighter then.

Video 1: Mouse Sensor Under Power

Video 1 shows the mouse PCB and lens under power. Because the LED is pulse-width modulated, and the frame rate is fairly low (the original was shot at 20 frames per second), the LED appears to pulse on and off in the video. That is not apparent to the naked eye.

Finally

Our purpose in writing teardown reports for EEHP is to educate. Electrical Engineering comprises a broad range of disciplines; and a report of this type necessarily touches on many. We hope you take the time to follow some of the links contained in the report above, as they have been selected based upon their ability to extend the educational experience.

To Learn More

You can improve your learning experience by investigating the following resources:

Re-Use

Michael Stanley & EE HomePage.com, February 2008

The original publication of this report resides at http://www.eehomepage.com/report.php?report=20080214.

This report is licensed under a Creative Commons Attribution 3.0 Unported License.

We ask that you include a credit similar to the following:

Provided courtesy of Michael Stanley & EE HomePage.com

You can use the following HTML for the credit:

Provided courtesy of <a href="http://www.eehomepage.com">Michael Stanley & EE HomePage.com</a>

Welcome to EE HomePage.com.