For contact info, please see the Sci.Electronics.Repair FAQ Email Links Page.
Copyright © 1994-2022
Reproduction of this document in whole or in part is permitted if both of the
following conditions are satisfied:
1. This notice is included in its entirety at the beginning.
All Rights Reserved
2. There is no charge except to cover the costs of copying.
Although working on cameras is generally
less risky than dealing with microwave ovens, TVs, and computer
monitors, there is one component in every camera with an electronic
flash - even the least
expensive throw-away variety - that is potentially lethal. Specifically,
it is the energy storage capacito.
And of course even more so for separate electronic flash units
or "speed lights" with their higher energy.
This may charge up as soon as power is turned regardless
of whether flash is called for, and may retain a dangerous charge for hours or
days. If working inside a camera or flash unit, on one that has
had its case damaged exposing internal parts, it is essential that you
read, understand, and
follow all safety guidelines contained in this document and in the document:
Safety Guidelines for High Voltage and/or Line Powered
Reproduction of this document in whole or in part is permitted if both of the following conditions are satisfied:
1. This notice is included in its entirety at the beginning.
If there is no electronic flash, the greatest risk is torn flesh from sharp sheet metal or gear teeth. ;-)
We will not be responsible for damage to equipment, your ego, county wide power outages, spontaneously generated mini (or larger) black holes, planetary disruptions, or personal injury or worse that may result from the use of this material.
For the specific case of the (mostly older) flat head screws, fabricating a driver with an aluminum blade is recommended to prevent damage to the screw slots. These screws also tend to have long thin slots, which don't fit the typical screwdriver well at all. The special tool can be as simple as an aluminum roofing nail with its point filed down to a thin blade. Daamge to the screw slots is almost unavoidable with a steel screwdriver, and at the very least, looks bad. ;( The softer blade will get chewed up after a while but that can be repaired.
A very few screws in cameras have a left-hand thread. I've seen exactly one (1) case of this - in the Copal Square S shutter (used in the Nikkormat FT/FTN among others) - which secures the "Slow Speed Lever" or "Retard Drive Cam" depending on which manual is being referenced. But be on the lookout - if the screw seems to get tighter, try the other way. ;-) Being overzealous and breaking or stripping the screw shaft would likely be fatal to the camera both because finding a replacement screw would be impossible and the remains of the screw is solidly stuck in the hole. ;-(
In the end, the quality of your photos will depend more on care in composition, lighting, a steady hand, and other factors not part of the camera itself. Technology can help but it doesn't replace these. No matter the cost of the equipment, if the lighting is not balanced or the depth of field is too shallow, the photos will be poor.
40+ years ago I owned top-of-the-line Nikon film SLRs including the flagship Nikon Photomic FTN body with several Nikon fixed focal length (non-zoom) lenses. The FTN body was around $350 in 1970s dollars, which would be comparable to roughly $2,000 now accounting for inflation. In those days the only assistance was the built-in exposure meter. Focus and aperture were manual. I did my own darkroom work and ended up with a few good pictures and a lot of mediocre ones. Nowadays I have several low-mid level Nikon DSLRs purchased for either $500 new (D5600) or much much less than that on eBay (D70, D80, D3000, D5200, etc.) and several lenses, but what I use mostly is a Canon SX710HS point-and-shoot, primarily for Website photos. It was around $100 on eBay several years ago. A D70 (one of two each for $10 on eBay excluding lens but including shipping) has been used for most of the DSLR and lens dissection photos later in this document but it's not clear if the resulting photos are really any better on average than using the Canon.
So start out with your phone. If that proves to be too limiting, try an inexpensive point-and-shoot and explore its more advanced capabilities. After that, consider a low-end or older DSLR. One that provides many features of more expensive cameras can be had for $150 or less on eBay with a "standard" autofocus zoom lens. See if that provides benefits that are not offset by the hassle of lugging around several pounds of photo gear.
There is a D70 repair manual on-line. Search for "Nikon D70 Repair Manual PDF". It has many detailed photos with step-by-step disassembly and reassembly, some explanations, and parts identification.
I've actually come to like this camera despite having owned top-of-the-line Nikon F film SLRs many years ago as well as the newer D5600 and other D5xxx DSLRs. While the D70 is heavy and clunky (politely perhaps referred to as more "Solid") compared to newer Nikon DSLRs and has limitations, it is relatively simple to use with no excessive creeping featurism, excellent battery life since nothing is really running until the shutter button is pressed as there is no power hog live view mode, has a fast shutter response, and, uh, also takes decent pictures. ;-) This one has a shutter count of around 12K, so it's really only a teen as these things go. ;-) See my general comments on a "Selecting a Type of Digital Camera". You may be surprised at my conclusions.
The photos may be viewed at: Nikon D70 DSLR Dissection. (This opens in a single new tab or window depending on how your Browser is set up.) Some of the photos may be rather gory. So send the kids and pets to another room. ;-) These shots start with an intact camera similar to the one being discombobulated, and then the core, various covers (including the back one with the LCD), the microcontroller PCB - essentially almost everything that can be detached with only the use of a screwdriver and by unplugging cables. Reassembly would be straightforward, at least in principle with adequate notes, closeup photos, and some luck. Beyond this point, except for removing the CCD assembly, wires have to be unsoldered or cut. As can be seen, this has now commenced as the necessary chants and incantations to the gods of dead cameras have been issued and notarized. ;-) And yes, a close examination of the photos will reveal that a pair of buttons did disappear before they should have during the disassembly and I didn't notice. Live with it. ;-)
And as noted in the introduction, to complete the circle, the photos were taken with another D70. ;-)
Here are the descriptions:
First, the battery was removed since various pins on the connectors will be live even if the camera is OFF. The high voltage on the electronic flash components should not be anywhere near this area of the camera so that should not be a concern. The PCB on the bottom of the D70 has the main microprocessor, non-volatile memory, and RAM. The firmware is probably stored in the 29LV160TE 16M bit flash memory IC next to the chip with the Nikon label. So in principle, it could be swapped, but that's above my pay grade. ;-) Replacing the PCB is only a matter of screws and connectors. And the donor PCB had already been removed from its host during the dissection. Of course, nothing is ever quite so simple as there are at least a half dozen screws of several different lengths and their heads look identical, so either (1) care must be taken to arrange the screws in the correct relative positions after each is removed or (2) they can be compared to the unmodified D70. All the screws around the perimeter of the bottom cover must be removed along with the one on the bottom of the front lens housing, but not the inner ones that secure the metal base/shield inside the cover. Then cover can be angled up and slid off of the USB connector on the PCB.
There are 5 ribbon cables that need to be unplugged. For all except the large one at the end next to the Nikon chip, the black fasteners flip up; for the remaining one it slides out. If a wrong move is attempted something may break and it may not possible to assure the cables make good contact with the connector pins. Once the fasteners are released, the cables will slide out. The 4 large-head silver screws securing the PCB can be removed and the PCB will unplug from a white connector underneath and slide out of the USB housing.
Reassemble in reverse order. There are one or two cushy gray conductive pieces that technically should be replaced but they popped out when the PCB was removed and I could not determine where they went. So be it. ;( ;-) Taking the bottom plate off the other working D70 to check is not going to happen.
After reassembly, it was possible to upgrade the firmware so the 2 working D70 are now similar.
Then I was looking at the camera and realized that the reason the firmware would not upgrade was probably that it was actually a D70S, NOT a D70, though the V1.00 firmware may still have been out of date. There is no on-line way to upgrade the D70S firmware even though it appears as though the current revision may be something like V1.30. And some further digging revealed that the D70 V2.00 firmware is probably very close to the latest D70S firmware. So if that being totally confusing, it's staying the way it is until a reason appears to justify ripping the camera apart again. ;-) Since all functions I've tested seem to work with the D70 brain board (with V2.0 firmware) in the D70S camera, my conclusion is that there is no difference in the firmware.
The Franken-camera appears to work correctly and the photos look similar to the those from the other D70. However, what is not known includes whether there are actual physical differences between the D70 and D70s, and if there is a CCD defect map stored in a chip on the mainboard, which case it would not match. Nothing obvious has appeared but who knows? Most of the pics linked from here were taken with this camera so it appears to work well enough. ;-)
But a while later when attempting to set up a separate LCD monitor for viewing the photos after shooting, this camera does not recognize that a video cable is plugged in while an original D70 worked as described in the manual. It is not known whether this is a preexisting condition, damage caused by the transplant, or something else. Since the video plugs directly into the brain board, which is from a D70, it should behave like a D70. Lack of video is not a great loss though since the output from the camera is low resolution with mediocre quality and would be just barely useful anyhow - perhaps to confirm that the picture is framed correctly but not much else.
Don't panic as the previously recorded photos should still be present. But a suitable USB memory card reader may be required to recover them. Multi-format memory card readers are available for a few dollars on eBay and elsewhere if your PC doesn't have the capability built-in. Confirm that the model you select supports the memory card format! Many may not support the old CF format of the D70. Write capability is not necessary as it should be possible to format it in-camera, or in a Canon camera ;-) if that doesn't work. Formatting is recommended after the photos have been recovered to assure the card's file system is not corrupted.
There is a D80 repair manual on-line. Search for "Nikon D80 Repair Manual PDF". It has many detailed photos with step-by-step disassembly and reassembly, some explanations, and parts identification.
The $25 D80 selected for the tearup has problems with the gears driving the mirror and displays "Err" in the top LCD after valiant whirring attempts to reset it. This is repairable based on various Web videos, but requires an almost total teardown ;-) of the camera to be able to replace the motor assembly and/or large white gear. So while the tearup will reach that point, reassembly is probably not going to happen. ;-)
All photos were taken with the same D70 used for its portraits. ;-) I do have a working D80 but that may eventually be sold. Since its shutter count is over 66K, adding to that was not desirable. The shooting conditions are similar to those for the D70 with the same settings for Web Album Generator.
The photos so far may be viewed at: Nikon D80 DSLR Dissection. (This opens in a single new tab or window depending on how your Browser is set up.)
Here are the descriptions:
There may be more photos to come.
If what you want is entertainment with a bit of useful information, check out the YouTube video Prime Studios - Destroying a Nikon Camera or Web page with still shots PetaPixel - Step-by-Step Teardown of the Nikon D80 Shows You What's Inside a DSLR. Thankfully, both of these are the same D80 and it had already been fatally damaged before he got a hold of it, so the gore is tolerable. ;-)
For the specific problem this D80 has, namely the whirring gear error, there are a pair of more serious YouTube videos at Nikon D80 ERR Split Gear Part 1 which covers the disassembly to access the gear motor and Nikon D80 ERR Split Gear Part 2 which covers the installation of the replacement and then reassembly of the camera. And of course since there are a lot of shots of the camera in various stages of discombobulation nearly to the bare bones, it also serves as a decent dissection, though attempting to keep track of what screws were removed at each step may be rather challenging.
For this camera, the black gear attached to the motor shaft is indeed fractured so the motor spins with fully doing what it's supposed to do, but whether that happened on its own or was the result of excessive torque driving the mirror / shutter mechanism due to some other issue such as a faulty encoder position sensor is not known and I'm not really inclined to go to all the trouble of replacing the gear and reassembling the camera to find out. Sorry. ;-) Though the expense at least wouldn't be much as the gear in my dissected D70 is the same. Even if I didn't have that, there are over 100 listings on eBay for the gear, some under $3. It must be a common failure. And for someone with an attention to detail in keeping track of everything during disassembly (especially the locations of the solder joints for the dozen or so wires that need to be disconnected), repair should be straightforward if not cost effective. ;-)
And to top it off, I accidentally removed the motor mount (not just the motor and gearbox itself via the two large-head screws) and lost one of the rollers without even realizing it until the mirror would not come all the way down. Miraculously, I did find the roller later on the workbench and reinstalled it, but that's the reason why the mirror is in the fully up position in the Mirror Box photos rather than down as would have been preferred. ;( :-)
Coming soon. Photos of an intact D3000 may be viewed at: Nikon D3000 DSLR Dissection. (This opens in a single new tab or window depending on how your Browser is set up.) But that's so boring.
There is an internal flex cable running from near the center of the back panel assembly inside the rear cover of the camera to the LCD itself. It attaches to a "zero insertion force connector" - the ones with the thin lid that has to be flipped up to insert or remove the cable. In this case the cable is short and just barely reaches the connector. So even though additionally secured with a piece of tape, it apparently pulled out over time or more likely was never inserted quite correctly in the first place as that is a bit challenging. Voila, nothing on the LCD, only the back-light.
The rear cover of the camera is secured by 2 screws on either side, 2 screws near the viewfinder (partially hidden by the rubber eyepiece cup if present), 4 screws along the back edge of the bottom, 1 screw further in, and 1 screw under the rubber cover next to the battery compartment latch. There are several different size screws so make sure to set them aside labeled as to their origin. Once the screws have been removed, the rear cover can be popped off, perhaps with the aid of a thin blade. CAUTION: It is connected to the main board via another zero insertion force connector near the bottom so take care not to rip it.
With the rear cover separated from the body, the problem will be obvious. Remove what's left of the small piece of tape, flip up the latch, and carefully insert the cable so that it extends underneath the edge of the connector as far as it will go, and then flip the latch down. Then add a larger piece of Kapton or similar tape to help secure it. Or a bit of 5-Minute Epoxy. Camera operation can be carefully confirmed with the back in place but before installing the screws.
The photos were taken with the second working D70 / D70s following its brain transplant. ;-)
The photos so far may be viewed at: Nikon D5300 DSLR Dissection. (This opens in a single new tab or window depending on how your Browser is set up.)
Here are the descriptions:
More to come, perhaps.
The only repairs that aren't totally unreasonable for a first attempt are to replace the plastic Nikon F bayonet mount on smaller lenses that can get damaged easily (but even this may risk breaking a wire, tearing a ribbon cable, or losing a tiny part), or to replace a bashed front or rear lens group (though on many, doing the latter requires realignment, which is not possible without the proper high tech equipment and training), or the rear housing. The mounting ring may be available at reasonable cost but anything else is likely to have to be cannibalized from a similar lens. So what's the point other for the challenge or excitement value. ;-) If you do decide to take the plunge on a more involved repair, take photos at every step even if there is a repair manual available evem of how it goes back together seems obvious when taken apart. It may not be obvious an hour or day later. Label parts that fit together with "match marks" of some sort like coded scratches or white paint. Many assemblies may appear to have 3-fold (120 degree) symmetry, but that doesn't mean they will work correctly if the wrong choice made. Segregate screws as well since not all the teeny-tiny screws are identical. Work on a surface where tiny parts won't go bouncing off to oblivion. There are unique screws for plastic and metal, and of different diameters and lengths. A padded surface may be useful as well as a magnetic pad to "store" screws and such. Many of the threaded holes are into soft plastic so stripping them is always a risk. If a screw seems tight, it's probably the wrong size. A set of quality Jeweler's or miniature precision Philips screw drivers with magnetic or magnetizable tips is a must.
If a dozen of the identical model lenses need repair, then after the first 8 or so, this will become straightforward. ;-) But "identical" really means the same. For example, the AF-S 18-55mm and 18-70mm (both described below) are totally different beasts.
These aren't like your great uncle's SLR lenses - and even those were basically impossible to repair without proper precision tools, most excellent eyesight and/or a microscope, and a steady hand. There are serious high tech parts in Auto Focus (autofocus or AF) Vibration Reduction (VR) zoom lenses including a miniature motor and possibly a gear train, angle encoders and other sensors for zoom and focus, MEMS gyros and voice coil actuators, PCBs with highly integrated ICs including a microcomputer, and many fragile flex-cables and connectors. In short these are complex intricate electro-mechanical systems, not just a bunch of optics! Though even the basic "kit" AF-S DX Nikkor 18-55mm f/3.5-5.6G VR zoom lens has 11 individual optical elements including one that is aspherical, and others like the Nikon AF-S DX 18-200mm f/3.5-5.6G ED VR II may have 16 or more.
And while modern lenses in this class are amazing feats of mechanical design with sophisticated electronics, they do NOT have the look and feel of older "dumb" lenses with milled aluminum focus and aperture rings on well lubricated tracks. ;-) Rotating parts in these lenses are nearly all molded/formed plastic constructed with mostly sliding parts with minimal lubrication and a few rollers (but without ball bearings) in a few key places. There is no precision machine work to admire and forget silky smooth operation. Even on a brand new lens, this is obvious when rotating the zoom ring. Having said that, they work remarkably well and provide features like autofocus and vibration reduction that one could only dream about with older gear. I do not know whether high-end modern lenses are constructed any differently, but these are what most of us can afford. ;-) And even they typically have a cost if purchased new of several hundred to over a thousand dollars.
Where there is a Nikon repair manual available on-line for the specific model lens, a set of search terms is provided to find it. Should they fail to return anything useful because of decayed links (which is unlikely), I have copies available for the asking. If there is no exact match, a repair manual for a slightly different model lens could prove useful even if the details are not same. But these "repair" manuals are really just disassembly and reassembly manuals with little to no diagnostic information. They are partially in Japanese and make many assumptions about the user's level of expertise. And info on alignment assume the use of Nikon proprietary jigs and software. But there are numerous diagrams and photos. However, none provide the level of detail below with respect to AutoFocus (AF) and Vibration Reduction (VR) implementation.
The largest collection of on-line repair and parts manuals for Nikon equipment appears to be at Learn Camera Repair. That Web site can be searched which will probably be quickest and the manual downloads are all free.
Further, many lenses like the one in this section are made largely of plastic. The only major structural part made of metal is the cylinder with tracks to guide the extension of the focus and zoom lens groups. (Right of center in the photo above.) So it's easy to break parts or strip threads for tiny screws. In fact, the guide rollers which control the extension of the focus and zoom lens groups are plastic, and at least one was found to be fractured in the discombobulated lens used for the photos. It can be seen among the pile of teeny hardware bits.
But these are very high tech devices.
Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Photos and Description
Most of the photos referenced below are also available as a Web Album (though possibly at slightly lower resolution) at Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Parts Web Album.
There are a pair of Epson MEMS-based "XV-3500CB High-Performance Low-Power Yaw Rate Gyro Sensors" to provide the signals for pitch and yaw from which the feedback electronics can drive the VR actuators. The two 8 pin chips can be seen attached to the large cylinder physically 90 degrees apart, near the right-side in the photo above.
Angle encoders for focus and zoom consist of gold-plated patterns with spring contacts to tell the micro-brain what the approximate settings are. These would be used by the control system to know which direction to be able to move and to adjust feedback parameters like loop gain. They can be seen attached to the same assembly as the MEMS sensors are on. The patterns are designed so that no more than one transition can occur simultaneously to avoid ambiguity, though they do not use the common Gray code. This was probably done to minimize the number of patterns or stripes and thus conductive brushes. Those in the Nikkor AF-S DX 18-55mm VR or VR II lenses have 6 data bits but only 2 or 3 conductive stripes (plus a common stripe) providing fewer than half the possible states. More on this below.
For reference, here is the standard Gray code for six bits:
Dec Binary Gray Dec Binary Gray Dec Binary Gray Dec Binary Gray -------------------------------------------------------------------------------- 0 000000 000000 16 010000 011000 32 100000 110000 48 110000 101000 1 000001 000001 17 010001 011001 33 100001 110001 49 110001 101001 2 000010 000011 18 010010 011011 34 100010 110011 50 110010 101011 3 000011 000010 19 010011 011010 35 100011 110010 51 110011 101010 4 000100 000110 20 010100 011110 36 100100 110110 52 110100 101110 5 000101 000111 21 010101 011111 37 100101 110111 53 110101 101111 6 000110 000101 22 010110 011101 38 100110 110101 54 110110 101101 7 000111 000100 23 010111 011100 39 100111 110100 55 110111 101100 8 001000 001100 24 011000 010100 40 101000 111100 56 111000 100100 9 001001 001101 25 011001 010101 41 101001 111101 57 111001 100101 10 001010 001111 26 011010 010111 42 101010 111111 58 111010 100111 11 001011 001110 27 011011 010110 43 101011 111110 59 111011 100110 12 001100 001010 28 011100 010010 44 101100 111010 60 111100 100010 13 001101 001011 29 011101 010011 45 101101 111011 61 111101 100011 14 001110 001001 30 011110 010001 46 101110 111001 62 111110 100001 15 001111 001000 31 011111 010000 47 101111 111000 63 111111 100000
Here is a closeup of the actual focus encoder color coded and labeled with the binary value at each position: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Focus Encoder Coding. While it has 6 bits, there are only 29 states. Interestingly, the codes for the near and far end-zones are complements of each-other, which is probably a coincidence but perhaps not. ;-)
The Zoom encoder is longer because the zoom ring moves over a larger angle than the focus ring, but it's also wider. Here is a closeup of the actual zoom encoder color coded and labeled with the binary value at each position: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Zoom Encoder Coding. Like the focus encoder, it is a distance-1 code (but not the same one) with 6 bits but only 31 states. The common contact is on the opposite side of the coding strips for some unfathomable reason, the relative lengths of each code segment along its length may also differ, the code sequence is not the same (even accounting for the strip position swap), and there is a larger number of states. Aside from all the differences, everything is identical. ;-) Note that since the photo of the zoom encoder strip was taken in situ and it covers a large angle, the areas near the ends are squashed by perspective - the codes should be more spread out than they appear there.
What isn't clear is why the patterns are non-uniform. It would make sense for them to not have a relative spacing based on the sensitivity to change depending on the focus or zoom position. But all the encoders so far analyzed have seemingly arbitrary variable spacing. Further, in comparing the patterns for several AF-S lenses, no two are the same (even accounting for the required length and available space). Since the code is almost certainly mapped through a lookup table, the absolute values should not be relevant. Go figure. ;-)
Movement of the lens elements for autofocus in these AF-S lenses doesn't use a conventional motor with magnets and coils like in the good old days. ;-) Rather it uses a motor based on the piezo-electric effect, which is property of some types of crystals to deform when a voltage is applied. The most commonly used crystal is lead zirconate titanate (Pb[Zr(x)Ti(1-x)]O3), usually abbreviated "PZT". This is the same stuff used for the obnoxious beeper elements found in all sorts of electronics. PZT may also be used as an abbreviation for PieZo Transducer, the device itself, which is convenient. ;-) The focus motor is driven at a frequency beyond the range of human hearing and thus similar technology is called "Ultrasonic" by other manufacturers based on the name for the actual motor technology. (Some are called "Hypersonic" which makes no sense at all.....) So focusing should be very quiet - except perhaps for the bats you're attempting to film. ;-) (The "S" in AF-S supposedly stands for "Silent", so all AF-S lenses should use similar technology. However, even if the motor itself is silent, the gears still makes a sound and other parts of the lens may not be quite silent. Nikon calls them: "Compact SWM","Gear SWM", or "Micromotor SWM" where "SWM" stands for "Silent Wave Motor". Much more on this below. Upon first examination, the motor in the AF-S lens appeared to be very strange without any detectable magnetic field and with the rotor almost locked in position, held in place against a plate by a strong spring, more like a clutch than a motor. Furthermore, all combinations of the red/yellow/black wires tested open and applying low voltages to them resulted in absolutely nothing, not even smoke. ;-)
There are two types of SWMs used in Nikon lenses: Gear motors (like this one) and ring motors (which are almost the diameter of the lens). Both are piezo motors and have advantages and deficiencies. For a brief description with photos (but little explanation), see Two types of motors 'Nikon SWM' (RADOJUVA).
Gears are still needed to provide adequate torque to rotate the focus ring reliably, around 3:1 down from the motor to the white pinion gear. The ratio is around 1:3 up from the motor to the slotted tachometer disk. Count the number of teeth on each gear for precise values. ;-)
The capacitance of the each of the two phases was measured to be approximately 0.6 nF. Attempting to drive the motor with a function generator and 1 nF capacitor in series with one of the inputs to shift its phase resulted in absolutely nothing happening at the ~20 V p-p maximum output of the function generator over a range of less than 1 kHz to greater than 100 kHz. So actual high voltage is probably needed. Measuring the voltage on this lens is regrettably not viable since it's in bits now and was never in a usable state even before it was disassembled. If another sacrificial AF-S lens came along, that would definitely be high on my to-do list. However, guessing that there would be enough leakage from the high voltage components, a sense coil was constructed using a 3 turns of hookup wire and connected to an oscilloscope probe. It was then possible to easily measure the PZT drive frequency for a similar lens. The coil was simply placed under the lens - which turns out to be approximately where the piezo motor and POW PCB are located. A distorted triangle wave with a constant frequency of around 77 kHz and amplitude of over 80 mV p-p appeared on the scope only when the focus ring was moving. The waveform shape changed slightly based on direction, but who knows how the coupling (which is probably mostly from the ferrite transformers) affects it. So 77 kHz does indeed qualify as ultrasonic. ;-)
Using a sense coil and scope may be a decent diagnostic test to determine if AF-S autofocus failure is due to a mechanical problem or lack of the PZT drive signals. If no signal can be detected, or if there is only one phase, then it's probably an electronics or control problem.
Perhaps the key to understanding how this works lies in the "W" of SWM - Silent Wave Motor. This type appears to be what's called an "Ultrasonic Motor" operating at the resonant frequency of the structure, around 77 kHz as measured above. Each set of 4 segments connected to the same input lead if duplicated around the stator would result in a total of 8 more or less equally spaced equal length segments. The two sets appear to be physically offset by 90 degrees. Each set if driven by itself would create a standing wave around the stator. But if both sets are driven the result is a traveling wave whose direction depends on their relative phase with a net torque in that direction. Another way to think about it is with an analogy to those magnet and coil motors, specifically two-phase induction motors. These have two sets of coils that are offset from each-other physically by 90 degrees. The relative electrical phase of the power applied to the two coils then determines torque and direction of rotation. The reasons for the conclusion with respect to this piezo motor are (1) the diagrams and photos of ultrasonic motors from a Web search have a similar slotted architecture, (2) they also have only 3 connections, and (3) the explanation in Survey of the Various Operating Principles of Ultrasonic Piezomotors, K. Spanner seems to fit. But the details are still somewhat of a mystery. Nikon has a bazillion patents on autofocus using the SWM, mostly Japanese though Google does a half decent of translation. Two US patents are US8035906B2 and US20200350838A1 which seem to be close in appearance to this motor, even down to the number of segments on the slotted disk and precise pattern of the PZT elements. These come up or are referenced when searching Google Patents for: Nikon "Silent Wave Motor". Reading patents is always a treat though. :( ;-) If anyone can shed more light on the principles of this motor, please contact me via the Sci.Electronics.Repair FAQ Email Links Page.
Interestingly, the repair manuals for the 18-55mm and 18-200mm AF-S lenses found on-line make a specific note to not touch the area of the piezo motor. With high voltage involved, that may be good advice, but this in reference to disassembly and re-assembly where the lens is un-powered. So, it probably really refers to avoiding contamination which would cause electrical leakage and a reduction in drive voltage as well as possible actual damage after running a long time like what is seen here. Perhaps that is why AF-S lenses have a reputation for autofocus failure - with no sealing of any kind, dampness is getting in over time providing a leakage path so the motor eventually stops working reliably. This lens almost certainly had that failure.
As if that wasn't exciting enough, here is the good (VR) stuff. ;-)
Left photo: Solder globs for the fat traces to the two coils that move the lens are visible at the bottom and left. The set of 6 fine traces appear to be for a pair of Hall-effect sensors (stuck under the Kapton bumps at the top and right positions).
Right photo: Solder globs for the fat traces to the locking coil (see below) are visible at the top. The metal plate is one of the magnet pole pieces.
Left photo: The two coils that move the 3rd group lens element are readily visible. There is a rare earth magnet under each of the coils. A quick test with a 2 V power supply confirmed that the lens does indeed move from side-to-side in two axes as expected, but doesn't tilt. Two springs to keep the lens in place are also visible at the upper-right and lower left. The metal squares at the top and left are glued to rare earth magnets to bias the Hall-effect sensors which would be located directly behind them if the original flex-cable was still in place. But one of them fell off without any provocation. Gluing metal to plastic doesn't work well. ;( ;-)
Right photo: The large metal plate acts as a pole piece for both of the lens element positioning coils. A third coil using an additional pair of magnets hidden under the pole piece is used to rotate a collar to keep the movable lens centered and to prevent it from wobbling on its own due to vibration when VR is OFF. The leaf spring to keep it locked without power is visible at the right. The yellowish metal block at the upper-right is probably just a counterweight for this assembly and is probably made of brass to be non-magnetic.
There must be some pretty fancy footwork going on in the algorithms on that CPU board to actually implement the VR. And the lens element must be maintained centered electronically when VR is on using the Hall-effect sensors for feedback. There is no restoring force so it will tend to sit at the lowest point. When VR is off, it will be more or less centered using the locking collar. The springs only keep it against the surface of the front cover.
And you thought camera lenses were boring. ;-)
Out of curiosity, I puchased a supposedly defective sample of this lens where the trim ring had popped off as shown in: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens with Loose Trim Ring. It looked bad and for that reason the price was right, but was easily repaired. See the section: Repairing Loose Trim Ring on Nikon AF-S DX Nikkor 18-55mm f /3.5-5.6G VR II Zoom Lens.
While I search for another one to dissect, measurements of the SWM frequency may provide a clue as to the design changes. A 6 turn sense coil was placed under the lens, which as before was determined to be the optimal location for a strong signal, probably in close proximity to the ferrite HV transformer(s). Rather than 77 kHz as in the original AF-S lens, the frequency for the AF-S II is around 290 kHz (!!) which was originally assumed to indicate a much smaller motor (though this turns out not to be the case, more below).
When the focus ring is moving, the pulse width may be up to around 700 ns. But there is still a detectable very stable ~100 ns spike at the same frequency with similar amplitude for several seconds after the camera has beeped indicating that optimal focus has been achieved, not only while the focus ring is moving. So it is likely from the same source and not just pickup from the logic, perhaps a sort of dither to minimize stiction in the mechanism between focus operations.
Then I found photos supposedly of replacements for the AF-S VR II SWM and also one with the gear train in eBay listings and they look similar to those in the non-II version, but the gear train at least must be slightly smaller as the original would not fit the VR II lens. The size of the SWM has probably not changed though. And a listing for a replacement PCB showed that at least one of the PCBs is a ring rather than the rectangular as in the original. An eBay search using the terms: "nikon 18-55mm vr ii lens (motor,parts,pcb)" should turn up listings with these photos. It's not likely they will go away give the ridiculous prices the sellers are charging for replacement parts. ;-) Based partially on the photos, my conclusion now is that the higher frequency may have been selected primarily so the size of the electronic components like the ferrite transformers could be reduced, because it permits faster focusing, or because of complaints from local bats. ;-). And that the physical size reduction of the VR II lens is accomplished by three primary changes:
So mostly clever packaging. ;-)
Here is a summary of the Nikkor 18-55mm AF-S DX f/3.5-5.6G VR II lenses I've acquired so far:
Note that for the following, the Main Barrel, Fixed Shell, Metal Ring/Shims/Insulator/Spacer, and Bayonet Mount Plate are all one structure secured from the back by the 3 larger screws. The Zoom Ring rotates around the this and the lens groups move with respect to it.
Most of the photos are now present in the Web Album. Except for the first 2 stock photos, all are of lens #6 except for the few showing the back of the lens with the bayonet mount and associated parts removed.
Other than the front of the 7th lens group and back of the 6th lens group which has some mottling (which was easily removed with 90% isopropyl alcohol), the interior including the PCBs, connectors, SWM, gears, and everything appear to be pristine with no evidence that contamination caused the lens's failure. So perhaps it was zapped by static.
Some conclusions so far: The VR II is somewhat cost reduced but in a good way. ;-) For example, only one wire needs to be unsoldered to totally disassemble it to the level of major components. Most flex cables plug into the CPU PCB (PCB1) and are the push-in type with locking levers to break. Many have a small hole near the end of the cable for a thin instrument to aid in removal. There is a minimal number of pieces of double sticky tape or adhesive which would need to be replaced if re-assembled.
Focus implementation is generally similar to that of the AF-S VR lens but the SWM gear train is part of the main structure and NOT a separate assembly. The tachometer and its gears are gone. Speed sensing has been implemented with a magnetic strip as in some other AF-S lenses. The good news is that the SWM itself and the two remaining gears are replaceable without major disassembly.
The focus encoder coding is of course different than the one in the AF-S VR and all other lenses I've checked. But is that a surprise? ;-) See Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens Focus Encoder Coding. This encoder has 6 data bits with 22 states. The spacing almost makes sense being fairly similar in the central region and further apart near the ends. The extra tracks in the photo are pass-through for the magnetic read head.
The Zoom Encoder does NOT run axially as in the the AF-P 18-55mm VR lens (which has a similar lock button). It runs around the perimeter of the inner surface of the Fixed Shell and is extended to cover the locked area with a constant code. (Take care not to damage the brushes if the Fixed Shell needs to be reomoved.) At the location where it becomes unlocked, there is a very small length where the common conductor extends to the next one over, but that would not result in a different code, so its purpose is mystery. See Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens Zoom Encoder Coding. This encoder has 8 data bits with 36 active states and the locked state. In the main portion of the active zoom region, their spacing is similar but not precisely the same. For the locked region, the code is a constant 00110001. Note the 8 data bits rather than 6 data bits for all the other lenses analyzed so far. Yet again, we have a change to the encoder coding and now even the number of bits for no fathomable reason. 6 bits would be plenty. It almost seems as though the designers of these lenses (1) are not privy to documentation on other lenses, (2) have a serious case of NDITD (Not Developed In This Department) syndrome, (3) never heard of lookup tables, (4) are smoke'n sump'n, or (5) all of the above. It makes no sense whatsoever.
The VR II VR assembly has changed slightly compared to the AF-S VR (I) lens. In addition to being smaller and similar to the one in the AF-P VR lens, there are two sets of connections to components hidden under the moving armature. My conjecture is that these are replacements for the Hall sensors in the original AF-S VR lens, but they each have only two wires, so perhaps some sort of magnetic affair to detect changes in the VR lens group position. The AF-P VR lens has something similar.
The state of affairs as depicted in the Web Album is about as far as I intend to go in the disassembly. There are still a few individual parts that have not be removed or reduced to their individual pieces, usually due to issues of reversibility of the procedures, but all key sub-assemblies have been removed and documented. This applies for example to the SWM and VR assembly. There is little point to taking them to bits as that has been done for the 18-55mm AF-S VR and AF-P VR lenses, and they are similar. While nothing had been damaged, restoring the lens to its original condition is probably not going to happen. While there is a small probability that simply reseating the connectors will have cured the "not recognized by the camera syndrome", I'm probably not that determined to find out. And if the CPU PCB (PCB1) is dead, swapping in one from another lens will not work well as VR parameters and other settings need to be optimized for each specific lens using the custom specialized Nikon test instruments and software. However, physically putting it back together seems possible without having mastered a 27 level Rubik's Cube though there are a couple of steps that could be dicey like replacing the Fixed Shell.
Explanations of the dissection photos will be forthcoming so stay tuned. ;-)
And should anyone actually read this before the warranty on the Universe expires and has specific questions or requests, I may be contacted via the Sci.Electronics.Repair FAQ Email Links Page.
The trim ring is held in place only with a strip of ~1/4 inch tape wrapped around the entire lens. But while the original AF-S VR lens has a decent width area around its entire perimeter for the tape, the AF-S VR II trim ring only has six very narrow tabs that after awhile (or from an impact) can come loose as was the case with the first AF-S VR II lens I acquired (ID #1). See Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens with Loose Trim Ring. It looked really bad and the lens was destined for dissection if repair was not practical. But fixing it simply required removing the rubber grip by lifting it away from the body pulling it off, and then replacing the tape. But Nikon must use tape with super-strong adhesive as ordinary tape probably won't hold for long. So far though it's been behaving with just some Kapton tape, and many of the photos for the camera dissections have been shot using this lens. It might also be possible to use a few dabs of adhesive like 5 Minute Epoxy or industrial strength rubber cement between the trim ring and the barrel it seats against, but this has not be attempted - yet.
And the racing stripe (which is purely cosmetic) is just a length of really thin metal-coated tape which often detaches at one end. Or corrode as in the case of the lens ID #6 used for the dissection. It can usually just be peeled off if desired, which is a lot easier than attempting to neatly glue it back in place.
This procedure may apply to some other Nikon lenses. But specifically NOT to the similar AF-P lens where removing the CPU contact block screws results in the individual contacts popping out all over the place. ;-(
A #00 or #000 Philips screwdriver is required for the 3 sizes/types of teeny screws:
See Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens Bayonet Parts.
Reassemble in reverse order. And don't force anything! All of these are tiny screws so stripping holes is possible. When reinstalling the bayonet mount, carefully insert the aperture tang straight into the plastic receptacle that moves the iris diaphrapm. Unless the bayonet mount is replaced without changes, the setting of the aperture tang may need to be adjusted. It's secured with 2 screws and sealer. I do not know what the official procedure is, but in lieu of that, make sure the aperture tab (that's activated by the camera) just touches the the edge of the cutaway in the inner ring. Then after securing the bayonet mount, confirm that the tab moves smoothly by hand and the aperture goes through its entire range.
This lens (Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II ID #4) works perfectly in all respects from its closest focus to around 2 feet. It will go back and forth between those distances all day without issues. But beyond there, it will never move the focus ring to get closer, and at some point it will jam against the end-stop beyond infinity.
This is a works in progress but here are some observations:
My current hypothesis is that the encoder that reads focus position is either not working correctly or has perhaps become disconnected. So the lens's microbrain is getting confused, poor thing. ;( ;-) Among other things it's not being programmed to provide the necessary torque to focus in from near infinity, and it's overshooting infinity, hitting the end-stop and getting really stuck there. Monitoring of the SWM waveform appears to show that it is trying but almost certainly attempting to rotate the focus ring in the wrong direction. With the high gear ratio, it should have no problem backing away. The only way to unstick it if jammed is to flip the A/M switch back and forth. However, even if not jammed, it will still not focus in from more than around 3 feet to infinity. It will focus reliably from the closest spec'd distance to around 2 feet, though it may overshoot dramatically but doesn't jam. All these lenses sometimes overshoot especially if there little detail in the focus zone.
Since there is a definite boundary beyond where it screws up, this would appear to rule out an SWM or gear train problem. The SWM rotor and all gears except the one on the focus ring itself go through multiple revolutions over the focus range. The rotor of the SWM itself goes through ~2.75 revolutions for each revolution of the focus ring drive gear and that goes through nearly 2 complete revolutions to move the focus ring from end-to-end.
So this is probably a control problem. The only inputs to the microbrain are the tachometer (a magnetic strip with read head) and the focus encoder (which senses the absolute position of the focus ring). It's possible the magnetic pattern the tachometer uses has been partially erased, but the most likely cause is the focus encoder: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens Focus Encoder Coding with the bits color coded and labeled with the binary value at each position. More on these encoders can be found in the section: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens. But interestingly, the focus encoder in the Nikon AF-S DX Nikkor 18-70mm f/3.5-4.5G ED IF Zoom Lens also has 4 conductive strips and 6 bits, but the coding differs, which really doesn't make any sense unless each new Nikon designer is required to change something as a test. ;-)
If you're not totally confused, you weren't paying attention. ;-) But to reiterate: I believe that the focus encoder is either feeding bad readings to the microcomputer or they are being misinterpreted.
No doubt disassembly of this lens will be required at some point, to satisfy my curiosity if nothing else. Hopefully the problem will be something obvious like a damaged focus encoder brush, loose connector, or some dirt or soda residue on the encoder strip. ;-) However, based on the dissections (See above), this requires fairly major surgery, so it may be postponed until Major Medical is available. ;-(
So AF-P lenses return to motor technology with magnets and coils. ;-) As noted above, the "P" is supposed to stand for "Pulse", which kind of applies. They are claimed to be even quieter than AF-S lenses and that is probably true. But the manufacturing cost is also much lower. ;-) And there have been comments on various forums about AF-S autofocus reliability, which is quite credible given their complexity and opportunities for contamination to get to the motor. So perhaps a little of both. Replacing the piezo motor with a stepper motor also allows the AF-P lens to be more compact since the motor itself is less bulky and the high voltage drive components and gear train are eliminated.
However, AF-P lenses are not compatible with the D5100 or D3200 (or earlier) cameras that are happy with AF-S lenses. Others like the D5200 may need a firmware upgrade (but that is a free download). And since there is no VR switch on the AF-P lens, VR is always enabled on these cameras since there is no electrical contact in the camera body to control it. (The AF-P version has 8 contacts compared to 7 on the AF-S.) But this incompatibility is almost certainly due to a business decision for planned obsolescence. There would not appear to be any reason why an AF-P lens could not have been designed to look like an AF-S lens as far the the autofocus commands are concerned. Or at worst, with a way of selecting the mode via a switch
The AF-P lens is also significantly narrower than the AF-S VR lens and slightly narrower than the AF-S VR II lens which could be in part due to the more compact drive setup. The piezo motor has a relatively large diameter (almost 1/2 inch) and the gear train also takes up space. For smaller lenses like these, the only option is to increase its overall diameter. The stepper motor with its direct worm drive can greatly reduce the required space.
The AF-P lens destined for analysis is definitely well worn. The lock button doesn't work properly and in addition, one of the three tabs on the bayonet mount is broken off. Nonetheless, it still seemed to work well enough on a camera. But from the start, its days were numbered. ;-)
After starting the dissection, I had other suggestions for the "P" in AF-P: "Pathetic" or perhaps "Plastic". Nearly everything structural is made of plastic except the screws and a few tiny brackets. However, having said that, the AF-P lens is much simpler and may be more reliable than its AF-S cousin. Autofocus has only two moving parts - a stepper motor with worm gear shaft which moves an internal lens group over a total distance of around 7 mm using low voltage drive. Compare that to reduction gears in the AF-S lenses along with the possibly tempermental ultrasonic piezo motor. The manual focus ring generates signals to the microbrain that then controls the same motor - it is not directly coupled to it: "Focus by Wire". Vibration Reduction (VR) is simplified as well with no Hall-effect sensors or lock mechanism. As a result, the electronics are also much less complex. In fact, as will be seen below, the electronics is perhaps an order of magnitude simpler in terms of the number parts compared to the AF-S version. This may be largely due to the lack of need for the high voltage piezo drive since the large ferrite transformers and drive components are eliminated. But may also be due in part to the higher level of integration available at the time of its design. And there are no critical surfaces to get contaminated as with the ultrasonic piezo motor. So I officiatlly retract "Pathetic" because the AF-P lens should be functionally at least as capable as the AF-S version, and more reliable without the SWM, high voltage drive, and gear train.
But it almost appears as though this particular lens must be assembled from the inside-out. :( ;-) For example, in order to get to access any internal parts, the curved strip with contacts that make connections to the camera body must be disassembled down to its individual contacts, which then pop out all over the place. It isn't self contained with the flex-cable as in the AF-S. So if the plastic bayonet mount gets damaged (as would seem to be quite common even though this is a small light-weight lens), replacing it requires some serious manual dexterity. Nikon must have saved 3 cents. ;-)
Taking it to bits non-destructively isn't that bad, though putting it back together without detailed instructions would be like solving a 10-level Rubik's Cube blindfolded. ;-)
One mystery is solved though with respect to the silent propulsion system for autofocus. As expected and noted above, there is a very small stepper motor (~3/8" diameter) whose shaft has an integral worm gear and no other gears. That rests in a Nylon U-shaped bushing enabling the entire focus assembly with the 3rd lens group to be moved back and forth by around 7 mm with an opto-interrupter as a limit sensor at one end. The focus ring works in parallel with the manual focus electronically: There is an incremental encoder consisting of spokes on the perimeter of the focus ring with a pair of nearly microscopic opto-interrupters in quadrature to sense their movement. So, the stepper motor can be driven either by the autofocus electronics or focus ring essentially at the same time. It's "Focus by Wire". ;-) But manual focus will not work if power is off, which is only of academic interest unless the lens is used in an incompatible camera or for another application. This is fundametally unlike the AF-S version of this lens where the focus ring actually moves a lens group on a spiral track and the A/M focus switch selects (1) whether it is coupled to the gear train and (2) lets the microbrain know.
Without a gear train, this should be quieter than the AF-S. The stepper motor itself may make a detectable sound but sliding noise will reduced and there is no gear train to whine.
Nikon AF-P DX Nikkor 18-55mm f/3.5-5.6G VR Optics
The AF-P has 6 lens groups (unlike the AF-S that has only 4), though some may be single lens elements.
The position of the 1st and 2nd-6th (in the same relative position) lens groups move independently depending on zoom setting. The position of the 3rd changes relative to the others depending on focus setting controlled by the stepper motor.
Nikon AF-P DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Photos and Description
Most of the photos referenced below are also available as a Web Album (though possibly at slightly lower resolution) at Nikon AF-P DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Parts Web Album.
An absolute encoder for focus is not needed since the stepper motor is digitally precise: The position of the 3rd lens group for focus is proportional to the number of steps and direction referenced to a limit sensor at startup. The focus ring at the front of the lens is incremental and does not control the focus setting directly.
Nikon AF-P DX Nikkor 18-55mm f/3.5-5.6G VR Disassembly Procedure
Not all of the following is needed depending on whether this is for repair or curiosity. Do NOT do this if the future of the Universe depends on getting the thing back together in a functioning condition. ;-)
Remove the 3 screws securing the electronics PCB.
Remove the 3 screws securing the 6th lens group and set it aside along with the screws and washers.
The VR assembly is much simpler than on the AF-S lens. The Hall-effect position sensors are gone as is the lock mechanism. Three really tiny springs as well as two blobs of some elastomer goop hold the lens element more-or-less centered regardless of orientation. There are only two coils and their associated pairs of rare earth magnets, each with U-shaped pole pieces.
With no lock to restrict the movement of the VR lens, when VR is OFF (on cameras that support that), the only ways of minimizing movement are to effectively short out the coils which will reduce movement due to eddy currents or for the VR system to drive the coils with appropriate anti-VR waveforms to effectively lock it in place. None of these would be available on cameras without an electronic VR-OFF setting. Or perhaps that elastomer goop serves a similar purpose. ;-)
That's basically it. There are now a pile of parts where there used to be an AF-P lens. ;-) Reassemble in reverse order, left as an exercise for the student or masochist.
I have been in fact been unable to reassemble it into a fully working state - even mechanically. While it's reasonably straightforward to get the major pieces screwed into their proper place, fitting them into the appropriate combination of grooves and slots in the cylinders that control how far each one moves as a function of zoom is a challenge. There are probably match-marks in conjunction with jigs that to the trained (Nikon) eye would make this intuitively obvious. The closest I came was to get it to move the front of the lens back and forth in what appears to be the correct way based on zoom, but that was through random chance and could not likely be reproduced. Whether the other parts move correctly is not known. There are 3 separate assemblies that move based on zoom that need to go into their respective grooves and slots, and also need to be correctly oriented with respect to the 120 degree symmetry of the lens, so among other things, the zoom distance labels, and mark and lock line up correctly.
The more I look at these, the more they appear to be marvels of engineering down to the casting/molding of the numerous circuitous groves, slots, holes, posts, blocks, and other structures in plastic. It's probably just Zoom Lens Design 101 but still impressive to the uninitiated. ;-) Unfortunately, sometimes they aren't strong enough as will be seen with the next lens. :(
Autofocus on this lens is quick, but not necessarily quieter than on the AF-S lenses using the small motor and gears. The motor and lens has sliding surfaces which still make some sound.
But the sacrificial victim makes abnormally loud grinding noises and fails to be able to focus correctly - either manual or auto. :( ;-)
The cause became obvious as a huge part - the entire 2nd lens group - was loose inside the lens not attached to anything just bouncing around. ;-( Figuring that the 1st lens group would detach like the others - by unscrewing it after removing the label, that was attempted first. But either it has left hand threads or it is really tight and I don't have the needed spanner wrench, so it remains securely attached. No matter. ;-)
Plan B was to go in from the back, where the action is in any event. This turns out to be quite simple and even reversible. Removing several screws around the side of the bayonet mount and the back allows both to be removed without damaging anything. The electronics PCB is then exposed and its cables can be unplugged easily along with the A/M switch revealing the full diameter autofocus ring motor. The contacts remain safely inside their housing.
Nikon AF-S DX Nikkor 18-70mm f/3.5-4.5G ED IF Zoom Lens Photos and Description
Most of the photos referenced below are also available as a Web Album (though possibly at slightly lower resolution) at Nikon AF-S DX Nikkor 18-70mm f/3.5-4.5G ED IF Zoom Lens Parts Web Album.
Being simpler than the VR lenses, there are fewer photos for this one, but there is always the on-line repair manual to refer to:
According to the repair manual, alignment will be required if the 4th or 5th lens groups are removed. So, avoid doing this if possible. Or, at least, as each of the 3 screws is removed, take a closeup of the exact relationship of the screw hole to the clearance hole.
Prior to disassembly, the piezo motor did work well when focusing, moving from one end to other in 1/10th of a second or less. However, due to the broken tab on the 2nd lens group cell, it didn't do anything useful. :(
That gray spot where the flex cable center conductor attaches to the PZT stator ring (left-most piece in the left photo) almost looks burnt. But it's just a bit of conductive adhesive to assure good contact. There is a small hole in the flex-cable underneath it.
The rotor (right-hand pieces in each of the photos) consists of 4 parts: The aluminum ring with a lip which is what's in contact with the stator, a thin steel ring next to it, a segmented aluminum ring next to that, and the black plastic shell molded around everything, The sliding surface of the rotor is an insulator, possibly an aluminum oxide coating, and there is also no electrical contact between the metal rings embedded in the plastic shell or to the rotor itself.
To check the drive signal, a 3 turn coil was wrapped around the working AF-S 18-70mm lens. Not surprisingly, its frequency is much lower than for the small motor in the AF-S 18-55mm lens, clocking in at only around 28 kHz. The frequency is still ultrasonic, but not by that much. ;-) And as noted with respect to the AF-S 18-55mm lens, above, this would be a good way of determining if autofocus failure is due to lack of drive or a mechanical problem. But the sensed amplitude is lower, at around 20 mV p-p since the source is more buried. Use more turns if desired. ;-)
This lens appears to be more repair-friendly than the ones above especially if there is no need to go inside the assembly with the 3rd-5th lens groups. There should be no need to unsolder any wires and the flex cables detach easily. As noted above, just keep track of everything with photos, notes, and added match marks.
However, note that there is a magnetic strip and magnetic pickup that provides a signal in place of the tachometer in lenses that have the small SWM with gear train. Not only is it delicate and damaged easily, but ferrous tools can cause the magnetic pattern to become corrupted, which needless to say, would not be good.
Modern lenses are much more sophisticated and no one would want to go back to the fully manual older ones, but silky-smooth operation is not one of their features. And it's easy to see why. Most of the moving parts are made of plastic and a zoom lens has many of them. For example, see Major Moving Parts of Nikon AF-S DX Nikkor 18-55mm f/1:3.5-5.6G VR Zoom Lens. These all move when changing zoom or focus. The three cylinders at the top of the photo reside nested and rotate or slide with respect to each-other with a large surface area in direct contact. The center one is made of anodized aluminum with precisely milled slots that determine the required movement of multiple lens groups with respect to each-other when zoom is adjusted; the other cylinders are formed or molded plastic. The straight slots in the upper-left (outer) cylinder guide those moving parts that must not rotate. Pegs or rollers (without ball bearings) restrict their movement to the AND of the slots in the upper left and the other cylinders, but also add friction. Some parts reverse direction as the zoom ring is turned, adding additional friction/resistance at that point. It's all rather intricate and I bet Nikon has a really nifty CAD package for zoom lens mechanical design. ;-)
As an example of a common much larger lens, see the diagrams in Nikon AF-S DX Nikkor 18-200mm f/1:3.5-5.6G VR Zoom Lens showing Lens Group Positions at 18 and 200 mm. Lens Group 3 and the VR assembly move together, but those and all the others move relative to one-another and relative to the lens structure attached to the Nikon F mount.
Even on a brand new lens, there is detectable roughness and varying resistance over some parts of the zoom range. Over time, the lens will be exposed to dust, moisture, and contamination either from normal use or from being tossed in a storage bag, it gets worse as there are no real seals. And plastic is subject to wear. The good news is that for the most part, none of this makes any real difference in picture taking performance. That is, until the thing seizes up completely or falls apart. :( ;-)
But if the lens is dropped or whacked, parts like those pegs can get broken off or may dig into the tracks with varying degrees of damage. In minor cases, the roughness will just become worse but with enough trauma, major functions will stop working. When considering the purchase of a used lens, carefully check autofocus, manual focus, and vibration reduction, as well as for correct operation of the aperture at all zoom settings. Don't accept a lens where manual focus doesn't work reliably even with a discount (as I once did) because the seller said no one ever uses it. While that may be partially true, unreliable manual focus can be a symptom of more major problems. One thing that is often broken on used lenses though is the Lock button if there is one, used to secure the lens in a compact state for storage. The internal lip that the button engages is made of plastic and users often attempt to twist the zoom ring without realizing it's locked, so that lip gets damaged and the button non longer works properly. That alone is probably not sufficient reason for rejecting a lens - but perhaps it may be leverage to negotiate a discount! While Lock may not work or work well, the end-stops for the zoom ring should not be affected. But this should be confirmed as bad things may happen on some lenses if the zoom ring is rotated beyond the normal range.
See the Web Album at: Copal Square S Focal Plane Shutter Mechanism. (The Web Album photos are scaled to fit within 1024x768 pixels but the full size originals have the name under the thumbnail with a ".jpg" added.) The first 4 photos are of a beat up Nikkormat FTN in various stages of disassembly starting with most of the pieces of the lens mount in place to revealing the Copal Square S shutter in situ. These are followed by closeups of another similar shutter. The primary difference between them is the use of a less expensive Nylon gear for the speed select compared to the highly polished brass one, and some slotted head screws in place of Philips head screws. Since there is no real stress on that gear, cheaper is just fine, thank you. ;-) My black dot on the white gear lines up with the post for the 1 second setting. In the interest of full disclosure, I have swapped the gear and screws to make the separate shutter mechanism more photogenic. And in the interest of expediency, the screws that secure the body parts have been left off. ;-)
Two manuals relating to Copal Square S Shutter repair are known to be available on the Web and hard copies may be purchased on eBay and elsewhere. Both Copal Square S Shutter Repair Manual and Copal Square S Shutter Repair Guide are interesting reads, but they may not enable you to be able to do much in the way of repair. The first one does have a 75 (!!) step procedure with diagrams for assembling a shutter. ;-) Aside from the intricate nature of these mechanisms, special jigs and instruments are required for some of the procedures. However, cleaning and lubrication of specific parts may be possible. This will involve the use of solvents like alcohol or naptha along with an ultrasonic cleaner if available, followed by lubricating specific bearing points and surfaces ONLY with the tiniest speck of special oil or grease as appropriate. A shotgun approach of simply sprayng it with degreaser and adding oil anywhere that looks appropriate will likely result in a nice paperweight. DO NOT even think about allowing WD40 or anything similar near a precision mechanical device like this! ;( ;-) A Web search will turn up suitable procedures but take them all with a grain of sand.
Control of the shutter bears similarity to that of mechanical leaf shutters, but it needs to determine the timing of the pair of blinds rather than opening and closing a set of leaves. For the Copal Square S There are three (3) regimes of timing:
And as a matter of interest, operation of the shutter in a fully mechanical SLR and specifically the Nikkormat is as follows:
For the "B" setting, everything is the same except that a tab on the shutter linkage prevents the shutter from closing until the button is released. The actual shutter speed is probably forced to 1/1000th second so closing would not be delayed no matter how quickly the button is released. The Nikkormat doesn't have a "T" setting, but for that operation would be similar but there would be a simple escapement that would require the button to be pressed a second time to close the shutter.
It turns out that a variety of types of glass and plastic may be used and the optical elements may be either ground and polished or molded. Sometimes the lens specifications will include some information on the material thought probably NOT the fabrication method if the Marketing Department thinks it will help sales. For example, Extremely low Dispersion Glass (ED Glass) and aspheric are pointed out in the info for lenses like the AF-S DX Nikkor 18-55mm or 18-200mm. They probably won't state anything if plastic. ;-) If not specified, the material can be any either common optical glass (BK7, crown, flint, etc.) or plastic. Aspheric elements are probably molded since individually grinding and polishing them would be cost prohibitive.
There is no easy way to determine the material and fabrication method non-destructively (or at least without some damage) on an intact lens as they appear identical. But even if the lens is disassembled into the individual lens groups it's a challenge. Glass is several times more dense than plastic so the weight of a lens group can be a tip-off, especially for the larger ones. Ground and polished lenses will generally have frosted edges while molded ones will have smooth lips and perhaps even tabs. But the overall appearance of the individual lens elements is essentially identical in terms of surface finish and AR-coating.
Some possible causes are:
Diagnostics: Test with another AF-S lens. If that also fails, the camera is likely at fault.
All cameras tested that would turn on worked correctly with AF-S lenses. Dead cameras don't count. ;-)
Diagnostics: Cycling the switch back and forth a few times may clear any contamination. But on lenses with the small SWM, this also engages and disengages the gear train, so that can result in the gears moving a bit with respect to each other and avoiding a damaged tooth or becoming unjammed. Set the switch to M and confirm that the focus ring moves smoothly by hand. Also confirm that there is normal free play in the focus ring with the switch set to A. This should be a few degrees of rotation or around a mm at the perimeter. It should not be tight with no free play. Then listen carefully for any sound while the camera is attempting to focus. If present, there may be a mechanical problem in the gear train. If there is none, then it's an SWM or electronics problem.
Diagnostics: In a quiet room, listen for mechanical noise when the lens should be focusing. If the SWM is seized, there would likely be none as the drive frequency is beyond the range of human hearing - typically between 25 kHz and 300 kHz, though it might be audible to your pet bat. ;-) If you can hear anything which sounds similar to the sound made during normal focusing but the focus ring isn't moving, a mechanical problem is likely.
See the info above on lens ID #4.
Diagnostics: Cycle the A/M switch and move the focus ring. Determine if there is focus distance or range of distances where autofocus works consistently. and if there is a preference for focusing in or out.
On one Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II zoom lens (my ID #4), AF was intermittent and just cycling the A/M switch sometimes helped. But rotating the focus ring with it set at M always seemed to allow the lens to focus correctly for several shots. There was no binding or detectable resistance so this suggests a damaged gear. When it stopped working, repeating this would get it to recover. Since the gear ratio is so high and the motor and immediate drive gears are disengaged with the A/M switch set to M, the only gear likely to be relevant is the ring gear on the focus ring. The pinion gear that engages it goes though more than one revolution and all the others go through multiple revolutions over the focus range. And when rotating the focus ring manually, it also rotates the pinion gear. Chipped or damaged teeth would be detectable. For more info, see the section: Autofocus only Works Over Limited Range on Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II Zoom Lens.
Diagnostics: If an oscilloscope is available - almost any type, even an antique - it's simple to check for the presence of drive signals, though determining if both are present might be challenging. All that's needed is a simple sense coil of a half dozen turns with a diameter of around 1.5 inches. The sensitivity with the typical AF-S lens is very roughly 10 mV p-p/turn. It doesn't have to be beautiful. ;-) Using a 10X probe but no other termination should be adequate. For all the lenses I've tested, a good place to position the coil is directly underneath the lens slightly near the camera. But feel free to experiment. ;-) With the scope's vertical sensitivity turned up, there should be a very distinct signal while the lens is attempting to focus, and for some cameras, possibly even after it had succeeded or given up. What is probably being detected up are the fringe magnetic fields from the ferrite step-up transformers on the SWM driver PCB. A lack of this signal would probably indicate an electronics problem. See the sections on the specific AF-S lens models for details on testing.
On one Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR zoom lens (my ID #1, the lens documented in the section: Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens), AF was totally inoperative and there was obvious damage stator and rotor of the SWM as can be seen in Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR Zoom Lens Autofocus Drive Assembly and Piezo Motor Parts.
Diagnostics: Similar to those for the SWM, above.
This has not been seen except where someone damaged a cable or lost parts of a lens. ;-)
-- end V1.07 --