Quote from Theafro :

BTW mine's set at 540, it's a nice compromise for all the cars in LFS, te single seaters just feel a bit wrong at 900/720

I use different settings for each class of car...720 for road cars, 540 for GTRs, 450 for single-seaters and 270 for the MRT. That matches the in-game steering for all those cars.

Right, I'm about to tear my hair out...
I tried my GFX card in a friend's PC (replacing his 256MB FX5500) and it worked fine on both the games he tried. He has a 400W PSU in his PC so I reckoned on getting the same for my PC. As it turned out when I got to Maplin, they had a special offer on a 500W PSU in the same range as the 400W I was going to get, so I got the 500W. Also got a new case to improve cooling (was thinking about replacing the old case anyway).

Swapped all the components into the new case/PSU and tried my old GF2MX card...everything worked fine.
I then swapped in the FX5600U and...got the same weird colours and artifacts as before. If anything, the problem is worse. I have checked the temps using NVTempLogger and the core hit a maximum of 61 degrees (the new case runs 10 degrees cooler than the last one), so I doubt it's a temperature issue.
I've tried the latest 81.98 drivers and some older 77.72 drivers I had, both with the same problems. I've tried a fresh install of LFS and that didn't help. I've tried ToCA RaceDriver and that gives the same problems.
I ran DXDiag to see if there were any issues reported there, but it found nothing. The DirectX 7, 8 and 9 tests worked fine except that the textures on the spinning cubes were slightly distorted. It wouldn't let me reinstall DirectX because I already had the latest one installed.

So, any suggestions? My friend reckons that it might be a motherboard problem, the AGP port not providing enough juice to power the card. He has a point...the mobo must be at least 6 years old, but it supports AGPx4. I think the mobo is the same as the FIC AZ31 but since it was originally from a Compaq PC I can't find any information on it.
I've tried looking for new SocketA motherboards but can't find any which take SDRAM and I don't want to have to replace my RAM just at the moment.

Last edited by StewartFisher, Sat 18 Feb 2006, 20:16.

Quote from AndroidXP :

I'd really try the temp logger though, I mean, there's no harm in trying, isn't it? I left the nVidia window open just as you did, and in the time it took windows to render the desktop again the card was already cooled down quite a bit.

Oh yes, I will try it when I get in tonight, I was just saying that I was measuring the temps when LFS was still running. Thanks for the suggestion

Quote from AndroidXP :

Hmm, thats not good. You should try the app "NvTempLogger" to really log your temperatures.

My 6600GT for example ran at 92° (before I replaced the crappy cooler), but in the time it took me to switch out of the application and look at the nVidia temp bar, the temperature already dropped back to about 70°C !

When I started using the templogger I finally saw what was going on and the temp graph nicely showed how fast the card cooled back down to values that seemed quite normal. Don't let yourself be fooled by the standard temp gauge, while it displays the temperaute correctly it cannot possibly show the temp while you're actually running the appilcation.

Yeah, I noticed that the card cools down very quickly after I stop running LFS but I left the temp display open and switched between full-screen and windowed mode to see the temperature while LFS was running.

Quote from AndroidXP :

Forget about the critical slowdown. You will never reach that value - your card will be roasted far before it can even become nearly as hot as that. Personally I have no idea why they set the "critical slowdown" threshold so ridiculously high, it makes no sense at all.

I was wondering about that...it seemed very high to me but I assumed they knew what they were doing...when I installed the original 44.XX drivers it was set to 134!

Quote :

From my experience, a card will still run when it's in the 90° range, but it won't do that for very long. Anything below 70° is fine.

The 65 degree temp was measured with the side of the case off, but since it looks like I'm going to have to get a new case/PSU anyway I won't sorry about trying to cool the current one.

Quote :

Btw, how did you measure the GPU temp? NvTempLogger?

I assume so...there's a screen in the Nvidia display properties section which gives GPU core temp and ambient air temp.

Quote from glyphon :

Is the card (or system) overclocked at all? texture problems and a other graphical glitches can occur on systems/cards that are overclocked too much.

No, not at all, all the speeds are set to the defaults...I'm going to try underclocking it tonight though, to see if it makes a difference.

Quote :

and you mentioned that it was only 65 degrees. i'm assuming this is in reference to the video card. even though i can't find the operating temperature range for the 5600u, this temperature is most likely in the safe range, but it is higher than most of the reviews i found. those were in the low 50s.

Yeah, the GPU temp was 65 deg, still in the green range. The yellow range starts about 70 deg and the critical slowdown is set to 120. I've put so much extra stuff into the old Compaq case I have that it runs quite hot...about 40 degrees generally.

Quote from avih :

could be faulty ram or not enough current for the board or overheating in general. also, in my 5600u there's a direct cable from the psu to the card itself. did u plug it in?

Yeah, and I had to sacrifice my DVD-ROM to do it! The cables on my PSU aren't long enough to reach the GFX card unless I unplug one of my optical drives.

Quote from TiJay :

Your PSU's not good enough. I installed a regular 5600 in my old 250W Advent and it killed the PC eventually.

I think the GFX box recommends about 320W...

I think you're probably right...I've arranged with a friend to test the card in his PC on Friday but I'm fairly convinced I'm going to have to shell out for a new PSU. Trouble is, I think my PSU is riveted, rather than bolted, into the case...looks like my 'cheap' graphics update is going to turn out to be rather more expensive!
I did look on the box to see if there was a recommended power output but didn't find one.

I could try that, yeah. I am using DirectX 9.0c already but reinstalling might help, thanks.

Quote from XCNuse :

hm.. it may be something wrong with the card.. ask wherever you got it from that theres .. obviously something wrong with it, and if you get another one, if it does the same then we can say that its definetly not the card, but rather something else

I might ask him about it but the fact that I've seen other people with similar problems (with brand new cards) makes me think that it's a non-gfx card related problem.

Quote from AndroidXP :

I dunno, have you tried any other drivers? The newest aren't always the best for older cards, just try a few different drivers before you go out and buy a new PSU. I've had such artifacts with my older cards too, and IIRC these were always driver related.

I've just downloaded the latest drivers from work (slow connection at home!)...I'll try those as soon as I can. I've had similar problems which I've cured with drivers so we'll see what happens.

Quote from avih :

i have fx5600 ultra too. i use drivers 81.87 but it worked well with any previous driver too. try to open the driver settings (desktop -> right-click -> settings -> advanced ->gforce 5600 ultra -> performance and quality ) and disable anti aniasing ans anisotropic filters. then turn off all optimizations and set the image settings to "performance". see if it helps.

I've tried all that, it makes no difference Could you give me some information about your hardware, since you've got the card working?

Quote from the_angry_angel :

Consider downloading another copy of LFS, if you can, just to test. It looks almost as if all the textures have some how become screwed - as unlikely as that is.

The problem is quite intermittent though...the screen flickers between that shot I attached and a normal view. Sometimes only the trees go funny, sometimes the whole screen, sometimes just the fonts, occasionally it all looks normal, then it will all go weird again, etc... In any case, it happens in any game with 3D graphics.

Quote :

It is possible that there not enough current getting to the gfx card causing its memory to blip causing random-ness, which is possible as the audigy can suck up a lot of power...

I tried unplugging everything I could from inside the case but it didn't seem to make a difference.

Quote :

See if you can pop it into a mates machine for a little bit, see if it does it in anything else funky.

Yeah, I tried to get hold of a friend last night to see if I could try it but he wasn't around...will have to try again!

Thanks to all for the replies, I'll keep digging around

Last edited by StewartFisher, Mon 13 Feb 2006, 08:30.

OK, I know this card is a bit of a dinosaur. but I don@t have a great deal of money to splash around on new PC stuff at the moment. I managed to find one (practically unused) cheap and installed it, replacing an ancient GeForce 2MX.

With the original 44.XX drivers the display was fine in 2D but when I entered anything using 3D (ie LFS driving screen) it crashed and I had to restart the PC. I then installed the most up-to-date drivers I had on my HDD (77.72) and tried again. This time, a few seconds after I start LFS, the screen goes red and psychodelic. It doesn't crash though.

I've attached a screenshot of the problem in LFS, though it applies to any game which uses 3D graphics. The card isn't overheating...the temperature when this shot was taken was about 65 degrees.

I've tried to search through some newsgroups looking for answers but I get a bit lost when people talk about IRQs and mainboard voltages There seemed to be a consensus that low PSU power might be a problem but before I buy a new PSU I thought I'd check here. The current PSU is only 250W so I can accept that it's a bit weak.

System Spec:

AMD Athlon 1GHz
416MB SDRAM (256 + 128 + 32)
GeForceFX5600 Ultra (128MB)
Audigy 4 sound card
Windows XP SP2

Last edited by StewartFisher, Sat 18 Feb 2006, 18:20.

Quote from Bob Smith :

Yeah OK then, cornering is more like changing the value of g for different sides of the car than actually moving the mass. But that makes it much harder to understand the data.

So, thinking about the sprung/unsprung masses bit that Colcob brought up, wouldn't we then technically have two frequencies, which could be seen with a spectral view? Actually we'd have two moving bodies, both of which could have more than one frequency. I'm just thinking about sound, since when a speaker cone moves it's moving one path but can be moving at many different frequencies.

More thinking: if it is suspension travel is what was used to create the graphs, then even though weight transfer won't affect frequencies, it will still cause the suspension to compress on one side and extend on another (i.e. roll). So this would still affect the graph even without mass transfer occuring. Yes/No?

Since the 'suspension travel' is measuring the extension of the spring, you'd have one frequency for the vertical motion of the unsprung mass and one frequency for the vertical motion of the sprung mass. You'd also probably get frequencies corresponding to the roll and pitch frequencies as both of these will affect the ride height on an individual wheel.

Quote from Bob Smith :

The relationship is that the spring stiffness is fixed (the value you see in the garage), while the frequency depends on the mass on the spring. Weight transfer is taking off load from one spring, and putting more on another. So while the spring stiffness never changes, the frequency should drop on the additionally-loaded spring, and increase on the less-loaded spring (during cornering, this is the inside wheel).

I disagree. Following the discussion about suspension frequency change with downforce loading I went right back to the differential equations for a spring-mass-damper system and confirmed that spring load does not affect the suspension frequency. Only the mass supported by the spring and the spring rate (and, if underdamped, the damping constant) change the suspension frequency. During cornering the weight distribution changes but the mass distribution doesn't. This is ignoring change of CoG with chassis roll obviously, but that should be fairly small.

I did all the cars a while ago after Bob told me where to find the numbers...here's what I got:

UF1000: 160/50 R12
XF GTi: 185/50 R15
XR GT: 185/55 R16 front, 220/50 R16 rear.
XR GT Turbo: 225/45 R17 front, 245/40 R17 rear
RB4 GT: 215/40 R17
FXO Turbo: 240/35 R17
LX 4: 195/60 R13 front, 215/55 R13 rear
LX 6: 205/50 R15 front, 245/40 R15 rear
RaceAbout: 205/45 R17 front, 225/45 R17 rear
FZ50: 235/40 R18 front, 295/30 R18 rear
MRT5: 165/55 R13
XF GTR: 215/35 R16
UF GTR: 215/45 R11
Formula XR: 225/50 R13 front, 285/45 R13 rear
Formula V8: 275/55 R13 front, 390/40 R13 rear
FXO GTR: 335/35 R18
XR GTR: 335/30 R19 front, 335/35 R18 rear
FZ50 GTR: 300/35 R18 front, 370/30 R18 rear

Sizes are given as width/profile and rim diameter in millimetres percent of section and inches respectively.

One thing you might want to be careful of is the change of Reynolds Number caused by the change of fluid.
The equation for aerodynamic (or hydrodynamic) drag is
F = 1/2 * rho * Cd * A * V^2
Moving from air to water changes the value of 'rho' from 1.225 kg/m3 to about 1000 kg/m3 but there might also be a change in the value of Cd.

Reynolds Number is defined as
Re = (rho * v * length) / mu
where 'mu' is the dynamic viscosity of the fluid. If you look at a graph of Cd vs Re for a sphere there is quite a variation due to the difference between laminar and turbulent flow and the resulting boundary layer behaviour. If the Reynolds Number is high enough then Cd is pretty much constant but at low Re it might vary by a large amount.
Oh...I see FrankMD has already mentioned this...ah well, since I've already written it I might as well post it
Incidentally, Frank is also correct about air being incompressible below Mach 0.3. By the time you get to Mach 0.7 air is very definately compressible.

Quote from Scawen :

About pressures - what problem have you experienced?

I've got :

NM2_per_PSI = 6897.1126
PSI_PER_NM2 = 0.000144988
PSI_PER_BAR = 14.7

atmospheric pressure = NM2_PER_PSI * PSI_PER_BAR

I don't know exactly what problem Bob has seen but you're assuming that atmospheric pressure is 1 bar (100000 N/m2). The accepted atmospheric pressure at sea level is 101325 N/m2 (1.01325 bar). Your numbers put 1 bar at 101387.5552 N/m2 which is wrong.
ConvertIt.com gives the following values for the conversions:

(N/m2)/psi = 6894.75729316836
psi/bar = 14.5037737730209

Which gives 1 bar = 99999.99998 N/m2 (almost exact!)

This wing works because in a finite-span wing there are vortices shed at the tips of the wing in counter-rotating directions. Have a look at the rear wing tips on F1 cars in damp conditions and you'll see condensation in the vortices. These vortices rotate to cause an upwash at the centre of the wing and a downwash outboard of the wing tips. By splitting the wing in two you create a downwash in the centre, hence the 'centreline downwash generating wing' name.

However, there will now be two upwash zones at the centre of each of the smaller wings. Since the aspect ratio of the wing will be reduced (shorter span for the same chord) the upwash will be a lot stronger than that currently seen in F1. So the whole picture has a downwash along the car centreline then two strong upwash zones directly behind each wing then a downwash zone outboard of each wing. This would seem to me to be a much worse aerodynamic condition than having a single rear wing. Especially when you consider that, according to those crude diagrams, only the centre section of the front wing will benefit from the downwash. The outboard thirds will be in the upwash from the two rear wings. I'd really like to see a more complete CFD analysis of the new wing before I said I'd be happy with it. A single stagnation pressure distribution along the centreline really doesn't show much. Especially since it looks to me to be a highly sanitised 'show them the benefits but none of the drawbacks' diagram.

There's a thread active in this forum at the moment in which Bob Smith has just explained how to calculate the suspension frequency of a car:

http://lfsforum.net/showthread.php?t=471

The whole thread is quite informative but can be summed up by saying that downforce levels don't affect the suspension frequency...the bits you probably want are currently at the end of the second page.

Dunno about Bob (I assume he'll be along soon enough ) but I've calculated some of the unsprung masses using RAFs.
Basically, park the car on a flat place and leave it to settle, recording the fuel load. You can then export the 'Tyre Load' values from the RAF using F1PerfView. These seem to be spring loads rather than tyre loads so adding them (and dividing by 9.81) gives you the total sprung mass of the car. Subtract the fuel mass (using fuel tank size and density of fuel = 0.74 kg/litre) and you have the empty sprung mass. The difference between this and the LFS total mass is the unsprung mass
Here is what I've got so far...hope Bob can confirm this, otherwise one of us will look silly

XF GTi: 50.9kg
XR GT: 60.2
XR GT Turbo: 67.6
RB4 GT: 61.1
FXO Turbo: 64.6
LX4: 51.7
LX6: 57.7
Raceabout: 63.9
Formula XR: 59.7
Formula V8: 79.0
FXO GTR: 93.9

Hope that helps...I've not done all the cars yet!

Last edited by StewartFisher, Fri 14 Oct 2005, 00:34.

I've found the relevant byte offsets in the RAF output (using the description of the file format on the LFS website) but I can't convert the 4-byte hex codes into decimal. I assume they're IEEE single-precision floating point numbers but the conversion doesn't seem to work.
For example, the FOX has wheel widths 66 66 66 3E and 85 EB 91 3E but when I convert those to decimal I get 2.72x10^23 and -2.22x10^-35.
The converter I'm using is this one:
http://www.h-schmidt.net/FloatApplet/IEEE754.html

Any ideas?

BTW, I don't doubt your numbers, Bob, I just like to know how this stuff works for myself

Quote from Bob Smith :

280/55/13 front
390/40/13 rear

Just out of interest, how did you get this data?

Ah, I see...makes sense. I got confused because the only diagram I could find supposedly showing multi-link suspension looked suspiciously like double-wishbones to me!

Quote from Shotglass :

some touring cars (the gtrs in lfs) have multilink-suspensions and i cant say for sure but just guessing from the sheer amount and geometry of the wishbones formula cars use these days i assume those are multilink-suspensions too

There must be something I'm not understanding here...are double-wishbone suspensions not classed as 'multi-link'?

Sinbad, I think your definition of random is slightly different to mine (and geeman's).
Of course there is a reason for every failure, but the way you analyse manufacturing defects is through statistics, hence you can define the probability of a part failing in normal use without it having been abused by the driver. Those are the sort of 'random' failures people are talking about.

I don't want to see this in LFS because it would make the game pointlessly irritating. I really would like to see driver-influenced failures though.

Quote from Shotglass :

the first part of this suggestion should be fairly easy to implement with the current physics engine of lfs as we already have fully simulated double wishbone suspensions and altering their geometry a little in the setup would be an interesting addition (always thought it was a thing only rc cars had but looking at rbr apparently real cars have that setup option too)

I would imagine that this adjustment is more to do with the design of the car than the setup, but I could be wrong.

Quote :

and the second part is something that would most probably require a lot of work (my guess is its already on scawens list anyway) but since lots of race cars use multilink suspension and judging by the look of the formulas suspensions we already have at least two multilink-suspension cars in lfs a proper simulation of that suspension type would be appreciated

I don't really understand this. The suspension models in LFS are pretty much what you'd expect for each of the cars. Here's what they use:

UF 1000, XF GTi, FXO Turbo, XF GTR, UF GTR: McPherson Strut front, trailing arm rear.
XR GT, XR GT Turbo, RB4: McPherson Strut front and rear.
FZ50: McPherson Strut front, double wishbone rear.
All others: double wishbone front and rear.

What's missing from this model?

Quote from xaotik :

Nope - go to Fern Bay Club, damage your car and then drive off straight after the first chicane (where the ads are), you'll get dropped back into track with the same car state (tyres/damage).

Correct, but if you hit space (or whatever key you've assigned) your car will be reset with fresh tyres and all damage fixed. The automatic car reset works differently to the manual one.