Thadiun Okona

Super Lube was my initial goto grease but eventually realized it does a poor job over time. It evaporates critical ingredients and becomes sticky/ratchety like most greases do. Molykote is ok as a lube it will last a long time and is smooth but Nyogel 767a kicks it's ass because it's all that but also a heavy damping grease... the effect makes the Warthog -extremely smooth- and much less prone to overshooting inputs. Also lasts a really long time... checked my hog the other day and the Nyogel I put in it 5 years ago(!) is still completely viable with no signs of evaporation or degradation.

When banking an aircraft the forces push you straight down in the seat unless you are flying uncoordinated. Too little rudder (slipping) you feel like your slightly falling inside the turn, too much (skidding) and you feel like your falling outside the turn. Subtle roll cues can recreate this with astonishing fidelity and SFX 100 has 100mm travel to work with. Sure it can only replicate forces to a certain level, but it doesn't take a lot to accurately mimic most phases of real flight. Accurate suggestions of acceleration are the name of the game. Pitch roll and heave like are enough to accomplish this, even with 100mm travel actuators but those 360 carnival platforms are for kids and arcades as they don't' even come close to replicating what flying feels like.

Uhhg... just had a fully written reply wiped by a finger flub... second attempt Nice truncation of your pitch pulley, I've also embraced a partial pulley design for my final drive to further reduce the footprint/form factor, though mine is applied to a double reduction where the first stage is full round but 4" dia. Whether cable or belt, the same radial relationships generate the same mechanical advantages so the ultimate size is still influenced by this. With cable you can only drive it with so-small of a shaft due to cable properties and with belts the lower limit is by tooth engagement (smaller drive pulley increases tactile pollution and belt wear). My planned belt gimbals will only be a 1/2" bigger all around than my current ms hack (6x6x5 enclosure) even though the motors are twice the size. After playing 'musical cockpits' for the last 10yrs I've grown an appreciation for things that are compact and self contained :p As to pulleys/belts not allowing enough torque I think it's just a matter of selection. There are combos rated for more torque than it takes to rip our arms off, at least more than what's safe for a little baseball bat mounted by our privates. GT belts are rated at higher values of HTD due to the efficient tooth engagement by the modified design and are also quieter (lower tactile report). Pulley's limits are more determined by fastening type ie: set screw or screws (ok) or split-collar (better) or collet (best) but I'm starting simple with set scews since they're cheapest. If I have issues I can divot or cross-drill the shafts and permanently mount them, if that fails I can move up the food chain using standard options I can buy and make. Here's a neat chart comparing belts by type/size I can't find a similar chart for the polychain belt version, but they are the type rated for high torque at low rpm that use the same pitches/teeth so there are better low speed options if powergrip is insufficient, but my belt driven ff simracing wheel uses powergrip and works like a champ, which is what sold me on modern belt drive capabilities. Here's a Gates pdf on polychain belts, which are similar but tailored for applications under 500rpm https://www.ahrinternational.com/PDF_catalogues/Gates/polychaindesignmanual17595.pdf (p25 chart shows 8mm polychain rated ~.8hp and 14mm rated ~6hp at 10rpm) I fly irl and used to obsess over achieving 100% force value, which in most circumstances is pretty low but corner cases can be quite high for sailplanes but applies to most general aviation. Warbirds and big boys of course generate much higher forces but I scaled back my ambition after observing how readily the human brain accepts lighter forces. So long as they accurately resemble what my brain expects based on flight condition it's quite convincing. Same applies to motion of motion platforms... timely suggestions of accelerations that are short of full value, so long as accurate, are enough for your brain to accept. I worked for E2M Technologies for a few years, and got to build/play with high end belt driven electric ball-screw motion platforms rated at 3000kg (they make much bigger ones but switch to direct drive much beyond 3k and those weren't built at our facility). Their control loading hardware was all direct drive (didn't get to play with any but motors were yuuge), however my boss there worked on the Air Force One Blackhawk simulator some years back and that utilized (wait for it...) shaft winding for some of the ff power transmission :) Got to spend a lot of time picking his brain and talking shop which was helpful for adjusting my priorities. Kinda rambling here I guess so to summarize I don't see torque limitation of belts as an issue, even for powerful sticks and my belt driven 6Nm Fanatec simracing wheel regularly tries to break my wrist without any tactile artifacts. You do have to select the belt type/pitch/width suited for the application though. In my case it's 3mm GT2 9mm (may use 8mm polychain) but a warbird setup might benefit from say 5mm GT2 15mm (or 14mm polychain) for example. Shaft winding is good too of course, and could be made compact with similar tricks if desired.

Thanks for the kind words Miles. Ironically using the shaft winding technique was my original starting place on this project and I designed a stick around it in 2008, though never got around to the actual build. With improvements in HTD (rounded tooth) type belt drives, notably GT2/GT3 (rounded tooth modified for better engagement under load) the most compelling reason imo to use shaft winding, the lack of tactile pollution, was no longer an issue. To that end, I switched to belt drive due to off-shelf components making one more part of the build accessible, especially now that printing is viable for the larger pulleys. True it's not practical to achieve huge ratios without a double without or triple pulley setup, there's also a practical limitation on how small a shaft you can wind around (and hence size of final drive) before doing damage to kevlar with constant cycling. The other thing is you can't go crazy with drive ratios due to inertial influences. The MSFFII engineers chose 24:1, which seems a great tradeoff of power vs unwanted inertia from spinning armatures, though I haven't done any real testing yet to see how much is too much. At 24:1 the motors go through 2 full revolutions when moving the stick from stop to stop with the stock mechanism. I do prefer kevlar family ropes to aircraft cable though, it's much less elastic... more than an order of magnitude! Steel aircraft cable stretches like 20% @ 50% the break strength whereas kevlar stretches 1-2% at the same mark. It's harder to work with for terminating cleanly and tensioning installations, but there are enough advantages to be worth employing solutions. For me it was in the zone of diminishing returns in the face of GT3 belt though since they behave so nicely and are easy to work with and I don't desire super high ratios. Being able to print the large pulleys makes it economically sound. Here's a drawing of my shaft winding design from 2008 long before I knew now to use cad lol. That iteration had the pitch motor going for a ride, but mounted directly on the roll axis (with a miter gear) to mitigate its swinging mass. My current form factor has the roll axis moving with the gimbals and is near the axis but also augments the counterweight and after thinking about it, pitch is not operated vigorously enough for it to matter so the previous iteration was a solution looking for a problem. My new design is far more compact. I bought 1/4 hp Yaskawa 3ph ac servos/drives for this and had a full diy scheme for ff but now that msffii is hackable it's a silly wheel to reinvent but that was before Roland pioneered the mods. May still employ those solutions for ff pedals and possibly airbrakes down the road, but that's a whole nother can of worms :p

Thanks, the counterweight does a nice job keeping pitch balanced with the offset grip. I kept it all as light as possible so reduce the mass needed to keep it upright and can adjust the weight as needed. Not sure if you saw the link to the album I started but it has more pics and descriptions under each one https://imgur.com/a/4PHsM3V I just printed in PLA, however none of the critical load paths rely in its strength other than the 45deg elbows, so they were printed at 100% infill and have really thick walls. The base of the shaft has a brass tube to reinforce the fulcrum and there are printed bosses that engage the stock mechanism that have m3 screws running down the center of them. I might make an updated version in abs or petg but probably not since this is just my stepping stone stick before I build the 'real' one using motors the size of coke cans and a belt drive, though will retain the form factor and will basically be the same stick but on steroids

Here's a shot I just got after some bench testing a few minutes ago. I still have some wiring/soldering to do. It will have an enclosure and properly mounted boards soon but the important parts are all present and work nicely together. I'm making a breakout board to connect the grip to the shaft using standard 2.54mm plugs/wires/boards. I'm retaining the led/photodiode circuit, but it will be mounted in the sailplane cockpit it was made for so the stick will be active if I'm sitting in it vs mounted in the grip. Here's a link to an album that has more pics and descriptions: https://imgur.com/a/4PHsM3V I don't want to drift this thread too much by dumping all the images here and again will be making a proper thread for it here soon. Heh, thanks, but mostly I have the advantage of riding the shoulders of giants like Roland van Roy, who has kindly helped me with other related projects in the past. At one point I came up with a new scheme for full homebrew diy ff which was tested to good effect by Roland, but once the MSFFII hack became viable it makes a lot less sense to reinvent so many wheels when a good effects library of DirectInput can be extracted and utilized to good effect by such inexpensive hardware

The stock psu might not even put out what the motors can drink with the stock value at max draw which is 1.4A @24vdc. Doubling the current draw makes them capable of accepting up to 2.8A. Not sure the rated output of the stock unit is but the input is 1A @ 120vac. My 3A psu draws 2.8A @ 120vac to produce 3A @ 24vdc using similar circuit architecture. As to the psu being separate, while MS put them in the same enclosure the power supply is a discrete component so yes it's replaced. I'm not using the stock enclosure because my gimbals are a completely different form factor so it was not a huge leap from there. Even with the stock psu I was quite surprised how much more powerful the stick was after the resistor mod though. Unless you're holding both axes way off center (not a realistic flying condition), the one that's displaced more is getting more juice so in actual use the stock psu is likely fine but it seemed low hanging fruit to bump it up after coming this far :p To that end I have a 7A psu that I'll be using to power my newer conversion with the Pittman 14203 motors which should equate to somewhere ~3.5Nm final drive. My estimate based on similar sized motors/draws of the stock ones with the 24:1 ratio is ~0.8Nm, so the simple resistor mod we did should bump that to 1.6Nm or so which seems plausible. I'll try to get some pics tomorrow, it's still a wip but at this point is finally cohesive enough to see what's going on. I plan on doing a proper thread here on it once it's in its 'final' form. The gimbals are a combination of printed parts with bronze bushings pressed for pivots along with aluminum parts for mounting the motor assy's, using the stock pots for simplicity

Roland van Roy figured this out and has good documentation if you want to do this. The stock circuit can be made to run up to 400% the stock torque by swapping some transistors and adding some caps/diodes/resistors. It works. http://www.simprojects.nl/ms_siderwinder_ff2_hack.htm As to motors, he has guidelines on motor selection. For one you probably want a brushed DC with a skewed armature to reduce magnetic cogging. Beyond that you select them based on the voltage drop at full load using Ohm's law. You want the voltage drop to be 12v at full power. I'm making such a mod now and have settled on 38.8v Lo-Cog Pittman 14203 motors (about the size of a coke can and have skewed armatures), which also need 25w 1/2 ohm resistors to bring the voltage drop to close to 12v since the terminal resistance was a bit low. My ideal motor would be a 48v Pittman 14204, a bit stronger and has the correct native terminal resistance. Pittman motors are not cheap new but can be found as surplus if you have patience. I paid $60 for a pair of 14203 recently. Using a 25:1 belt drive for power transmission (5:1 into 5:1). I currently run a modded MSFFII using the stock motors running at double the power by halving the resistance in the current sensing circuit and a 24v 3a psu to supply them with full power. Much stronger than I expected it to be. My gimbal is custom for my sailplane cockpit and pitch axis is 50mm longer than roll, but I kept the stock motor/gear assy's. Printed a glider grip and made a carbon fiber/printed elbows stick to achieve a full length offset stick that closely replicates typical sailplanes. This iteration of the stick is just to tide me over until I finish the one with the Pittman motors.

Might have already been mentioned but I didn't see it but in VR the effective resolution is basically half of its stated value because the left/right screens (or sides of the panel) are generating nearly the same image for stereoscopic 3d. Between that and pixels per degree of fov is why VR requires so much higher resolution to be on par with seemingly equivalent monitors

Neat to see this shaping up, the gui looks fantastic and I have a few questions. What is the torque of the motors, as in holding torque and peak/transient torque? Being direct drive, knowing that that would allow anyone with force questions to easily extrapolate how much to expect in their configuration like length of shaft and grip type. The base seems pretty compact so presumably the brushless drives are external? If so, how big are they and are x/y mounted in the same enclosure?

If you have an Nvidia card you may need to install their hd audio drivers but eitehr way it sounds like a driver issue

Haha, forgot to check which side the notch in the ball was on when you greased your Warthog? Easily repaired honestly, you can solder the wires back and put tiny pieces of heat shrink on them (slip a piece on before soldering). I've done this repair before without splicing in new wire and it's fine.

If you have your desktop up in the menu (pull up desktop in menu while controllers are still on) you can use kbm or just mouse

Damping makes it much easier to put the stick where you want without overshooting.

${1} e: what's the trick to making youtube links work here again?

There seems to be some confusion about the 'sweet spot' on Index, likely stemming mostly from a misinterpretation of what Norm from Tested brought up. The sweet spot he was talking about was the relation of your eye to the lenses and how much wiggle room there was for the hmd to get out of position, not edge to edge clarity. The dual element lens system requires your eyes to be properly aligned, so if the headset shifts on your face it can more easily go out of the desired location. However with the headset actually in that location, apparently edge to edge clarity (what most people are talking about when they mention sweet spot) is extremely good. The dev from Hotdogs Horseshoes and Hand Grenades goes into more detail about this (around 5:30)

The Lighthouse has to excite a minimum of three photon sensors on the headset in order to triangulate its position and orientation. The ability of Lighthouse to track is derived from the time differential of when the beam hits the different sensor locations in relation to each other.

It's not a strange quote there is just more nuance to this because the fov is not fixed due to moving parts that affect the results. The Index has 20deg more fov potential than Vive. Both Vive and Index can move the optics closer/farther from a users face. While Vive is spec'd at 110deg, that is at the ideal distance but if you have glasses or super long eyelashes or whatever and run the optics way out it's going to be less for you. With Index it's the same but with 130deg being the best case scenario instead of 110.

TARGET does more than kb/m emulation, though I've encountered scenarios where it was the only emulated mouse a game accepted (MWO) which was vital to my control scheme. For kb output, you can also natively script macros, sequencing, timed events. Once I scripted a pulse-width-modulation scheme to make a digital axis in a game behave like an analog one, etc. It also can return digital events from analog inputs with a high degree of control of the behavior. You can set it up to be able to alter axis parameters on-the-fly, it can be used to initiate 3rd party software. It has the option for shift layers and on and on. Didn't know VPC gear can be recognized by it, that's a big deal for people that want the power of TARGET with the quality of VPC hardware, especially for those that rely on combining devices into a single virtual controller which is mighty useful for games with awful joystick support

Reverb is said to be 63mm with +/-8mm adjustment

Hehe, no not to put in the hydraulic damper, to use instead of one or to supplement it so that you can run the hydraulic at minimum resistance. I have one of those hydraulic dampers, and also make adjustable tension friction rub joints and the quality of the kinematics of a generous amount of surface area is actually better than the hydraulic damper. I assumed my damper was a lemon, so did some homework and bought some synthetic 10 weight motorcycle fork oil after reading up on motorcycle forums. Mine also has sticktion, that gets worse with higher resistance. I'm hoping the bleed/refill makes it start behaving, otherwise I'm sticking with my mechanical dampers. Here's an album of a positioning joystick I made for MWO in 2012 https://imgur.com/a/ixi64. Towards the middle of the album is a breakdown of the simple friction joints. I used it emulating absolute mouse, and it was totally viable playing against kbm warriors, due to the precision of the damping and Nyogel made it from from good to great.

Use real damping grease. Nyogel 767a is not like anything you can buy at any hardware store (I've been making damping mechanisms and exploring greases for this purpose since 2012) and makes mechanical damping mechanisms work nearly as well as real hydraulic ones. One of the things that make it very different form other greases is the lack of stiction under these conditions, despite being really heavy stuff. Breakout force is no different feeling than the kinematic stroke and it doesn't evaporate away the way normal greases do.

What a waste that would be... just swap the sensors because the TARGET capable original board that also can be tuned with Cougar Control Panel is worth its weight in gold by comparison. If only the Warthog or any other joystick board were as useful as the Cougar is... OP, if still available I'll gladly take it off your hands!

I love projectors, but not for $5k! I have an Epson 8350 I bought as a refurbished unit from them and it's lasted years now. It's also a 3lcd/3chip (no spinning wheel...) projector, so image is very nice and no flickering. 1080p, very bright and short throw, though not ultra short. I gladly paid $950 for mine, but they have them for $675(!) now. https://epson.com/Clearance-Center/Home-Entertainment/PowerLite-Home-Cinema-8350-1080p-3LCD-Projector---Refurbished/p/V11H373120-N The 8350 is very low latency for a projector, under 20ms which may not be up to 'gamer' standards for a shooter, it's more than adequate for normal people and sims. I use it hours every day since 2014 and am on the original bulb, which only costs like $50 to replace. I too use my Rift for simming/playing though, but it's nice when friends are over they can see what's going on in-game and I like watching others playing too. Bottom line it's a sweep projector, that seemed to outperform the units made to replace it as well.