New releases: Geneva professional and OnAir: Airline Manager

Jump to content
nobicus

FSX and Max's CAISET gauges

Recommended Posts

The polar that my vario uses is embedded in the gauge as half-a-dozen speed/sink pairs.

Hi Ian,

RW glide computer use frequently a parabolic approximation of the polar curve with three coefficients, a,b, and c. These are found for many RW gliders on the web, or in polar libraries of glide computers.

vs = a * TAS^2 + b*TAS + c

Afaik, the best STF is usually found by finding the tangent from the origin to the polar curve. Influence of metereological lift, wind, and McCready setting are considered by shifting the polar curve up and down by current lift and, additionally, down by the selected McCready number. For considering wind conditions the curve is shifted left (head wind, positive), or right (tailwind, negative), so an "effective polar" may be defined as (v: ground speed):

vs'' = a * (v+w)^2 + b*(v+w) + c + s + mc

Then the tangent from the origin to the effective polare curve has to be found. The point defines the ground speed to fly. The tangent point is ruled by:

dvs''/dv = vs''/v

For the parabolic approximation this can be solved as (s: meterolical sink, mc: McCready number):

vE'' = sqrt( (a*w^2 + b*w + c + s + mc) / a)

Head wind (w) is usually guestimation, s (metereological sink) can be retreived by subtracting the polar sink from the total sink shown by the VSI. The result is the ground speed, so STF has to be adjusted by wind:

STF = vE + w

There is another option, if only a speed command increase, or decrease is required. The critical determination of the metereological lift is not necessary. Instead, the current total sink (vs) is used. The optimal speed to fly is established, if the tangent equation is fulfilled. Thereby only the left side is evaluated with the polar coefficients. The right side is shown by the vario and ASI (with wind estimation):

2a*TAS + b = (vs+s+mc)/(TAS-w)

-->

2a*TAS^2 + b*TAS - w*(2a*TAS + b ) = vs+s+mc

Unfortunately, there is the product TAS*w, which prevents from expressing the wind influence independently form the other values as another offset to the sink speed. Basically this is the formula which is implemented by a McCready ring. You know, that wind is normally considered by offsetting the ring by some value.

For zero-lift, the correct extra McCready-Offset can be determined by extrapolating the glide ratio for the best STF backward by the amount of wind speed.

2a*TAS^2 + b*TAS = vs+s+mc + mcw

Cheers,

Peter

post-24548-034239400 1294071486_thumb.pn

Share this post


Link to post

What is the method for storing multiple tasks in the CAISet? (or is this RW feature not implemented?)

I have this in my GPS-NAV.dat (partial)

0066,44:11:26N,077:18:54W,320F,TA,Belleville  ,Gravel        

0067,44:11:30N,077:19:38W,3937F,S,BlvlSTART  ,

0068,44:19:13N,077:19:46W,410F,T,Scuttlehole ,

0069,44:24:12N,077:45:34W,630F,TA,Campbellford,Turf        

0070,44:11:05N,077:06:28W,298F,TA,Tyendinaga,Hard     

0071,43:00:17N,082:18:56W,593F,TA,Heritage,Hard        

0072,42:54:39N,082:31:43W,649F,TA,St Clair CO ,Hard        

0073,42:36:46N,082:49:54W,577F,TA,Selfridge AB,Hard

0074,42:35:26N,082:22:37W,600F,T,WWorld      , 

0075,42:48:20N,082:27:58W,600F,T,HPlant      , 

A0,0067, 123.4,,,,

A1,0068,  14.3,,,,

A2,0069,  35.4,,,,

A3,0070,  57.2,,,,

A4,0066,  16.5,,,,

B0,0071, 149.1,,,,

B1,0072,  20.2,,,,

B2,0073,  41.3,,,,

B3,0074,  37.3,,,,

B4,0075,  25.0,,,,

B5,0071,  25.3,,,,

It allows me to select Task A, no problem, but Tasks B to J show nothing, and selecting Task B just defaults back to Task A after going through the Declare? Start? clicks.

Share this post


Link to post

What is the method for storing multiple tasks in the CAISet? (or is this RW feature not implemented?)

It allows me to select Task A, no problem, but Tasks B to J show nothing, and selecting Task B just defaults back to Task A after going through the Declare? Start? clicks.

Just give me a couple of hours to get my C/C++ compiler running (I've just downloaded a new one) and I might be able to give you some real answers... :)

:cheers_s:

Share this post


Link to post

Thanks very much Peter - I've read through your post with interest - I'm aware of the use of quadrilatrals to approximate polars and it was helpful to see your dvs''/dv = vs''/v solution which was enough to prompt me to wade though Reichmann where he ends up with Vstf = sqrt((c + airmass_sink - avg_climb_rate)/a). I can see how the quadrilateral polar makes some of the computation simpler. There is an important practical issue that it easier to tweak an FSX polar to get a netto gauge to read zero thoughout the range by adjusting individual sample points than it would be in a real glider. I.e. I turn weather off, fly the FSX LS8-18 at 100 knots, see the error in the netto, and adjust the polar at that point by that amount. You can get the values right in pretty much one pass even when the polar isn't a parabola.

You lost me where you included head/tail wind in the STF calculation though, as afaik the wind has no bearing on the stf calculation (ignoring a last glide to a turnpoint or a landing out). It's the first time I've heard mention of including the wind - what's the rationale? I have Riechmann's book and don't see it in there. The physical rationale to ignore the wind is that the optimum STF moves you through the airmass at the maximum overall speed, irrespective of whether there's a headwind or tailwind, and you won't improve on that by (say) cruising faster because there's a headwind.

I used the ( maccready + airmass sink) adjusted polar with the tangent to derive the STF, but thanks for your comments on that approach. I actually tested this with a ruler on a drawn out polar which shows there is quite a wide range of acceptable speeds to fly for a given (sink+maccready) that result in a similar cross-country speed, and given that classic MacCready doesn't take into account the lower probabilty of finding a thermal the faster you fly I've calibrated the LS8-18 stf profile to be fairly conservative in recommended speed (with no penalty on cross-country speed).

Cheers

B21

Share this post


Link to post

You lost me where you included head/tail wind in the STF calculation though, as afaik the wind has no bearing on the stf calculation (ignoring a last glide to a turnpoint or a landing out).

I learned speeding up in headwind almost 30 years ago, while acquiring my gliding license. Don't know why Reichman didn't had this, but his book is older actually.

If you do the mathematics with the parabolic (not sure about the meaning of quadriliteral), approximation, you find out that actually you have to speed up in headwind even to have to best gliding range (not only for timing). It can also be deduced by the graphical approach. The tangent to the sink-vs-ground-speed polar curve has its best angle at high speed, because of the bending of the curve.

Indeed, frequently there is a little sensitivity on varations of the speed as the tangent and the polar are close together over a substantial speed range.

Cheers,

Peter

Share this post


Link to post

I learned speeding up in headwind almost 30 years ago, while acquiring my gliding license. Don't know why Reichman didn't had this, but his book is older actually...

It's mentioned in my copy, 4th printing 1988:

In Part I on page 57 (adjusting the ring for wind)

And in Part II, page 97ff. (moving the polar to account for wind)

However, he does point out (as did Ian) that this only applies when you are trying to maximize distance, not speed. (Page 99) If you are trying to maximize speed, then wind can be ignored, "Since the best cruise airspeed will result in the best groundspeed as well".

Share this post


Link to post

just to make sure: mcw should read like mc + w ?

Hi Max,

great to see you hanging around here again. mcw is the extra McCready offset that accomodates for the wind, which is not mc+w. See in the diagram the blue set of lines. Where the extrapolation intersects on the left side the axis with TAS=0 (for headwind this is where GS is already negative), there you can read the offset in m/s.

Obviously the offset will depend on the actual sink and the McCready number for expected lift (s+mc). The latter is known and could considered, however the first one changes all the time, so you could only reasonable assume a zero average lift and hope for the rest.

We used to have graphical glide computers (circular slide rules) matched to a particular aircraft, where the McCready offset was shown against the headwind component and McCready setting (assuming average lift or zero during glide).

Of course, a digital glide computer could do more sophisticated things.

Cheers,

Peter

Share this post


Link to post

Hi Peter,

thanks for making it clear - yes, it's fun to dig out FSX again, installing some add-ons (like CumulusX and WinchX - thanks for those too) and trying to soar again.

I was rather suprised to see that old CAI set is still being used and thought that perhaps it was time for an update. It would be nice to finally get the bugs out.

So far I've dug out my Reichmann books - although my German is a little rusty, rediscovered those old source files,

installed a new compiler and right now I'm trying to understand what I've done 10 years ago... :blink:

Been dreaming about a virtual Cambridge 302 connected to a real Pocket-PC already. :rolleyes:

Cheers,

Max.

Share this post


Link to post

Hey Max it is good to see you're thinking about flightsim gauges again - your LNAV is a modern classic...

B21

Share this post


Link to post

Hi Ian,

it's fun indeed, if only the Visual C++ Express compiler 2010 would be a little user friendly...

LS-18%20screenshot%202.jpg

But it's nice to see what a change of font can do. Only problem is it ain't a default Windows font.

:cheers_s:

Share this post


Link to post

What is the method for storing multiple tasks in the CAISet? (or is this RW feature not implemented?)

...

It allows me to select Task A, no problem, but Tasks B to J show nothing, and selecting Task B just defaults back to Task A after going through the Declare? Start? clicks.

Hi Bulau,

(btw.: is Bulau your first name or just a nick? First names are so much nicer to use here...)

the answer to this one is: no, this hasn't been implemented yet. The reason for that is a simple one: the Viewer creates only one task.

I'm sure I can put multiple tasks in the CAIset (rather easily) but what about the Viewer ? Copy and paste ain't my idea of easy setting up.

The most urgent to me seems to be the matter of "right" virtual polars - it would be great if we could solve that one.

Let me know what you would consider the most important issue(s) right now.

:cheers_s:

Share this post


Link to post

Hi Max,

before you start with this have a look at:

http://www.soaringpi...GPS_LOGpage.htm

and:

http://www.xcsoar.org/

All what is needed for those is a GPSOUT facility for FSX, as FSUIPC, or Xconn (which does not work with V64).

Cheers,

Peter

Hi Peter,

YES - that's exactly what I was dreaming about.

These are even better than the Cambridge solution - they are free.

Now all I have to do is to connect my brothers iPaq to the GPS-NAV or a Cambridge 302.

Thanks a lot.

Share this post


Link to post

...is Bulau your first name or just a nick? First names are so much nicer to use here...

Bulau is screen name. Real name is David Billo

I'm sure I can put multiple tasks in the CAIset (rather easily) but what about the Viewer ?

Not sure I understand the question. The Viewer seems to only handle one task at a time. I'm just learning how to use these tools, so was just seeing if that feature would actually work. My thought was to make up a master .dat file for each soaring region I would fly, and have it preloaded with 10 tasks of my choosing (by cut and paste method). You are right, though, that it does not take much time to set a task in the viewer and send to to the CAISet. So, not a priority.

The most urgent to me seems to be the matter of "right" virtual polars - it would be great if we could solve that one.

Let me know what you would consider the most important issue(s) right now.

I'm completely new to FSX soaring, so probably not the best person to ask. I come from a couple of years of Condor Soaring Sim, so I'm used to having everything there in terms of soaring software, PDA, moving map, STF, Arrival Height, etc., all automatically adjusted for ballast on board...maybe a tad unrealistic, I don't know.

I'm following with great interest the work of Ian and now, your renewed interest in the CAISet. I would agree that the polar issue is the priority...we need correct virtual polars, and a simple, reliable way to handle them. In fact, I'd be quite happy with what the CAISet already provides if I could count on making, or getting, a correct virtual polar for it. It shows everything I need in clear, easy-to-read format. Beyond that, anything else you or Ian produce is going to be icing on the cake.

Those who have been at the FSX soaring game for many years may have different thoughts.

Share this post


Link to post

Re: XCSoar:

YES - that's exactly what I was dreaming about.

These are even better than the Cambridge solution - they are free.

Now all I have to do is to connect my brothers iPaq to the GPS-NAV or a Cambridge 302.

I used XCSoar with Condor, but not on a PDA, rather the PC version, running on a second monitor. I was trying it because Condor currently does not handle AAT tasks.

This reminds me of another class of problems with FSX soaring tasks, and those are concerned with creating and sharing the task in the first place, and then to score it afterwards. I learned from Ian that FSX Missions were tried at first, then just simply using the saved Flight Plan (four files which can be distributed) scoring by submitted IGC. I've tried the latter, using Plan-G for the initial task layout, and it works fine...easy enough to distribute the task.

The CAISet offers another method, working from a .dat file, setting a task in the Viewer, send it to CAISet, and distribute the resulting GPS-NAV.dat file. This method also works fine, and I tried it just last night with my friend in a multiplayer session. Some added difficulty presented itself because we set a task using our own turnpoints, which are not in any .dat file. This was handled with Plan-G, followed by a little cut-and-paste and hand editing into a custom .dat file, loading that into the Viewer, etc. It would be nice if there were a simpler route from Plan-G to a .dat file, e.g. if the Viewer could load either a Plan-G file, or an FSX .PLN and then save that to CAISet, or if Plan-G could load and save .dat files. I don't know...just thinking out loud...

Some other aspects of typical soaring tasks which are not currently handled well in FSX, although they can be dealt with at the IGC level in the software that displays them:

1. Start line, or start sector, and its dimensions

2. Maximum start altitude

3. Turn point shape, (FAI sector or cylinder) and dimensions

4. Finish line, sector, or circle and its dimensions

5. Minimum finish altitude.

Here, my inexperience with RW soaring tasks hinders me. All of the above is handled very simply in the Condor Flight Planner, but not AFAIK in FSX, except maybe the maximum start height, which can be set as the elevation of the starting waypoint. Both the FSX and CAISet GPS seem to have a default waypoint, which is a 500m radius cylinder. What does a RW soaring pilot do, using the Cambridge instruments, to account for a 3km long start line, perpendicular to the course line? Modern PDA soaring software handles it, but in FSX all starts are 500m radius cylinder, same for finish.

Sorry to ramble, but this is where I'm at with my very limited experience in FSX soaring tasks. I offer this, not so much as request, but for discussion.

Share this post


Link to post

Hi Dave,

Not sure I understand the question. The Viewer seems to only handle one task at a time. I'm just learning how to use these tools, so was just seeing if that feature would actually work. My thought was to make up a master .dat file for each soaring region I would fly, and have it preloaded with 10 tasks of my choosing (by cut and paste method). You are right, though, that it does not take much time to set a task in the viewer and send to to the CAISet. So, not a priority.

Because the Viewer can only setup one task is exactly the reason why I didn't develop the GPS-NAV to handle more than one task. It's just that no one ever asked me to include more than one task. So I think your idea is a very good one - I'll keep it mind.

I'm following with great interest the work of Ian and now, your renewed interest in the CAISet. I would agree that the polar issue is the priority...we need correct virtual polars, and a simple, reliable way to handle them. In fact, I'd be quite happy with what the CAISet already provides if I could count on making, or getting, a correct virtual polar for it. It shows everything I need in clear, easy-to-read format. Beyond that, anything else you or Ian produce is going to be icing on the cake.

I'm working on that one - it's a complex gauge that's placed on a test panel - you just paste it to any sailplane you want to fly and you fly it in a situation without weather and very low simulator settings - let it fly by itself - handsfree - and voila - after 10 minutes your polar is ready. I'll have a working prototype ready in a couple of days...

Cheers.

And yes - that XCSoar looks perfect for what I have in mind too...

Share this post


Link to post

Here's some cake icing on the LS8-18, coming soon. This is pretty much the final graphics. Programmng done. Just testing :

ls8-18_panel.png

1. Winter Vario - Total Energy

2. Slave vario to the 302 displaying Netto

3. 302 computer/vario: needle displaying Speed-to-fly, digital data is Arrival Height, MacCready setting, Climb Average. Click at 9-o'clock gives L/D ratio.

4. Airspeed Indicator. Click face to have a 'speed-to-fly' indication appear on the rim.

5. GPSNAV, improved for readability

6. Moving map display with flightplan

7. Altimeter

8. Ballast % indicator

9. Gear up/down indicator

10. Trim indicator

11. Radio

Plus I've implemented a new TrimWIzard (press trigger on joystick (i.e. brake) resets the trim to current joystick position). All instruments metric/imperial depending on FSX settings.

B21

Share this post


Link to post

Here's some cake icing on the LS8-18, coming soon...

Yum!

Share this post


Link to post

Hi,

I could do with a little help right now... B)

I decided to try and update the Caiset for FSX. So I dug out the old source code and run it through the compiler.

And mr-fs9-caiset.gau changed to mr-fsx-caiset.dll - here's a download link: http://www.xs4all.nl...-fsx-caiset.dll

You'll need a new font too: http://www.xs4all.nl...wnloads/LCD.TTF . Please download and install.

Put the .dll into the FSX \ Gauges folder and please use Wofgang's LS8-18 to test it.

Change the entries in the panel.cfg from


gauge09=mr-fs9-caiset!lnav,   441,484,139,142

gauge10=mr-fs9-caiset!gpsnav, 456,353,116,117

gauge11=mr-fs9-caiset!caf,    346,369,100
to

gauge09=mr-fsx-caiset!lnav,   441,484,139,142

gauge10=mr-fsx-caiset!gpsnav, 456,353,116,117

gauge11=mr-fsx-caiset!caf,    346,369,100

like this.

The problem: it's reported the gauges won't show up on Win 7 and I need some confirmation.

As I still run Windows XP and don't have access to a Win 7 PC I'm up the creek. :rolleyes:

Cheers,

Max.

btw: those without Win7 are welcome to the update too. :D

M.

Share this post


Link to post

They worked for me in Windows 7 64bit, and I didn't have to load up the Font cause it was already there, except for an error I ran upon while just tinkering with the L-NAV buttons (I have to be honest, I just use this guage in it's basic form or less)...

lnaverrormessage.jpg

It happens when I hit the right arrow on the L-NAV about about the 5th-6th time (Compensation or Std. Vario).

Scott

Share this post


Link to post

Hi Max,

The .dll works fine on my system with W7 32-bit. I didn't need to install the LCD-font.

I cannot reproduce the error that Scott reported earlier on his system (I tried it several times and everything kept on working, also wheb I switched from std. vario to slow and/or fast and back).

Bert

Share this post


Link to post

Interesting - it was me that tried the DLL with no success on my 2 PC's - both my machines are Win 74 64 bit - I wonder what's different about my systems ??? One is FSX Deluxe (whatever that is) upgraded SP1->SP2 and the other is FSX SP1->SP2->remove SP2->Acceleration, plus the FSX SDK. Both machines have VC++ 2008 in case that makes a difference. Both machines running the mr-fs9-caiset ok.

I've tried running FSX in admin mode or not - makes no difference - the mr-fsx-caiset gauges just don't appear on the panel (although FSX asks me to 'trust' the new DLL, so it's finding it ok).

Is there anything wierd about Win 7 keeping track of executables it has downloaded from the internet, the way 'exe' files sometimes always prompt to check it's ok to run ??

B21

Share this post


Link to post

Thanks guys,

the score is 2 - 2. What I don't understand is that both sets use the same code and where the fs9 set don't give any problems at all the fsx set has difficulties. Somehow I'm under the impression the last build is less stable but now I'm not so sure.

Scott's debug error is just a program error and quite easy to solve.

I would like to have Ian's problems solved before I continue...

The only way I could make the gauges disappear on Windows XP was by removing the SOARrec.dll from it's folder.

Cheers,

Max.

Share this post


Link to post
Guest
This topic is now closed to further replies.

×
×
  • Create New...