The intracies of modern camera lens repair (2024)
salvagedcircuitry.com216 points by transistor-man 16 hours ago
216 points by transistor-man 16 hours ago
The TPS62140's 30ns propagation delay is not enough to blow a fuse. The first rule of fuses (which many modern engineers do not understand!), is that fuses are not there to save your parts, and they simply will not do that. Fuses exist to prevent fires.
Even a fast fuse is very, very slow compared to semiconductors. I've seen transistors blow up to "protect" fuses. They're for stopping fires and preventing the slaughtering of batteries, nothing more, nothing less.
Only partially true.
In well-designed microelectronics, they will.
The standard circuit involves a fuse, a fast Zener clamp, and sometimes a small resistor (e.g. 1 ohm) and/or capacitor. The design parameter is that, with the current limit from the resistor, the Zener should not blow out before the fuse.
The resistor needs to be small enough to not lose a lot of voltage under normal operations, but to still protect the Zener during the short surge during which the fuse blows. For most microelectronics, that's not hard. A 0.5W USB device might have 100mA of current max, which across 1 ohm is 100mV, so negligible for most purposes.
With high-power devices, it gets more complex.
Of course, consumer devices (a) will never be fixed (b) don't sell on this (c) every penny counts, so there's no market pressure to do things right.
But that's how we used to do it, and how it's still done many places where things count. If I'm building a one-off or few-off, it definitely will have proper protection.
That's a good design for input overvoltage protection (assuming I've drawn the correct schematic in my head). But it doesn't help against most other fault conditions, particularly anything to do with the load/downstream going short.
Really, if you care beyond "blow the fuse if something shorts", you need active current limiting. For common cases like USB ports, you can buy chips that do it cheaply and efficiently. There are also some textbook circuits, though they each have their pros and cons. No matter what, if you've got a pass transistor or switching transistor that's about to go seriously overcurrent, you have to do something about it with active parts -- fuses aren't going to get it done.
But defense in depth is always a good strategy, and fuses play a key role there. The active circuitry saves the rest of the design; if it can't get the job done, the fuse makes sure nothing burns.
Yeah, fuses are more of a overload protection kind of stuff. For cases where the load is trying to kill itself and you with it you need current limitation circuits, either on a converter or a latch current limiter.
Orrr you can design your circuit to survive a short condition for a bit longer than a fuse takes to blow. For old cars this was common, let whatever is short take all the current it wants for a tenth of a second and forget about it.the battery may lose a bit of useful life but those things were gonna fail early either way.
Just want to mention how much I appreciate this discussion and the opportunity to learn from it. This is what I come to HN for (nowadays there are also really interesting YouTubers who do informative teardowns of power electronics and other devices, too e.g. Labo de Michel, Watch Wes Work, etc.)
It's a scale problem; fuses can be set up protect components or machines in a car or factory context; and often this is what is taught (or what people have first hand familiarity with). It just doesn't scale down to semiconductor level.
Even in a car, fuses are there for safety. A car battery will easily start a fire if there's a short anywhere in the wiring or accessory devices.
> It's a scale problem; fuses can be set up protect components or machines in a car or factory context;
Induction motors use overload relays to protect the motor, fuses are strictly for overcurrent protection to stop the wire from catching fire. A VFD will have a fused switch or circuit breaker on the line side for overcurrent protection and a set of overload relays after the contactor to protect the motor itself.
Switchgear uses ground fault protection relays to protect the equipment, not fuses (or humans for that matter, GFP relays are set around 30mA IIRC)
Some motors also have phase loss protection relays that will power down the motor if a utility phase is lost.
A car is the same, fuses are to protect the wiring and prevent fires. Fuses do not act quickly enough to protect equipment, even fast-acting ones.
Relays act fast enough and deterministically enough to protect equipment, fuses aren’t fast or accurate enough to do so.
For those who haven't followed the camera world for a while, at this point a lens for a mirrorless camera will have a USB-C port to receive firmware updates.
Tamron lenses for instance will allow a wired control or a wireless dongle to communicate with an app/computer and change the lens behavior, switch what the physical buttons and rings do. Potentially you can manage stepping through settings for stop motion like effects, time lapses or stacking.
We're far from the days a lens was just metal and glass※. There are obvious downsides, but in practice it's actually a huge stepup IMHO. Every photographer is different and does different things, being able to fully adjust your gear is a godsend, especially as we need speed and reactivity.
※ there are still plenty, and plenty more will be designed and produced anew, but I don't think it's the major trend.
I would contend that if you were to pick a trend line in the last ten, twelve years it is actually the return of high quality, cutting edge manual focus lenses for stills photography, which was a negligible market by 2014.
Mirrorless has restored the utility of manual focus lenses in such a dramatic way that many of the significant advances in optical quality at every price point are happening in manual focus, and there is a real return to the understanding of the value of lens "character".
Time will tell how much of this is a stepping-stone to Chinese manufacturers moving to AF lenses (which is definitely a part of it) but many of the best new lens designs are coming from the likes of Cosina-Voigtländer and various cinema-adjacent stills brands.
Rokinon nailed their timing in particular in the 2010’s
LK Samyang, yeah, doing some great stuff. They also make some (all, maybe?) of the native lenses for the Cambo Actus, which is another mirrorless era winner.
This is only true for photography.
For some reason, cinema lenses are still - for the most part - purely mechanical. For film and TV, most camera operators still focus manually - often via gears attached externally to the lens.
Coming from modern photography, manual focusing is inconvenient and difficult to learn. But there's something very old-school cool about cine lenses. They feel great.
Autofocus is very nice for photos, especially when it works.
Autofocus in moving pictures isn't so great. It might be nice when you're not filming, but while filming, a focus change should really be intentional; auto focus isn't that. Might depend on what you're filming though.
> a focus change should really be intentional; auto focus isn't that. Might depend on what you're filming though.
As a solo operator, autofocus is great. Maybe the right metric is the number of crew per camera. 2-3 crew per camera? Manual focus is fine. 2-3 cameras per crew member, like solo filming a podcast or a theatre show? I'd choose autofocus every time.
Well.. The automatic part comes from the camera directing the settings mostly. The lens would be motorized focus/aperture.
For motion picture cinematography, I've seen remote controlled focus anyway. I don't see why you could not have a good motor built I to the lens and remote control it. If the external motor focus is quick and precise enough, then the internal motors should be as well.
If you're just filming a/your talking head autofocus with face recognition is very reliably and helps when people unintentionally go out of focus ;)
I think in cine it’s a lot more important to have smooth focus and to be able to control the focus speed, hence the MF lenses
I also think cine lenses have the budget to continue making high quality mechanical interfaces. Consumer lenses must have AF and so are incentivized to reuse that functionality if it would reduce the BoM.
Hm, I'm not sure about that. I suspect autofocus motors are much more expensive to manufacture than mechanical lens throws. They're almost certainly more expensive to design. I don't know about other manufacturers, but recent sony lenses contain 4 autofocus motors, and they can snap autofocus in tens-to-hundreds of milliseconds depending on the distance. Where do they even put those motors in the lens housing?
Its probably a scale thing. Photography lenses make up for the design, engineering and manufacturing costs with scale. Everyone who takes photos needs lenses. But far fewer films are made, and cine lenses are often rented. So they really can't be manufacturing that many units in total. I suspect they don't manufacture cine lenses in high enough volume to justify the engineering costs of fitting complex microcontrollers and motors into the lens housing. And if the production can afford to hire a focus puller anyway, autofocus just isn't that valuable.
A large focus throw and more importantly, focus consistency, is paramount for cinema lenses. When you rack focus, the lens needs to show the same field of view every time, no questions asked. The early red pro cinema lenses had a design flaw where the focus rings would come loose over time and ruin shots. Someone clearly forgot the threadlocker.
I think people in cinema have (and want) more control over the take. For a photographer, autofocus is quality of life, for a cinematographer it can get in the way very fast in everything that's not following subjects.
When telling a story through film, changing what depth is in focus is a great way to move the viewer’s attention, without the need for cutting to a different angle or camera movement.
Sure. But most shots have a deep depth of field or just follow one person. Even when autofocus isn’t needed, you can just turn it off. It’s just kinda weird that they don’t include the feature at all.
Indeed! I've had the privilege and honor of collimating a Cooke s4i. It's a thing of beauty, joy forever.
Because typically we use (usually wireless) whips to control lens elements. The 1st AC ideally is on a dedicated monitor with the setup in hand to pull as needed while the cam op handles the physical camera and focuses on framing and any movement.
You also have the sonar/radar thing on the camera which can measure distance. Can that be hooked into focus so it's automatic?
If you’re going to go through all that trouble you may as well just use a lens or cam body with a quality, built in AF. Way cheaper and the tech has matured greatly. Plus even with lower to midgrade cameras it’s easy to find a capable system with a solid codec you can color match enough with the rest of the movie. Even my GH6 I travel with can shoot 5.7k pro res
> For those who haven't followed the camera world for a while, at this point a lens for a mirrorless camera will have a USB-C port to receive firmware updates.
Iam not sure if this is a general truth. I recently bought a canon rf 24-70 f/2.8 which is pretty SOTA and it does not have an USBC port.
Third party lenses that cannot be updated by the camera body will include a USB port.
Why would this be a good idea to break away from the norm of what has been done before? The mechanism of updating the lens through the camera exists. Why reinvent the wheel? It only increases the BOM for the lens to include the USB and the electronics involved.
Because third party lenses cannot leverage the camera body to update.
Sigma has a dock that allows updates to their lenses in this fashion however.
I suppose it depends on the system? I have updated Sigma, TAMRON, and XiaoYi lenses on my Panasonic and Olympus MFT bodies, as well as Panasonic and Olympus with each other: https://support.jp.omsystem.com/en/support/imsg/digicamera/d... (Sadly not an exhaustive list. I have firmware for several more lenses stashed away in my archive, but the upgrade mechanism is the same.)
I shoot the m43 system (have since the GH2, then E-M5, E-M1, and now a G9 II) and didn't realize this.
I only have one lens tight now (I tend to stay small on my system but this is a low point) but I'll keep it in mind.
Thanks for the info!
You can update sigma lenses through Sony camera bodies but it requires running a program on your desktop with the camera plugged in and it’s a bit of a pain. Especially on macOS where it requires enabling kernel extensions.
Would have been nice if Sony just let you drop a file on the sd card to load an update.
Only for their old "Art" lenses, the modern "Contemporary" lenses can leverage the camera body update mechanism and only need a file on the SD card containing the firmware update.
"Contemporary" lenses aren’t more modern that "Art". The monikers were introduced at the same time, along with "Sports". Rather, "Art" is Sigma’s high-end line, similar to Canon’s "L". "Contemporary" on the other hand is a somewhat euphemistic term for "consumer" or "affordable".
First party lenses will be handled by either camera firmware or by the camera - for your lens (I'm in the Canon RF system heavily) you can do the "download to SD from canon.com" and the firmware update on the camera will recognize it as a lens firmware, not a camera firmware. And sometimes camera firmware will also bundle lens firmwares that get updated when you attach the lens (usually these are reserved for issues that have a potential to damage the lens or camera - like excessive hunting wearing the motor, etc., not 'nice refinements').
Yep. Sony lenses are the same. You can update the firmware on a sony lens by attaching it to a sony camera.
> For those who haven't followed the camera world for a while, at this point a lens for a mirrorless camera will have a USB-C port to receive firmware updates.
Besides the slightly interesting stuff Tamron is doing, why on earth would I want firmware updates for a lens? Also, this seems like it would be much more readily accomplished by the camera itself… if you’re doing weird stop motion racking and whatnot, why would you rely on the camera and lens being separate? Seems like kind of a pain to me.
The lens is communicating with the camera body, and you might want to adjust for newer bodies supporting more things for instance.
I'm not aware of what exactly is changing, but I've already seen it happen with newer Sony bodies getting released, and an update going to Viltrox lenses to fully support thems.
On the camera and lens being separate...in an ideal world you could ask the camera to do absolutely everything. In practice that's a tough order for a single company.
The bright side is also that you can use a mildly older body while benefiting from a very flexible lens, or have different profiles for different lenses and not have the body care about which lens needs what.
I can't imagine Nikon be bothered to properly operate a software ecosystem TBH.
Because to control the autofocus motor and other features it makes sense to have a microcontroller in the lens. If you have a microcontroller in the lens you have software in the lens and if you have software in the lens you're going to need to update it.
You could argue that the camera should do firmware updates but the manufacturers for (semi) open mounts like the ones Tamron is making lenses for don't want to have to design a protocol to do updates for third party lenses through the body when the lens manufacturer can just slap a USB port on the lens and call it a day.
The port is also useful for customizing the lens functions. For third party lenses the camera can't be expected to manage those functions.
There have been plenty of motorized lenses in the past that relied on the micro-contoller inside the camera body for control. What does having the controller live on the lens permit that the pattern we've used for years doesn't afford?
You don’t _have_ to update lenses. The updates pretty much just fix edge case autofocus issues usually with specific cameras and settings. Before update mechanisms you just had to deal with it or buy a new lens.
> If you have a microcontroller in the lens you have software in the lens and if you have software in the lens you're going to need to update it.
No, no you shouldn’t. There’s no reason why a microcontroller should ever need its firmware updated. The only reason why you would need to update the firmware is to add features, which I guess is mildly interesting for the tamron, but like I said… you could handle all extra fancy focusing things in the camera body itself. Just give me a dumb lens that does exactly what the body tells it to do.
> Just give me a dumb lens that does exactly what the body tells it to do.
It's hard to convey, but for instance you can reverse the fo us ring rotation, which probably only resonates with people who had to deal with it daily.
Or you can adjust the ring travel.
Or make the focus ring work as an aperture ring instead. Or straight disable annoying buttons.
Up until now you'd have to go hunt the perfect lens where the maker was 100% on board with your preferences. Or adjust your whole shooting style to every gear you use.
Sure a camera body could handle every single settings of third party lenses. But given what we've seen these 2 last decades, I don't think there was any chance of that happening. Tamron is partly Sony owned and they still did it on their own.
Most third party lenses reverse-engineer the AF protocol. They're going to need updates.
Lenses are only tested on bodies which are available at the time of manufacture. They might focus hunt on newer bodies, because of tweaks to AF algorithms (for example, the speed at which instructions are transmitted). Sometimes in ways that can potentially cause damage long term.
> why on earth would I want firmware updates for a lens?
One reason is to update the lens to work with new camera or new camera firmware.
> Besides the slightly interesting stuff Tamron is doing, why on earth would I want firmware updates for a lens?
Improved algorithms for focus hunting, diminishing chromatic aberration (most of it is in the glass but some positioning can tweak it).
I get it, there's not a lot that will happen there, but some of that stuff will be useful on an investment that can easily be several thousand (I don't get into the wildlife telephotos, but two of my lenses were $3,300 or so - RF 85/1.2 and RF 28-70/2).
> at this point a lens for a mirrorless camera will have a USB-C port
Ideally, camera bodies should support firmware updates via the body in a non-discriminatory way, but until then I wish manufacturers support firmware updates via USB-C.
Looking at you Samyang Lens Station. I think users have been sufficiently upset, and they're adding USB-C to newer lenses.
> USB-C port to receive firmware updates.
So do you mean that even camera lenses now ship unfinished?
How did the intricacies become "intracies"? ;-) That's the question.
Placing disassembled screws on double sided tape is such a great idea, I am going to try that next time I disassemble my electronics (as opposed dropping everything in a container and spend time finding the original size screws during reassembly).
There’s a very useful putty for watchmakers called Rodico, it can serve this purpose well and can also be stuck to a screwdriver to hold onto the screw while placing it, instead of relying on magnetism (no good in watchmaking to magnetize your screws, I figure cameras are similar, you can increase resistance to movement if magnetism pulls things together…)
I’ve also seen a clever technique of drawing a diagram of whatever you’re working on on cardboard, and stabbing the screw into the cardboard to keep track of where everything lives
Thanks for the input about Rodico! It sounds pretty useful. I've used mounting putty / blue-tack when assembling PCBs, but not much more than that. I've seen the card board trick used many times to ease in painting engine valve cover bolts.
It also has an interesting capability to absorb smudges and oils, I use it to clean watch dials, just dragging the putty over the surface.
I don't use anything sticky, but if it's my first time opening basically anything, one side of my working space becomes a map of the screws.
Roughly in the layout of the way they came out. Top to bottom, in layers, as it comes apart.
A habit from working on laptops in a warranty center. I might have to try some tape.
Maybe the best idea is to make a video of the disassembly process, so you know exactly where what part went and in what order and orientation.
Thanks! After many lens teardowns with screws of varying lengths and magnetism, tape is truly king!
That was a fun site to browse. I really enjoy fixing / hacking on non-disposable equipment (lab, test, optics, etc.) and there are some well-done write-ups in there.
The author says PH screwdrivers may be used on JIS screws, but in my experience they strip every single time.
This is incredible work, though.
JIS screwdrivers are 100% necessary. It may seem PH fit at first, but it's _a tiny bit off_ enough to cause damage. The point of JIS is shorter and squared.
Thanks for the kind words! The ifixit screwdrivers I linked are JIS compatible.
The PH-0 and smaller screwdrivers from Good suppliers like iFixit or Wera are actually JIS because there's no downside.
But if you happen upon a real PH-0 it will destroy the JIS screws.
One of the better teardowns I've seen in some time.
Glad I only fix old manual lenses. The Nikon AI/AI-S era ones are less scary :)
The Nikon AI/AI-S era lenses are fun to repair, but this sigma was not too bad. I've repaired some very intricate panasonic / olympus consumer lenses that were far more complex.
I did some work recently while I was between jobs, designing and assembling camera lenses. It was so fascinating, and satisfying. Strongly disappointed they couldn't afford to pay me to stay.
"Someone" should correct the typo in the submission.
If you mean dang or tomhow the fastest way to have that actually done would be to email hn@ycombinator.com (which _anybody_ could do).
If you mean the submitter, they only get a short interval of time to make such changes ... shorter than the six hours it has been up already.
This is the kind of thing I'd love to see robots take over. Companies don't want to service these things much; the labor is too costly. They could just distribute the repair manuals and we could get robots to do it.
How it currently works is if your stuff breaks, you don’t pay someone to fix it, you just sell it as is on Facebook marketplace and people who love fixing stuff as a side hustle buy it off you, fix, and then resell as a working item.
That way someone can specialise in just fixing Xboxes for example and make a nice profit.
Getting robots to do this would be incredibly difficult because it’s too custom and too fiddly. They don’t even use robots for most of the work originally manufacturing these lenses.
If companies could just distribute the repair manuals I'd be happy as a clam :D I'm all for automated repair!