Rfc: Tilt/accelerometer - Would You Want One?

[jdh2550]

Hi,

I've been looking at options for adding a tilt/accelerometer device for the Pandora. As with everything there's a number of ways to skin this cat (apologies to felines everywhere). This poll is my attempt to gather some data on how best to proceed. Hmm, it's the 4th of November - there's another poll I'm supposed to vote in today but I just can't seem to remember.

BTW, some pertinent facts about what I've discovered:

(a ) to build a professional version in a 50 piece batch would cost around $10 for the parts (for a USB OTG solution), $6 for the PCB, $16 for the assembly. These figures are based on online research only - could change based on getting an actual quote. It doesn't yet include the price of an enclosure - more googling required to find somewhere I can get an instant online quote for that (any suggestions?)

(b ) accelerometer chips that output voltages start at about $7 in single quantities, whereas I2C based chips start at $14. Note that the I2C chip alone is more than the parts for an entire USB OTG solution.

(c ) An I2C only solution would also be less portable. Given b & c, I'm thinking USB is the way to go - but, hey please vote your conscience

(d ) If we do USB we can simply expose ourselves to the system (stop snickering at the back!) as a joystick and this means no special driver required. Even a 3 axes accelerometer should be OK because there are various three axes joysticks right? (I'm thinking of things like the sidewinder that you can twist the stick for the 3rd axis). Note that I'm NOT saying the implementation would work very well as a 3 axes joystick - but if someone wanted to use all 3 axes for some other software task data from the third axes would be available.

(e ) Even though no driver would be needed for a joystick replacement to increase it's usefulness I'll figure out how to have it emulate a keyboard (i.e. tilt left to get key Z, tilt right to get X for a really simple game that just requires left and right).

(f ) The design will be based off of other open source efforts and will be fully open source itself (both software and hardware). The costs above are what I'd want to charge to cover the cost plus a small amount extra to cover the extra work and risk to make it a "product". I won't be getting rich!

I think that covers it. I've likely left out some important stuff or important questions. Please let me know what you think (as if you wouldn't - that's why I love the Pandora community! )

 

[OrR]

I voted I2C before reading your remarks... Looks like USB is more practical.
If it's done as a kit, I think it would be awesome if you could decide yourself whether you want it internal or external. I imagine that would be easier to realize with I2C, right?
Three axis is cool but how about a chip that also includes gyro in at least one direction? The DS Motion does that and it's pretty cool for some stuff. Are there chips that also include a compass? If so, that could be great if coupled with an external GPS (via Bluetooth or USB). I think the chip should provide as much movement information as possible for a reasonable price because most of the cost is going to be design/assembly/packaging/shipping anyway.

 

[parag0n]

I would like a device that plugged into the big usb port, and contained 2 usb (2.0/1.1) ports, and a usb-connected acelerometer.

 

[jdh2550]

I voted I2C before reading your remarks... Looks like USB is more practical.
If it's done as a kit, I think it would be awesome if you could decide yourself whether you want it internal or external. I imagine that would be easier to realize with I2C, right?
Three axis is cool but how about a chip that also includes gyro in at least one direction? The DS Motion does that and it's pretty cool for some stuff. Are there chips that also include a compass? If so, that could be great if coupled with an external GPS (via Bluetooth or USB). I think the chip should provide as much movement information as possible for a reasonable price because most of the cost is going to be design/assembly/packaging/shipping anyway.

Adding that stuff would likely push up the price by another $20 or $30. What does a gyro give you that an accelerometer doesn't (sorry if that's dumb question)

I2C for internal would be easier because it would take less space. But until we get our units we won't know how small it needs to be.

 

[Dunny]

Couldn't vote, as although I chose "not interested", I still have to choose how I'd like it implemented!

How are you going to get a good idea of how many people *don't* want this if they're gonna have to vote that they do?

 

[paolo]

Using a pandora as a Wiimote emulator? COOOOOL but we need also a hole for the wirst-strap then!

or.. just imagine a mercury_meltdown clone where you can block your metaballs from falling helping with your finger on the touchscreen

 

[jdh2550]

I would like a device that plugged into the big usb port, and contained 2 usb (2.0/1.1) ports, and a usb-connected acelerometer.

That's a good idea. For an internal USB solution I was wondering if I could make a 2port hub fit inside and use one port for the accelerometer and replace the other port external.

I much prefer your idea though. I need to read up on how one makes a USB hub. This might significantly change the design (because currently I was planning on software emulation of a USB client device rather than using a dedicated USB chip).

Couldn't vote, as although I chose "not interested", I still have to choose how I'd like it implemented!

How are you going to get a good idea of how many people *don't* want this if they're gonna have to vote that they do?

Yeah, oops! Sorry about that. But, as an approximation I can look at the total number of people who read the thread - and use that as the baseline (i.e. if 20 people voted "Yes" and 200 people read the thread then I can assume 200-20 = 180 people aren't interested. It's only a rough approximation - but it will have to do.

 

[polaris]

Personally I'm tired of playing games with my iPhone's accelerometer

 

[VRAndy]

After playing "Space Invaders" for the DS with the "paddle controller" (spinner knob), I would say that I'd love to have a "paddle controller" for the Pandora. More than I'd like a accelerometer.

 

[TaG]

I would like a device that plugged into the big usb port, and contained 2 usb (2.0/1.1) ports, and a usb-connected acelerometer.

That's a good idea. For an internal USB solution I was wondering if I could make a 2port hub fit inside and use one port for the accelerometer and replace the other port external.

I much prefer your idea though. I need to read up on how one makes a USB hub. This might significantly change the design (because currently I was planning on software emulation of a USB client device rather than using a dedicated USB chip).We can't really vote for an internal device until we know there is room, and from previous posts I doubt it. I like the idea of the hub/accelerometer, but I'm wondering what else I would dare to have connected whilst using rotation/movement as a controller. I'd be interested if someone fleshed out their thoughts on some of this.

I haven't voted as I don't really know if I want 3-axis, or how the bells and whistles option affects the price. Again, what would people use 3-axis for.

Couldn't vote, as although I chose "not interested", I still have to choose how I'd like it implemented!

How are you going to get a good idea of how many people *don't* want this if they're gonna have to vote that they do?
Yeah, oops! Sorry about that. But, as an approximation I can look at the total number of people who read the thread - and use that as the baseline (i.e. if 20 people voted "Yes" and 200 people read the thread then I can assume 200-20 = 180 people aren't interested. It's only a rough approximation - but it will have to do.
It doesn't work like that - one person could read the thread several times as they keep coming back to read the one or two new posts, views goes up each time.

 

[DAP]

An I2C solution could be internally mounted since the I2C is a bus that allows many devices to be attached to it (provided there is room).

A USB solution would have to be externally mounted, or would require sacrificing one of the external USB ports (there are no internal USB ports, and USB allows only one device per port).

The accelerometer would give you 3 axis (up/down, left/right, away/approach). adding a gyroscope would add 3 axis of rotation. You need both if you wish to do any kind of inertial guidance stuff.

I like the idea of building this with a 2 port hub (one port for the internal accelerometer/gyro, the other external for something like a GPS). Putting it in the main USB port also gives it a bit more mechanical support.

 

[jdh2550]

An I2C solution could be internally mounted since the I2C is a bus that allows many devices to be attached to it (provided there is room).

Unfortunately it's not as simple as that. Which pins will drive I2C? I believe MWeston said that UART3 could be configured that way (or was it UART2?). These pins are going to be exposed on the external connector. So, what happens if I configure them for I2C but someone else configures them for some other purpose? I imagine it can probably be solved by the software driver - but I'm not sure it can be done automagically (i.e. it would require user intervention).

EDIT: Or is there already an I2C bus on the board somewhere? Perhaps for controlling other peripherals. If so and I can get access to that then this would change everything!

QUOTE

A USB solution would have to be externally mounted, or would require sacrificing one of the external USB ports (there are no internal USB ports, and USB allows only one device per port).


Unless one were to use a 2 port USB hub chip - connect the new hardware to one port and the pre-existing connector to the new port. This would also solve the "hi-speed only except through a hub" issue. But it would be quite an undertaking to shoehorn all that stuff in there and to cut and reroute the signals to the existing port.

The reason I favor external over internal is that you don't have to futz with opening your pandora and voiding the warranty.

The reason I favor USB over I2C is that even though the chip count is higher for USB the cost of the device is lower (because USB is ubiquitous)

QUOTE

The accelerometer would give you 3 axis (up/down, left/right, away/approach). adding a gyroscope would add 3 axis of rotation. You need both if you wish to do any kind of inertial guidance stuff.


Ah, speed of rotation - of course! But I don't plan to be doing inertial guidance of missiles

AFAIK, an iPhone only has accelerometer - what about a Wii remote? Does that have a gyro and an accelerometer?

QUOTE

I like the idea of building this with a 2 port hub (one port for the internal accelerometer/gyro, the other external for something like a GPS). Putting it in the main USB port also gives it a bit more mechanical support.


Yep - I was looking at some usb port chips - it looks somewhat doable; but it pushes the complexity up higher and I'm a noob at hardware design so it might actually be stretching my meager capabilities way too far. But I'll keep looking. I haven't looked into gyro chips yet.

 

[jdh2550]

We can't really vote for an internal device until we know there is room, and from previous posts I doubt it. I like the idea of the hub/accelerometer, but I'm wondering what else I would dare to have connected whilst using rotation/movement as a controller. I'd be interested if someone fleshed out their thoughts on some of this.

Right, we certainly don't know if we can put it internal yet, but a minimal I2C solution would very likely fit - take a first class stamp and cut it in half, that's the biggest it would likely need to be. If you could have your way would you have it internal and limited to use on the pandora or would you want it external so you could use it with other devices?

As far as what might be plugged in at the same time and what you "dare" to plug in - I don't think this device will be used like a Wii remote. With the Wii the screen is fixed in front of you so you can wave your Wii remote around with abandon (I still love the site name wiihaveaproblem.com!). However, with the iPhone and what I'm envisioning with the Pandora is that because the screen is moved when you move the device that it will be more gradual tilts around and about. Does that make sense?

QUOTE

I haven't voted as I don't really know if I want 3-axis, or how the bells and whistles option affects the price. Again, what would people use 3-axis for.

3 axes doesn't add much cost or complexity - but if everyone had said "it's just a joystick (2 axes) then the design could be slightly simplified. I haven't looked at gyro's or compasses yet.

QUOTE

Couldn't vote, as although I chose "not interested", I still have to choose how I'd like it implemented!

How are you going to get a good idea of how many people *don't* want this if they're gonna have to vote that they do?
Yeah, oops! Sorry about that. But, as an approximation I can look at the total number of people who read the thread - and use that as the baseline (i.e. if 20 people voted "Yes" and 200 people read the thread then I can assume 200-20 = 180 people aren't interested. It's only a rough approximation - but it will have to do.
It doesn't work like that - one person could read the thread several times as they keep coming back to read the one or two new posts, views goes up each time.

Yeah, I know so I'll have to scale down the number of page hits to approximate the number of unique visitors. But it'll have to do because I screwed up the original poll. At the moment about 5 or 6 hours after initially posting I have over 30 positives - that tells me this is likely worth pursuing further. But we'll see.

 

[TaG]

Based on what you have said, I voted 3-axis external USB, subject to cost.

It did occur to me that if you could create a 'nice' consumer unit for a low enough price it would also work on pc etc. Price is a case of finding the sweetspot. Size / form factor is also likely to be important.

 

[phrosty]

Where's the "I'm too afraid I'll drop my panda when waving it around" option? hehe. I'll be using a Wii controller for motion stuff.

 

[TaG]

Where's the "I'm too afraid I'll drop my panda when waving it around" option? hehe. I'll be using a Wii controller for motion stuff.

I thought that originally - but not much fun for on the move with a wii-mote strapped to the back/bottom of your Pandora. But that probably determines a maximum price point.

 

[Drack]

I wanted to just vote no on the first poll and not vote in the others, but the forum only let me vote if I filled in an option for all polls.

Just a heads up, the results might not be as accurate because of this flaw.

 

[.Gogeta§§J4BR.]

just realized we could actually make the wiimote contact via bluetooth and the appropriate drivers
sweet

 

[absolofdoom]

This would be awesome, even if you could only use it as a simple joystick replacement. I mean, imagine playing wipeout on the psx emulator by tilting the pandora. The wiimote doesn't work too well with a handheld. I've tried it with my n810. You have to put the device in your lap or something, and it's just annoying. It'd be better to tilt the actual device. I hope you do decide to do this, I'd assuredly pay up to $75 for it.

 

[DAP]

A USB solution would have to be externally mounted, or would require sacrificing one of the external USB ports (there are no internal USB ports, and USB allows only one device per port).

Unless one were to use a 2 port USB hub chip - connect the new hardware to one port and the pre-existing connector to the new port. This would also solve the "hi-speed only except through a hub" issue. But it would be quite an undertaking to shoehorn all that stuff in there and to cut and reroute the signals to the existing port.

The reason I favor external over internal is that you don't have to futz with opening your pandora and voiding the warranty.

The reason I favor USB over I2C is that even though the chip count is higher for USB the cost of the device is lower (because USB is ubiquitous)

I don't think you will find room for this inside the Pandora, Pandora is just too small.
QUOTE

QUOTE

The accelerometer would give you 3 axis (up/down, left/right, away/approach). adding a gyroscope would add 3 axis of rotation. You need both if you wish to do any kind of inertial guidance stuff.


Ah, speed of rotation - of course! But I don't plan to be doing inertial guidance of missiles


Who said anything about missiles? I'd like it for tunnels, city canyons, and inside of buildings where GPS does not work well. It would allow continuous GPS tracks even if part of the path was not covered by GPS. As long as the resolution of the accelerometer and gyro were good data from them could be used to fill in the gaps of the GPS path.
QUOTE

QUOTE

I like the idea of building this with a 2 port hub (one port for the internal accelerometer/gyro, the other external for something like a GPS). Putting it in the main USB port also gives it a bit more mechanical support.


Yep - I was looking at some usb port chips - it looks somewhat doable; but it pushes the complexity up higher and I'm a noob at hardware design so it might actually be stretching my meager capabilities way too far. But I'll keep looking. I haven't looked into gyro chips yet.


Look at www.SMSC.com for Hubs. I have not used their parts, but an off the shelf 7 port USB hub using an IC from SMSC did not have the same problems I ran into with the philips 7 port hub IC that I have used.
NEC makes USB parts, but it is difficult to get data sheets from them.
For USB slave devices, look at cypress. They have a USB High Speed micro-controller for about $10. It uses an 8051 micro-controller architecture that has some annoying quirks, and is very slow, but it can also be set up so the data transfers bypass the 8051, and in that mode can operate at the full USB High Speed 480 mb/s throughput.