Explaining Hall-effect, TMR, and other new types of "mechanical" switches

Some keyboards with contactless switches go further by letting users program keys so that one full press results in two inputs. For example, you can set a key to input “A” if you press it down 0.5 mm and then “B” if you press it down another 0.5 mm. Some keyboards with contactless switches also support analog input, which lets the keyboard detect how far a switch has been pressed and adjust the input accordingly, giving gamers a joystick-like experience. Traditional mechanical keyboards don’t support analog input. Contactless switches are becoming more common largely because of their potential appeal to gamers. Vendors often claim keyboards with contactless switches register input faster than traditional mechanical keyboards, mostly because the switches don’t require debouncing. Debounce delay is the time between when a key on a mechanical keyboard is pressed and when the keyboard accepts the keypress as an input. The delay is necessary for standard switches because their metal leaves can rapidly bounce off each other when they make contact during a keypress. Debouncing ensures that the keyboard’s microcontroller doesn’t interpret these bounces as additional inputs (which would cause pressing “a” to register as something like “aaaa”). Keyboard firmware often handles debouncing, but hardware, such as an FPGA, can also do it. Most of us never notice debounce delays, and manufacturers rarely disclose how much debounce delay keyboards use, though keyboard enthusiasts often point to 5 ms to 20 ms as common. Cherry says some of its MX2A switches require “less than 1 ms” for debouncing. And keyboard manufacturers occasionally implement longer-than-necessary debounce times to compensate for switch degradation over time. But keyboard latency, or the amount of time it takes from when you start pressing a key to when you see that key registered on-screen, depends on more than just sensing technology. Other factors include the keys’ travel and actuation times and the keyboard’s polling rate. A Bluetooth keyboard with full-height optical switches and a 133 Hz wireless polling rate, for instance, will still display more latency than a traditional mechanical keyboard with low-profile switches and an 8,000 Hz polling rate through a wired connection. Again, for many of us, none of this matters. It’s primarily professional gamers who need a keyboard with ultra-low latency, and that’s assuming they’ve already reduced latency in more essential areas, like their GPU, CPU, and monitor. And because contactless switches frequently target gamers, there are way more linear options than tactile or clicky ones. Optical switches, which rely on an infrared (IR) light beam to actuate, first hit consumer keyboards in 2016. Optical switches work differently depending on the manufacturer, but generally, each switch contains a light beam that is redirected when the switch is depressed. This causes the beam to either hit or stop hitting a dedicated photoelectric sensor, registering an input. When the spring resets the switch, the light beam returns to its original position. Hot-swappable, prebuilt optical keyboards are rare. And you can’t install optical switches in a hot-swappable keyboard made for traditional mechanical switches. I asked Razer’s spokesperson if the design of optical switches precludes Razer’s optical keyboards from hot-swappability. Since optical switches aren’t subject to the same amount of degradation as traditional mechanical switches, the representative argued, users won’t need to swap them over time. But that argument is about longevity, not customizability. “While the switches themselves are not soldered on, the keyboards are engineered as a fixed assembly to maintain the precise alignment required for consistent and accurate performance,” Razer’s rep told me. The Type-S Topre silent electrostatic capacitive switches I’ve used have had remarkably smooth travel as they depress and reset, making every part of the keypress feel predictable and intentional. Additionally, Hall-effect switches have various magnetic flux ranges, usually measured in Gauss, that indicate the minimum and maximum magnetic field strength the switch supports. This matters for some users, particularly those who want their keyboard to detect very light presses. And a switch with a broader magnetic flux can support a larger analog range. So one company’s Hall-effect keyboard may have north-facing magnets with a magnetic flux range of 120 Gs to 750 Gs, while another has south-facing magnets with a magnetic flux range of 102-905 Gs. Swapping the two keyboards’ switches could cause compatibility issues, as one keyboard’s sensors or firmware may not support the new switches’ magnetic flux range. Magnetic flux and polarity specs are hard to find because there’s no standard for measuring magnetic flux, and many users don’t seek or need this information. Monsgeek, one of the companies that provides magnetic flux specs for switches, told me that “most casual users don’t need to worry about magnetic flux numbers, but for enthusiasts who like to fine-tune actuation distance, rapid trigger, or overall switch feel, these specifications can be helpful and make the keyboard easier to customize.” Keychron doesn’t provide magnetic flux specs for its magnetic switches because magnetic flux “is an internal engineering parameter of magnetic switches, not a user-facing performance metric,” Paul Tan, Keychron’s COO, told me. Keychron’s keyboard firmware and launcher calibrate the magnetic sensor to each switch so “users experience consistent performance regardless of minor variations in magnetic field strength,” Tan said. TMR keyboards became available in 2024 and are marketed as using more advanced sensing technology than mechanical, optical, and Hall-effect keyboards. Like Hall-effect switches, tunnel magnetoresistance (TMR) switches use magnets, but TMR keyboard PCBs use a different type of sensor. In fact, some Hall-effect keyboards have a higher peak sensitivity than some TMR keyboards. Other factors, including the sensor and its placement and the keyboard’s polling rate and firmware, can affect a keyboard’s peak sensitivity. On rare occasions, electromagnetic noise or external magnetic fields can make magnetic-switch keyboards act erratically around other electronics, as seen in the video below.