2DOF Build Log Thread

fpwm is at 15? the arduino can be set to 25. is there a reason for this frequency setting?

The VNH2SP30 chips that are on the MMs say they're only rated for operation up to 20KHz, see page 1 of this doc:

http://www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00043711.pdf

These datasheets are always on the safe side, and I'm sure 25KHz is fine for 99.9% of cases.

I had it set conservatively for troubleshooting, I'll try turning it up to see if it feels better.

It's the arduino your communicating with. It is capable of 25hz.

EDIT: that's kHz sry.

Right, the arduino can send out PWM at up to 25KHz, but the monster moto chip is only rated to receive it at 20Khz or less. The MM doesn't send data to the arduino, it only receives it. Like I said before, these ratings are usually pretty safe, and 25KHz operation is probably fine in 99% of cases, you could probably go even higher and still not have any problems.

The clamps on my g27 broke

I suppose it's my fault, I probably should have built a little flange to use the hard mounting points on it. I'm gonna see if some JB weld epoxy will hold them on for now.

I've been trawling ebay for replacements but no luck as of yet.

Made some changes today! My HTC Vive arrived last week so I've been getting my setup prepped for VR. I added bolt flanges to the wheel mount so it can be taken off. I also made two mounting posts for the HOTAS that I purchased, and bought a few castor wheels so that I can easily move my sim in and out of the VR area.

​

On another positive note, my G27 was still just under warranty when the clamps broke, so I was able to warranty it, and since they don't make it anymore, they moved me up to a g29. The RMA process has been a tad awkward though, as for some reason when they shipped my replacement they only sent the shifter and AC adapter . The full set is on its way, but I still might not have access to a wheel for another week or so.

So after Logitech screwed up my shipping order once, I finally got my replacement wheel in. It was working fine for about an hour, and then it started to slip off center when making fast corrections against the FFB. It turns out this is a known problem where the rotary encoder, where it comes from the factory broken/not fastened properly. This issue has been present since the G27, further confirms that the G29 is nothing but a price-hiked version of the G27 without the shifter. I could open it up myself and fix it, but if the encoder itself is cracked then I have to wait another two weeks to have a laser cut one shipped from china for $50.

Seriously disappointing. This will be the last time that I buy a logitech product.

My g 27 did this after a couple years. The true problem Imo is the motors get too hot causing failure in the plastic that connects the sensor to the motor. I opened mine up, hot glued... Lol yes hot glued the sensor to the motor and installed a couple small fans. Now mine was well out of warranty by the time it broke so it was fix or forget for me. I now have countless hours on it and it is as good as a g27 can get. It's many years old now not sure of the year when I bought it but it was when the g27 was first released. Gl hope they square up with you.

Sorry to hear that Tim, a buddy mentioned today that he bought a used G27 and was told it was a simple calibration issue. He too had to order a new encoder from China

I'm hoping support can come through for me again and ship me a new wheel. I'd open it up and have a look but I'm nervous about voiding the warranty now that the wheel costs $500 CAD...

On another note, would anyone be interested in me uploading a 3D model of the ebay pgsaw gearmotors? I'm designing a new base platform for my sim and had to whip up a quick and dirty sketch of them for measurement purposes. Might be useful for new guys modeling their builds.

Here's the model. I've got it saved as solidworks files as well as a .STEP file, so anyone should be able to open them. Do you guys know where this would be best uploaded elsewhere on the forum?

Thank you

Alright, well the idea has been gnawing at my brain for a while now, and I think I want to add a third axis to my sim.
I'm re-modelling the base plate of the sim to be more portable and a bit heavier (it's jumping around a bit without extra weights put on it as it stands now). I'm going to opt to add heave instead of traction loss like most people would for a couple of reasons. Traction loss usually adds a fair amount of footprint to the sim, and I'm already space limited. Secondly, I use the sim for flight as well, and I want to have an axis that's usable in both flight and racing. Not to say that I won't add a yaw/traction axis eventually, just that it won't be my next addition.

With that out of the way, here's a rough sketch of what I'm thinking:

​

I want to build a module that I can just swap with the current U joint I have sitting under my platform right now. The motion would come from a 50:1 gearmotor with a very short CTC mounted under the plate. The motor would be assisted by four springs which would ideally hold the sim at half heave height when a person sits in it (I'm likely going to have to add some sort of pre-loading mechanism to the springs to account for differences in driver weight, most likely an adjustable bolt threaded to the outside of the rod that the spring goes around). I still have to do most of my calculations for lever length/spring rate, but everything I have so far seems feasible. The top platform would likely weigh about 300-350lbs (125-150kg) worst case, so I probably won't have a ton of travel total if I'm using the standard pgsaw motors, but since I primarily use the sim with VR I don't actually want large axis movements.

The whole top platform would slide up and down on 4 linear bearing blocks (http://www.aliexpress.com/item/2pcs..._6&btsid=9a3f2d79-86f1-4382-9b8d-a1927f3ea5e1), bolted on to the plate with angle stock. I'm not sure what the exact size of the bearings will need to be, that's one thing I still have to figure out. I'd have to add hard stops to either end of the bearing rods to stop the platform from travelling too far in either direction.

Anyways, suggestions and critiques are welcome! If anyone else with a heave axis could let me know how much travel they have on their sim I'd appreciate it, as it'd make some of the guesswork in calculating things a bit easier.

I've got some rough calculations down:

Assuming the motor outputs around 180W of power to the shaft at 75rpm, that means it should be generating about 23.74Nm (210lb-in) of torque (T = P/RPM) running at full speed.
If I end up using a 30mm CTC that gives me 235.6mm/s of movement at 791N (177lbs) of lifting force in either direction.

Now, I don't want the motor to be doing any static lifting, ideally the springs should be taking care of that. Optimally, the springs should be exactly halfway compressed when a driver is sitting in the sim. I also want to be using as long of a spring as I can fit under the mechanism. This reasoning behind this is somewhat tricky to explain, but I'll do my best. Springs will produce an output force in proportion to how far they are compressed, and their spring rate. So if you compress a spring with a spring rate of 30lbs/inch by 2 inches, the spring should exert a force of 60lbs. For the sake of the calculations, let's assume that my sim has 2 inches of heave travel, one inch up and one inch down. If I use a shorter spring, with a higher spring rate, say a 2.5 inch long spring with a spring rate of 280lbs/inch, then the sim would sit halfway down the spring (1.25 inches) which is good. The problem arises when the sim moves up and down on the spring. At full upward heave, the spring is now only compressed by 0.25 inches, which means that it's only providing 70lbs of static lifting assist, which is far too little.

If I use a longer spring, say a spring that's 12 inches long, with a spring rate of 58lbs/in, then the sim would still sit halfway down the spring when static, but at full extension the spring would be providing 290lbs of static lifting support, at which point the motor should be more than capable of picking up the slack.

In either case, it's actually the ratio of spring rate to length that matters most to me. So if I were to use a single spring, the rate-to-length ratio would need to follow the formula: Spring rate * Spring Length = 700lbs, or more generally, (Spring rate) * (Spring Length) = 2 * (Total weight of platform+ driver).

I still don't have a good source for springs in my area, so that's one thing I'll still need to figure that out.