Board index FlightGear Development Aircraft

CRJ700 family goes canvas

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: CRJ700 family goes canvas

Postby tdammers » Tue Nov 27, 2018 8:22 pm

it0uchpods wrote in Tue Nov 27, 2018 7:46 pm:Hi tdammers,
you are still missing this commit in master: https://github.com/tdammers/CRJ700-fami ... 1a4323f8ac

Merging from nl256 or cherry picking into master will work.

J


It appears that you are right. I think I forgot to pull in the nl256 branch from github before merging into master. Should be fixed now.

Now putting my teeth into the standby instrument; it's kind of useless right now because it drifts way too fast, and I want the "CAGE" button to do the right thing (set the attitude indicator to "level").
tdammers
 
Posts: 293
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: CRJ700 family goes canvas

Postby Octal450 » Tue Nov 27, 2018 9:34 pm

Are you sure that the CAGE button should do that? then it would be off, as the CRJ does not sit level on the ground. In the Airbus, it aligns it to current attitude when pull the knob. (in the old AI at least, the new IESI automatically aligns to current att)

Josh
What I do: Flight Dynamics, Systems, Canvas Displays, Autoflight, FlyByWire, Cockpit Animations
Aircraft I currently develop: MD-11, A320-family, Secret

My GitHub|MD-11 and ITAF Development Discord|Airbus Development Discord
User avatar
Octal450
 
Posts: 4537
Joined: Tue Oct 06, 2015 12:51 pm
Callsign: WTF411
Version: next
OS: Windows 10 x64

Re: CRJ700 family goes canvas

Postby tdammers » Wed Nov 28, 2018 12:38 pm

it0uchpods wrote in Tue Nov 27, 2018 9:34 pm:Are you sure that the CAGE button should do that? then it would be off, as the CRJ does not sit level on the ground. In the Airbus, it aligns it to current attitude when pull the knob. (in the old AI at least, the new IESI automatically aligns to current att)

Josh


I don't actually know what it does in the real aircraft; the current behavior seems wrong in any case.

A comment in the documentation mentions that in the real CRJ, you have to keep the aircraft level while caging the standby gyros, so from that I concluded that the CAGE button will simply calibrate the attitude indicator to a level position. This makes sense in that it would allow you to still calibrate the standby instrument after a main instrumentation failure, as long as you have a visual horizon reference available.

Further googling unearths this: https://www.studyblue.com/notes/note/n/ ... k/11048093

Assuming that the standby instruments in the 700 and 200 are more or less the same, this would suggest that the CAGE button would align with the horizon as indicated by the other attitude indicators, but because the standby gyros are subject to acceleration-induced errors, the calibration should be done in level, unaccelerated flight. So you are probably right; CAGE should just reset the attitude indicator.

So the question is what a realistic error model would be. Apparently the standby instrument uses piezoelectric gyros, not vacuum-driven mechanical ones, so I guess it might not drift significantly. It will, however, be off in accelerated or decelerated flight; and depending on the types of electrical failures, it might stop working anywhere between 0 and 11 minutes.

So maybe what we should do is this:

- A property rule that applies the acceleration-based error on the fly (add scaled longitudinal acceleration to pitch and roll indications)
- A Nasal job that periodically adds a small amount of drift to the standby gyros
- A binding for the CAGE button that resets the standby instrument's drift to negative acceleration error (simulating what would happen if you cage the instrument during accelerated flight)

And on top of that, we might want to model electric failures for the gyros, treating the standby instrument separately (i.e., keeping it available longer than the PFDs).
tdammers
 
Posts: 293
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: CRJ700 family goes canvas

Postby D-ECHO » Tue Dec 31, 2019 10:20 am

Hey,
I just wanted to ask whether the CRJ700-family is still actively developed. I think about porting it to canvas (which has already been partially done https://github.com/DanielMeissner/CRJ70 ... -efis-beta ) and to JSB (I think this has not been started yet). But considering there are 9 forks on Github with each having a huge load of branches and different developments, I'm kind of confused as to who still actively develops what.
Regards and thanks to all those that have already made this an incredibly great model :)
User avatar
D-ECHO
 
Posts: 1958
Joined: Sat May 09, 2015 12:31 pm

Re: CRJ700 family goes canvas

Postby tdammers » Tue Dec 31, 2019 4:40 pm

I'm still kind of working on it, though I haven't done anything in the past couple of months. I tried consolidating those branches at some point, but quickly got into merge hell; I have since given up on the canvas branch and the JSB attempts and instead just kept using what I have. It serves my needs well, and I think what's unique about it is that the autopilot works well, including SPD CLB and SPD DSC modes and turn anticipation; I've also put quite a bit of work into making the (non-canvas) MFD and PFD look better (getting rid of z-fighting and other visual artifacts). Octal did some work on the autopilot itself, made a working yaw damper, and implemented yoke backdriving. Another thing I fixed is that I changed the FADEC logic such that the FADEC mode is now set by moving the throttle lever, which I think is more in line with the real thing (lower half of the throttle range sets manual thrust between idle and max manual, upper half has CLIMB, TOGA and MAX detents). The autothrottle only affects the manual range; setting the throttle lever to IDLE, CLIMB, TOGA or MAX overrides the autothrottle (but doesn't disable it). I've also changed the elevator trim to use stab incline instead of fake trim on the elevator itself.

So I'm going to keep flying my version, unless someone braver than myself manages to merge it with changes from those other branches. Canvas ND and PFD would be lovely, at least if it's not at the expense of performance (unlike most of the other airliners, the CRJ performs rather well on my measily Intel Graphics), and I would really really appreciate a working FMC, proper air data calculations, and autopushback. There are also still some weight-related bugs in the FDM (not all models have all the weights properly implemented).
tdammers
 
Posts: 293
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: CRJ700 family goes canvas

Postby D-ECHO » Wed Jan 01, 2020 7:54 am

oof, complicated xD

But that sounds quite good. I think I'll try developing a JSBSim FDM, based on your fork of the CRJ700-family. Would you want me to do this in an own fork or should we work on the same repository?
User avatar
D-ECHO
 
Posts: 1958
Joined: Sat May 09, 2015 12:31 pm

Re: CRJ700 family goes canvas

Postby danielHL » Wed Jan 01, 2020 1:27 pm

There is a lot done that's sadly unreleased as of now. Henning has all EICAS pages perfectly done in canvas and the PFD/MFD is also coming along nicely. I suggest you get in touch with him about that - he's mainly active on the mailing list and I think you can reach him best via email.

I have a basic jsbsim fdm done started, I need to look up where my latest branch is. I'd really like some collaboration on this, if possible. Integration with the rest of the CRJ700-family is done but we aren't sure about how to use systems (general xml vs. jsbsim), where a lot of code is already there for the YASim models which could be made redundant when using jsbsim.

We really need some integration as it seems there is a lot of work done in separate branches. I still love this plane and would like to see everything integrated and a lot of things being worked on. Next to a milestone release integrating the work already done what the project needs the most is 3d/texturing work. This would bring the plane up a few levels...

Cheers,
Daniel
danielHL / GVA0387 / D-FMPW / (UA83DAN)
danielHL
 
Posts: 122
Joined: Fri May 02, 2014 6:23 pm
Callsign: UA83DAN, GVA0387
Version: git
OS: Linux

Re: CRJ700 family goes canvas

Postby D-ECHO » Wed Jan 01, 2020 4:38 pm

Hey Daniel,
great news there :)

I'll write Henning about this.

I also have a (very basic) JSBSim FDM set up by now, mostly aeromatic with as much tweaks as necessary to make it flyable, although far from good. I think it would be best if we could agree on one repository where we part by part integrate all the diverged work-pieces. Then we can also compare our JSBSim tries and cherry-pick to get the best of both :) In terms of what to keep in property rules and what to port to JSBSim, I think it would be best to put all systems related to the FDM e.g. the flight controls (CRJ700-flight-controls.xml or so) into JSBSim to have this consistent while keeping the others that only need to run as fast as the framerate in normal property rules.

Will keep you updated :)
User avatar
D-ECHO
 
Posts: 1958
Joined: Sat May 09, 2015 12:31 pm

Re: CRJ700 family goes canvas

Postby jsb » Wed Jan 01, 2020 10:45 pm

Hi all,
on vacation until 4th of Jan. Sorry for the delay but last year was kind of a mess on different levels. The EFIS is almost complete but one of my goals is to extract the canvas EFIS framework I wrote and bring it to FGDATA so others can make use of my work (last 1.5y ?)
I hope to push this next week.
Have a good start into the new year. :)
jsb
 
Posts: 249
Joined: Sat Oct 25, 2014 8:17 pm
Location: Hamburg, Germany
Callsign: D-JSB
Version: next
OS: Win7/Linux

Re: CRJ700 family goes canvas

Postby D-ECHO » Thu Jan 02, 2020 8:56 am

Good to hear :) Sounds like a good way to go.

To ease bringing the forks together, I've created a kind of "collection" repository here: https://github.com/CRJ700-family-team/CRJ700-family in the organization CRJ700-family-team where you three are added, I hope this way is fine for you. If it's okay, I'd put my JSB flight model in a branch there, maybe @Daniel you can upload yours as well so we can compare and put them together?
Have a good time
User avatar
D-ECHO
 
Posts: 1958
Joined: Sat May 09, 2015 12:31 pm

Re: CRJ700 family goes canvas

Postby tdammers » Thu Jan 02, 2020 11:49 am

I'm all in favor of joining efforts. Only reason I forked things in the first place because so little activity happened elsewhere, and I really wanted a couple of things working that were important to me. The organization is absolutely a good idea, I just joined, and I'd prefer for development to continue there. (I was never super keen on that "maintainer" role anyway, just ran my public repo because nobody else seemed to).
tdammers
 
Posts: 293
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: CRJ700 family goes canvas

Postby tdammers » Thu Jan 02, 2020 12:07 pm

Oh, and unless anyone objects, I'd like write access to the "team" repo. I promise I'll behave ;)
tdammers
 
Posts: 293
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: CRJ700 family goes canvas

Postby CaptB » Thu Jan 02, 2020 12:49 pm

I can't offer 3D help, since I am and will be busy with the A320 and other projects, but I can offer some mentoring(Blender) in that regard if anyone is interested in picking up the 3D.
CaptB
 
Posts: 565
Joined: Thu May 23, 2013 6:36 pm
Callsign: EKCH_AP
IRC name: CAPTB
Version: 2018.1
OS: Xubuntu 18.04

Re: CRJ700 family goes canvas

Postby danielHL » Thu Jan 02, 2020 1:17 pm

Okay that's good to hear. My JSBSim work can be found here: https://github.com/DanielMeissner/CRJ70 ... jsbsim-dev

The main problem with my FDM at the moment is performance. You have too much power (in excess of 5000 fpm climb on half MTOW) after takeoff, and not enough lift at maximum altitude. I suspect a reynolds-dependency on CL should fix this but other than guessing, I don't know how to do this at the moment. The rest flies quite nicely for my taste, maybe I've exaggerated on the flaps a bit but that can be easily tuned...

Echo, where is yours? I can't find a branch for it.
danielHL / GVA0387 / D-FMPW / (UA83DAN)
danielHL
 
Posts: 122
Joined: Fri May 02, 2014 6:23 pm
Callsign: UA83DAN, GVA0387
Version: git
OS: Linux

Re: CRJ700 family goes canvas

Postby D-ECHO » Thu Jan 02, 2020 2:01 pm

Hi Daniel,
I have not yet pushed mine, I will push it to the jsb-echo branch of the team repo now :)

Will test yours and report back xD
User avatar
D-ECHO
 
Posts: 1958
Joined: Sat May 09, 2015 12:31 pm

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: YandexBot [Bot] and 4 guests