SimHawk's motion simulator

Yesterday I got my valves and ran a few tests with a single speed setup.

The sim works in a way! ,but I was a little bit disappointed as I could not get the sim to follow the target value properly. :-( With low pressure (1 bar) it was not able to follow fast enough and felt too weak to toss around my body. With more pressure 2-3 bar it is strong enough, but overshoots wildly and the result is that i am thrown from end position to end postition.

I also tried an air flow choke to regulate the cylinder speed. I suspect that it has to do something with that X-sim think it has three speeds and I have mapped all to the same valve output so far since I do not have air flow choke's for all valve inputs. Profiler think it uses slow speed and can shut off the valve just before the target is reached but the sim is going far beyond do to its single speed setup. I see there are settings for how far before the target value profiler shut of the valve on each speed.

:?: Anybody has experience in tuning a pneumatic setup and could give me some info? I will try more tuning this evening.

For 3 speeds with 5/3 valves you need 4 of them.

I know. I do have four 5/3 valves connected and wired up ready for 3 speed setup, but because I lack airflow chokes that should be on two of them it would not work like intended. I'll try to make some air hose squeezers today to realize the three speed setup.

:?: Another question:
I see that profiler do not follow the target value continuosly, but needs a motion on the target value to enable the valves. Is that correct?

:idea: And a last some pneumatic thoughts.
When I realized that reducing the air pressure made the cylinders go slower but also weaker, it struck me that regulating air pressure is not the same as regulating the air flow. It should be possible to run the sim on relativly high pressure to make it strong but regulating the air flow to get the speeds slow enough.
Kind of similar to voltage/current in electronics. Am I on the wrong planet here or do this make sense? :?

EDIT:
I found this on the web explaining the best way to control an air cylinder:

You can read it here:
http://searchwarp.com/swa78890.htm

I see that I have a lot to learn!
If I understand it correcly to get an optimal setup I should restrict the exhaust air out of each cylinder instead of the inlet air and do it as close to the cylinder as posible. That means that the t-split used to join the two valves (for each speed) should be located close to the cylinder and have seperate air hoses down to the two valves. I think I need to buy some needle-check-valves! :razz:

Here is a link to a collection of pneumatic knowhow:
http://www.about-air-compressors.com/flow-controls.html

Here is the valve one should have to control cylinder speed:

You can adjust the sensivity in the Profiler (smoothness). As well as you can adjust the speed range of every speed related to the offset values.

imho you should apply the chokes on the air-intake side.

Btw, if you want a higher airflow with lower pressure you need a 3/8 system or higher.

@egoexpress - thanks for the tip, I found the settings in the profiler.
Actually I want higher pressure and lower airflow. :razz: Strong, but not catapult speed.

A video of the first tests of my SimForce :
For some reason DiRT worked alot better than LFS due to a update rate of only 2Hz on the data sent from Force Sender and LFS. Here is DiRT :


http://www.youtube.com/watch?v=1NgQwUUCoHc

What you dont see is that it do not follow the target very closely. I will look more into using needle-check valves to control speed so I can realize the three speeds. The video is showing only a single speed setup.

Cool :gut:

*hint* Without chokes you can realise at least 2 speeds

If I do not misunderstand I cannot with my setup. It is not restricted by airflow so opening one or two valves makes no difference. I tested both, but without chokes it is unnoticable difference in speed.

But what the heck - it was great fun even if it was not perfect!

Please tell me some closer what you mean. If it would be follow the target at any time, your seat would not stop shaking even when you are driving straight forward. Please note that Thanos' interface has 1024 speeds, and your pneumatic setup just 3.

Could you please make a vid of the Profiler calibration process. Does it calibrate properly and does your seat return to center after calib?

Hi egoepress. That is what I meant, but I see the problem with the seat shaking. It just becomes very visible when my seat overshoots and is far off the target value. When the target is calm after an overshoot then my seat is permanent out of position until the target has enough movement for the profiler to start regulating again.

I see that this will be a minor problem when I get my speed-control valves(next week) . Hopefully I will get rid of my big overshoots. But maybe the profiler could just check how far off the target-seat was before it stopped regulating? Could not do any harm.

I think my seat is calibrated correcly because it moves like intended, just a little too much depending on my air-pressure setting.

I calibrated the seat without the valves, just by hand. So I do not know if it returned to center. The registry values makes sense though, almost zero/255. The seat also moves like intented when moving it with the profiler sliders.

:?: Could profiler have a deadzone checking of the difference between the target and sim - not a minimum movement on the target alone?

Anyway - I think most of this trouble goes away when I innstall speed regulation valves.

It should center everytime. I saw the effect you are describing in your vid. I think it isnt a valve problem.

Either it is a wrong calibration or perhaps you selected the wrong effect numbers. Havent installed Dirt yet. Dont know which are the right ones.

You should try it with LFS with some tested effect numbers.

EDIT: Just in order to know, if your system is calibrated correctly. Put your seat in full extended position.
Click default setting and then move Simulator.

If your seat does not go to center position, then your seat isnt calibrated properly.

Thanks for the tip, I'll try that when I have my next available timeslot in between nappy changes and other baby stuff.

Have fun

Preliminary Conclusion

Ok, now all the parts are assembled, valves tuned and I have had a few rides. Time to sum it up!

Pneumatics was a completely new field for me so I have learned a lot. One of the things I leared was that this is not cheap. Every fitting or tube was pretty pricey. One other thing I learned was that it is difficult to get the pneumatics to control the sim with precision. Somewhere in the forum I read that pneumatics is too soft. That is not the case. If I use more that 3 bar the sim gets pretty strong. At 6 bar it is not gentle!

I Use needle valves at the exaust ports of the cylinders and it works good. Crank up the pressure to get force and close the needle valve to reduce air-flow and cylinder speed.

After installing the needle valves the sim is acting as expected. It follows the target values from x-sim pretty closely. The drawback with pneumatics is that it is only three speeds. The valves shuts on and off fast and generates a shaking movement, not smooth like an electric ride. For gravel rally it is perfekt because this shaking could very well be simulating an uneven surface. A smooth tarmac ride around a circuit is not so well simulated. The valves turns on and off more seldom, but the on off feeling is not so good. I usually turn down the pressure a bit on tarmac stages to reduce the shaking.

I drove the buggies in DiRT and it was great fun. I feel the bumps in my back and I also noticed the deaccelleration when I lifted the throttle. Turns and braking/accelleration are also well simulated. I will upload a video when I get time.

So my preliminary conclusion is that the sim is great fun! Pneumatics has the speed and the force but not the smoothness and precision. Most suited for gravel rally stages. A little bit annoying with the air compressor and air hose from my garage. The compressor makes some noise and I am worried for my neighbours! :twisted: I myself cant hear the compressor while I drive. Considered the price of the pneumatic and the extra mess with the compressor, valves and air hoses I would considered an electric solution if I could started over. (I kwill eep an eye on the projects that build electric frex's at this forum - using tronic's pwm interface)

So now after completing this sim I will lean back and enjoy it for a while and maybe I will try an electric version in the future. But not in a while. :wink:

Video, video, video

Does your seat center, when clicking default and then move simulator?

Yes, give us some video with DiRT!!!!! :top:



Thanos

I will record some videos soon and post them here! Living pictures tells more than a hundred words.... :razz: In a few days!

I tried that now just with profiler and force sender and no air - no it does not open any valves to return it to center. But as soon as I move the x or y slider a bit it starts to move towards the blue dot and shut of the valve when I move the sim to the target value.

If the blue target is dead still the sim stops. But in practise this is never a problem since there is always movement on the blue target dot. Only the slightest movement makes the sim start. Should I be worried? :/ Everything else seemes normal, the registry values are close to 0/255, the sim follows the target as good as possible with my system. I cannot see any drawbacks right now. I tested this on my single computer setup.