Rodent Mark II

[mkawa]

are you sure that _all_ the sensors output only movement data? in the g9x it looked like the laser receiver camera had a pretty wide bus to the MCU, enough that it could just be a raw image sensor that just scanned its photon wells and dumped the raw charge data onto the MCU. the MCU was certainly fast enough to handle all of that data.

Yes, in normal operation mode it outputs Δx/y.
You can capture a single frame and you could use that to get a constant stream of the frames (afaik somebody did that with some older optical sensor).
I don't know if the bus would be wide enough but bear in mind that the SPI is 2 Mhz and the sensor operates with up to 12.000 FPS (images are 30*30 in grey-scale).

it operates over SPI? FACEPALM

12000 fps means USE A WHOLE BUNCH OF GPIOS OR DEDICATE A CORE ON A SOC. so backwards i don't even know where to start.

… acceleration …

Hi Bullveyr

I was discussing the acceleration of the 9500/9800 with mkawa. For anyone not knowing what kind of acceleration this is, here is a short explanation by Cyro: http://www.teamliquid.net/forum/viewpost.php?post_id=18189604

Do you know of any way to find out whether the acceleration is a result of sensor architecture or SROM?

It looks like it's a general problem with the architecture and not a simple SROM bug, it's probably not fixable but Avago never really cared anyway.

replace with a philips sensor -- hope it can give you raw camera output

The twineye works with doppler effect, so there is no raw camera ouput.
Besides, it also just outputs Δx/y

It also has it's own problem in the form of z-axis tracking.

this is why we need open standards and firmware and to adopt one of the current arm prototyping volume socs (the beaglebone cortex has a SWEET simd unit). hell, it's not that hard to build an optical sensor. you buy up some tiny high speed camera sensors in bulk from sony, you buy some binned LEDs and you go to town with a piece of plastic to aim them. the hardest thing by far is the optics, actually. you can't print that, and you can't build a small injection molding tool for it either. a molded optic would need hand or machine finishing etc. etc. ideally you could reverse engineer the avago sensor protocol and tap into the image feed by force but that's very much easier said than done.
 

[blueslobster]

You can always buy 3rd party lenses, like from Kingsis. Although that can influence tracking.

[mkawa]

yep, mouser has a whole category for led optics. however, it's VERY hard to model these things and basically requires someone to devote a couple months to putting a package together even from premade parts.

[damorgue]

there's definitely a difference in feel between microswitches. fightstick guys are serious about their microswitches. i'm surprised that the cherry b1aas are better than the high end omrons though. i quite like the omrons

Don't drink the Omron koolaid... Cherry is where it's at, and everyone I know who have personally tried the DG23-B1AA love them. Cherry DG23-B1AA are a bit heavier than the Omrons that most people are used to, but after you get used to them all those Omrons just feel wrong. They're just so CLICKY! They're the buckling spring of microswitches.

Judging the qualities of switch manufacturers by comparing a 140/150g force switch to a 75g one seems a bit weird.

They are both around 140g. The Cherry ones with a flat lever are labeled around 45g and the ones with roller lever are at around 60g simply because those are measured on the lever. The actual switch is the same for all of them at 140g though. The lever will be popped off if I get ones with them anyway. They are all actually fairly similar in force.

[damorgue]

The new, although a bit weaker, simpler and smaller sockets for the nuts. Switchmounts etc are hidden.

[Bullveyr]

@mkawa

That just won't happen.
Besides, making a working sensor might be easy but making a competitive high end sensor isn't.

there's definitely a difference in feel between microswitches. fightstick guys are serious about their microswitches. i'm surprised that the cherry b1aas are better than the high end omrons though. i quite like the omrons

Don't drink the Omron koolaid... Cherry is where it's at, and everyone I know who have personally tried the DG23-B1AA love them. Cherry DG23-B1AA are a bit heavier than the Omrons that most people are used to, but after you get used to them all those Omrons just feel wrong. They're just so CLICKY! They're the buckling spring of microswitches.

Judging the qualities of switch manufacturers by comparing a 140/150g force switch to a 75g one seems a bit weird.

They are both around 140g. The Cherry ones with a flat lever are labeled around 45g and the ones with roller lever are at around 60g simply because those are measured on the lever. The actual switch is the same for all of them at 140g though. The lever will be popped off if I get ones with them anyway. They are all actually fairly similar in force.

Although the are also 150g variants of the Omron switches the ones used in mice are 75g.

[damorgue]

Although the are also 150g variants of the Omron switches the ones used in mice are 75g.

Are you absolutely certain that it didn't have a lever? Do you have a link to a spec sheet or something? I have found all the 40-75g ones to have levers and their non-lever counterparts have been in the 130-160 range.

[Bullveyr]

Omron D2F Datasheet

As you can see D2F-01F, probably most common switch people mod into their mouse, has 75g.

[damorgue]

Seems to be true then. I suppose I will still have to get even more and compare them. The ones I got hare D2FC-type and not D2F-
D2FC-F-7N appears to be the most common in mice and also the ones I got.

[The_Ed]

Judging the qualities of switch manufacturers by comparing a 140/150g force switch to a 75g one seems a bit weird.

It's interesting that you are trying to twist my opinion down to being purely based on the actuation force. Have you personally tried the Cherry DG23-B1AA microswitches and compared them to the standard mouse's Omrons in both feel and electrical reliability/longevity? Don't make accusations about people and products you don't even know.

Although the are also 150g variants of the Omron switches the ones used in mice are 75g.

I see you are also lumping all mice that use Omron switches together as using the same actuation force. That is obviously not true since the only reason to make different actuation forces from a manufacturer's prospective is gaining more sales.

Oh and by the way, you measure the force it takes to actuate a switch in cN (centi-Newtons), N (Newtons), or gf (gram-force) as those are measures of force applied.

[damorgue]

Lets all calm down a bit. If anyone has any of these micro switches, I would appreciate it if you would send them to me. That is probably the only way to compare them and find out how they differ. If you know of a good source who sells them in low numbers without an enormous shipping fee to Sweden, let me know. I am having trouble sourcing many of them.

I have:
TIAIHUA ... unknown model#
Omron D2FC-F-7N

Looking for:
Omron D2F-D
Omron D2F-F-D
Omron D2F...
Cherry DG13B1AA
Cherry DG...

[The_Ed]

Why would you look for a DG13-B1AA? DG23 series are what work with mice, and DG23-B1AA is what I recommended and gave a link to.

[damorgue]

Why would you look for a DG13-B1AA? DG23 series are what fit into mice, and DG23-B1AA is what I recommended and gave a link to.

I have listed that I am interested in other Cherry DG series switches as well, chill down. I have particular interest in the DG13B1AA model simply because I have found it. The only difference is that it is rater for higher currents. It fits mice as well.

Edit: To be able to compare the feel of them, I am particularly interested in Cherry DG*3-C*** as well.

Edit2: Are Omron D2FC discontinued? Their specs are harder to find.

[damorgue]

A this point, I don't care all that much about what type of pins they have either. They will work fine for testing to see whether I want them and if it is worth the effort getting the correct ones. I can probably make any of them work if nee be.

[The_Ed]

This is why you can only use DG23: [Contacts: Silver Alloy (DG13, DG43), Gold Flash, Gold (DG23)]

Gold flash is desirable for the small amounts of current that a mouse passes through it. If you use standard silver alloy you will likely run into problems of missed clicks or multiple clicks. Gold flash is reliable for lower (signal) amounts of current, while silver alloy is reliable for larger (power) amounts of current.

Also it should be pointed out that 140gf is not available on DG43, and 75gf is not available on DG13.

Why would you look for a DG13-B1AA? DG23 series are what fit into mice, and DG23-B1AA is what I recommended and gave a link to.

I have listed that I am interested in other Cherry DG series switches as well, chill down. I have particular interest in the DG13B1AA model simply because I have found it. The only difference is that it is rater for higher currents. It fits mice as well.

Edit: To be able to compare the feel of them, I am particularly interested in Cherry DG*3-C*** as well.

Edit2: Are Omron D2FC discontinued? Their specs are harder to find.

I mistakenly put "fit into" instead of "work with" in my last post, I have corrected it.

The C model Omrons are the Chinese ones, but people usually seem to prefer the Japanese ones though.

A this point, I don't care all that much about what type of pins they have either. They will work fine for testing to see whether I want them and if it is worth the effort getting the correct ones. I can probably make any of them work if nee be.

Yes, because you are direct wiring the switches in your custom housing any of the 3 directions of pins should be fine. But the straight ones are usually the cheapest and easiest to find in my experience.

[The_Ed]

By looking at that Omron spec sheet that was linked I found a similar piece of info regarding gold vs silver for current:

Item	D2F	D2F-01
Specification	Crossbar	Crossbar
Material	Silver alloy	Gold alloy
Gap (Standard value)	0.25 mm	0.25 mm
Minimum Applicable Load (See note)	100 mA at 5 VDC	1 mA at 5VDC

Note: Minimum applicable loads are indicated by N standard reference values. This value represents the failure rate at a 60% (λ60) reliability level (JIS C5003).

The equation λ60=0.5 x 10-6/operations indicates that a failure rate of 1/2,000,000 operations can be expected at a reliability level of 60%.

So you'd need a minimum of 100ma at 5v for a silver contact Omron microswitch's actuation to be reliable. Yikes!

EDIT: I found more in that Omron spec sheet

Using Microloads:
Using a model for ordinary loads to switch microloads may result in
faulty operation. Instead, use the models that are designed for microloads and that operate in the following range

{graph below}

[damorgue]

Good to know that the high current ones won't work in the final application. That doesn't matter much now though. Do you know if the 13 and 23 differ in feel? As I said, I have found the 13 and intend to evaluate how they feel if the alloy in the contact doesn't affect it much. I am quite far from using them in mice. You seem to have misunderstood what I meant about pins. Yes, what you say is true, any sort of pins will probably work for me since I will solder wires to them, but frankly, if I could find ones without contact pins whatsoever that would be fine for me as well since the purpose is to test and compare them. After that, I will see if it is worth searching for the exact variant I want of that particular switch. This way, I will have far easier finding candidates locally for now.

[The_Ed]

I would expect different contact materials to affect the feel slightly, even though the actuation force will be the same. But I have only tried DG23-B1AA, so I don't know for sure. If I were you I would just try to find ANY DG23-B*** variant to try out what they feel like.

But yeah, only Cherry DG23 or Omron D2F-01 series will be reliable for mouse operation.

[Bullveyr]

Judging the qualities of switch manufacturers by comparing a 140/150g force switch to a 75g one seems a bit weird.

It's interesting that you are trying to twist my opinion down to being purely based on the actuation force. Have you personally tried the Cherry DG23-B1AA microswitches and compared them to the standard mouse's Omrons in both feel and electrical reliability/longevity? Don't make accusations about people and products you don't even know.

Calm down.

I never tried a Cherry DG23 but I also never said anything bad about them or judged them in any way.
You said the common Omrons are bad compared to the Cherrys but I still think judging the general feeling or clickyness by comparing 2 switches with totally different actuations forces is a bit weird.
Out of curiosity, have you ever tested a 150gf Omron?

Although the are also 150g variants of the Omron switches the ones used in mice are 75g.

I see you are also lumping all mice that use Omron switches together as using the same actuation force. That is obviously not true since the only reason to make different actuation forces from a manufacturer's prospective is gaining more sales.

I don't know of any halfway common mouse that uses 150gf Omrons.
These switches aren't only used in mice, having different actuation forces (and actuator types) opens it up for more applications.

Oh and by the way, you measure the force it takes to actuate a switch in cN (centi-Newtons), N (Newtons), or gf (gram-force) as those are measures of force applied.

Do we really have to go down that road when it's rather common practice (even on this forum)?

[The_Ed]

My mice over the years have taken more or less force than each other, but I haven't opened them all up and written down the part numbers of the microswitches, so I can't be absolutely sure that I have or haven't tried 150gf Omrons.

You seem to think that the most important aspect of feeling is the actuation force for some reason. Why are you hung up on that? The Cherry DG23-B1AA microswitches are the loudest, toughest, most clicky, most tactile, and most crisp to actuate microswitches I have ever had the pleasure of using.

The first 4 DG23-B1AA I got were from the mouse buttons on 2 NIB POS boards.

Just because terms are commonly used wrong does not mean it's OK for those who know better to perpetuate those wrong uses of the terms. It irks me to no end when people use grams for springs and switches, it's not that hard to just put an "f" at the end to make it "gf"...

[Bullveyr]

My mice over the years have taken more or less force than each other, but I haven't opened them all up and written down the part numbers of the microswitches, so I can't be absolutely sure that I have or haven't tried 150gf Omrons.

The shell also has some impact on that.

As I said I don't know a single mouse with high force Omrons.
In the meantime I asked a friend who works in the industry, he also knows none and would find it weirdly speced for a (gaming) mouse.

You seem to think that the most important aspect of feeling is the actuation force for some reason. Why are you hung up on that? The Cherry DG23-B1AA microswitches are the loudest, toughest, most clicky, most tactile, and most crisp to actuate microswitches I have ever had the pleasure of using.

I don't think it's the most important aspect but don't you think that having an actuation force of 75 or 140/150gf has quite some impact on the overall feel of a switch?
I do indeed think because of that the actuation force is one of the first thing to look at when choosing them for a DIY mouse project, although there is nothing wrong with testing some switches with a much higher (or lower) actuation force.
If I would want to test out I Cherry switch I would rather go for a DG23-C1AA because it's in a similar range to what I'm used to.

[blueslobster]

If anyone has any of these micro switches, I would appreciate it if you would send them to me.

The same switch can feel quite different. Dependant on the production batch and who knows how many parameters. So don’t make a final verdict based on a sample size of two switches. ;P

I have heard that D2F-01F are selected in Japan — but I don’t know if that’s true and I don’t know what they are selected for. Cherry claims to produce in Japan, but they don’t give you detailed information if you ask them directly.

[damorgue]

I think I'll draw the line right there at batches of switches feeling all that different. It isn't that important to me.

[blueslobster]

I understand. I wouldn’t underestimate the degree to which a nice click can enhance the perceived quality of your trackball, though.

[damorgue]

The difference is quite notable. I am still waiting on the Cherries but so far I like Omrons the most when compared at equal forces.

[damorgue]

The TIAIHUA has more of a cheap plastic feel to them, very short, sort of a simple feeling of buckling plastic.
The low force D2F and D2FC felt very similar except that D2FC was crispier and a bit louder. They were both far better than the TIAIHUA


The TIAIHUA, the Omron D2F-series, the Cherry DG-series and pretty much all other micro switches have two holes 2mm in diameter. The Omron D2FC series does however have 1.4mm diameter holes just to mess with the system. I actually preferred the feel to the chinese D2FC but by the looks of it I will go with D2F-01F.

[damorgue]

I have to look into those fancy scroll wheels of Logitech where you can flip an electronic switch and turn free-rolling scroll on and off. If I were to guess I would say that since the switch is electronic and not mechanical they use a small coil and a little magnet which moves in the field created. When the field is off the magnet moves through it without resistance from the coil. It doesn't seem reasonable though that it should draw current through a coil, especially since many of the mice which have this feature are wireless. I think quite a lot of power would be required to create forces strong enough to cause this effect.

Since the switch to turn this feature on and off is electrical, if the actual mechanism is mechanical then it has to have a small motor or solenoid of some sort to engage and disengage a lever of sorts.

I need to get my hands on one or some detailed pictures of an opened one. Ivan linked an image which did mention a motor but there might be several designs for this.

[The_Ed]

You'd have to have the switch change the scroll wheel delta on the teensy from regular (120) to smooth scrolling (as small as 1).

EDIT: Though if you are making a regular scroll wheel into a smooth scrolling scroll wheel you wouldn't have to do that as it's resolution is still the same.

[damorgue]

Yay, just got a sensor. I am also awaiting a delivery of teensy3.0 and 2.0 but I can start working on the mount of the sensor now.

[damorgue]

I recently helped a friend with a project where they used parts from a cheap mouse. They didn't need the scroll so i took it for now. I have to mention how crappy that things was built. I never imagined cheap mice to actually be that badly designed inside, other than using cheap parts. To block out the light from the LEDs which light up the shell they put some black  electrical tape over the sensor chip are. The switches were soldered so bad, and a few of the pin holes where out by the edge and had been intersected by the edge for no reason, there was plenty of space. Probably the worst of all was the scroll assembly. It consists of two components, the scroll wheel and a tiny rotary encoder-like thing. This electrical component is soldered to the PCB and the scroll wheel is inserted into a hexagonal hole. The other side of the scroll wheel where the other end of the axis is just hangs free above a switch unsupported by anything. When you press the wheel, the entire assembly flexes and the free hanging end presses the switch. What actually bends is the tiny parts in the rotary encoder where it shears the axis a bit and causes quite some unnatural stress on it. This is built to fail fast.

I will use the rotary encoder-like component for now, but I will never make such a solution. Elrick seems to speak highly of some Mionix and Roccat scroll mechanisms. I wonder if Mionix would be interested in helping since I believe they are Swedish.

[damorgue]

Teensies have been a bit delayed, but at least I can move forward with the mount for the sensor. I have been worried about the internal construction of the sensor+pcb+lens since the lens is completely loose and can move around a fair bit. John tells me that it supposed to be like that and I suppose I have to trust him on it, but it does seem weird to me.

The distance from the tracking surface to the lens is greater than that between the lens and the sensor.
1/S+1/S=1/f
The above formula would indicate that the tolerances of the sensor to lens distance has to be greater than that between the lens and the tracking surface. Weird

[damorgue]

Some pictures of the switches I have tried: from left to right:
YSA, Huano, TIAIHUA and two different Omron


Some stuff that will be used, a few parts still missing and the scroll assembly will have to be changed to a whatever I decide is best once I have tested some more assemblies:

[damorgue]

Progress is slow, I vote GH get a second printer as has been discussed in the GH tooling thread

I made a quick bottom plate for now:



There will be a circular hole in the middle which will fit a little assembly to hold the lens and chip together as well as allow its position and rotation to be adjusted somewhat.

(I also countersunk the holes on my Ergodox and replaced those portruding allen socket heads while I was at it.)

[MOZ]

Is it possible, to make a ring and index finger rest for the Rat 3/5/7/9 for the right side as an attachment to replace the current replacements.

I really like the RAT 7 overall, but love teh rest on the CM Spawn, the kind of rest in your design

[damorgue]

Is it possible, to make a ring and index finger rest for the Rat 3/5/7/9 for the right side as an attachment to replace the current replacements.

Of course it is. It is just a matter of complexity. I don't own one but if you could take some pictures I might be able to help you. The only pic I could find of its right side is the one below which sort of already has a rest though.

[MOZ]

I'll get some pictures up soon.

[damorgue]

I'll get some pictures up soon.

 

[MOZ]

Been busy with some RL stuff and GBs. Will try to get them next week.

[wasabah]

Man, you're living my dream. Just some weeks ago I considered harvesting an old mouse and somehow creating my own shell. I'm not very good with either my hands nor do I have access to a 3D scanner.
So I basically gave up...

I'll be watching this closely and with great interest. Maybe I'll find a way to do my own after all, somehow.

Oh and btw, new replacements for the RAT3/5/7 is a cool idea too!

[damorgue]

Man, you're living my dream. Just some weeks ago I considered harvesting an old mouse and somehow creating my own shell. I'm not very good with either my hands nor do I have access to a 3D scanner.
So I basically gave up...

I'll be watching this closely and with great interest. Maybe I'll find a way to do my own after all, somehow.

Oh and btw, new replacements for the RAT3/5/7 is a cool idea too!

Not that hard really. You can make decent scans with a camera using camera tracking.

Another way to get the shape you want into the a digital system would be to take Xmm thick cardboard or foam board and stack on top of one another to form a block. Then you shape and sand it to the shape you want after which you separate the layers and label them. Once you scan them, you can build them up in software as Xmm extrusions. You don't need very high accuracy here as you can continue sculpting the digital model once you have a rought draft from the physical world.

Don't give up, just keep trying. I failed a few times as seen in my tests with vinyl but that is another way to go which remains viable although a bit more tedious to get right.

[wasabah]

I'll try camera tracking!

What's the advantage of scanning the several layers separately?

[damorgue]

If you use thin layers and you are able to separate them while keeping their relative position, the data will be very accurate. It is tedious though.

Using a camera and tracking software is easier and the workload is moved to the algorithm which is to analyze it. The resulting data will converge towards a shape with increasing amount of camera data, up until a point where additional images just makes it move back and forth in an interval where you have reached information saturation.

A third option for rather cheap 3D scanning is done through silhouette photography. This can be achieved by submerging the object in a contrasting liquid or drawing lines on it with a laser pointer.

A final option is to take pictures, measure the model and model it yourself from scratch.


Note that my original scan was made with your average cheap-o point and shoot camera.

[wasabah]

Thanks for the explanation!

May I ask how you plan to implement the mouse buttons in your shell?

[damorgue]

There are slots for the switches on the inside where they can be mounted. RMB and LMB will essentially be part of the shell and flex since there are U shaped cuts around them. The side buttons will be small parts which are screwed to the shell and then portrude through holes in the shell. I may exchange parts of them to metal foils in order to avoid fractures in the bending plastic.

The only thing remaining now is really the scroll wheel where I can't decide how much of it I should do myself, what design to go with and the best way to mount it.

Edit: Found an old 3D scan done with a point n' shoot: http://p3d.in/grQOo

Edit2: It is a good idea, or depending on method even necessary, to include a reference dimension in the scan. The scans will be roughly scaled the same in X, Y and Z but at the very least one scale so that you can ensure that the dimensions are correct. This was you don't need an absolute scan and a relative scan can suffice once you scale everything so that the reference object is correct.

[HotKillerZzz]

is it possible to use mx switches as the left click and right click?

[damorgue]

is it possible to use mx switches as the left click and right click?

Short answer: Yes

Long answer: Yes, it is possible, but you will probably not like it. A few have tried and I have yet to hear of anyone who liked it. The throw is too long and the hysteresis is also not suitable for fast precise clicking in my opinion. I don't think that excessive movement and decrease in precision yields any notable benefits other than being able to state a bit gamer-gear-gimmicky: "my mouse has mechanical MX switches"

Feel free to try though.

[damorgue]

5 points to the first one to guess what now works in my mouse.

Edit: Just noticed that there is still a bug in there somewhere as it switches from 00 to 11.

Edit2: Well, I'll just use a library for this anyway, this was just a test to see if it worked. I have confirmed that I will not need a logic level converter for the this particular scrolling mechanism and that it can be used. I will have to check some nicer scroll wheels as well. If anyone has a broken mouse with a nice scroll wheel feel free to PM me.

[damorgue]

Ok, this encoder I found and described above doesn't appear to comply to the usual specs. The output above actually is how it works, which means that no library will work with it and I wrote some of my own. I am still able to sense direction of rotation, but with less redundancy.

[damorgue]

Well, it works just fine, although with less redundancy. The advantage is that I have some code which will work with other encoders if they behave the same. I intend to try a few better more elaborate scrolling mechanisms. I have a certain interest in Logitech and their wheels where you can set them to free scroll or discrete steps. If anyone has any info on those it would be greatly appreciated. Even better if you have a broken one which I can take the mechanism from and save me some work creating it from scratch.

On a second note, I have discovered that most of razers mouse use the same really terrible scroll mechanism as the dirst cheap mouse I investigated. They actually just mount the wheel to the rotary encoder with one side of the axis hanging free above a switch. When you press the wheel, the entire axis bends a bit along with the rotary encoders internal structure as well as the solder joints. That does not give a good impression of their build quality and I bet the encoders aren't built for that type of load.

[damorgue]

That scroll was bad, and it should feel bad.

But honestly, that scroll mechanism was terribly designed, relying mainly on deflections in a critical part not built to deform with an unsupported axis hanging above a switch. Apparently most Razer mice use this mechanism. I am about to buy an MX Revolution which I might be able to get really cheap since it is mislabeled at a local auction site. It even has two scroll wheels and at least one of them is really fancy with the ability to free-spin.'

Edit: It even looks quite nice.