Author Topic: Why Hall effect sensors used so rarely in human input devices (HID)?  (Read 3388 times)

0 Members and 1 Guest are viewing this topic.

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Why Hall effect sensors used so rarely in human input devices (HID)?
The only two i know that (probably) user hall effect sensors:
  • Thrustmaster t16000m - a joystick, only uses hall effect sensor for the x and y movement of the joystick, it also supports rotation of the stick, but that's handled with a potenciométer, it needed a contact cleaner after only 2 months of use (i own one), i'm not sure isn't a magnetic sensor used for that
  • Haken continuum - a very expensive musical instrument I haven't had the chance to try one, they are super expensive, starting at $3390

If you'd say, it's because magnetic fields can disturb the device, don't forget that we are using magnetic data store for the most part of the century.

I'm planning to make a (commercial) device that would utilize hall effect sensors, but i'm wondering if there is some serious downside of using them, since so few devices use them.

The use case i'm considering:
Having input keys with a decent travel length, where can track their position with relatively low latency and at at least ~256hz update rate,

Alternative solution i could utilized but ruled out:
  • Mechanical potenciométer - not the cheapest option, and will degrade with using, fairly quickly in this use case.
  • Optical solutions - i'm not sure if i would be able to get it manufactured in a sealed fashion and with time dust and dirt would compromise it's operation, this applies to both encoder and the Adomax Flaretech (wooting) switches
  • Force sensors - They are pretty expensive as i see, and would require fairly complicated mechanical setup

So after this options, Hall effect sensing looks like a very nice option.
TI DRV5053 looks ideal for the job, at $0.3 at 1ku it's price is acceptable. The magnets required would cost less than $0.1 each.
For the physical assembly a cherry mx style switch could be used, since the have an extruding shaft, the magnet could be mounted on the end of it.
They are the most resistant to any external effect, unless you put a very strong magnet next to it, but even that could be detected in software and the user could be warned that it wont work correctly in the current situation.

Adomax Flaretech sound nice, but they are not really time tested, they are not very well protected. These "switches" are basically just optical modulators, you need to have an LED and a light sensor on the pcb, the switch itself is really just a "mirror" that changes the amount of reflected light based on how much it is pressed.

Am i missing someting? Please let me know!
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8240
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #1 on: February 01, 2019, 03:26:39 am »
The main reason is cost. Apparently you can find keyboards with hall-effect switches now, but they're still quite rare.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #2 on: February 01, 2019, 03:55:36 am »
If you'd say, it's because magnetic fields can disturb the device, don't forget that we are using magnetic data store for the most part of the century.

High density magnetic storage relies on incredibly high field strength generated over microscopic distances.  Typical handheld magnets will have no effect on modern magnetic storage but will completely disrupt an unshielded hall effect device.

Quote
I'm planning to make a (commercial) device that would utilize hall effect sensors, but i'm wondering if there is some serious downside of using them, since so few devices use them.

In the past hall effect sensors had poor reliability but I am not sure about integrated hall effect sensors.  They are not precision devices which matters in some applications.

Quote
Alternative solution i could utilized but ruled out:
  • Mechanical potenciométer - not the cheapest option, and will degrade with using, fairly quickly in this use case.
  • Optical solutions - i'm not sure if i would be able to get it manufactured in a sealed fashion and with time dust and dirt would compromise it's operation, this applies to both encoder and the Adomax Flaretech (wooting) switches
  • Force sensors - They are pretty expensive as i see, and would require fairly complicated mechanical setup

I would also consider reluctance (inductance with a moving core) sensors and capacitance sensors.  In the past and even now in high reliability applications, LDVTs are still used but they are more complicated.  I think LVDTs should be used instead of hall effect sensors in human safety applications like drive-by-wire vehicles.

 

Online Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #3 on: February 01, 2019, 04:05:15 am »
If you'd say, it's because magnetic fields can disturb the device, don't forget that we are using magnetic data store for the most part of the century.

High density magnetic storage relies on incredibly high field strength generated over microscopic distances.  Typical handheld magnets will have no effect on modern magnetic storage but will completely disrupt an unshielded hall effect device.

Ever pulled the magnet out of the seek arm actuator of a 3.5" HDD?  That is one serious bit of magnetic material, yet doesn't seem to be a problem for data integrity on the nearby disc.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 6694
  • Country: nl
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #4 on: February 01, 2019, 05:08:27 am »
You could do a rotational sensor (at the pivot point) based on capacitance between overlapping electrodes.

Main problem I see with Hall effect sensor is cross talk.
 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #5 on: February 01, 2019, 10:41:38 pm »
The main reason is cost. Apparently you can find keyboards with hall-effect switches now, but they're still quite rare.

I heard about that, although it's a bit different since that is only an "on/off" button, while i need an analog readout.

In the past hall effect sensors had poor reliability but I am not sure about integrated hall effect sensors.  They are not precision devices which matters in some applications.

I'm not really worried about the precision of it, i only really need 8 bits out of it.
Re-calibrating, even on the fly is not an issue and can be easily done, assuming that the amount of change, as the key travels does not change.


I would also consider reluctance (inductance with a moving core) sensors and capacitance sensors.  In the past and even now in high reliability applications, LDVTs are still used but they are more complicated.  I think LVDTs should be used instead of hall effect sensors in human safety applications like drive-by-wire vehicles.

Reluctance sensors only sense a changing magnetic flux, processing it would be pretty heavy, especially compared to a hall effect sensor.
LVDTs are nice but they are expensive, bulky and i don't have any human safety application.
For some reason i did not think of capacitive sensing, so i'm going to investigate that.
It's a bit funny because the keyboard i'm using every day for the last 3 years has capacitive buttons. Nevertheless topre capcitive switches construction are pretty involved and not easily done.

Thanks for the suggestions!

You could do a rotational sensor (at the pivot point) based on capacitance between overlapping electrodes.

Main problem I see with Hall effect sensor is cross talk.

I'm not really sure what kind of rotation sensor do you mean, but there are no pivot points, it's a linear up/down motion.

Crosstalk is the least thing i worry about. I only need to detect ~3.5mm of up/down motion, while they are at least 18mm apart, the change in the position nerby magnet between fully pressed and released is 0.34mm, that's a 2% change, i can easily live with that.


I only said magnetic storage, didn't say "modern HDD". I know they are obselate but floppy disks had pretty long lifespan. Also tapes can be erased with a magnet aswell.
 

Offline MT

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: aq
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #6 on: February 01, 2019, 11:54:30 pm »
As Hakken is an quite old design why not a Linnstrument which i personally think beats them all, unfortunately to a price, i recall its capacitive and/or resistive sensing. http://www.rogerlinndesign.com/linnstrument.html


Also there is Roli Seaboard. I wonder what kind of sensing they use since tangents/sensors seams to be gelly like!


Also the competition is fierce for past 3-4 years in this area one can by quite chap general purpose capacitve surfaces
in which the top is replaceable with whatever layout you want.Here just 2 examples.
https://www.play-joue.com/en/


https://sensel.com/
« Last Edit: February 02, 2019, 12:14:17 am by MT »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #7 on: February 02, 2019, 12:01:02 am »
Reluctance sensors only sense a changing magnetic flux, processing it would be pretty heavy, especially compared to a hall effect sensor.

...

For some reason i did not think of capacitive sensing, so i'm going to investigate that.
It's a bit funny because the keyboard i'm using every day for the last 3 years has capacitive buttons. Nevertheless topre capcitive switches construction are pretty involved and not easily done.

Both reluctance and capacitance can be used in a simple oscillator and frequency is one of the easiest things for digital circuits to measure.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8240
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #8 on: February 02, 2019, 01:32:32 am »
In the past hall effect sensors had poor reliability but I am not sure about integrated hall effect sensors.  They are not precision devices which matters in some applications.
Do you have any evidence of that? They've been used in electronic engine ignition systems for decades, switching dozens to hundreds of times per second, and likewise in brushless fan motors. It's rare for them to fail.
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #9 on: February 02, 2019, 01:55:14 am »
Why Hall effect sensors used so rarely in human input devices (HID)?
  :-// video game controllers use them
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #10 on: February 02, 2019, 01:58:26 am »
In the past hall effect sensors had poor reliability but I am not sure about integrated hall effect sensors.  They are not precision devices which matters in some applications.

Do you have any evidence of that?

Personal experience.

Quote
They've been used in electronic engine ignition systems for decades, switching dozens to hundreds of times per second, and likewise in brushless fan motors. It's rare for them to fail.

The engine ignition pickups I am familiar with are reluctance sensors.  I have had plenty of non-integrated hall effect sensors in brushless motors fail.
 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #11 on: February 02, 2019, 11:59:56 am »
Why Hall effect sensors used so rarely in human input devices (HID)?
  :-// video game controllers use them

Wow, i didn't even think of that really, and it seems that it's true for the last 20 year. I guess if it's good enough for Sony and Microsoft, it's good enough for me!
Although that's not a lot of different products, just huge volume.

As Hakken is an quite old design why not a Linnstrument which i personally think beats them all, unfortunately to a price, i recall its capacitive and/or resistive sensing. http://www.rogerlinndesign.com/linnstrument.html

Also there is Roli Seaboard. I wonder what kind of sensing they use since tangents/sensors seams to be gelly like!

Also the competition is fierce for past 3-4 years in this area one can by quite chap general purpose capacitve surfaces
in which the top is replaceable with whatever layout you want.Here just 2 examples.

I really don't like the idea of these input devices.
They have absolutely no feedback to the user, there's a reason why button grid controllers are more popular amoung musicians (among anyone who would need specialized keys) than touch controllers.
I think an instrument needs to have some physical feedback, espeically when there is concrete notes, just a range. Any string instrument have tons of physical feedback to the player.
This is really far from what i want to achieve anyway. I just want to have buttons that have an analog input range, that can be read at any time.


Both reluctance and capacitance can be used in a simple oscillator and frequency is one of the easiest things for digital circuits to measure.
For reluctance, i think that would require to have connected component on the moving part. I'd like to avoid that, since that's something that's very hard to do in a a relaibly way.
Capacitance sensing might work tough.
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #12 on: February 02, 2019, 12:07:29 pm »
IIRC, there were several (many?) keyboard designs that used Hall-effect in each key.
I don't think the initial premise of the question is correct.
 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #13 on: February 02, 2019, 04:23:05 pm »
IIRC, there were several (many?) keyboard designs that used Hall-effect in each key.
I don't think the initial premise of the question is correct.

Care to show a few examples? The german RAFI company did manufacture some a long time ago, at an extraordinary price, but they don't seem to do that anymore.
Recently there was a keyboard buy on massdrop, for a hall effect keyboard, where the overall rating for it is under 1.5 stars out of the 5.

Even if there are a few, they are incredibly rare, especially compared to how many rubber membrane, mechanical switches or even optical switches there are.
 

Offline MT

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: aq
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #14 on: February 02, 2019, 05:02:04 pm »
I really don't like the idea of these input devices.They have absolutely no feedback to the user,
Really!? yet you seams be for capacitive sensing who is the worst of them all, but i would like to see poly Theremin interface for sure.

I havent studied Hakken in detail but it seams the only feedback it gives is sponginess, Roli seaboard could be said to be the same as Hakken also giving back a sponginess/clayess kind of feedback but i cant see any particular muscular or motorik, vibrationall (mobilephone, personal pager, massage stave) feedback in either one.
Quote
there's a reason why button grid controllers are more popular amoung musicians (among anyone who would need specialized keys) than touch controllers.
I assume you meant non touch responsive buttons, then i disagree since most musicians go for multi directional touch sensitive interface silicone button , branch MI sales figures points to that. Point are, you can bild anything with any features there are but the first thing a musicians will complain about is the asking price, as have been shown time and again.
Quote
I think an instrument needs to have some physical feedback, espeically when there is concrete notes, just a range. Any string instrument have tons of physical feedback to the player.
Well peoples milage vary, piano (string hammer) players little of tactile feedback at all while church organists(standing wave in tube) have lots of vibration feedback despite players contact only interface.
Quote
This is really far from what i want to achieve anyway. I just want to have buttons that have an analog input range, that can be read at any time.
you want both feedback and non feedback?!, well then its classic resistive foam/esd bags/conductive rubber sheets cut into size.  :-//



« Last Edit: February 02, 2019, 05:33:40 pm by MT »
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #15 on: February 02, 2019, 05:49:43 pm »
IIRC, there were several (many?) keyboard designs that used Hall-effect in each key.
I don't think the initial premise of the question is correct.

Care to show a few examples? The german RAFI company did manufacture some a long time ago, at an extraordinary price, but they don't seem to do that anymore.
Recently there was a keyboard buy on massdrop, for a hall effect keyboard, where the overall rating for it is under 1.5 stars out of the 5.

Even if there are a few, they are incredibly rare, especially compared to how many rubber membrane, mechanical switches or even optical switches there are.

Quoting from your very citation:

Quote
The Hall sensor, for which this innovative keyboard is named, was popular in keyboards of the 1970s and ‘80s thanks to its reliability. Frequently used in nuclear reactors, missile silos, and aircraft cockpits, the sensor steadily became less popular due to high cost of production.

https://www.massdrop.com/buy/apt-hall-effect-mechanical-keyboard

You said "used so rare" and your citation says "popular in keyboards".  So you can understand my confusion.

Quote
Hall-effect keyboards are extremely reliable and can accept millions of keystrokes before failing. They are used for ultra-high reliability applications such as nuclear power plants, aircraft cockpits, and critical industrial environments. They can easily be made totally waterproof, and can resist large amounts of dust and contaminants. Because a magnet and sensor are required for each key, as well as custom control electronics, they are expensive to manufacture.

https://en.wikipedia.org/wiki/Keyboard_technology#Hall-effect_keyboard
 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #16 on: February 02, 2019, 07:01:06 pm »
Even if all the nuclear power plants, aircraft cockpits used hall effect buttons, that's pretty still extremely rare. But obliviously not every one of them used hall effect switches, and it appears that they don't do it anymore since you can't really buy an industrial hall effect keyboard or keys, most references to it trace back to the 80's.
Also, compared to the number of other input devices, that amount is marginal at best.
 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #17 on: February 02, 2019, 07:18:11 pm »
Really!? yet you seams be for capacitive sensing who is the worst of them all, but i would like to see poly Theremin interface for sure.
I mean capacitive sensing of a buttons linear position, not sensing touch, that's completely different use case.
Quote
I havent studied Hakken in detail but it seams the only feedback it gives is sponginess, Roli seaboard could be said to be the same as Hakken also giving back a sponginess/clayess kind of feedback but i cant see any particular muscular or motorik, vibrationall (mobilephone, personal pager, massage stave) feedback in either one.
Feedback doesn't have to be 'real', but you need to be able to feel it. Touching down harder on a capacitive surface does not offer feedback, you can't build muscle memory for that, while you can easily learn press an organ's key partially and since it has enough travel you can learn it to control easier.

Quote
I assume you meant non touch responsive buttons, then i disagree since most musicians go for multi directional touch sensitive interface silicone button , branch MI sales figures points to that. Point are, you can bild anything with any features there are but the first thing a musicians will complain about is the asking price, as have been shown time and again.
I have no idea what are you saying here
Quote
Well peoples milage vary, piano (string hammer) players little of tactile feedback at all while church organists(standing wave in tube) have lots of vibration feedback despite players contact only interface.
Quote
This is really far from what i want to achieve anyway. I just want to have buttons that have an analog input range, that can be read at any time.
you want both feedback and non feedback?!, well then its classic resistive foam/esd bags/conductive rubber sheets cut into size.  :-//
I'd argue that the piano key offers a lot of feedback to the players. Playing a real piano feels way more connected than playing on a cheap midi keyboard.
I want feedback, although that's might be the wrong word for it.
I want it to have a feel.
A smooth, flat surface has absolutely no feel, while, for example, a piano key does. The point is that the user should have an easier association of what's happening in the sound, for what physical motion.
The resistive foam type of sensor would probably die after a few hours of use.
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #18 on: February 02, 2019, 08:05:24 pm »
I used many ordinary QWERTY hall effect keyboards in ordinary office and computer room settings.  Never been in a nuclear plant, etc.  But then perhaps I am older than you and have first-hand experience of the era.

Certainly the reason they are not seen anymore is because new technology allowed simpler, better and cheaper alternative ways of making keyboard switches, so like many technologies Hall-effect was made essentially obsolete by progress.
« Last Edit: February 02, 2019, 08:06:57 pm by Richard Crowley »
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 6694
  • Country: nl
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #19 on: February 03, 2019, 09:37:00 am »
I'm not really sure what kind of rotation sensor do you mean, but there are no pivot points, it's a linear up/down motion.

This kind of sensor.

Why would you want do a pure linear motion? You need lots of rigidity and tight tolerances to make a large key move evenly up and down regardless of where you strike it. A hinged key such as a piano key is easier in that respect and allows you to measure rotation rather than distance.
 

Offline SilverSolder

  • Super Contributor
  • ***
  • Posts: 6126
  • Country: 00
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #20 on: February 03, 2019, 03:59:23 pm »

You can get some pretty cool hall effect products these days, for example Rotary Position Sensors:

https://ams.com/as5132

Pretty mind blowing alternative to an optical rotary encoder!

 

Offline hun_yetiTopic starter

  • Contributor
  • Posts: 25
  • Country: hu
  • Distortion is not always undesired.
    • My personal webiste
Re: Why Hall effect sensors used so rarely in human input devices (HID)?
« Reply #21 on: February 03, 2019, 05:42:35 pm »
This kind of sensor.

Why would you want do a pure linear motion? You need lots of rigidity and tight tolerances to make a large key move evenly up and down regardless of where you strike it. A hinged key such as a piano key is easier in that respect and allows you to measure rotation rather than distance.
I want to use cheap, and widely available cherry mx style switches, the shaft extrudes from the housing when the button is pushed down, so i'll just stick a tiny magnet on the end of the shaft.
I want it to be compact, at least for the first version.
I like the hinged design, and it it can have a better feel as well, but they required way more space, and the need to be stacked on each other if more rows are desired, while plain linear switches can have multiple rows while laying flat.

That kind of sensor sounds really nice to use instead of optical encoders.

You can get some pretty cool hall effect products these days, for example Rotary Position Sensors:
https://ams.com/as5132
Pretty mind blowing alternative to an optical rotary encoder!

cool, but at $4 1ku price they are defiantly not a way to lower your bom cost.
If i'd need to do something like that i'd use 3 cheaper sensors. If i wanted to have a digital output, it's just ad a cheap PIC mcu or something, it would still be less than third the price.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf