Sigma 45mm f/2.8 Lens Repair & Analysis | Salvaged Circuitry

Sigma 45mm f/2.8 Lens Repair & Analysis[05.12.24] I have a camera gear collection problem and as part of my personal 12 step plan, I restrict myself from purchasing functioning lenses. This sounds illogical, and it frankly is, but it's very hard for me to resist heavily discounted lenses. To keep my hat, I tend to only bid on lenses that are less than 1/4 of the going used sale prices and have little to no mechanical damage. In this case I've been eyeing the recently produced sigma I-series lenses that feature mostly aluminum construction. A broken 45mm f/2.8 lens popped up on ebay in January for a song and a dance, and I simply could not resist. The auction was listed by an ebay seller that tends to have regular inventory of broken modern camera gear which is great from a repair perspective. Occasionally the seller will teardown equipment and sell parts, leaving me a bit uneasy about the internal state of the listed items for sale, but I took a chance and went with it. Arrival The lens came well packaged and on initial inspection, featured zero mechanical flaws. No scratches on the barrel or lens elements whatsoever. To properly inspect the outer lens elements, I use my oil free air compressor and thoroughly blow off any debris from the lens and properly clean both front and rear element with a kimwipe and lens cleaning solution. Eye glasses cleaner from the drug store is adequate for most external lenses. Isopropyl alcohol is another good alternative, but don't use on plastic lenses. This is broken??? I mounted the lens to my Lumix S5 and with seemingly too much force, it clicked into place. The camera booted up fine and even displayed a live image, but no electronic controls worked whatsoever. None of the dials or switches on the lens responded to user input. The control dials on the camera did not register movement. Clearly, there was something electrically wrong with the lens. The control PCB is usually found on the rear of the lens nearest to the rear lens contact block. It would also provide a good time to investigate the very stiff lens mount.

Tools The barriers to entry on this repair are low. Most of these tools are pretty standard and generic at this point. The biggest expense besides the lens is filtered air, but even a compressed air duster can suffice. Note: Since most of the camera industry design folks are centered in Japan, the JIS screw is standard. Using a Phillips will work but it tends to wear down the heads on the JIS screws faster. Here are some of my go-to tools: Kimwipes / lint-free lens cleaning wipes Spray Isopropyl Alcohol (ipa) Eye glass cleaner Microfiber cloth Nitrile gloves Highly filtered shop air / oil free compressor Tape Sharpie Scalpel Plastic Spudger Magnifier / optic JIS x 2.5mm / Philips #00 screwdriver JIS x 3.0mm / Philips #0 screwdriver Disassembly For disassembly, I orient the lens with the aperture mark facing me and the table front edge. The rear plastic beauty spacer around the rear element is removed first along with (3) black machine screws. Following, the two nickel plated screws that fixture the side of the plastic lens block terminal interface to the metal lens mount are removed. The screws are placed on double sided tape in an orientation that matches the lens orientation. This makes future reassemble substantially easier. Next order of business, the lens mount bayonet and shims. The orientation and order of these shims matter so they deserve their own bit of tape. This lens had trouble mounting to a camera body so I thoroughly inspected the shims, bayonet mount back and the lens body for imperfections and surface contamination. I cleaned all surfaces with ipa and moved on. Note: be extra careful when handling the lens contact block flex cable. At this point in the disassembly, the lens contact block can be freely removed from the control PCB. The contact block for L-mount features 10 terminals that connect to the control PCB via a flexible polyimide cable.

This flex cable has a tendency to tear easily, especially if not handled carefully. Before continuing further with the teardown, take a multimeter and check the continuity of each trace. If there are visible tears in the flex cable, repair that first before continuing to diagnose any problems. I have another guide on how to make your own flex cables here. The flex cable metered out to be flawless, so I proceeded the teardown. The rear CNC machined aluminum shell of the lens is up next. Two grounding straps are fixtured to the rear shell using nickel plated machine screws. The straps are oriented around the 2pm and 7pm clock position. A push-in switch flex connector is positioned around the 11pm position that can be wiggled out using a pair of tweezers. There are (4) black oxide self tapping screws that mate the shell to the center plastic lens module. The rear shell can now be safely lifted up and set aside, with the aperture mark positioned toward the table edge. The control PCB and remainder of the flex cables are now easily accessible. Three black self tapping screws mate the PCB to the plastic lens module at the 2pm, 7pm, and 10pm clock position, respectively. Once the flex cables are wiggled loose, the control PCB can be freed from the lens body and more closely scrutinized. PCB Analysis The C-shaped PCB looks right at home with the dozens of other lens control PCBs I've analyzed. There's a main microcontroller, DC-DC controller, motor controller, crystal oscillator, and a slew of passives. The reverse side is adorned with FPC (Flexible Printed Circuit) connectors, test points and an 8-pin SPI flash package directly below the main microcontroller. Inspecting any unknown PCB for faults can be quite intimidating, but I find it easiest to start with tracing the input power lines first. Where is this board supposed to receive power from? Where do the V+ and Gnd traces first begin on the PCB? What is the first component that receives power on the board?

PCBs can be a very complicated mess of layers and jumped traces, so feel free to write down simplified schematic notes on scrap paper to keep a simple PCB power order-of-operations. To start, trace the input power from the lens terminal block. The thicker flex PCB traces are most definitely V+ and Gnd. Follow those traces on the PCB and use the continuity mode of your multimeter to find out where the power traces on the PCB lead to. In this case, tracing the input power is tricky because the large traces associated with the flex cable are hidden beneath the FPC connector and passed through the PCB to the opposite side through vias. The power traces were then fed to a small square black chip, a staple in small scale electronics: the DC-DC converter. A telltale sign of a DC-DC controller is the presence of an adjacently placed properly-chunky tan, beige or black colored block that dwarfs the size of the power controller. The photo below indicates a circled "2R2" or a 2.2 uH (micro Henry) inductor. This close proximity inductor arrangement is universally recommended by semiconductor manufacturers in effort to reduce radiated emissions / noise.

In this case, the 16-VQFN package TI TPS62140RGTR Buck converter, labeled as "PA71 TI 18i" is used in this sigma lens PCB. The layout engineer heeded the advice of the TI datasheet and implemented a very similar layout but added a few components to keep things spicy. Looking over the layout recommendation, "C1" connects Vin to Gnd, and serves as the main input filter capacitor for the dc-dc converter. Inquisitive minds can easily determine that the unknown "N" labeled package adjacent to C1 on the input voltage rail is indeed a fuse to protect the dc-dc from damage. A quick check with the multimeter confirms that the fuse was open, so it took one for the team and saved the DC-DC from destruction. Searching online for an "N" labeled fuse does not lead to many promising search results but it did bring up a suggested 2a rated smt fuse on aliexpress.

The TI TPS62140RGTR datasheet quotes a 2a output current, and while there's operational quiescent current, 2a is likely an appropriate value. Being familiar with the Panasonic Semi smt fuses used all throughout the Lumix camera line, I picked up a 2amp 32v fast blow fuse part# ERB-RE2R00V. The lumix gh3, gh4 and gh5 cameras use a mixture of 32v 2.5a and 1.5a fuses, so I knew I was in the right ballpark. I have found that in camera electronics, 2 terminal resistor-looking packages with arbitrary single letter notation tend to be smt fuses. They sometimes have scalloped terminals. The fuse implemented was sized as 0603, making repair possible with less expensive and precise equipment. 0402 and even 0201 fuses very much exist. The layout engineer also left space beside the fuse for ease of access with repair tools. There have been many times when a fuse was put in a pocket or crowded section of a PCB requiring a nearby component desoldering to gain access to the failed component. An example is the Lumix GH3 / GH4 mainboard that features the battery input fuse sandwiched between an SD card slot and a protruding battery connector. SMT tweezers make this repair a breeze, but wielding two soldering irons also works in a pinch. Desolder the failed fuse, clean the pads, position the new fuse, hold down the fuse and solder one terminal at a time. Fuse Investigation As to why the fuse failed, I have not discovered a specific failure point. Certain conclusions can be arrived at, such as specific edge cases where someone leaves the lens in AFC mode (auto focus continuous) and sets the camera to hunt for focus for hours / days on end. It's possible the lens was never designed to be used 24/7 for racking focus and caused the buck converter to pull more current than than 2a fuse could sustain, causing it to open. An interesting operational condition in the TI datasheet may provide clues to the failure point.