Last time, we figured out that the phidgets controlled everything and that there were all of these phidget files... sigh...
(If you haven't been reading this series on the CRJ, please read 1 thru 3 please.)
Return of the Phidgets
Rick (my lovely programming engineer) wrote a small but significant program that allowed us to find all of the serial numbers and what button connected to which phidget. So, we literally pressed every button in the simulator and figured out where they all hooked up... and which phidgets were dead. Huge spreadsheet later and all of the buttons are working and we know where they are! Now onto updating the phidget text files...
In the phidget text file, there is a serial number, a channel number and a variable name. Where in the world does the variable names come from? Well, figuring this out was fun. Most of them are FSUIPC variable names for items such as the landing gear, throttles, and the like. Most of them are variable names created for Project Magenta by Project Magenta with help from FSUIPC.
Here is the rub... the phidget text files on the original computers are a mess (10 years worth of messes and replacements). The serial numbers in the files do not match what is in the current system! Some of them do, so taking a logical approach to this, I matched them up. If a set of buttons were grouped together and matched what appeared in the phidget file, I changed the serial number. Slowly, we got it down to the just the auto-pilot and the radios which had Project Magenta variables and I had no clue as to who went where.
Rick managed to clean up the airplane logic. Thank you Rick!
Next step... New Computer Systems
Moving everything to new computer systems. This was NOT fun.
We have a tendency to build very nice and robust computers for the simulators. As a result, this insures that the computer systems will last as long as possible (decades). We build them in-house and typically are not purchased through a name brand like HP, Dell and the like. We like to install hardware from Intel, Asus and nVidia.
Therefore, the new PC being the latest hardware, it will only run Windows 10. Thank you Microsoft and thePC manufacturers for keeping us in mind that we might need a better OS (like Windows 7 64 bit and a non-UEFI BIOS - wish that the manufacturers would get them figured out and running properly!)...
So, Windows 10 is installed, hardware drivers are installed and working. And Project Magenta HATES it! Crashing and burning small blazes... (marshmallows anyone?) back to square one with the glass cockpit. At least, the EICAS is working so that I can make sure that all of the screens work (again).
One of the radio displays is dead... thank goodness for Amazon and the Raspberry Pi movement... tiny high quality displays are cheap! That's fixed.
All of the computers are working. Prepar3d is running. The instructor station works. Found the lesson plans for the system and moved them over to the new computers for the customer. Old computers are off line and set up for recycling.
Now that all of the hardware is properly running and we know where everything is...
Project Magenta Update
Time to call Enrico at Project Magenta and get the glass and logic working. Working with Enrico was a pleasure. Even with the 6 hour time difference and only being able to remote into the computers, he managed to get the last little items to drop into place over the course of a week.
The last items that we have to get working is the auto-pilot and the radios...
The phidgets for the auto-pilot and the radios have to be connected to the computer that the software is running on. At this point, every phidget is connected to the main server computer. But, while the phidgets can run over a network, Project Magenta's software to talk to them does not. So, I had to figure out which phidget was running what set of buttons, split them off to the correct computer, copy the phidget file to the correct directory on the computer and reboot everything.
Enrico checked all the phidget files and phidgets (I was pushing buttons). The auto-pilot is running! The radios were next and that took about 5 minutes to start up and run. The radios are running!
Documentation... and the lack thereof
In all, if we had documentation (which we have now), fixing this simulator would have been so much easier. I cannot stress enough that simulators need to be documented and any changes or fixes to the system need to be documented. This documentation needs to stay with the system where ever they go. While the documentation for Project Magenta was good, it was not enough to remotely get the system running. There was nothing about phidget files. The documentation for FSUIPC is just.... And Paradigm... well, I found them on LinkedIn and tried to connect with them... nothing but crickets. Not even a hello.
So, in the end, the simulator is working. We flew the CRJ simulator from Orlando to Chicago on auto-pilot and landed. My six year old even flew it. It was a blast and I look forward to fixing many more in the future.
Next time... Installation on site of the CRJ!
For more information on our products and services, please head over to our web site at www.servos.com. For more immediate help, click on the button below and fill out the form and someone will contact you ASAP.