Jump to content

MilesD

Members
 View Profile  See their activity

  • Posts

    968

  • Joined

  • Last visited

 Content Type 

Everything posted by MilesD

  1. MilesD
    Hey Deano, That is in DCS and it varies depending on the aircraft. There will be several user programmable options for the mouse wheel using the buttons. center and forward button pressed together = mouse scroll down center and aft button pressed together = mouse scroll up In the future I should look at making it a universal mouse wheel function as you described Thanks for jumping in and picking up the slack for me. Miles
  2. MilesD
    You are correct. I only did basic proof of concept testing, While the mouse cursor did move with the cockpit there is still a lot of work to do on that. Thanks for the question, Miles
  3. MilesD
    Hey Guys, I am in the final push here, The Vive and Steam VR Base Stations caused quite a delay and I have a 95% working solution. The only time you see any interference is when the base station and the FCU are both aligned with the center of the HMS. Essentially, if you look and point directly at the Base station, the cursor will ghost, both conditions must be met. I had a conference call with the sensor manufacturer yesterday and we talked about future hardware solutions. The web site is not up at this time, I am pushing all info through this thread. Thanks and I will keep you posted, Miles
  4. MilesD
    Todays Work: Fixture for FCU Assembly, and 3D printing out FCU enclosures. Miles
  5. MilesD
    Theres that cheeky Imperialist humor, @Hansangb, Thanks for the offer of sending the Pimax for fitting, I certainly appreciate it, but I don't think it will be necessary. I will try to find some data on the 5k dimensions and make a test mount and send it off to you to test fit. I just shipped out a Vive-Pro mount to JoeS to test fit because I dont have the Vive_Pro either. I am focusing on production fixtures now and kicking out a few units for final testing. I hope to delivers Los's in a few days for his extensive testing and feedback. Los is my adviser on all things Flight Sim. Here is today's work, the FCU and APB (Aux Push Buttons) fixtures for hand soldering. I did not feel comfortable with these buttons in a re-flow oven.
  6. MilesD
    Now I am a little disappointed in you Ratcatcher. I had a good laugh at what I certainly thought was a "smart arse comparison", and a good one too. Except that 3 wheeler has better lines than the FCU. I rather enjoy a bit of cheeky wit, especially when its grounded in the truth. :) and there is my first ever emoji, just for you Ratcatcher. Miles Edit, I just realized, your concern for the looks of the FCU, may be because you named it.
  7. MilesD
    I agree, It's a much more immersive when you can feel what what you are physically looking at in VR. I think that why while using PointCTRL is convenient, and a definite step up, it still feels like "no big deal" when you use it. Miles
  8. MilesD
    Hi Harlikwin, Just a quick response, Thanks for your comments, I would agree with your 95% of your post. I only meant PointCTRL to be a interim device until integrated HMD tech was viable for our flight sim purposes. The problem is, it's been taking so long for me to start production, the sales cycle may indeed be short. I would expect the VR manufactures are working on an integrated solution that will render all of the aftermarket tracking gloves and devices obsolete. I certainly don't think this will be all for "naught" though, because I am manly doing this because I enjoy it, and I do feel PointCTRL will be a viable simple solution for current HMDs. I haven't forgotten about the conventional big and multi screen simmers either and I will transition to that as soon as I can. Thanks Miles
  9. MilesD
    Thanks everyone for your comments and feedback. I have been following along with them. Today I finally completed all the FCU and HMS enclosure changes and tweaks for the initial version, and I am printing production units as I write this. The FCU is function over form for now. It does look high and boxy, however longer or wider would present interference issues with those that have smaller fingers. The good news is that all components are modular, so if a new slimmer battery becomes available you can just replace your battery and enclosure with a new slimmer one. The current shape presents no issues when you’re in VR, it only looks cumbersome, but feels fine. My take away from most of the comments is “just get the friggin thing done”. I think some of you have surreptitiously contacted my wife, because she has been telling me the same thing. Thanks for the feedback and suggestions, I will respond when I have a few more minutes. Here is a pic showing the FCU’s, Rift, Vive, and Vive Pro Mounts.
  10. MilesD
    I had originally planned on mounting the HMS on the face like Leap Motion as you suggested, but a few things changed that. 1) The IR LED is located 1" or 25mm above your finger so to keep the relationship of the sensor to finger it needs to be mounted above the center line if the HMD. 2) Though testing, a slighter higher mounting location on the HMD gave a better line of sight to the IRLED when interacting with switches that are lower and on the sides of the cockpit. Specifically, shoulders would block the sensor when you were reaching behind you for a switch. This also benefits those of us whose like Beer, and result thereof can block the line of sight to a switch mounted low in the center panel. 3) While not a big issue, the IR Sensor is designed to mounted at an 90 degree angle to the PCB. As far as trying it on different devices of the same model you can certainly use Velcro to mount it, Great Idea. Miles
  11. MilesD
    Okay Guys, If something sucks, please feel free to tell me. I truly want your feedback, positive and negative. I was not pleased with the HMS mounts, so I am sure many of you said "ugg", but were to polite to say anything. Just as RatCatcher asked about smoothing edges on the FCU, I appreciated his feedback and I am going to try and increase the radius on the FCU based on his feedback. My mistake on the HMS mounts was trying to make them quick change for users with multiple devices. In the end end the whole thing just got to big and ugly. With this new design the whole mount stays attached to the HMD and the PCB and sensor can quickly be removed to put into a different device. These are still rough prints that need some more refining but I think its going in a better direction. All of PointCTRLs components will be available to purchase as a set or individual pieces to include just enclosures or PCB's. In the pics are the Vive and Vive Pro mounts and the graph squares are 1/4", and 6mm. Thanks, Miles
  12. MilesD
    I received 2 Steam VR 2.0 lighthouses today for interference testing with PointCTRL. They are really amazing tech. There is no problem filtering out the coded lasers carrying the positional data. My only concern is PointCTRLs IRLED interfering with the lighthouse comms with the vive pro. Since it operates on a very high carrier freq, I don't see it being a problem as long as PointCTRLs PWM dimmer is tuned to a different freq. Here a few pics of the Vive HMD mount. Thanks, Miles
  13. MilesD
    Hey Guys, Just a quick update. After much trial and error, I have the HMS enclosure and HMD specific mounts complete. Nothing fancy to look at. I just made it as small and functional as possible for now. The HMS enclose snaps into the HMD specific mount which is attached with double face tape. The 3D prints look better in person than in the pics and the finish is still a WIP. @ RatCatcher, I am not entirely happy with the box look of the FCU either, but we are really limited by LiPo battery. I may be able to put a bit more of a radius on the upper edges. My goal was to make the FCU narrow enough for people with smaller fingers and have it not interference with the joystick. In the future I may look at a lower design profile that extends further up the hand. @Marineer3302. I am currently using a XYZ Davinci 1.0. I think its about 4 years old and I did not have much use for it until now. I have 4 CNC machines in my shop and never thought much of 3D printing until I got into this project. The sheer speed to go from CAD to print is remarkable, The CAM phase, and setup for CNC operations is time consuming. Ther is also much more design flexabilty with 3d printing than traditional milling. I am sold on it, and still deciding on my requirements for a production 3d printer. @ FoxDelta, I certainly would consider sharing stl files with customers in the future. Every day we get a little closer. Its just taking much longer than I anticipated. I am working on PointCTRL full time, but it’s just me, thanks for your patience. Please feel free to share all observations and feedback, to include negative ones that can may contribute to a better product. Thanks Miles Here are some rough prints that don’t represent the final product surface finish.
  14. MilesD
    Update: I just wanted to share a pic of the FCU in black, as well as my productive fails. I ordered some grey ABS also to see how that looks. This is without any post processing. Working on the CNC mill fixture right now to mill the slots for the attachment straps. When I integrated slots in the 3D print I was not satisfied with the results. Questions: I have asked myself that many times. Good suggestion on updating the first page, I will do that. Thanks Thanks Deano97 for answering Q1, Q2, and Q3 I think so, I have to get over to Los’s to deliver Serial Number 1 and test it. Thanks for the link Kirk, That’s looks really interesting. I must look into that tech some more. Thanks everyone for the questions and support. Miles
  15. MilesD
    Hey Guys, I have been working away on PointCTRL non stop. There just seems to be endless final things to do before production, and I really want to send it out right. I am surprised that I had a working proof of concept device working in a week and here I am 8 months later still refining things for production. Anyway, we get closer everyday. I assembled some production PCBs for fixture design and testing and all went well. I have been focusing on the enclosures the past week and can say I have the final design production files for the FCU's. The HMS enclosures is tomorrows project. It should be much simpler than the FCU's. I have a basic concept that allows for quick changing between various HMD manufactures. After a lot of research and testing 3d printers for production I decided to go that route. The prints in the picture were printed for speed, and I will be applying some post 3d printing processing such as vapor smoothing. Thanks I will reply to the post questions tomorrow, Thanks for picking up my slack Dino, Miles Edit: The protrusions on top are for the mounting straps, I will post some pics tomorrow showing how the system works.
  16. MilesD
    My neighbor who has very limited vision in one eye tried my rift and said it was great, and felt 3d to him. I was surprised, so I closed one eye and it had a similar 3d feel, especially when moving your head laterally. I would do what Strike eagle suggest to see if its a good fit for you. Good luck Miles
  17. MilesD
    Hey Guys, Let me know if I am oversharing, but there seems to be interest in following the development of this. The last 24 hours I have been on this Vive issue non stop. I setup up a mock lighthouse that strobed 850nm light every 8 ms. It interfered just like it did on Los's Vive. I was able to integrate a photo-diode to read the flash and was about 50% successful in timing the exposures when the strobe was off. After a few hours analyzing the programmable registers on the sensor I may have a more effective solution that will also benefit Rift users. The short version is, any new sources of IR light are classified and tracked, but not assigned to the mouse cursor, unless certain parameters are satisfied. The first test with my IR strobe simulator had no effect on the cursor. Additionally sometimes in my pit I would catch a ceiling light that was on bright and higher than the cursor and it would take control. That no longer happened on my initial test, I am sure a few issues will pop up, and I need do add some more parameters, but I feel this is on the right track. I forgot I said I would post a few pics of the latest enclosure. Everything fits in there nice now, I just have to pretty it up a bit. Thanks, Miles
  18. MilesD
    Looks like the polarizors are out, anything close to economical will not get into the IR range that we need. Im working on capturing frames only between strobe and laser sweeps of the light house. It will reduce cursor FPS to 60, but it probably wouldn't be noticeably. Edit: Without getting to far into the weeds I believe I have a workable solution. The Vive tower actually strobes every 8ms followed by the laser sweep of either X or Y axis which takes 8ms also. I only need about 1.2 ms exposure for the PointCTRL sensor. My intent is take a frame after each strobe firing. As long as the lighthouses are pointing approximately toward the Vive the exposure will be completed before the laser reaches it. These numbers also put us back up to up to 120 FPS on the mouse. I hope to write some code and add a IR photo-detector to the HMS tonight. I do have to confirm that both lighthouses fire at the same time though. Even if they don't fire concurrently I believe I will still have enough time to capture the frame. @Los, When ever you are free I will come by with another bottle of wine for another testing session. Miles
  19. MilesD
    I think I will start with polorizing filters 90 degrees apart. One on the lighthouse and one on the HMS..
  20. MilesD
    Thanks FireCat for your time and effort, it was interesting to see how many people are interested in PointCTRL. We are probably good on the list for now. After the slow roll, I will post here when the rest of the units are ready to ship and they will be available through the website when its completed. Update: I have been working to complete the enclosures and attachment methods. It took significantly more time than I anticipated to get them to where I was satisfied. I should have some pics up tomorrow of the final product. Issue: All of my testing has been on the Rift. It was recently bought to my attention by Joe S. that the Vive lighthouses put out an IR wavelength. I took pointCTRL to my friends Los's house tonight, and we tested it on the Vive. The results were not good. The lighthouses interfered with the PointCTRL sensor and caused the cursor to jump around. I am trying to get a few lighthouses to run test on now. After analyzing the lighthouses output, I am hoping I can filter out the interference. I will also have to do some more testing to confirm that PointCTRL LEDs do not interfere with the Vive tracking diodes. Los didn't mention it, but it was something I thought about afterwords. So thats it for now ,the 3d printer has been running non stop making refinements to the enclosures, but I think we have it finalized. My wife was a big help sourcing and testing many different materials for the finger strap, and I like her final solution. Thanks everyone for your patience, Miles Los testing PointCTRL in his Pit, and the death ray lighthouse.
  21. MilesD
    Update: First proof of concept print, I will CNC mill the front and rear endcaps from sheet ABS tomorrow. The 3D printer does not do well on thin flat parts.
  22. MilesD
    @ Deano87, Good question. The buttons can be configured to your preference. Since we don't use the center mouse press in DCS much, I have that configured the farthest back button, as well as a long press to activate it, Short press is the radio menu. Mouse wheel is also emulated with center button hold followed by forward or aft for mouse wheel up or down respectively. I am stil working on the UI for user options. Update: Its been a few years since I used my 3d printer so it took me all day to get it working again. I am printing the FCU enclosure as I write this. I could not find the black ABS so the first prints are in white. The Black ABS will be here Monday. If the production molds are holding things up for the "slow roll" I will ship with the 3D printed enclosures and then ship the production enclosures for replacement of the 3D printed ones. Thanks for the list, I think we have what we need for now. It was very helpful to get an idea for the initial production. I have cleared out my message box so if any of the early posters to this thread would like to be on the "slow roll out" drop me a PM. If all tech questions could be asked in the forum that would be helpful. If one person is asking, I am sure others have the same question. As always, thank you to everyone for your support on this project. I feel its getting close. Miles
  23. MilesD
    I got you Immel
  24. MilesD
    Hey Guys, Thanks for making the list, Its kind of a interesting to see how many people are posting or interested in PointCTRL. I am almost ready to ship a few prototypes out. I plan on a slow roll-out after I get feedback on the prototypes. The slow roll-out is just in case there are any unforeseen issues that need addressing. FireCat, you seem to be the man with list, if you want to start a list with anyone who would like to be on the initial roll-out, that would be great. Active, and tech savvy users with the time to abuse PointCTRL would be a plus. I made some more observations and changes to the sensor, so I hope to put up another video soon. Thanks again everyone for your time and support. Miles
  25. MilesD
    Hey Ratcatcher, Issues with a "wobbly" finger shaking the cursor usually means it's time for drink. There are a few things going on in the video that are contributing to what you observed. The first one is because of the wide angle lens that is needed for the horizontal FOV only utilizes about 30% of the sensors resolution in the vertical. I am addressing that now with optics. Your dead-zone suggestion is a good one. I spent a significant amount of time with cursor smoothing and dead zones. Since I have already written the code for both, I will probably include them as another user option in the programming interface. I personally found using a dead-zone made the cursor feel jumpy and made me feel a bit disconnected. The smoothing option worked great with only a slightly noticeable delay when the FCU was moved quickly over long distances. Unfortunately, this same smoothing was applied when you moved your head while attempting to press a switch. With head movement, the cursor initially moves away from the switch and then settles back over the switch 200ms later. In the end I found it better to have a slightly shaky cursor that responded instantly. And quite honestly you are on and off the switches so quickly your hand does not seem to shake that much. In the video I was moving the cursor in calibration mode. In calibration mode the cursor movement is slowed down because it actually flashes three positions separated by 150 pixels with a time delay. This is to let you know that you are in calibration mode and which axis you are currently calibrating. At one point I forgot I was in calibration mode and was moving switches which made the cursor jump as its supposed to. I think I may have referenced that mistake in the vid. The cursor also appears to jump around on the canopy frame because it jumps from 3d surface to 3d surface, just like a laser pointer reflecting off objects at different distances. Anyway, thanks for pointing it out, I always appreciate your feedback. I will talk about the resolution in the next video. Thanks again Miles
×
×
  • Create New...