Tekkx

copy

Just one short question in between: (Did some thinking while I repaired our kitchen sink today) @All: Have you ever touched a real Magnetic Hold Switch? Does it have the same "klick"-effect as a normal switch? Specially if it NOT powered - means if it doesn't stay in ON position? Meanwhile I ordered some copper wire for extra coils. And my brain is already over some special design problems... Stay tuned.

Hey Rocco. Ohne die Sache wirklich zu verstehen, will ich Dir meinen Ansatz mitteilen: In den letzten Tage hatte ich etliche Versuchsreihen zu meinem Centipede-Problem mit verschiedenen Arduinos (Leonardo, UNO, MEGA) laufen und bin mitunter auf das gleiche (oder ein ähnliches) Problem gestoßen. Der UDP-"Strom" zeigte nichts oder nur Datenmüll an. Ich konnte dies lösen indem ich die Baud-Rate erstmal auf 115200 senkte und (ganz wichtig) es muss überall das Gleiche eingestellt werden. Also im Gerätemanager, in der Batch-Datei und in der *.ino. Dann lief meine Kommunikation in beide Richtungen. Mit dem COM-Handler bin ich auch nicht weiter gekommen. Geht bei mir nur in Richtung DCS-BIOS -> Arduino. Den Centipeden habe ich allerdings trotzdem noch nicht am Laufen. (aber das nur nebenbei :) Das Projekt ist erstmal zurückgestellt, da ich Steuerkram machen muss und aus reiner Profilierungssucht "mein" Magnetic-Hold-Switch-Projekt ganz oben auf die Liste meiner freiwilligen Aktionen gesetzt habe). [THREAD=146392]look here, if you want[/THREAD]

If I look at.... that's correct. Even if it works, switch would be a little bit lacy or slacky like a limp daisy :cry: Mh. ATM I have no idea... My very first concept (long before this thread was opened) had such a "flip over". But this will bring other problems. If first physical tests will fail, so I have to wind a stronger solenoid which could override the flip over.... Inside the housing is much more space than a prefabricated solenoid can use!!! So I have to buy copper wire (different diameters), some iron tubes (out of the scrap-box of a plumber) to make a series. All experts I met at Internet calculated first and made a series of different solenoids. Seems not to be an exact science. Just Magic :D Meanwhile I had a concept in mind with an impuls drive like model railroad switchpoint. This impuls could be strong enough to push over the tilt point. No risk of overheating. But this would require special code: One single impuls for ON, another impuls (same solenoid maybe) for OFF. Ian (if you read this): Don't write any code in this matter since there is no hardware solution in sight :) Let me (us) do some thinking. Think, I follow for the present the Pull-and-Hold Taktic with an Flip-Over. But right now it's family time :)

Attached is my latest concept. It is "On Scale". Width is 40 mm. Travel of Lever is 30°, like Honeywell. Housing (2mm Aluminium), Bushing, Microswitch, Brass Tube and Coil I have already on Hand. It is not assembled yet! Microswitch (ON-ON-Type) will close if his lever is released. I hope to save power this way. Two questions are still open: - will available power be adequate to pull the switch? - will Solenoid overheat or not. Right now I have just this single Solenoid at hand. Yesterday I ordered (here is NO local dealer!!) an other one. Current used Solenoid has a designed Duty Cycle of 25% (on 10 Minutes) at 12V and runs now on "Testbed" for about 45 minutes at 7 Volts and is close to 50 Centigrades (stable) without extra Cooling. Edit: after 60 minutes still at 48 to 50,5 Centigrades. Room's temperature is about 28 centigrades. Air pressure (250 m ASL) and humidity N/A ;) Nxt edit: IF power isn't enough I could Solenoid fire with 12V and a simple electronic circuit limits current or voltage (depends on what's more applicable) to the minimum value after a few seconds.

next very small step Yes, indeed. I found a lot of typos. But I hope, dimensions are right. Cause it's my(!) only source where panels are exact verticaly to be seen. Yesterday I did a next step to our Mags. After I had a sketch (close to drawing) at right scale I was some shocked: There is really not much sufficient room. Some experiences in watchmaking would be helpful. :) My coil on stock brings some problems maybe. This adventure began in January with a less-than-20-Bucks-purchase of A10C.... Now I have to learn how to calculate and wind(?) solenoids... Is there still someone out there who says, gaming has nothing in common with education?

If you talk about the real Bird: You are right. And if you operate your Sim just from the SimPit you also right. What if someone clicks a switch in the virtual cockpit or uses automatic Startup procedure or mission starts hot on RWY? You WILL get an inconsistent condition that you have to catch with your code: Virtual Switch is ON, Device is ON, physical SimPit Switch is still OFF... Thats why I think by now, a Simpit NEEDS a Mag what acts in both directions. Or you have to deactivate clickable cockpit. Imagine: You hit the "Auto" button (I know, it's out of flavor) and around in your Pit clicks and cracks and blinks: very funny. Isn't it? @deadman: Thank you. I took the picture from here, cut the SAS, scaled it to requiered dims (5.75 x 3.36) .. Ready. So came out: A Mag must not exeed 30 x 40 mm (view from top). That was close: The already purchased Aluminium-Profile (planned for housing) is exactly 40 mm. :)

I think, my model would already work with current available code: DcsBios::ActionButton saspPitchSasLPush("SASP_PITCH_SAS_L", "PUSH", PIN); // from Micro Switch DcsBios::LED saspPitchSasL(0x1108, 0x1000, PIN); // to MOSFET/Coil Input Just "cold" thinking! Your objection, "... how to press against holding force..." in case of manually switch off is a strong argument. Let me manufact a prototype and we'll see. In worst case I have to add an additional miroswitch which opens hold-circuit in case of pressing lever into OFF direction. I'm still in a good hope, that good timing of lever position and opening of microswitch would do the service. If you move the lever just a few degrees, microswitch opens, DCS-device stops working, DCS-BIOS stops sending the LED-Signal, Coil's holding power is broken at once, spring snaps lever to OFF. Theory!!!

Thanks, Warhog, for your warm words. Now I'm under pressure :D Attached is my first approach. Just brainstorming. There are no special solutions to be seen. It is completely out of scale!!! Will try to explain my thoughts: Coil is fired by 12V, powered by a MOSFET with Logic Level (costs a few Cents pp) Main Problem will become, to give Coil enough Power to drag the switch without overheating. If you don't come under fire of AAA most devices working whole mission (several hours) I have a single one here for tests... it gets some warm. Have to keep that on focus. Maybe here are some experts (one is enough) who can calculate the right coil (I can't!). The bushing on top is a small part from a lamp-shop(?). It usally feeds a cable thru the housing. I just not drilled a single hole in this matter, so I don't know in which direction my first test will show. IF the coils dimensions are right, Switch will be able to snap in ON position if equivalent Switch in DCS is clicked. If not (and coil becomes too warm (or hot), a second contact could be added to reduce (by a resistor or whatever) current in HOLD position. Before I start sawing, drilling, soldering and execrating: Can someone provide me measures of a real SAS-Panel? So I will know in which borders I have to move :) SAS will be the mother of all Mag-provided panels :) What I forgot in my sketch is a Suspension Point in OFF-Position. This gives a notable KLICK! if it blows off :D

I disassembled already a circuit breaker. .... I live in a metric world and everything looks other than overseas :) The "Old World" provides nothing usable in this matter :( Right now at home. Digging my garbage stack for my sketches... Thank you for "... 5.75" ..." :)

75 Bucks for one rusty old - maybe disfunctional - switch... that's my budget for a complete panel :) Will post my first sketches and thinkings soon! (Right now at work)

Right now I make a Rough and Ready solution with lighted buttons. The code will be just the same as Gadroc's new approach. It doesn't matter if a LED or a Magnetic Coil is fired. Difference is only needed Power. So maybe I'll be one of your (Gadroc) first Test Pilots :) Gadroc: Thank you so much for your efforts in DCS-BIOS and also HELIOS (I love it). Break a leg! Tekkx PS: I already opened a Thread for brainstorming DIY Magnetic Hold Switches.

After Gadroc gave the right Keyword (look here, if you want) I had following question: Where do you Guys purchase your Mags (Magnetic Hold Switches)? After it takes me a few weeks to find out what to look for, I found just those Honeywell's for 1000 $ each (still can't beleave that). So I did some thinking, made some sketches and collected a box full of Aluminium-profiles, different brass-tubes, micro switches a.s.o. Is it worth to develope and build own ones? Or is there an affordable source anywhere? I have no real workshop. I do anything "on desk". I plan my Mags to build with an "extended" function, so they will snap into desired position if you klick a Switch in the virtual Cockpit. No matter, which direction. Real ones can't do so. If YOU (community) say, it's worth to do such switches, I will increase my effort in this matter and will share my solution. Right now I make a Rough and Ready solution with lighted buttons. By the way: The code will be just the same as Gadroc's new approach (Link top of this post). It doesn't matter if a LED or a Magnetic Coil is fired. So maybe I'll be one of his first Test Pilots :) Gadroc: Thank you so much for your efforts in DCS-BIOS and also HELIOS (I love it). Break a leg! Tekkx

Digged some new(?) information. Sorry for pushing this Thread, but I can't stay calm with a problem I dream about :) Since Arduino's IDE gives his Compiler's messages (I found the checkbox) I copied some relevant (so I mean) parts. Maybe some Experts find some time to analyse? BIOS_calls.h is nothing but a list of Definitions of Buttons like "DcsBios::ButtonCS cdu0("CDU_0", 6);" Wrote source after // Maybe it's irrelevant. But how will I know? There maybe some other interesting information inside this Log. So I attached complete output as a Textfile, althow reading this file is like counting stars... terrible. Compile-Errors.txt

This morning (think, I dreamed about before) I made some other tests. I remember my first breakthru in Arduino: Converting an Leonardo into a Keyboard. But I don't remember, what and how I did :( So I tried to reflash original firmware - without success. What if a Leonardo isn't a good choice - ... So I took an UNO-clone with CH340 Chip from the stack. Flashing CDU.ino (with ButtonCS-function activated) ran like hell. :) So my next step is to unscrew my CDU an replace Arduino Board. Or I find a way to reflash my Leo. (shorter way, cause there are a lot of screws and keys to remove) Edit 10 hours later: UNO and MEGA (that's what I have on stock) swallowing sketch without complaining... but after this there is no communicaton with anything. So I am at Zero. Again. Today - while working some stuff - I made some thinking: What if I2C uses same pins as DCS-BIOS? And this causes the crash? How could (I - ha ha, very funny) exclude those pins from been written (or red) by DCS-BIOS? Hands off of matters, you know nothing about. And learning curve is relatively flat ;)

That's exactly what I'm doing right now (instead of managing the dishwasher) :) I found out (by deactivating different parts of code) COM-interface (sorry for this term) crashes if pointer(?) shows at example!: DcsBios::ButtonCS cdu1("CDU_1", 50); If I comment just these examples compiling runs fine. I suspect a interference between I2C and Serial??? (Just a naiv thought of a novice). Maybe Centipede needs some special treatment. Copying class is the wrong way, maybe. And: I never suspected Leonardo itself. :)

My prayer for broken cables was rejected :( Sth's wrong with my sketch. After reset the Leo it apears a short time as Leonardo Bootloader at COMxx (different to recent). If I upload (timing is the goal!) an other shurely working sketch it behaves normal. Than I flash the A10C-CDU.ino and the board disappeares from the radar. Means: Device-manager. So has an old donkey to learn dancing. Means: Serial Communication. I'm now at Zero Level.

This I understood after reading your first lines :) Hope, you'll become a teacher or Uni-Professor. So the next generation will be safe :D After I did recommended changes, Arduino IDE compiles w/o error. But PC lost connection to Leonardo. I'm afraid it died this night. Have to make further checks. I pray for a broken cable :unsure:

Hello, dear Community. Since I stimpled my first steps into the amazing world of DCS-BIOS, I want to ignite next stage and bring my CDU into REAL life. Thanks to [FSF]Ian, HMA and agrasyuk, they helped and inspired me to this decission. This CDU worked slightly well as a Keyboard, but it made me never really smile :) Due to hardware limitations (I don't want to rebuild a 65buttons/20LED-PCB) I need all 64 IOs of Centipede and all IOs of Arduino (there will come an AAP soon) So I did the following steps: (most code is based on the UFC Master_Caution_Button Example) What you have to know: My programmer's knowledges are not much more than copy_and_paste. I operate very close at my brain's border :) Late 80s to early 90s I did some Turbo-Pascal and Basic. I don't know what a serial-parser really does... The matter: Project consists of A10C-CDU.ino - main part consists of code from the example and includes following files BIOS_calls.h - just a text file with commands copied from control-reference Centpede.h - library DCSBios.h - library I created a new class in DCSBios.cpp // ---- Centipede void ButtonCS::init_(char* msg, char pin, bool reverse) { msg_ = msg; pin_ = pin; pinMode(pin_, INPUT_PULLUP); lastState_ = CS.digitalRead(pin_); reverse_ = reverse; } void ButtonCS::pollInput() { char state = CS.digitalRead(pin_); if (reverse_) state = !state; if (state != lastState_) { sendDcsBiosMessage(msg_, state == HIGH ? "0" : "1"); } lastState_ = state; } // ---- End Centipede */ and in DcsBios.h // -- Centipede ------------------------------- class ButtonCS : PollingInput { private: void pollInput(); char* msg_; char pin_; char lastState_; bool reverse_; void init_(char* msg, char pin, bool reverse); public: ButtonCS(char* msg, char pin, bool reverse) { init_(msg, pin, reverse); } ButtonCS(char* msg, char pin) { init_(msg, pin, false); } }; // -- End Centipede ------------------------------- Althow a Centipede Opject is created by Centipede CS; in A10C-CDU.ino, it failed to compile by error: 'CS' was not declared in this scope lastState_ = CS.digitalRead(pin_); Meanwhile something strange happen: While my last attept 2 hours ago fails, it compiles right now (I just wanted to copy the error message) w/o fail ????? Edit: Found reason for strange behavior: Removed for debugging 'CS.' from CS.digitalRead(pin_) in DcsBios.cpp and forgot to redo it. Anyway. Arduino's pins operating as expected. Key(s) send commands to DCS, LED (backlight atM) lights as expected. I attached complete A10C-CDU.ino and example code Include-File. Due to Forum's limitations renamed to .txt Is there anybody who knows what I did wrong? A10C-CDU.ino.txt BIOS_calls.h.txt

Thank you, Hans. I found it. Even if it doesn't work so it could be a light to show the way :) Right now (before I found Your edit) I created a new class in DcsBios.cpp (sorry, Ian, for raping your code) void ButtonCS::init_(char* msg, char pin, bool reverse) { msg_ = msg; pin_ = pin; pinMode(pin_, INPUT_PULLUP); lastState_ = CS.digitalRead(pin_); reverse_ = reverse; } void ButtonCS::pollInput() { char state = CS.digitalRead(pin_); if (reverse_) state = !state; if (state != lastState_) { sendDcsBiosMessage(msg_, state == HIGH ? "0" : "1"); } lastState_ = state; } But it works not yet. While compiling comes this error: 'CS' was not declared in this scope lastState_ = CS.digitalRead(pin_); Due to lack of experiences (and knowledge) it'll take a while to debug. I suggest - if this of public interest - to open a new Thread. I did it: http://forums.eagle.ru/showpost.php?p=2452352&postcount=1

So, if I think right, I have to create a new function ... I can't recite, how Centipede's pins to poll, but I wrote it somewhere at home. So I'll try tonight. Thank you.

Hey Hans. Could you tell (to me and community), HOW you got centipede to work? Maybe Ian is right to say, "it should be fairly simple"... But his brain glows much brighter than mine ;) Edit: Uuups. Ian was more quick :)

... and simultan another question: I searched the reference and can't find output-code ("...LED..." maybe) for Instrument- or Panel-Backlighting. So I could synchronize my CDU-Backlighting with other Instruments. This would be the only one DCS-BIOS-Output in this CDU-Project.

Hey agrasyuk. Nice to meet you. I think, I red already almost all of your posts. Your CDU-Project has been one of my inspirations. THX alot :) My problem is a special DCS-BIOS-question. My CDU was already working (as I wrote), I want just to reflash it with new, real life. I don't want (and can) to move a single srew nor solder anything. Take a short look at, If you want. http://forums.eagle.ru/showpost.php?p=2411684&postcount=3 The missing link is: How to bring my centipede to talk to DCS-BIOS or v.v.. I tried some (wellknown) ways, but the outcome is silence :( Is there a special(?) library additional to Centipede's?

Hello Community. Due to personal treatment by Ian (I'll build a monument of You and put flowers there every day) I managed my first steps in DCS-BIOS. Amazing!! Have a look at if you want (german) Now I wanted to come to business with my good old CDU (started thinking in Feb 2015 - finished in July 2015). This thing works as a keyboard. The result isn't what makes me really smile :) (it workED as a Keybd - I flashed it already) Messing around with Keycodes of two different languages and modifiers sucks. Today (I broke my job and ran home) I did a first attempt to convert this CDU to a REAL device, using (you'll already know) DCS-BIOS. I'd need some input cause I stuck again... My hardware is Arduino Leonardo with Centipede. To read or write I/Os on a stand-allone-Board with DCSBios is now easy to me. But I have coplete no idea, how to map(?) 64 IOs of centipede. Origin IOs of Leonardo I'll need too. I could trickle around as a blind dog, but this here is a shorter way. Maybe. If this has been diskussed and solved at other place so give me a link. Please. Thank you in advance :) Edit: The layout of CDU's PCB prevents making a matrix. All buttons sit with one leg on ground. So I need really ALL Inputs of both boards (for CDU and AAP).