Silly(c)One's first 2DOF SCN5 racing sim

:thbd:

hope for you they offer cross shipping at least if you have to return them the actuator ...

You better call them on monday then sending an email ... they were pretty slow to answer my questions via email ...



Bleco

Good idea, thanks for the tip... I must admit it's pretty frustrating

I'll ask them to ship it with UPS. That's what Act-Labs did and UPS is pretty fast to go through the customs => my shifter was shipped on monday morning from Canada and was at my door on wednesday midday... I'm ready to pay the difference, at that point, I'd do pretty anything to get it back as fast as possible /me angry

Following your suggestion, I asked Mirai by email if they'd do cross shipping. They replied very promptly saying they never tried it but were willing to try due to my unfortunate situation. I'll pay the UPS shipping, though, won't come cheap, but I might have my SCN5 back at the end of the week or early next week

Communication is the key...

BTW, the last pieces for my gear indicator and speedometer arrived today : 3 led bar graphs (green, yellow, red) and 4 blue 7 segment displays (4.5cm/1.8 digits). I'm analyzing the specs of the MAX7219CNG IC to control all the leds at once with only 3 pins on the Arduino. My current question is wether the chip will be able to drive the different voltages of the led devices... To be continued...

The MAX7219CNG controls the current through each segment, voltage drop of each diode doesn't matter. As long as the LED is within the operating voltage range, it will be fine.

The led bar graphs have a forward voltage of 2.0v, which is perfectly into range, but the 7 segment displays are either 12V or 10V depending on size.
I think I'll have to cascade 2 MAX7219CNG : one for the bar graphs driven directly by the IC, the other for the 7 segment displays through an ULN2003A for the DIG0..DIG3 pins of the MAX7219.
Just trying to understand why Maxim uses an extra chip between the MAX7219 and the IRF540 used to drive higher power 7 segment display in the datasheet : http://datasheets.maxim-ic.com/en/ds/MAX7219-MAX7221.pdf at page 12.

This is the most difficult part for me as my knowledge in electronics/electricity is very shaky compared to my development skills... I can't wait to finish the electric connexions to dive into the code :

What 7seg displays are you trying to use with 12V? Are you making a store front display?
There are plenty of larger displays with a forward diode voltage of 3-4V, perfectly fine.

External components are used to source or sink the required current, above the 7219's capabilities. If you wan't to dig deeper, just look into these components' datasheets.

The digits I bought are 3 RED @ 12V and 1.8 blue @ 10V. The red ones are too big indeed but I like the size of the blue ones...

Apparently I can't use the ULN2803A to drive my bigger displays, but the MAX394 is much more difficult to source and I don't want to order new parts... Maybe I better go with a single MAX7219 for the led rpm bars and a combination of chained 74HC595's + ULN2003 for the digits (1 combo for each digit) and do the digits rendering by code... Easier for me to write a few lines of code than to learn electronics in details to do the same... after all, that why I chose the Arduino to get back to electronics: more code, less soldering

It works !!! :clap: :yippiee: :rofl: : :thbup:

I got my SCN5 back from Canada this morning... Nathan from Mirai told me they had to replace the whole PCB on my SCN5 (which they received on monday morning) and the cause was likely the RS485 adaptor. I was terrified to reconnect them in case... So I connected both adaptors without the SCN5s first, they were both recognized in Windows without any trouble. So I changed their COM ports (COM5 & COM6), and connected the SCN5s... :yippiee: everything was now running smooth !
I really don't understand what happened with my first SCN5. At first I thought it was a voltage spike, then a mechanical issue, then Mirai tells me it's probably coming from the RS485 interface... Will I ever know ? Does it matter now that they're running smooth ?

Spent the day doing the hardware adjustments, messed around with profiles until I found the one for Race On made by MasterYoda and tested it successfully on GTR Evo/Green Hell & Limerock Mountain/Formula Master & DTM Mercedes 2009... What a blast !!!

First I tried a few other profiles but had no reactions from the SCN5s... Next is to try Bernard's profiles pack with GTR Evo, rFactor, LFS, DiRT2 and F1 2010.

I found a great solution the last 2 days to calibrate my rev counter and rpm light bar by the Arduino. Instead of writing an interface plugin like symprojects (C++ in MFC with subexistant documentation translated from German, I don't think so ), it was the perfect time to test Titanium Desktop with Ruby. In 2 hours or so, I had a test interface (HTML5+CSS3), database connection (SQLite) and COM Serial communication (Ruby with a precompiled serial com gem - couldn't compile anything under that shitty Windows cygwin - damn I miss my OS X 10.6 !). Now that my techs are working, I have to implement everything, plus make a little circuit board and find a system to power my 10V 7 segment displays for the speed and gear...

On the bass shaker front, I still don't understand the bass shaker plugin very well, plus I'm not sure yet if I can run 2 DACs at the same time on Windows 7. Still no news of my Buttkicker Gamer2 from RennSimulatoren.de, it's been paid more than a month ago now... grrrr.
I don't have any precise filtering function on my cheap ebay subwoofer amps, so unless I get the TAmp-S150 from Thomann as I saw on sim-parts.com, I'll limit the bass shaker frequency to the RPM and use the Buttkicker for the general sound from the computer. Also, while the bass shaker is working as it should, I didn't find the effects very effective. Maybe it's because of the lack of a better filtering, or lack of power of the amps in the low frequency range (5Hz-50Hz), or physical placement of the shaker ? The last one I wondered while watching an SRT video on the Buttkicker, vibrating 15cm away from its mounting point... Maybe if I find a flexible but sturdy metal plate, I could try a similar setup, the DIY/cheap way

A few pics taken tonight :

The 911 Turbo S wheel, the RS Shifter 2006 with custom black chrome gear knob and DIY handbrake :

DIY foot rest plate (aluminium+MDF) :

The infamous SCN5 in their aluminium arms :

M10 fine thread (1.25) RH Rose Joint for the SCN5 threads and high misalignment spacers for maximum angle :

Left arm up close :

I had to cut into the dampening bar to allow maximum range for the lower joints :

I added a bar behind the wheel to attach my gauges :

The bass shaker not very far from the pivot :

And the cheap amps I found on ebay for my bass shaking experiments (10€ each incl. shipping ) :

A great tool I found during my adventures as an amateur accoustic guitar luthier, perfect to measure the right length of the arms at the horizontal position :

I'm back to it, cheers

Silly(c)One.

Here's a first quick vid of my brother racing a DTM on Limerock Mountain on GTR Evolution. I had initially the acceleration set to 400 but that was far too strainous on the SCN5s as they bottomed out once each until I found on Ledfoot excellent topic (kudos for documenting all your steps, that's been a tremendous help) that he was running them at 150 !
So back to the SCN5 plugin, a few changes here and there and voila, the SCN5 were back to their usual sqweek zwip sound, no harm done. (I don't want to damage my precious, so if anything seems abnormal, I immediately stop them and check them out with TBVST before starting the Profiler again.)
I'm now starting to understand more and more of X-Sim, SCN5 power and limitations and playing with profiles. It looked so obscure in the beginning I thought I'd never get it, but after hours of tinkering, X-Sim power is starting to be more apparent... The limit is the 8th gear

Thanks so much to all the initiators of this wonderful platform : SirNoName of course, as well as egoexpress, bvillers, nima, tronicgr and I'm sure many others. You guys have made a tremendous job :thbup:

But enough talk, here's the vid :
http://www.youtube.com/watch?v=yg9Sx1iEJ5M

I also did a video of the arms lower joint in action to show the dampening, but it's not very well shot so bear with me.

http://www.youtube.com/watch?v=16TfHc4GKiA

The dampening seems very soft/chewy to me, what do you think ?

Cheers,
Silly(c)One.

It should not be possible to bottom out your actuators (assuming a driver that is 250lbs or less) with any combination of actuator speed and acceleration if you have the proper pivot point location and actuator angles.

I would suggest experimenting with your actuator angles and pivot point rather than detuning your motion profiles.

You dampening bar looks fine to me. Keep an eye on the rubber to check for stress cracks. I tested many different dampeners. Most of them didn't hold up and eventually broke.

BRAVO pour ton simu, ça m'a l'air de très bien fonctionner !! :clap: :clap:
J'ai quelques p'tits problèmes avec l'anglais!! J'espère que tes problèmes sont résolus. Si c'est toujours d'actualité, n'hésite pas si je peux t'aider d'une queconque manière (à condition d'avoir un p'tit résumé in french please) :
:thbup:

Hello Bernard, thanks for passing by !

Good to know... I'm a bit lost however, as I followed your instructions on the proper dimensions for my setup with 150mm SCN5s to the letter :

If have precisely those dimensions setup, and the pivot is right under my butthole. The only 2 parameters I can change easily now are the distance between the pivot and respectively the dampening bar and the seat base, but both are consistant with your advised dimensions. It's quite difficult to imagine all the side effects of small adjustments, due to the quantity of variables involved : weight and size of the driver, pedals distance to the seat, pivot positioning, actuator angles... Do you think a few centimeters offset would change much ? I tried to follow your video about testing, but your profile was saved for an older version of X-Sim and I haven't understood everything yet to convert it to the higher resolution of the new Profiler.

That's a relief. I'm using german made silent blocks, 40mm diameter and 30mm tall. They don't seem prompt for stress cracks but I'll keep an eye on them nonetheless, thanks for the tips.

Cheers,
Laurent.

Hello Jo,

Oui, ça foncionne déja pas trop mal, mais apparemment je dois encore ajuster les dimensions et angles de la structure afin d'avoir le moins de stress possible sur les SCN5. Bernard est sur le coup, avec son experience on devrait pouvoir y arriver, mais merci pour la proposition de coup de main... Qui sait, quand je travaillerais sur la prochaine version en barre d'acier, je pourrais profiter de ton experience dans la soudure !

Bien a toi,
Laurent.

Haha!! I love that part

Can we add this to the official Wiki? Something like: be sure to precisely place the pivot point beneath your butthole

Sorry, I digress.

Back on topic....

Those were the words of Bernard, more or less

Ok so complete change of plans.
I think I understood the goal of the structure in terms of angles. Everything must be done to use maximum leverage. There are 2 axis : the first is the lever, from the pivot to the top ball joints of the actuators, the second is the actuator arm itself. From what I understood while brainstorming with a friend (architect), we must have the center of gravity of the driver as close as possible to the pivot or at least as close as possible to the lever axis, and the angle between the above mentioned 2 axis must be as close as possible to 90 degrees when the seat is in horizontal position (SCN5 at mid-travel).
I had the revelation when showing the beautiful work of Ralph (from Porsche Specialist in NL) to my friend to get ideas... I simply have to heighten the dampening bar and in consequence shorten the actuator arms ! Easy as f*** with the aluminium profiles, I had some spare with the exact dimensions needed. I'll have to wait for tomorrow before shortening the arms though, it's 3AM here and I don't think my neighbours would be very happy with the sound of the Festool saw cutting through aluminium

Here are a few pictures of the WIP :









I almost forgot : the buttons ordered from China arrived yesterday, so I dremeled an alu U profile I had at hand initially intended as SCN5 support and screwed the buttons on it. They will be connected to the miniPAC (http://www.ultimarc.com/minipac.html) along with the handbrake. At the top, you'll notice a switch, it will be configured as a Shift key in order to double the number of functions available on the panel. The 2 top buttons will be Esc and Enter, then a cursor <v^> followed by 8 buttons, in total 28 assignable keys. So I won't need a keyboard anymore. I thought of integrating a trackball to complete the computer interface but I'm out of cash for the moment so it's left for later. Here it is :



As always feel free to comment

Cheers,
Silly(c)One.

A quick drawing on known bases to illustrate my line of thought :



The red line is the lever. The purple one is the actuator. The most important angle is inside the orange circle.

The most important variables that we can act on are :
- #1 : distance between pivot and actuators lower attachments ;
- #2 : height of actuators lower attachments ;
- #3 : length of the actuator arms ;
- #4 : distance between pivot and the T behind the seat ;
- #5 : height of actuators higher attachments behind the seat.

Ideally :
- #1 should be as short as possible to reduce the sim length behind the seat and allow the shortest possible #3 ;
- #2 is relative to #5 in order to obtain an angle of 90 degrees between #3 and the lever ;
- #3 is the length of the actuators + dampening + a short arm allowing adjustments in length + 2 ball joints ;
- #4 is relative to the seat dimensions, angle of the back rest and the driver's CG ;
- #5 should be as high as possible so the lever is longer, but not too long in order not to reduce the travel too much.

All those observations are only applicable to a normal horizontal driving position. I believe a few observations should be adapted for a laid back Playseat F1 style position and this is not the direction of my study.

If the pivot is positioned in such a way that the driver is stable with the seat in horizontal position without the actuator connected, that means that the SCN5 will have a load of 0kgf at mid-course (75mm in our case) and the maximum loads at the extremes will be balanced and still quite light.

Think of when you balance yourself on a chair (I have ADHD so plenty of practice at the back of the classroom ) : when you're at your point of balance, that's when a feather like push in one direction or another is the most effective.

It's also when the actuator makes an angle of 90 degrees with the lever that it has the most torque... Hence the idea of mounting the arms higher so they can be shorter while still having a great angle and thus the length needed behind the seat can be limited further.

I don't know why it wasn't clearer to me at first, when you know what my principal means of transportation is :

Those are my pride and joy : Kris Holm KH36 and KH26 munis (muni = mountain unicycle)

But I disgress... back on topic : am I completely out of my mind or does it make sense to you ?

Cheers,
Silly(c)One.

Your thinking is correct. I did the same thing on my very first sim. Unfortunately, in that case, the design was suffing from being extremely narrow. Raising the bar on the rear helped, but the design wasn't sufficiently wide enough. To answer your earlier question, yes, as little as an inch off in any direction can negatively impact performance. Somewhere I posted an example of this on a wooden testing platform a year or so ago.

I think in your case, the rear bar change may prove to be helpul.

My schedule is rather crazy right now, but I will try to get the testing profile and tutorial updated sometime before the end of the year.

Sweet music to my ears... Say it again, pppppplease ?
(damn I'm too cerebral )

If I may ask so, why did you forget that route ? It shouldn't be too complex to build with a TIG, like Ralph did...

I was afraid you'd say that. It means I'll have to spend hours adjusting this single top pivot bar and I'm already mad in anticipation (ADHD turns repetitive tasks into brain torture...)
But hey, I want the very best for my SCN5s, so...

The theory of physics is on my side, at least, and I find the solution to be quite elegant... Experimenting is all the fun anyway, isn't it

Don't bother, for me at least. I'll look into converting the testing profile myself (as an exercise for X-Sim) and will post it here once it's done. If you find it suitable, you can of course put in on your tutorial page in the wiki along the actual version (if I can have edit rights on the wiki, I can even do it myself - I've been administrator of my own community and wiki for 4 years, web developer for 10 years and I've got plenty of free time so if my skills can help the community, that'd be great!)
I'll also post my sketchup file so if my solution is workable, people should easily be able to copy all my alu profiles dimensions. It would be nice to have an online database of working platforms in sketchup format, don't you think so ?

Have a nice sunday,
Silly(c)One.

Okay so your test profile just has to be loaded in the new version of X-Sim and saved again.
No resolution change affects any values from the math section.
The only problem I noted is that the maximum values stay unchanged in the TestPlugin from Sender, so I had to match them to the math section extremes by hand in the TestPlugin window directly and then export my TestPlugin settings to a file so I can load them again.

Concerning the physics of leverage, I think I know how to calculate the pivot position under the seat, theorically at least. The CG must be as close as possible to the lever, ideally, cross it at its center, as low as possible on the lever. That explains why the seat must be as low and close to the pivot point as possible (and why Playseat owners have to remove the base of their seat )
So I just traced a line in Sketchup from the pivot to the SCN5 top joint, and moved the pivot relative to the seat until my CG crosses that line. I haven't changed anything yet, but on paper, that should put the pivot approx. 14 away from the back of the seat base, which is approximately your dimension. I'm currently 5cm (2) away from that position, hence the extra effort given by the SCN5s...

Also in the process of heightening the dampening bar, I noticed the top angle between the SCN5s and the seat base was much bigger... The rose joints were at their angle limit and were putting strain on the SCN5 arms, so I had to enlarge the top of the seat base T. It's now 36cm instead of 27cm long, and the angle is back to the normal range of the rose joints.