JVC Camcorder (1994) – part 2

I realize I am already on part 2 of this teardown and have not even removed the main panels.  So the side panels are now unscrewed to show the tape mechanism on one side.and the optics and PCB boards on the other side

Optical Lens Assembly

Lets start with the optical path. The whole assembly an be unscrewed and the PCB board has a single 16 pin flex connector to the main board. The CCD sits in a plastic socket. There is an elaborate lens assembly, with is a flex lead that wraps around the assembly connecting to three stepper motors.


Opening it up and you can see it is a four lens system,  with fixed lens at the front and one in the middle with two movable lens that are driven by the stepper motors




The third stepper motor drives a variable aperture that is directly in front of the middle fixed lens.



The front of the CCD PCB is covered in 13 capacitors (Decoupling capacitors?) together with what I think is a small 4 pin surface mount voltage regulator in the corner (JVC B19.069) and the socket for the CCD

The CCD is housed in a ceramic DIP package with a plate that screws into the lens assembly. A rubber casket holds a glass (infrared) filter, and a plastic piece provides a light seal to the lens assembly.



I haven’t yet mastered how to decap ceramic packaged CCD without damaging them, so I shot the die photo with the die still in the package.  This limits me in magnification to 10x objective (Higher magnification objectives have a depth of focus that is too small to see the die through the glass).

The CCD is a MEC 3716, MEC is Matsushita Electric Company (Panasonic today).  I was able to dig up a datasheet for this sensor. Its a 6mm (1/3 inch) interline transfer CCD with 270k pixels (542h x 494v).  In this image I have masked out the core filter area (I am not able to stitch together the continuously repeated pattern).

Here is the pixel core edge, showing the pixels are not square (7μm x 10μm) and unlike the vast majority of sensors today (which have RGB Bayer filter array) this has an unsual CYGM filter array  that is Cyan, Yellow, Green, Magenta

CCD Drivers

The rear of the board contains three IC’sThere is a Matsushita MN3111H which is a vertical driver IC for the CCD.  A JCY0035 for which I could find nothing (One problem with vintage stuff is the documentation is a lot more sparse) possibly the CCD Front End A/D encoder,  and a small 8 pin device 2018S.

This is the MN3111H after depot, a 48 pin die 4.7 mm x 4.7mm and consisting mostly of a collection of large multi-fingered MOSFETs  (Driver transistors).

The chip is made on a 5μm CMOS process which is 1970’s/80’s technology. Looking at one of the driver transistors at higher magnification you can see the polysilicon gates are 5μm long. (Thats the channel length across the gate, the distance the electrons or holes flow from source to drain.) Each finger on this transistor is ~250μm wide, and on this transistor there are 40 gate fingers so the whole transistor is 10,000μm (10mm) wide.

The other primary device on the CCD board is the JCY0035 this is the die after depot.  It’s a 64 pin gate array, built on a two metal 2μm CMOS process.

The gate level transistors are laid out in a repeating array and the logic depends on how they are interconnected by the Aluminum wiring. The colors of the dielectric over the NMOS and PMOS FETS are different and show up clearly here.  Each block of transistors (Called Active Area) has two polysilicon gates crossing and a small piece of active area below where a well or substrate connection is made. In this image the top NMOS and PMOS gates are connected up the bottom gate is not used.

Where the circuitry is a bit more dense it is harder to make out the gate level as the first level metal is usually covering the gates as in this imageTo keep the blog size manageable, I will pause this here, there is still the main board and and a look at the mechanical and tape read heads still to come.

This entry was posted in Teardown and tagged , , , , . Bookmark the permalink.