Universal 6DoF Recessed Simulator

Hi

I understand what you are saying, what are you using, Arduino? The crash filter can also filter bumps given a different set of parameters. I have only spent about 1-2hours on the profile, I'll only really concentrate on it when the seat is in and I can feel it, but a combination of profile tuning and firmware filtering can go a long way. Simtools is a bit limited, for certain DOF outputs even on the 2D motion math you might want a different positive and negative scale.

if you are using Arduino or STM send me your code and I'll see if we can do a filter, I'm not sure if it's possible or if it will work well on Arduino but worth a try?

I use Thanos's AMC.
Thank you for the offer.

I think that unfortunately there is nothing to do with simple.

it is the G forces that are problematic.
it's not really suitable for moving on a physical axis on a simulator like ours
the management in position for the force G is not correct, that imposes very fast displacements towards "the target" which is finally very short to generate an impression of strength to the body.
it turns the simulator into a violent rodeo machine ...
the faster the simulation is, the longer axis is violent and difficult to adjust.

it could be a very big resolution to compensate and a management of the engine speed especially for G forces.

besides, some pro have understood the problem!

Agree, it's why I have a G system fitted to my 6DOF. Although 'Washouts' can also help if implemented and setup properly

Hmmm, should I have studied at Tongji University instead of ETH Zurich⁈

Okay unfortunately it is closed so you can't filter anything. Like I said I can modify the crash filter to reduce the effects of bumps, for instance we can reduce the position instead of reducing the speed when detecting bumps, but I'm busy logging some data and then we'll see what we can do.

yes but reducing the position decreases the length of the axis and decreasing the length of the axis, the sensations become "rough"
a lot of information in little axis length, it becomes confused in sensations. (Dbox style)

it's very complicated.

I've got 12-axis capability in the new controller I'm working on.

No I don't think you understand. I mean if a bump normally would lift the machine 100mm, we can reduce, and only this one bump by 50% if it meets criteria, the rest of the motion isn't affected.
And I think I have a better solution in any case, but it will need a bit of time to work out the details.

hi
I said you that your platform was not strong enough, you said me I was wrong, now you made big reinforcement, am I still wrong?

You said the epoxy isn't strong enough, it is still epoxy, so I guess you are wrong.

So this would be something along the thinking, so basically a high pass noise filter, this is on sway with 8-bit output, channel 2 is unfiltered simtools output and green is the filter I made. You can see where I touched the wall and another car, the two big spikes, those are the ones the crash filter takes care of if you want it to. But if you look at the other movements, you will note large reductions at low speed bumps and vibrations compared to the high speed movements with steering input, some low speed reductions are as much as 90% and on the largest crash it goes down to 0. Using a filter like this will allow you to use long travel at high speeds without holding on for life on the violent rodeo machine. When combined with a crash filter the axis scale can be further increased along with the filter strength so that you can use much more of the available travel. Obviously this is a very basic comparative filter and it can become much more complicated if done with code, but it does explain the concept. I tried the same with crazy settings in simtools by putting large axis movements and a large smoothing number which you can see in the videos I posted, but I don't think it is possible to do this correctly with the simtools settings, you need external processing either via software or on the microprocessor. I think it's quite similar to smoothing, but this method works a little bit better.

Wrong, smoothing is nothing like it, smoothing seems to be a bandwith reduction filter.

I did not really test the filters on Simtools, maybe a track.
for the moment the tests I did with the filters, reduced the small effects a lot.
I dropped the filters for the moment.

The simtools filters won't work for your problem. The only thing that can be fixed nicely with the simtools filters is the yaw "snap" at -180 and 180 points.

@Rider63: I noticed the new reinforcing too, the inner triangle around the driver? I couldn't imagine those joins holding either.
It seems those joints are not just in sheer and peel, but more in bend forces attached to long moment arms.

I'm developing a sunken cockpit idea very similar to this but I'll be using bent one-piece tubes which are far simpler, and probably stronger.

your bad faith is impressive...

read again my message, I said that the way you assemble is is not strongh enought, not strong enought for the configuration you have. of course if your double the structure or ad reinforcement, epoxy is enough, but when your make it at start, it was not strong enough. so I was right.

why not just said that you made a little fault ? the way you answer all people is juste making them hating your, like me now. good bye

I read it again, you said: "I see the way you wan't to join all tubes togethers, in my mind, this kind of bended join are not strong enought, glue or tube will break with the time, I'm almost sure..."
I guess I still read that as a statement to mean that the epoxy will not last over time, in your follow up you stated: "I understand but I thinks the surface is not wide enough, but it's just my feeling, we will see."
This further makes me believe you were specifically talking about the strength of the epoxy joints, I changed the attachment system at the bottom where it is taken apart, the rest of the joints I wanted to match to look the same, but it's quite a bit of work so I've redone it over for the next one, simpler to manufacture, simpler to assemble. There are many alternative ideas that can be tried that will work, but it's all a compromise of meeting the final requirements. Right now I'm only focused on electronics and motion, but at least I had some time to do some more fun development work.

Please don't assume what you see is what is happening or that ideas are posted, loads of things change and posts are usually a bit late. That said initially there was coupling in the dual layout, it was removed for aesthetic reasons, when the new bends came along it removed this coupling, the platform got weaker and a new coupling solution was then found, it solved the strength problem, but stiffness was still an issue due to the deep recess, the top section was added with a hinge to open up when you get in an out, it improves stiffness 100's of times, I prefer the look without but it's too much of an improvement to ignore. Just plain normal development and it's been on the back burner until I had issues with the electronics and decided to bring it onboard and see. The bends will stay because they are pivotal to the modularity, they have many more benefits than drawbacks.

So I ended up making the seat mould like so, I was told by the composite guys the dual skin seat would be too expensive and we should go with a single skin.

8 blocks, 3 slices per block, XPS, they come in 80mm slices so built it up on the machine.

Right now busy with this:

Keep posting , really enjoy watching what can be done in a well fitted out shop. Development is a iterative process - I'm sure you won't push a product out that you are not happy with.

Will do, I think you will enjoy the upcoming stuff on motion .