Hempstead

I don't have Crystal, but my Quest 2, and now Quest Pro controllers are just as precise as a mouse, if not more so. Works great! So great that I never opened my PointCtl. I will sell you my unopened PointCtrl if you want (if I can find where I put it. ;-). A simple strap work solves the problem, although I have to admit, the right stick hand with a monkey on the back is very annoying particularly in a dog fight. I don't hang the right hand controller though. I only use the right hand controller when I am on the ground doing cold start anyway.

Haven't posted for awhile. Due to family issues and a bit of health issue. Anyway, I finally decided to build a kinda pit... to get around my health issue. https://blog.hempstick.org/2023/09/a-pit-finally.html

Works just fine...

https://blog.hempstick.org/2023/07/price-comparison-between-shapeways-and.html I don't know about you... the price differences between DIY and Shapeways are just too big... PA12 on Shapeways costs > $100 than printing it yourself with PA12 + CF. But, it's an option for those DIY challenged... on the other hand, if you are DIY challenged, you should not be reading this thread to begin with. Unfortunately, PA12 burned fume is toxic. So, it will be difficult to get this mil-spec. approved.

https://blog.hempstick.org/2023/07/design-one-thing-test-many-times.html Don't you ever say I never test my designs!

https://blog.hempstick.org/2023/07/kids-dont-do-ldpe.html LDPE slumps too much... no good as a bag. But the useless good news is that it does not bond to Nylon.

https://blog.hempstick.org/2023/07/sous-vide-test-on-nylon12cf-controller.html Sous Vide Annealing Test is a success! Note that it's going to be difficult to do this with a Sous Vide stick in immersion mode, because it's difficult for them to reach 90˚C or even 100˚C Nylon12 needed. But for other plastics with lower glass transition temperatures, it might be viable. However, I am thinking, if it's exactly 100˚C that you need to hold.... you could basically red neck it by steaming it for 12 hrs... I mean buy a cheap aluminum steaming basket, a big pot... and steam it on your stove top for 12 hrs... it's just that you might have to top off the water several times.... maintaining 12 hrs steam is a chore all by itself.

https://blog.hempstick.org/2023/07/sous-vide-f16-like-control-stick.html Sous Vide my Nylon12 Carbon Fiber F16-like Control Stick 3D print!!! I mean, Annealing! I was shopping for an oven suitable for annealing 3D prints at the right temperature.... although in the back of my mind, I have wanted to do it in one of my ovens... or the jerky oven I have ordered. There are a couple of problems. 1. Although Nylon and Carbon Fiber are both non-toxic, that's not the same as food-safe, or food-grade. Particularly, this filament is made in china -- the land of gutter oil. Whether it's made in china or not is immaterial, the fact that it's not marked as food-grade or food-safe, it's a no go into my regular ovens for cooking food! 2. Regular ovens are not instruments... precision is kind of a concept that does not get talked about in culinary "arts." And my old built-in large oven is from the 1980s... analog control mechanical alarm/timer, and all that. My better oven... the Anova Precision Oven... all modern and all that and there is even an app for it to program and control through WiFi. But.... again, it's for food! 3. The new Jerky Dryer I ordered (still to be delivered) is meant for drying filaments, not for food. And it could only do about 80˚C, just shy of the minimal 90˚C required for Nylon. It can be used to anneal PLA and others but not for Nylon. So, it hit me.... when I narrowed down the model of oven to buy.... Why don't I just vacuum bag that thing in a food safe plastic bag and throw it in my Anova Precision Oven and do that bagless Sous Vide thing? Your regular Sous Vide stick in water immersion operation cannot do more than 100˚C, but this Anova Precision Oven can do higher than 100˚C bagless, non-immersion, Sous Vide! In it goes... We will find out if I ruined a perfectly good print tomorrow!

https://blog.hempstick.org/2023/07/carbon-goodness-or-bronze-goodness.html Carbon or Bronze?

https://blog.hempstick.org/2023/07/f16-like-controller-trigger-assembly.html Fit test for the trigger assembly groove.

https://blog.hempstick.org/2023/07/f16-like-stick-body-pla-matte-printing.html I can tell you that I am definitely going for the Nylon12 Carbon Fiber. It's very strong. Too bad that Nylon12 generates toxic smoke when burned. So, most likely it won't qualify for mil-spec in the cockpit. But PETG might pass... I have two spool of PETG Carbon Fiber, I should print one of those for study. Anyway... qualifying for mil-spec is not one of my goal. But it might be good for a simulator. Supposedly, Nylon 12 is not that much stronger than PETG... and it's less hygroscopic than Nylon. Printing Nylon is an annoying business... you have to dry the filament (oven for 12 hrs), and/or have to have a filament enclosure, or else. And it's picky on print plate material and glue stick you use... PETG is not that picky. 12 hrs printing... minimal post processing, on a USD $1,500 printer (raw material is about $25)... that makes it commercially viable if you are willing to setup a print farm. But... this is designed for home made, with just a printer and some simple home tools (well... plus some specialized tools and parts you still have to buy). But if you can afford a USD $1,500 printer. You can afford to make it. I will let others have it for free... if only I can figure out a way to share without getting my IP stolen and then be accused of stealing from LMCO, again.

https://blog.hempstick.org/2023/07/f16-like-control-stick-assembly-grooves.html

That's it... New stuff will be posted here, https://blog.hempstick.org/ The stuff is really hosted by blogger.com, by Google, not on my servers at home. I just set it up to use my own domain DNS entry for the redirect. So, even if I got hit by a bus tomorrow, the stuff will stay online...

WTH? Now that I figured out a way to generate higher resolution but smaller file size to fit the 500kB limit. Looks like there is a total upload limit.... 200MB. Should I got and delete all my past posts? Create another account, or just leave?

Looks like I am going to have to reconsider the plan on hosting my own blog on my own hardware/website for this stuff. That is... abandon this thread altogether. I can host the not-so-large pictures on my own servers/staticIP/DNS etc., and refer to them from here... but I hate that idea... as you well all know forums are filled with broken links pointing to pictures hosted on some previously-popular-pictures-hosting-sites-now-defunkt.

ED forum is now enforcing 500KB per file upload limit... So, you are now treated with the lowest quality pictures I can produce to fit that limit. For crying out loud... disk space is dirt cheap! Anyway... a fit test with previously printed PLA parts... fits just fine... well, mostly... It's a bit tight, and I would have to force it a bit.... and the top cover does not sit 100% flat (needs a bit sanding... I am guessing it's shrinking/warping that is expected of 3D printed parts.. particularly between two different materials). Also found another problem.... See the 3 rivets on the genuine OTTO trigger? Well... I didn't model the extrusion. And the stick was originally designed to be split into two halves... I did encounter a bit of a problem when doing the test fit for the split two halves. So, I designed in a recess for the top pivot rivet. And the two lower ones? Don't care... screw down and the plastic will yield... even better. But that is a problem with the one-piece print.... My tolerance was designed just right... I modeled the NozzleDiameter in it. So, for some critical dimensions, I did reference that NozzleDiameter variable, set to 0.4mm. This way, the trigger will sit in just snug... well.. now I have to design in the "assembly recesses/tracks" for the rivets. Or.... I can use a hammer.

100% infill Nylon 12 Carbon Fiber. Fresh off the printer, support tore off... no sanding yet. I am pleasantly surprised about the raw surface texture where the support touches main body... not bad... a little sanding would take care of that. This is extremely strong! Also, take note on the back of the handle where the main body bends... This is the top of a curved 2-directional curved surface, printed horizontally. Usually, with fixed layer thickness, 0.2mm, you will get a serious stair case isolines. I turned on the adaptive layer height printing, so those staircase lines are dramatically reduced together with drastically increased print time. But, I think this is worth it. It will make sanding a lot easier. But, like I said in the previous post... it took about 12 hrs to print it.

100% infill, speed adjusted for strength... adaptive layer heights, rotated to have the trigger part facing down for support... all the trimmings... It will take about 2x amount of filaments, 103g vs 207g. And it will take 11 hrs 50m vs. 4 hrs 49m. It's being printed, now.

Nylon 12 Carbon Fiber printed (damn, I forgot to get rid of the screw holes designed for fastening two halves!). The surface condition on the supported side doesn't look too bad, nothing a bit sanding couldn't solve. There are some surface defects on the top side as well... nothing a little epoxy potty and sanding couldn't solve. It's incredibly light (feels cheap), but strong. Next time, I will try printing it with 100% infill (and get rid of the unneeded screw holes). Now I just have to figure out how to get the support materials out of the core. Don't need to have perfect surface condition inside... so shouldn't be a problem. Next time, I might print it with the trigger side down. This way, the surfaces we see mostly have better surface condition.

Try the drop down "Show all Axis Command", and then see what else axes are bound to Roll/Pitch/Yaw and Throttle. Clear them all except the ones you want! DCS has a nasty habit of "autoidiotically" assigning any analog axis of all "new" controllers it sees to Roll/Pitch/Yaw/Throttle in a "common sense" fashion... all of them! Make sure you have one and only one axis of one controller bound to one of R/P/Y/T. Any time you plug in a brand new controller/mouse DCS has never seen before, it will again autoidiotically assign the new axes again... to all airrcrafts you have! So, any USB device reporting analog X/Y/Z/Rx/Ry/Rz axes will autoidiotically be assigned to R/P/Y/T, respectively, all of them! So, you bump your desk, and the idling mouse would suddenly try to assert its authority over your roll and pitch. But from your PoV... you have no idea where that "unintended" movement came from! It might win some and lose some against your TM Warthog. Or, you pull on the TM stick hard, and some yaw authority would come from nowhere and then quickly evolve into an uncontrolled nose dive. Any combination is possible. All kinds of weird things could happen!

The new Bambu X1 Carbon prints so much faster, but it's actually not as fast as the eye can see. The real comparison is that after slicing and actually printing one half of the thin stick mold... it's about 14hrs vs 8 and 1/2 hrs (Ultimaker 3 vs Bambu X1 Carbon). The thing about it is that both are core-XY mechanism. The Bambu print head moves and prints at eye blurring velocity, while Ultimaker 3 moves a strolling pace. I would have thought that it moves at least 3 times faster, resulting in 1/3 print time. But no. It's a tremendous time saving, but not the 2x 3x time faster as I was hoping for. 40% time saving is nothing to be sneered at, but it's not 66% or more! (My Phrozen Might 8K prints the same thing in about 4hrs!) My guess/observation is that Ultimaker 3 has dual print heads, 1 prints with the primary filament, the other breakaway support. For each layer, switching filament is quite fast.... basically heat up the idle print head which is already almost up to temp... a second or two, then go the prime tower to print a round or two... and start printing. On the other hand, X1 Carbon has one print head only and an Automatic (filament) Management System, which would retract the current filament, it would need to readjust the temp of the print head for different filament, heating up would need the similar amount of time as U3, but if it's lowering temp... that could take longer. Once the print head is "up" to the temp, it has to re-insert the other filament by feeding it all the way from the AMS to the print head, then slowly feed forward to eject enough into the purge chute (similar to your loading filament operation), clean up the print head, then start printing (starting with the prime tower). It has to do this twice for every layer that has both primary and support portions. That adds a lot of time. I have already did the sensible setting of only using support filament on the "support interface" instead of all support. Meaning... only a thin layer of the support will be printed with support filament... so for layers that do not have support interface (the few layers where support actually connects to the primary article) do not need to go through this time consuming unloading and loading filament operation for each layer. I do expect that if you use the same filament for primary article and also for support... the print speed would dramatically improve. But I do not like that, as for some of my more complicated designs, it's very difficult, if possible at all, to break the support off! In essence, X1C is much faster than U3. But not the 2x 3x faster you might expect in practice. Durability, reliability... don't know for X1C yet. But U3... it's been utterly reliable for me in the past 5 years. But, since Ultimaker is also a coreXY machine, i.e. both movement of X and Y axes are on the print head, its head design is also light although not as light as X1C. It potentially can move faster than it is now if Ultimaker also program in Input Shaping as Bambu does. So, U3 speed could improve, but with Ultimaker's current "corporate direction", I am not holding my breath. In reality... I will be using the X1C as more of my exploratory printer (if it holds up to my abuses), and U3 as the primary.... and the fleet of resin printers as the messy-only-when-print-quality-matters stuff, like the knobs in the pit. Or... printing plugs for silicone RTV molding (lost wax).

Bought a Bambu X1 Carbon, upper left in the picture of my 3D printing "nook." Not shown in the picture is my messy vacuum pump, pot, and hoses on the floor! Boy, is this X1 Carbon fast! Don't get me wrong. I love my Ultimaker 3. It's my go-to printer, despite my plenty of other "choices." Slice it, send the file over, as long as the damned auto-leveling passes, it will print and get me what I want.... just slow... but very reliable. What I have problem with.... is Ultimaker's direction -- enterprise-oriented direction. New Hardware has very little innovation, and charge an arm and a leg.... and worst of all, no upgrade path for faithful customers like me. No Ruby print head and abrasive filament resistant parts that are in the S3 and S5 but no upgrade path for me, and charge like USD $4,500+ for a new S3! So, I can't print carbon fiber with my Ultimaker 3..... ya ya ya... I can try to mod it and stuff... but thank you but no thanks for messing with my primary printer! So, I have ordered some PETG CF (Carbon Fiber) from Bambu... maybe that will be good enough (I doubt it!, but I got to give it a try, right? And, I am sticking to it as my "justification" for buying the X1 Carbon!).

I was finally able to make the experiment today using the borescope and a new my own design of wrench for bicycle Shrader valve with the Stretchlon 800 plastic sheet. The idea was to rely on Stretchlon 800's ability to stretch and vacuum pressure to have it stretch and compress the fiber/resin mix from inside without the "ugly" folds of plastic bags. The folds could sometimes cause resin blocks locked inside making extraction difficult and require a lot of post processing. The experiment result is that the vacuum pressure I was able to achieve with a vacuum pump purchased from Easy Composite in UK, it wasn't able to stretch that much as I hoped. I was hoping the stretching to be the main "shape conforming" component but with excess material folds to supplement it. Unfortunately, the experiment result shows that I will have to live with excess material folds to be the main component, with stretch of Stretchlon 800 as supplement. In other words, I will just have to make the diameter of the inner plastic tube big enough to minimize the stretching, instead of making the D as small as possible but rely on the Stretchlon to expand. I guess, next experiment is in order -- paint on silicon RTV as the inner "bag." That, for sure, will have no fold.... but it requires making a new inner plug (extra CAD and 3D printing time).... Oh well...

Ya... tried just casting it. that was the original original idea... hot hot hot... Hence the thin shell idea of about 1mm thick, maybe 2mm max (would prefer to keep it at 1mm or thinner if I could). Tried using woven sleeves (D=4.5")... difficult to get it to conform to the more extruded parts of the plug...(the palm rest, obviously cannot be done in 1 piece with woven sleeves). Plus, it's going to be difficult for later more shapely control sticks/throttles like AH64, F15, etc. In other words, I am searching for construction methods that is as generic as possible, yet easy to do at home. Hence, the one time disposable thin mold (commercially bad idea as it took about 30 hrs on an FDM machine, 4 on an SLA, instead of the lost wax process for metal casting, over hang all your want within reason, and no daft angle required, no need to split the mold correctly), and thin shell. With chopped tow... the more shapely curvatures necessitate compression (even the woven sleeve requires compression to hold it to the desired shape), -- either use two, or three, rigid mold halves to compress it (draft angle again), or... do vacuum bagging. Vacuum bagging is far more flexible than rigid molds. I am sure open lay up of two halves and then later glue or bolt together would be good enough for commercial efforts... but I am not looking for commercial venues. I am just looking for a better way of doing it at home. If you have ever tried the Solidworks mold tools.... trying to split the damned F16 throttle mold, you would know how difficult it is to use SW to split that mold... Oh man... took me two weeks of trial and error to find the "right" lines to split it off into several pieces. It will tell you you don't have enough draft angle here, there... there is always something that got in the way no matter how I split it... until I cut it into more than 2 pieces.... It was neigh impossible to do it in two-pieces casting... for wax mold. And since it's going to be a bunch of flexible hand made silicon RTV molds, SW's mold tools was just too "rigid." It was obviously designed for CNC milled solid metal molds. Even a simple F16 rudder pedals are difficult to split due to those little ridge bumps on the side of the pedals. It's just difficult.... if you are doing mass production, it may be worth the efforts to go through that... but one off, two off... it's just not worth it. Hence, I tend to like lost wax... you are free to design it in any shape you like.... no need to consider later construction methods and change the designed shapes to fit that construction methods... much.