Gadroc

Been to long and slow.. but more progress. I got in the first prototype batch of the EOS Panel Board. They turned out great and I've got my list of changes before I mass order PCBs to run the cockpit. Each board has the following: 16 Digital Inputs 16 LED outputs (non-dimmable) with builtin limit resistors 1 Backlight channel (dimmable) with 14 outputs with limit resistors 4 Analog in (can be used as servo outputs as well via firmware change) 4 100ma switched outputs After using the board I'm going to make a few modifications. 1) Drop 4 switched outputs only used in a few scenarios. I'm going to pull up the I2C bus and a few more unused pins to an expansion header. This will allow me to make a small daughterboard to drive LCD or 7segs. 2) Make it a little shorter to cut down on cost. Number 1 drops four transistors and resistors off the board. I also re-did my AHCP panel. I switched over to the NKK locking switches and made it run off the EOS Panel Board. Things got much tidier than the mess of wires going to shared bodnar boards. I've also put in place an mini-ITX computer which will run my HSI/ADI screens. I also hacked it's power supply to run all the EOS panels and in console monitors. Turn on the pit computer and the pit springs to life. I also dropped back from 3 46 TVs to a single one. Life is so much simpler setting up DCS on one external monitor. The right side rails are all in place with caution panel and CMSP complete and mounted. I missed my goal of being flight ready today, but I just need to setup my Helios profiles and new monitor configs. Next weeks plans: 1) Re-route new center panel to use 5" LCD for ADI. It's a much longer term plan to get the real ADI working. 2) Finish electronic panel.

New build of Helios is out. Once in cockpit Helios should be on top again, although main menu still prevents Helios from being on top.

Sounds like network drop outs on the packets. This could either be the LUA script blowing up out (Check the DCS log files for errors "My Documents\Saved Games\DCS\Logs") or networking problems.

You should have little to no frame rate loss with your specs. The probable cause is turning off aero desktop. Make sure your taskbar is transparent and that neither SoftTh or DCS is configured to turn off aero.

If I remember right (I'm at work right now). On the properties of the monitor (double click on the monitor in the explorer/preview and remove all selections by hitting esc) there is a keep on top check box. This will force Helios to periodically bring itself to the front.

Well first off you linked one LED Output card and one Joystick Input Card... so neither is "better" they do different things. I assume you meant to link the groovy gamer input controllers. In that case the bodnar cards have a great reputation, but to be honest there is probably not a whole lot of difference. Possible differences between joystick controllers: 1) Price 2) Number of inputs 3) Additional programming (Keyboard/mouse emulation, etc) 4) Speed of response 5) Accuracy of ADC Unless you are using them for primary flight controls (throttle, collective or stick/cyclic) there is going to be very very little difference aside from the number of inputs. For bodnar and most of the gpwiz products 2 and 3 are identical. With that being the case go with price as the deciding factor unless you can find definitive tests that show 4 and 5 are better on one vs the other and that makes a difference in your use case.

The library works for any AVR arduino board if you want direct board to PC communications. A bus interface board talks to the computer and manages communication with many other EOS boards. This allows you to hook up several boards with one USB port. Due may or may not work with the libraries as it is an entirely different animal. I do not have one nor have I tested it. I highly doubt it will work with the library as the EosBus object does low level UART control. I do that because the default Arudino serial libraries are not good when you're trying to control a TX/RX pin on a half duplex chip. The mega will work as a bus interface board as it has multiple UARTs, but again because of low level access the EosBus library is not compatible with the out of box Serial library. You would have to write the interface sketch from scratch. Anyways the Mega is way overkill for a bus interface as it has many pins that would be wasted. A leonardo is half or less of the price ($20 with headers $19 without).

Yes it is a library to be used inside the arduino IDE. I specifically have done that so it's easy to reconfigure a sketch to make a board specific to a purpose. The sketch itself for my panel board (see below) is only 223 lines including whites space and comments. It would have been easier at time to go bare metal and through out the arduino libraries, but that would have made it harder for others to extend. The library has the following so far: Handles the EOS Protocol (drives the UART directly for speed so replaces "Serial") Fast Digital Pin wrapper (digitalWrite is SLOW so replace it for things in driving UART, Steppers and RS-485) Wrapper to treat random set of pins on the ardiuno as a bank of digital input pins. Wrapper for treating an I2C connected MCP23017 as a bank of 16 digital input pins. Wrapper for treating an I2C connected MCP23018 as a bank of LED output pins. Wrapper for treating PWM arduino pin as LED outputs with brightness control Wrapper for treating digital arduino pin as LED output with brightness control Wrapper for treating arduino analog pin as analog input Stepper driver module which handles acceleration and slow down of a stepper motor (I have it running at up to 10Khz step rate) On the deck to: Wrapper for MCP23017 that does a 8x8 64 input matrix Wrapper to do servos The arduino leonardo boards are excellent Bus Interface controller. You just need to add a RS-485 driver and optionally 3 LEDs (rx/tx/error). It can do full speed UART to RS-485 bus and do a USB Serial port to the computer. I have written a sketch for it and updated Helios to be able to act as both a Bus Master or use the Bus Interface. This offloads polling to the micro-controller which is much better at tight timings. In addition the PC now only has to process packets that actually contain data it needs to act on. I have designed two circuit boards which are stand alone EOS boards not requiring an arduino: LCD Board - this was the board used in my CMSP. It drives a HD44780 compatible LCD. Has 21 digital inputs, drives up to 14 dimmable (as one channel) back light LEDs and has one analog input. Panel Board - 4 analog inputs, 16 digital inputs, 16 led outputs (not dimmable), 4 100ma digital outputs, and up to 14 dimmable (as one channel) back light LEDs. I will be putting all of the libraries, sketches and PCBs under an open source license. I know I've said it's coming several times, but this is a lot of work. Light is at the end of the tunnel.

I've been creating a library of code to make it easier to do that. Any day now on releasing it. I finally have my prototype panel boards and I'm debugging the last few things this weekend.

What software did you load onto the arduino? EOS is a bus protocol that Helios speaks. You have to write the firmware on the arduino to make it talk the same language.

Helios does not "sync" physical switches. That is tremendously more complicated than one would think, unless you dictate exactly how to hook up each switch there are a lot of "unknowns". It way be simple in your use case, but across any profile that people come up with it is very very hard. In reality it doesn't help that much except for cold starts and would actually prevent doing missions without cold starts, as in general you're aircraft would have all kinds of systems just switch on/off as soon as your mission started. You'd end up having to physically sync switches before starting the mission anyways. If you enjoy cold starts I would assume you enjoy the immersion of parking and shutting down the ugly duckling as well!

I would find some Thrustmaster to real side by side pictures very helpful. I've been debating printing some new parts to extend the levers on the Warthog to match the real throttle.

I didn't take it any other way. I'm aware of the state of Helios and DCS and relative pace of change. I'm keenly aware that people paid money to use the product originally and it hasn't had the lifecycle I originally desired. I told people I would do the right thing in this event, which in my mind is opening it up.

I'm working on getting Helios available as open source, which would enable enterprising individuals to support additional modules. Unfortunately my personal coding time is much less than it used be. I have updated Phidgets interfaces, and upgraded EOS support along with a few other bug fixes. Next up is support for 1.2.4 world. Combing through the lua for additional modules is time intensive, and technically "fiddley" so I'm not sure how many people will do it to the level I did for BlackShark and A-10C. I'll try to move that config outside the code where possible so it's easier for non-developers, but some of the techniques used are fairly specific coding for individual switches.

Looking good, can't wait to see it in action.

Ok. Ran out of time this weekend on getting the new EOS Libraries up. I ran into a few bugs which slowed me down. Have a new tester app which dumps packets off the bus and allows you to send all the commands to devices. Just need to button up a few last things and I can upload it. Hopefully have it and Helios updated with the new libs by end of week (less time for coding week nights).

Also Waxi give me a PM. I'd be interested in where you diverged from EOS and why. If there is something that won't work I'd like to discuss before I start ordering some PCB designs.

Actually you should be able to do it with an 8x8 matrix with just one 16 channel chip.

hehe... yea that's the short version agrasyuk! The one thing people need to keep in mind when comparing to things like open cockpits is market size. You can drive down some of these costs by pre-buying items in bulk to drive up your margin. The FSX / 7x7 cockpit builder market is probably much more sizable than the A-10C.

Wow. This is why I always caution forum goers about thinking about entering a business selling to this market place. This is going to sound like I'm defending frank, but it's not. I don't think a $700 item will sell, but I FULLY understand that it has to be at that price point right now. This forum is flooded at times with people promising super cheap parts. Mainly they do them out of a labor of love. They are gung ho on building their dream cockpit and start buying the tools to do it. Bright idea comes - "I'm going to let everyone else in on this awesome tool. I'll recoup cost of tool and maybe more all the while getting everyone else pits for cheap.". Problem is they price the product on what they'd want to pay, how easy it was to make for them and over estimate the market. Except something funny happens. After you make the third one you realize it actually took longer than you remember. All those money saving shortcuts traded out cost for your build time. But now that these parts aren't for your baby time goes much slower. Next you know you're spending all of your free time building other peoples parts and make about $1.50 an hour to do it. You start to dread opening up the forums as they feel like work and knowing there are still "orders to fill" makes you feel guilty about working on your own pit. Several of them have dropped off the scene because of this killing their passion. There is a HUGE difference from bodging together stuff and making something you can sell over and over again. There is an even greater difference between drawing something in CAD and actually making it. For those capable of making one already, buying a part will never be cheaper, because you are also paying for someone else's labor. Unfortunately those who are capable are usually the ones who speak up about how expensive something is. On to specifics of franks BOM and my interpretations of the decisions: PCB I imagine this is a fully assembled board from a PCB shop where it is delivered with all 64 TACT switches and all back-lighting LED's already soldered onto the board. This is mandatory for anything aside from building your own or a few buddies. Soldering 128+ joints per board is mind numbing and slow. If you actually want to remain interested in doing this over more than two weekends, plan on having this done or shipping as a solder your own kit. Solder your own is suicide by angry customer who thinks they know more about electronics than they do. TFT Screen w/ VGA Board While cheaper TFT's have been found they use composite input. This requires someone to buy an additional $25-$50 VGA to composite adapter. This drops signal quality and also adds more support calls. Much better in my opinion to have a TFT with VGA. As a matter of fact I would prefer one that has HDMI/DVI. VGA ports are getting much harder to find and many newer video cards only have one DVI-I port that can support a native VGA singal. Joy Stick Controller For off the shelf use this really needs to support joystick output. It's fool/full proof so you don't have to instruct people on setting up potentially complicated software (Helios can still be complicated, but it's still significantly less than many options.). That being said you are not going to find controller boards with 64 inputs cheaper than $100. Most joystick boards have <50 inputs and all cost between $30 and $60. Custom arduino shield is going to be nearly the same price $20 arduino Leonardo or teensy w/ $50+ custom shield (see soldering statements above) to make a joystick controller. There are many "home cockpit" boards that all would require significant work to make them consumer friendly. This adds a "second mouth to feed" with the profit as there are not many mechanical engineer / programmers combos that have enough of both skills. Writing polished software for this space is a non-winning challenge. You will not be able to support a software developer with the necessary skills. They will have to be "in the community" and the problem is they'll scratch their own itch for usability WAY before it's consumer friendly. (Even though I original charged what people considered an exorbitant rate for Helios I've only made back 1/8th my normal per hour wage in software development. That's not a solicitation for anything! Just stating facts as an example of over estimating a user base.) Laser Cutting vs 3D Printed Buttons Laser but buttons will have better light transmission and feel, but require you to glue a flange, paint and engrave. This translates to a lot of labor. Making the buttons for my CMSP panels was the biggest pain. I can definitely see the appeal of 3D printed buttons. Machine Time Charges This is an absolute must. None of the machines are cheap to do this work. They are not tools you run down to the local hardware store and replace for less than $300. The absolute cheapest one is a $1000 3D printer. And if you are using a DIY 3D printer like that there is "endless" tweaking to make sure it's printing well. You have to charge money for the wear and tear on these machines. If you don't budget and put that money away for replacing these machines at some point you have worked yourself out of business.

I have a new set of EOSlibraries which I'll be putting up on Github this weekend. Along with a set of C# libraries and command line utilities that interface with it.

Also would you mind sharing the mechanical setup of the oxygen switches? I have a real oxygen panel and haven't had a chance to think through how to make those work.

Continued excellent work frank! I can't wait to get some of the electronics and software sorted so I can get back to physical pit building.

Unfortunately I don't know of another user friendly board. The phidgets hardware is expensive because its designed simple to use. It limits current so you don't have to figure out limit resistors. It has full dimming on each output. It has voltage conversion to support easier power supplies. You can design specific solutions which are much cheaper, but you'll be doing a lot more electronics work and potentially software work.

Short answer... no. Long answer there are several folks on the boards who have laser cutters and periodically sell panels (including myself). There is not enough volume of jobs to make money on this via volume, so you have to charge a lot which lowers your volume even more. You may be able to recover some of your cost and then have a great tool, but don't count on this as a money maker. In addition there are many who want mass produced quality and plug and play out of anything they buy. There is a lot of manual work to do in small run productions like this which just erodes your profit again, unless you don't value your time.