Jump to content

rel4y

Members
 View Profile  See their activity

  • Posts

    969

  • Joined

  • Last visited

 Content Type 

Everything posted by rel4y

  1. rel4y
    Just uploaded the CH index finger slew sensor housing + hat to Shapeways. It is designed for an ALPS RKJXV1224. I also ordered all my latest models for a testprint. Now I just need to get the PCBs sorted. https://www.shapeways.com/product/77CM7HHHX/ch-throttle-slew-housing?optionId=134489379 https://www.shapeways.com/product/74N89QK2E/ch-pro-throttle-slew-knob?optionId=70902280
  2. rel4y
    Really good find and well analyzed mate! Please fix this ED!
  3. rel4y
    I had a chat with Merkin via PM. I believe this may be due to noise on the power supply or digital noise of the CH Throttle Mainboard. With a bench power supply I get no jitter whatsoever and in a Cougar the signal is also very clean. This seems to be a CH specific problem and I will have a look at it. CH sadly does use last century electronics in their devices, so my planned upgrade mainboard will probably solve this once and for all as well. Ill test if different type of decoupling Cs have any influence and if I can improve on grounding somehow. However Merkin said it is not noticible ingame at all, so it seems to be only minor noise. Anyway I will investigate this, Id like to get things perfect! You can also try some filtering. A simple 2x filter is in most cases more than enough to get rid of minor 1 Bit jitter, while it cant help with a badly spiking pot. This guy here wrote a nice script:
  4. rel4y
    At last I have finished the CH Throttle Slew sensor. The gameport CH Pro Throttle does not have an analog stick, so I made a nice slew stick in place of the index finger hat. However, theoretically that slew fits in all hat slots. I am really looking forward to this one. This is gonna be one big fat overhaul of my CH Pro Throttle. :D Big shoutout to Deltaalphalima1 who send me the STEP file of the ALPS analog stick, without which I couldnt have modeled this. :thumbup:
  5. rel4y
    Thank you for posting your impressions mate! Will have a look at their forums. :) Yeah, but then I made this to solve the problem. :D Like I said earlier I cant fit the PCB of the other sensors in the Throttle sensor mount, it is just too large. And making a new PCB layout solely for the Cougar Throttle is not going to be cost/time effective tbh. Maybe Ill do it in anyway in the future. It stopped the spiking for me at least.
  6. rel4y
    I have a question, are you simply referring to the bad pullup resistor (R10) on the blue STM32F103 boards? Or what did you change about the USB connection for EJoy v3? It looks like a awesome software, but can you please add 4021 shift register support? Without it, we cant use the abundant Thrustmaster hardware. It would really make EJoy so much more attractive for people! :(
  7. rel4y
    The TM TPR pedals have a crappy pincer mechanism, there is no comparison really. Go BRD or Slaw for amazing combat style pedals, for desktop foot on top style the MFGs are good as well.
  8. rel4y
    Hey guys, Sorry for not responding. I got food poisoning and was knocked out of action for a few days. I am just recovering, so Ill have to keep it on the slow side for a few days. Here is the promised Throttle Hall sensor which doesnt require my standalone adapter, but you will have to buy two extra parts and calibrate it yourself. Just to make it clear again, except for the Cougar Throttle sensor all my other sensors are plug and play and work with the original controllers and software. https://www.shapeways.com/product/KTB6AUAJX/cougar-throttle-hall-sensor-mount?optionId=70678300&li=shop-inventory What you will need is a N52 (or above) 4mm cube magnet and an Allegro 1324 SIP hall sensor. You can find these on ebay or much cheaper somewhere else. In electronics stores you can usually get them for under 2$. However here are some ebay links for the lazy: :smilewink: Allegro 1324 SIP hall sensor on ebay: https://www.ebay.com/itm/Allegro-Microsystems-A1324LUA-T-Linear-Hall-Effect-Sensor-5V-3-Pin-SIP/163123980014?epid=730000234&hash=item25faf266ee:g:OykAAOSwZrdbNlBS:rk:1:pf:0 4mm cube N52 magnet on ebay: https://www.ebay.com/itm/Super-Strong-Magnets-4mm-Cube-Neodymium-Rare-Earth-Block-Magnetic-N52/223137027440?_trkparms=aid%3D555017%26algo%3DPL.CASSINI%26ao%3D1%26asc%3D54463%26meid%3Dd0547255d5b44bf49feae78dd0d13100%26pid%3D100505%26rk%3D1%26rkt%3D1%26%26itm%3D223137027440&_trksid=p2045573.c100505.m3226 Assembled it will look like this. Finally put a piece of tape on the back to avoid shorting the hall sensor legs on the metal mount. Here is my current progress on the CH conversion PCBs. It turns out is incredibly difficult to measure the distance between hats precisely. What I did was basically trace the cutouts with a pencil, then mark their middle points and measure the distance with a caliper. If anybody has an idea on how to measure it more precisely please give me a hint. The plan is to install an ALPS analog stick as slew sensor (on the Gameport version) of the Throttle. The USB version does have an analog stick, albeit its not great. The second PCB is for converting a Combatstick into a Fighterstick. Dual trigger is coming as well, I already have a plan for it.
  9. rel4y
    Hey mate, I think I got something for you. I drew up a quick sketch of an analog hall sensor mod for the Cougar Throttle. This one is based on one of my earliest tries with the Cougar. Now as you may have realized I have a distinct antipathy for hall sensors, but here you go: What you would need is an Allegro 1324 hall sensor and a 4x4mm neodymium magnet cube, best N52 or above. Calibration involves only moving the hall sensor on one axis, so it is already miles improved over these expensive Cougar hall sensor kits other people offer. :smilewink: How far you insert the hall sensor in the slot sets your maximum range, the closer the larger the voltage range you can achieve. If you are going too close to the magnet you will actually lose range. So find the sweetspot or settle shortly before that and calibrate the last bit in Software. The latter is how the manufacturers do it. Finally put a piece of tape on the back to avoid shorting the hall sensor legs on the metal mount. I will upload the sketch to Shapeways and you can order it. Full disclosure, I think it would work on the Cougar Joystick as well. So I may be cannibalizing my own sensors here somewhat. But if you like real precision work then go for my magnetoresistive kits. Believe me, the ball bearing MR sensors are a league of their own! :thumbup:
  10. rel4y
    On the subject of resolution: No controller is currently using 16 Bit resolution, the marketing divisions are simply lying to you. The math doesnt add up. 16 Bit real resolution at the sampling rates possible would make your standard 8 Bit Throttle seem like a swiss clock. For reference: http://www.hempstick.org/download/articles/OnResolution.pdf
  11. rel4y
    I think you are confused Sokol. :huh: It does! It has completely analog outputs. It replaces the original sensors 1 to 1, pnp. So all my Cougar, CH and Saitek sensors can and should be used with the original controllers. The only digitally communicating sensor is the TLE5011 in the Cougar Throttle sensor, that one can only communicate with my USB standalone adapter. I just want to make new higher res boards for those who would like to upgrade their original 8 Bit controllers to 10 or 12 Bit. PS: I have soldered up the orders of the last couple of days and should be able to ship by tomorrow evening. The two CH sensors are still available, so feel free to PM me.
  12. rel4y
    Have I somehow attacked you? I wasnt even addressing you, so chill mate. Still all my points stand. :thumbup:
  13. rel4y
    If you guys could... really... that would make me like Normandy again! You guys can do it! All the best of luck! :thumbup: The comparable performance of the G-14 stays the same for all current planes, it will outurn and be slower than a Mustang and Dora, it will be outturned but faster than a Spit. What do we gain? Well, nothing really. Lets face it, noone who owns the K-4 would buy a G-14 for another 50$, its too similiar and a waste of model. Many Allied only jocks dont even own axis planes at all. So, whos the audience here? The only late G model that makes sense to me is a G-10 as addon to the K-4. Same engine, same everything, just slower and lighter. Remove a few 3 model parts, throw in MG151 code, remove a few kilos in the code, done! I am not even sure the A-8 will sell well, there is no real reason to choose it over the Dora except for nicer looks and supposed accuracy for the inaccurate Normandy map. I own the Dora and I dont think I will invest in such a similiar plane.The same argument as I made for the G-14 performance can be made for the A-8, but at least it has a different engine and is sexier. Id much rather have them set up a mid war timeline over Normandy, get rid of the stupid ALGs and add a few airports in Britain and up to Calais. If it could be community modable this map would go crazy! For the late war uberplane stuff a Bodenplatte type map or german soil map would be much more fitting. Although the Luftwaffe was basically done at that point in time anyway.
  14. rel4y
    Ok, good news. I am done on the CH hatswitches and uploaded them to Shapeways. Obviously still untested, but I am pretty sure they are good. I will print the hats in smooth fine detail plastic and paint them afterwards. The prototype Saitek pot 3D models are also done. I made 3 different versions, two of them with snap fits and one that needs to be screwed down. Lets see which works best. Here is one with snap fits. Plans for the next weeks are designing a CH and Saitek Flight Pedals upgrade mainboard with 10-12 Bits res and a dual stage trigger for CH Fighterstick/Combatstick.
  15. rel4y
    In July 44 over 700 G-14s and G-14U4s with AM motor were put into service. Around 400 G-14 ASMs went into service in August 44. Basically total 109 production capacity and lines stayed the same, just the model name changed, well and obviously the blower on the ASM type. But honestly while the MG151s would be amazing these models are really simliar to the K-4s, just a bit lighter, slower and lacking in high alt. What Id like to see is a DCS F-4 or G-2, those where simply the sexiest 109s ever and then release a Spit Mk V and it would be a perfect world. The Normandy map timeframe is nonexistant anyway, at no point in time is it accurate and for the closest it gets the LW would hold not a single airport in that map.
  16. rel4y
    Good price. I have a bunch of chinese copies laying around. At least they work so far, but I am curious for how long.. :music_whistling: Recently I also bought a bunch of "high quality" OpAmps from China. High quality my ass, they were so blatantly fake and crappy it hurt my head. Instantly opened a ticket on AliExpress and b**** better have my money! :mad: I think in future I will stick with Mouser for proper high quality, although I must say AliExpress has a good customer refund policy. I think Sokol is being a bit harsh on the CH gimbals. They do a good job, but as said rely heavily on their sturdy pot axis which are simultaneously the gimbal pivot points. That is why you need to somehow either keep that sturdy axis in there and modifiy the original pot, or replace the axis with an equally strong and low tolerance part. -> eg 3D print with a ball bearing :P Hey Sokol, do you by chance know which Saitek products use the same pots as the Saitek Pro Flight Pedals? I had two pots in my Pro Flight Pedals fail on me, which is a ridiculous. I did fix them with a hall sensor, but thats a pretty ghetto fix tbh. See here: https://theairtacticalassaultgroup.com/forum/showthread.php?t=27201 The Combat Flight Rudders seem to have the same pots. I remember there being troubles with the X52s as well, was it the same type of pots?
  17. rel4y
    Good job mate! I will add a bunch of encoders etc to my CH Pro Throttle as well. Yeah, MMJoy rocks! :thumbup: Well, I cant really make them cheaper. The 3D printing cost at Shapeways is 36% of that alone. But actually with my whole modding endouvors I spent over 600€ total in the last years, trying to get stuff right and 3D printing prototypes. Shapeways is really costing me a crap ton... If I sell enough Sensors I will buy a Cetus 3D printer from that money. :cry: Listen to Sokol, the gimbal is not made for this type of mod and hall sensors are a real hassle. I avoid them like the plague nowadays. There is a reason why eg Simundza of MFG got rid of all hall sensors in his Crosswind pedals quickly and switched to programmable sensors. But if your in for a bit of DIY work you could break apart the CH pots and glue a magnet to the axle, then stick a hall sensors to the back like in the picture from MegaMozg. However then the real fun begins, calibration. :doh:This will probably be the easiest method. I chose to 3D print the whole pot because it was the best option for pnp imho. Btw, what could these possibly be for? :music_whistling:
  18. rel4y
    Hey Sokol, I didnt know that. Do you have a link perhaps? I plan to do a board which allows you to plug the connectors of all the CH devices straight in and makes use of their button matrix. Since CH uses 0.1 in connectors I can just put a bunch of header pins on there. Or maybe I find these male connectors somewhere. Thanks! Well no, it wouldnt work with CH Control Manager. But the CH manager would be adequately replaced by some equally powerful manager software (as eg MMJoy).
  19. rel4y
    Hi Thasiet, How are you doing? Done with your project? :) The cost is 20€ per piece and I have only one left at the moment. Yeah if you use the standard CH mainboard there is no change from the 8 Bit resolution. But I still plan to make a higher res mainboard in the near future. However it is not yet clear based on which microcontroller. The STM32F1 seem interesting, so I have ordered a few.
  20. rel4y
    No you dont need any PCB, these are plug and play. The cool thing about these MR sensors is that they replace the standard CH pots 1 to 1, but contrary to them they work contactless. That means you can simply connect these sensors to your CH mainboard and done. You will never have to worry about deteriorating or dirty pots again. The CH Control Manager recognizes this sensor normally and you can calibrate it. I just finished making the first batch of CH MR sensors. But I just saw that Shapeways increased 3D printing prices, which I calculated to a 17% price increase on my models. Well great Shapeways... :doh: For installation you can simply pinch the original crimp connectors a bit with a pair of pliers, but I always recommend soldering! :smilewink: Then just stick them ontop of the header pins and your good to go. Your wires colors will be different since this is a gameport CH Pro Throttle. On the sensor PCB the pins are labled as GND, OUT and VCC. For the USB CH Pro Throttle green = GND, yellow = OUT and blue = VCC. While setting the range I also took a screenshot of the output in VKB joytester. This is uncalibrated so dont be suprised that it doesnt reach full range. Btw the original CH pot doesnt even reach as far as this MR sensor does in uncalibrated state.
  21. rel4y
    Yes your right Sokol :thumbup:, maybe I just upload a sketch for an easy DIY mount for an Allegro 1324 for the TQS on Shapeways. Then people can print it themselves. I tried fitting the MR Sensor but Id need to make a completely new PCB and stencil for that. Not worth the effort imho. So you are saying I am too cheap? Need to raise prices then I guess. :D PS: Tonight I was busy otherwise: CH Hats. As far as I know noone has done these so far and Id like to convert my Combatstick to a full blown Fighterstick. :smilewink: Switches are OMRON B3F-1022.
  22. rel4y
    That Bi-Tech 6127 HALL Sensor is prohibitively expensive for a simple hall sensor.. I can make a 3D printed Hall Sensor for half that price, but as plug and play option. My MR sensors are basically the same as a MagRez Sensor, just in a sealed housing and with ball bearings. But I cant make the PCB small enough to fit in the Throttle pot holder.
  23. rel4y
    After acting as a human pick and place machine for a few hours I have just finished the second batch, half of which is already reserved. If your interested shoot me a PM. :thumbup: I have also pretty much finished on the CH pot replacement. These pots fit the CH Fighterstick, Combatstick, Pro Throttle and Pro Pedals. Just the trim pot needs to move to the side and its perfect. My next project will be a combi-shield usable in the CH Combat/Fighterstick + Pedals + Throttle based on the 32 Bit ARM STM32F1 controllers also used in Drones. That microcontroller is much more powerful than the Arduinos and has internal 12 Bit ADCs. I plan to make that one open source and upload it to OSH Park. These sensors would be a perfect fit for such an upgrade from 8 to 12 Bit, as the standard CH pots just dont cut it at 12 Bit. Yeah, thats pretty much what I am doing. :smartass: What sort of pot do you need? Do you have a picture?
  24. rel4y
    Well, in theory if both types of sensors are set up perfectly there should be no discernible difference. In the real world however you have to mount the sensors somehow and neodymium magnet strength varies by a huge amount. I tested a batch of N52 magnets and found that they varied up to 30% in magnetic field strength. As I have mentioned before, hall sensor output is a product of their sensitivity for (eg the very sensitive Allegro 1324 5mV/G), the magnetic flux density they experience and the angle they are set to the magnetic field. So if your flux density varies from magnet to magnet by 30% you will experience huge problems making a fixed mount work. You have to set up every hall sensor at the perfect distance for each individual magnet to get correct readings. Not to mention they should also be perfectly on axis of the magnetic field, to make the output as linear as possible. Now magnetoresistive sensors alleviate that problem since their relative independence on magnetic flux density and robustness to axis offset. They usually are also of higher sensitivity. It makes my job of setting up the sensor much easier and makes for a more consistent product. A few drawbacks however are, you need to make a PCB with a high quality op amp in your circuit (= higher price) and the temperature stability is much worse. But since we use these sensors indoor in our living rooms and problems start upwards of 70°C, its a non issue. I need to check if I can get the circuit of the magnetoresistive sensor fit within the Throttle mount, since the trim pot I need for setting the center position is fairly large. Fitting a hall sensor in there is no problem, but has the disadvantages I mentioned above.
  25. rel4y
    Ok I understand. Like I said, the main problem is not making it, but the two 3,3 and 5V versions of the Throttle. People need to figure out which one they need and that will cause trouble. If you are interested I can probably make one the next weekend. Would you want a simple hall sensor one or a magnetoresistive one? I think I still have a bunch of Allegro 1324s laying around. But for that you need to tell me which PCB version you have, check here: https://forums.eagle.ru/showpost.php?p=2863739&postcount=130
×
×
  • Create New...