230 replies to this topic

[Foust]

The original one, when you plug it into the PC does it show up as your controller? Have you added any USB devices? Are you using the same USB port as you did when you programmed it originally?

Sometimes USB devices get pissy and you have to play games to get them to straighten up.

[Foust]

So I have just a few things to show for the pit. I said I decided to take seperate the foot pedals into two push buttons. I had two foot switches from a hard DDR pad I built many years ago. A bit of MDF and a hour or so and I ended up with these.




I believe I also owe a few shots of the pit in its current condition.

Looking at the in progress dashboard.


Front of the new seat.


Side/back of the seat


You might notice the old PC case holding up the dashboard. Originally that was just to get a feel for height. Now though I think I like that idea and will investigate actually placing the PC there. Seems like a good idea right now.

This week I am hoping to get that center console area worked out, finish roughing out the dash and get the monitors moved over to the pit. It is a ambitious goal to be certain. Feeling a bit motivated. With a bit of luck and garage time maybe I can make it happen.

[Ulric Kell]

Foust, on 17 March 2013 - 06:56 PM, said:

So I have just a few things to show for the pit. I said I decided to take seperate the foot pedals into two push buttons. I had two foot switches from a hard DDR pad I built many years ago. A bit of MDF and a hour or so and I ended up with these.




I believe I also owe a few shots of the pit in its current condition.

Looking at the in progress dashboard.


Front of the new seat.


Side/back of the seat


You might notice the old PC case holding up the dashboard. Originally that was just to get a feel for height. Now though I think I like that idea and will investigate actually placing the PC there. Seems like a good idea right now.

This week I am hoping to get that center console area worked out, finish roughing out the dash and get the monitors moved over to the pit. It is a ambitious goal to be certain. Feeling a bit motivated. With a bit of luck and garage time maybe I can make it happen.

Interesting Idea with the foot pedals. I would think however you'd get a better system with some leaf switches but if it works cool beans.

I'm not really sure what's the fascination behind building your own seat in simpits. I see people do it all the time and i just don't get it. The Vulture pod's seat was a thing of beauty but I know personally I wouldn't be able to game on a seat like that for more than 30 minutes without being in pain.

For my pod I chose a GMC Conversion VAN bucket seat, which is super comfortable and a pleasure to game on for hours. It would be fun to build sure but I'm not an ergo-metric engineer and I would spend days trying to tweak it just right to the point of where it would drive me insane.

[Foust]

Originally I had one of those Vrocker seats in the thing, but when I went to iteration 2 I just couldn't work it into the thoughts I had. Plus those seat parts are directly from the akers-barnes pit I have been cannibalizing. My wife will sew me up some nice pads for the seat and extra comfort will be achieved.

If not I have a fiero seat out in the garage that should bolt right up. :-)

[Ulric Kell]

The Vrocker seat imo wouldn't go with any pit so +1 for taking that out. I also went with a car seat because it already had a built in adjustment system. I know you're building the pit for yourself but you never know if you want someone else to go into it so don't forget to take some time on building a track system to move the seat or at least look into it.

Edited by Ulric Kell, 18 March 2013 - 05:04 AM.

[Foust]

Yeah the seat is on casters, will be tracked with variable locking points. Foot switches will be movable. Got a growing Mechwarrior in training to accommodate.

[Foust]

Sides to the dashboard and mount. Monitors mounted. Joystick testing and tweaking.



Locnar will be happy to know that I am considering a T16000m... The throw on that F-22 stick is really heavy, and I can not seem to get the response from the stick to be as quick as I would like.

My current line of thinking if I end up with a new stick is to reprogram the Teensy changing it to more non-matrix buttons, removing the unused axis's. I have two other sticks I can try without major rewiring jobs as well however I am not going to hold my breath. One is a FFB TflightX and the other is a microsoft precision pro. Maybe Ill be surprised.... Maybe Ill end up with a T16000m.

Edited by Foust, 25 March 2013 - 05:51 AM.

[Ulric Kell]

Looks good. I like the framework.

I see you're rolling triple monitor. I hope it works out for you. Personally I couldn't deal with the distorted side monitors and menu system. Was not enjoyable at all.

[Foust]

I have been thinking about how I'm going to translate the switched from the in game cockpit to functional switches. While the number and location/types of switches in the in game cockpit are nice ascetically, they are not really layed out in any functional manner. At least in the Catapult dash. Also they have no labels.

So I think I am going to have to take some creative license with the switch locations and lay out some more functional panels, while keeping some of the ascetics.

[Loc Nar]

There's a little pit type out there, called a SimLight, which I found quite inspirational.
(http://www.xflight.de/pe_sim_sml.htm)
I needed something that could be which could be used at my normal computer work desk/battlestation, without overly dominating my small apartment and seeing the SimLight gave me a ideas and I made a whole new cockpit, which has already gotten much use. I didn't need mine to fold up like a lawnchair, so it actually doesn't look like the one above, but this is what got me going and I think I see some family resemblance, at least when I look at your seat, although mine has a separate component for the headrest. Our current pits look similar... I'm stuck in cycles of 'one more thing' before I get proper pics of it and make a post, and keep not taking pics but will do it soon and make a new post about it.

On homemade cockpit seats/ejection seats... the ejection seat thing I made has got to be the most comfortable seat I've used in a long time. I'm not against using a good bucket seat from a car, but don't discount what can be achieved by paying a little attention to the actual compound curves of the backs and bottoms of real ejection seats. The contours are not arbitrary, and if followed and then covered with good foam, the results are excellent. My has never been happier in years, and I marathon play in my pit.

Quote

Locnar will be happy to know that I am considering a T16000m...

:) ...the degree of control you have over the axis makes it worth it alone. I have 2 different types of MWO scripts I am using now, but only one of them can be used with the t16000. No biggie, as this is actually the one I use with my own hardware as well for reasons below. The first one I was using (which only works on Cougar/Warthog) has x/y sensitivity tied to an otherwise unused analog pot on the throttle, but also relies on an override switch that needs to be tied to an actual toggle switch, which the T16000 does not have and I never figured out how to remove that feature without killing the rest so this is why there is no T16000 version of this script.

No matter. I wrote fresh script for my stick that uses fixed point sensitivity and axis type, since I don't switch between relative and absolute, and find myself using the same sensitivity settings. The biggest advantage of fixed point settings, is the ability to make the sensitivity automatically dynamically scale with zoom in addition to being able to easily manually snap between sensitivities, much like any gaming mouse worth it's weight, so in practice is more practical than the pot dialed version, and this one does work with the T16000.

During this process I also figured out the x and y axis actually need to be scaled at different value sensitivities from one another in order to be equally scaled to each other in-game. This sounds like a small detail, but it's a big deal for aiming with a stick. With my first script I was using, the x/y are equal to each other with no easy way to assign an offset between them (which of course can be done), and while I can dial the combined sensitivity of them to useful levels with the pot, the y axis is always too sensitive compared to the x axis and aiming is more challenging than it need be.

What does this mean for you? If you end up with a T16000, I already have scripts that do the above, in addition to a complete scripts to drive MWO with the T16000 by itself. Additional features added to all after last patch:

• center legs
  
• toggle arm lock on (automatically *increases x/y sensitivity)
  
• arm lock off (normalizes sensitivity)

*current MWO implementation of arm lock reduces mouse input sensitivity, which my TARGET script compensates for. If lower sensitivity is desired while arms are locked however, the normal or reduced sensitivities can be selected as well, offering 3 levels to choose from in this mode

[Foust]

Loc Nar, on 27 March 2013 - 12:12 PM, said:

During this process I also figured out the x and y axis actually need to be scaled at different value sensitivities from one another in order to be equally scaled to each other in-game. This sounds like a small detail, but it's a big deal for aiming with a stick. With my first script I was using, the x/y are equal to each other with no easy way to assign an offset between them (which of course can be done), and while I can dial the combined sensitivity of them to useful levels with the pot, the y axis is always too sensitive compared to the x axis and aiming is more challenging than it need be.

Yes, very much this. Last night I tried with the Afterburner FFB stick (basically a tflightx). Had a few issues with it, mainly that windows 8 and it don't get along real well and even after I got some ancient drivers to install, it doesn't center well so I always drift. The microsoft precision pro installs without issue and is rock stable. (it uses halls) My Issue with it and the potential T16000m is that in comparison to the F-22 stick and throttle, they don't really look the part. That and I don't have a matching throttle for either of them, so I would be mixing and matching between devices...guess Ill have to get over that.

Thankfully there is the testing grounds, I looked like a drunken monkey trying to aim with that... I did manage to get the x axis "feel" about right but as you stated above, the y axis is just way to fast in comparison.

I also investigated JoytoKey that I saw mentioned in some other thread. I only got to spend a few minutes with it, but is able to do a few things that I think will be useful. First you can set it to emulate mouse control, and you can scale the level of input per axis. The scaling bit I did not get to play with and is on the list of things to do.

Also on that list (that never seems to shrink) I need to raise the strakes. In part for usability and comfort and in part to closer match the in game cockpit. Need to bring them up somewhere around 8 inches or so. I also will be building some type of removable table to span the strakes so that I can run a mouse and keyboard for playing games other than MWO. I am leaning toward using the throttle from the Tflight with either the MS Precision Pro or a new acquired T16K. If I do that I can then repurpose the Teensy++ out of the joystick role and into a massive button control box. Or I go about modifying the gimbals on that F-22.....

[Propnut]

I would suggest upgrading the F-22 to hall effect sensors (not really that hard to do).

[Foust]

Propnut, on 28 March 2013 - 08:16 AM, said:

I would suggest upgrading the F-22 to hall effect sensors (not really that hard to do).

That has been done my friend. My issue with it is the heft in the centering springs, compounded by the volume of wire coming out at the base of the stick rubbing on the base plate. I'm planning on taking the base plate off and standing the base on something to see if that is what is causing the heavy throw or if it really is the centering springs.

I have read on some modifications that involve removing the torsion springs and replacing them with expansion springs run out to the four corners of the stick. It is supposed to remove that "catch" when moving from one axis to the other, and it is supposed to provide a consistent force through out stick travel. Looks like this.


I also am considering replacing the halls with force sensing resistors and removing the majority of the travel component from the equation. To do that though I have to spend 30 some odd bucks on sensors, and a T16000m is 40 something...

[Pht]

I wonder how hard it would be to get a collimated screen setup going in one of these DIY cockpits for Mechwarrior...

[Loc Nar]

Tricked me for a second, I was marveling how much the F22 stick looks like a Cougar with the 'evenstrain' mod, but it is a Cougar with the evenstrain mod and this was just a sample pic...

Unfortunately there's no easy fix for your gimbal problem as far as I can see, and unfortunately even a perfect working pitch/roll gimbal is still quite challenging to pilot with, even if using dialed settings, absolute inputs and no spring centering, although that's a major improvement. I know I've harped about the T16000, but I should also mention some of it's drawbacks. While it uses the same 16bit x/y sensor assy as the Warthog, this is not a normal x/y HALL deal but a combined one that has it's own stupid logic codes and can't even be snipped out and replaced by simple pots, nor used with anything that's not firmware compatible etc so you're stuck with that sensor, which is fine if you are using the stick as-is or scavenging it for another pitch/roll arrangement, but it can't be hacked into pitch/azimuth, which is extremely limiting IMO. That sensor package has only 1 hot and ground going to it, the other 3 wires are logic voodoo wires.

The other 2 pots are normal however, but also only 8bit resolution, with throttle coded at 0-255, and the twist pot is funny in that it's coded to 0-128-255. These pots are the ones that could be hijacked and re-purposed as x/y gimbal, and one of the axis on the 16bit pot used for throttle, or throttle/turning. Another potential issue is it it uses those pressure button thingies, but at least in the case of those each button has it's own individual wire going to it, so would be no problem to redirect those wires to any other switches so no issues there.

My honest recommendation? Keep your eyes on ebay for a roached out Cougar and fabricate a new gimbal for it. I live on less than a low budget, but buying a broken Cougar was one of my best investments I've ever made due to the joy it brings me every time I use it to play MWO after a little sweat equity. I have a lot of joysticks laying around too, but they're all toys, and besides, eventually will all be used for *something. From a hardware standpoint alone this things a keeper, but on top of that it can be controlled by TARGET, and that is what makes the difference between a setup that merely works and a setup that actually works well. After my last mod to the stick, I played for pretty much14hrs straight the other day >_> (waring: X-Wing piloting is highly addicting!)

Due to the poorly manufactured gimbal on stock Cougars and awful pots they used, it's common to find them at good prices. Saw one sell for $120 after shipping the other day, with the typical wobbly gimbal and spikey pots. New gimbal not actually hard to make (I guarantee it will take less time than all that soldering ), and I've already seen enough of your work to know you not only have the tools, but the know how. Up until my last mod I just posted about that required a lathe, all work was done with a drill press and other normal shop and hand tools, and there was no real urgency to do the final mod, I'm just kinda fanatical.

As long as you use a pitch/roll gimbal though, you are definitely experiencing MWO on 'hard mode', and even with perfect dialed settings will not entirely escape the drunken monkey syndrome you describe, unless you stick exclusively to heavies and assaults and play on a team like most other stick users that haven't switched back to a mouse after getting pummeled repeatedly. Most meds and all lights are no where near what I would call controllable with relative inputs; you either dumb it down enough to aim, which leaves makes your reactions far too slow for survival, or you set them high enough attempt to keep the pace of movement but now can't hit sh*t. I say this not only based on my own somewhat extensive probing on this subject, but with many matches comes much chance for me to spectate, and watching other stick users is generally not inspiring, especially once a round picks up pace when the lights come out to play...
<--oh noes!

While on on a roll, the force sensor idea -I see as moving the opposite direction from optimal, further forcing your hand to relative inputs which I've already harped enough on about being at the center of the problem. Those types of sticks are meant to fly twitchy jet fighers pulling high g's to avoid uncommanded inputs, and while I'm sure it's neat to use in a sim that calls for it, a tank/mech uses a very different type of scheme, one that is built around being able to use large absolute inputs, not tiny relative ones. For the same exact reasons this stick type would be less than stellar with MWO, it's a nightmare to try to pilot aircraft in a sim using a mouse with it's absolute x/y Cartesian inputs, which is what MWO is engineered around from the ground up.

Thought exercise. Picture a 12" square sheet of glass and a marble sitting on it. Using relative inputs is like trying to get a marble where you want it to be on a glass sheet by tilting it back/forth and side/side, and quickly leveling it again when it's where you want in x/y coordinates, with deadzones keep you from spamming inputs and accidentally moving the marble. Using absolute inputs is like using your hand and placing it right where you want it with it level the whole time, and there is also no deadzone to speak of.

Luckily a mouse is not required to be able to use relative inputs, nor is one required to have a device that moves the natural direction of travel in game. Every degree of separation from absolute inputs and natural range of motion adds unnecessary challenge. Sorry for another long rant/wall of text, you just seem to be at a crossroads and have not yet committed to a solid direction yet for this crucial aspect so I feel obligated to share. For me, the controls came first since they can be mounted in any box/seat/etc, and ultimately is the interface you use the most and what defines your experience. Great pit with awful controls still = awful in-game experience. Great controls in an awful or even with with no pit is still really fun to play, but naturally even funner when you build it into a nice pit/closet/clubhouse/whatever...

[Foust]

I read this friday night and stewed on it all weekend. In part because I had some friends in from out of town, part because we spent all day saturday celebrating international tabletop day and part because of the sheer volume of information a fog of cold medicine.

I have another experiment that I will look into this week. I have two of those MS precision pro sticks, one of them USB compatible the other not. I believe they share the same internal workings. To me that equals spare parts and a base to experiment with, while maintaining a working stick.

The MS stick sounds like it has a similar layout to the T16k in so far as it has one sensor suite so to speak that includes x/y and rotation with another for the throttle. It is all as you say voodoo for the sensor suite, but I'm thinking that I can leave that alone and perhaps tweak the physical parts of it.

What I mean by that is to create something similar to your friction gimbal. Lock out X movement, create some type of friction solution for Y and remove the centering for the stick rotation. I wont have the benefit of target, but I think I can hack a semblance of axis sensitivity adjustments out of a joytokey or a glovepie or the UJR.

Now I got all these ideas and things to play with and more pit to build and I will put most of it on hold to get some playtime in before the double XP expires since I really didn't get to take advantage of it this weekend.

It should be noted that when you actually have a dream involving joystick structure and handles of joysticks attached to a mouse, one should consider extra sleep and less cold medication.....

[Gevurah]

Pht, on 28 March 2013 - 07:24 PM, said:

I wonder how hard it would be to get a collimated screen setup going in one of these DIY cockpits for Mechwarrior...

Collimation can be done with a fresnel, but my difficulty has been finding a fresnel that fits a 16:9 screen.

[repete]

Ulric Kell, on 18 March 2013 - 04:45 AM, said:

I'm not really sure what's the fascination behind building your own seat in simpits...For my pod I chose a GMC Conversion VAN bucket seat, which is super comfortable and a pleasure to game on for hours.

Yup. $20 for a seat out of a Honda Civic. Best $20/5 minutes I've ever spent.

[Foust]

Message received, people like their car seats for cockpits. Understood. :-)

Did a few things to the pit last night. I built the risers for the strakes, which really brings the control surface up to where it should be. I then built a bridge of sorts to cross the strakes and give me a surface to set a keyboard and mouse at a comfortable height to do things other than play MWO on that hardware.

Then I decided to tear down one of the precision pro sticks to see what made it tick and to see if I could achieve the axis lock out and friction based rotation. In other words, shamelessly steal Locnar's design. What I found was that the stick does not use halls like I originally thought, instead it uses optical sensors. It is a pretty neat set up, but not really useful from a hack it up perspective. However, the board is labeled really well and has large contact pads for each of the buttons on the base. If I want to string them out to some other switches, it will be really really easy.

The stick uses a saturn ring to center the stick, which explains why I like the throw on the thing. The down side is I don't know how I would convert that to anything else the way it is laid out. I can get the torsion spring out of it easy enough for the rotation, but then I'm stumped on how I put some resistance into the thing. I thought maybe I could use a ring of dense foam to grip the shaft to hold rotation, but how long would that last before it needed replaced? As to the locking out the X axis, I could attach some thin strapping at the base of the stick to prevent travel in that direction. I did not really like any of those ideas.

Then I went after the F-22 stick. Thought, well maybe I could fit a friction setup on the factory gimbal and make that playable. So I stripped the gimbals out of the thing, and pulled out the springs. Started looking at the gimbal on the desk and decided that I hate it more than I hate a murder of 3L's. Well ok maybe not THAT much, but I really think there has to be a better gimbal, and I think the cougar community at large agrees with me considering all the replacement gimbal solutions out there.

Locnar, I keep thinking about your post above and the natural movement of the stick with relation to the in game motion. How much travel do you put on the stick for full traversal and at what speed is the cursor moving? Would you happen to have a video of you piloting with the stick up somewhere?

[Loc Nar]

Quote

How much travel do you put on the stick for full traversal and at what speed is the cursor moving? Would you happen to have a video of you piloting with the stick up somewhere?

Before this goes unanswered for too long while I gather the words for an expanded post more about the hardwarey bits, this is easy enough (for me) to answer in less than 500 words. I think. Currently there are no stops installed in my gimbal, as I wanted to see what range of motion felt best before committing, and yet another useful feature of TARGET is everytime the stick is plugged in you go through a calibration procedure first thing, that at least *digitally defines the full range of travel, so it has let me explore different ranges. The larger the range of motion, the easier it is to be precise, but only within the confines of ergonomics. One's wrist can only effectively bend so far or twist so much before you can no longer push buttons or effect precision inputs. Stick travel values:

Y-Axis - +/- 15°
X-Axis - +/- 22.5°

As to cursor travel, it's amazing how tiny it needs to be in browser mode for it to work in MWO. Now that I've found dialed settings, when I move my stick through it's range of travel in my browser (but while stick is active), the amount of movement is so slight it's hard to convey. I use a normal arrow pointer for reference. In the Y axis, it moves about 1 full cursor height above neutral and 1 below, for a total travel distance of about 300% the size of the cursor itself. In Y axis *X-axis it travels much farther, perhaps 4 or 5 times the width of the cursor, to either side for a total of around 900-1100% the width of the cursor. I'm not sure what it works out to in pixels, but if I taped off a box on my 23" 1080p screen, all cursor movement would take place in a rectangle about 9/16" tall and ** 1" 1/2"wide!

In absolute inputs, that translates to full range of motion in-game and this was my initial failure once I got to working hardware and was playing around with TARGET in the beginning. I didn't even know at the time that a mouse with normal browser sensitivity was way too much as well, so it was a double whammy. My first TARGET programming not only didn't reduce sensitivity, I increased it! After achieving the functioning hardware stage, step 1 for me was setting up my stick as a mouse emulator and use it, well... to emulate a mouse, like for normal browsing. I used it for a full day to gauge how it was going to feel compared to using a mouse to move the cursor while browsing, and it was surprisingly challenging. For that, I had the settings bumped up enough to allow me to travel the cursor the entire page in x,y, and then played my first MWO round with it. Ha. Ha. Ha.

I have an aversion to cameras, but I really should make some videos of it, cause it would make so much of this actually make sense to people that aren't already mechanically inclined/don't understand the terminology. I can explain it all day long and still get people no closer to understanding what it is, what it does, and what it means, however very few would still 'not get it' once seeing how stupid simple it actually works in practice.

*thanks CyBerkut

**corrected after measuring

Edited by Loc Nar, 03 April 2013 - 08:32 PM.