|
Post by paulg100 on Aug 2, 2016 18:34:13 GMT
do you mean 1kw (1.5hp) ? that is the next step up from the 0.75kw.
I think you will need bigger gear boxes if you step up? the 090 gear boxes are for 1 and 1.5kw. And inverters also? might be an expensive upgrade
Despite the issues I have got, errors due to acl/dcl have never been a problem. ive run 0.5acl and 0.2dcl without issue so far, I'm sure I could probably get acl even lower.
|
|
|
Post by Trip Rodriguez on Aug 8, 2016 3:39:57 GMT
I'm a bit concerned about certain dimensions and don't find the info without doing a bunch of figuring and math subject to error. 1. Rods- My rods (purchased nearly 2 years ago) are 22 1/4 inches long from center to center. The main thing I'd have thought a possible problem would have been clearance to the motors but I'm going to invert my gearboxes which should give me lots of clearance there. Still I would really like more info on rod length: Is 25" the ideal rod length or was it selected for another reason? Would this affect the effective final drive ratio (via change to total mechanical advantage)? I assume it would affect max angle and interference with the motors/gearboxes but no idea of the math required to understand the relationships between these things. My brain says that longer rods would give a larger range of motion but have less mechanical advantage but I'm not sure that is right. If so my crank arms longest setting (of 4 holes) is about 6.25" which I think is longer than most others have used with similar builds so I might be able to make it up there if having shorter rods but longer arms doesn't have any kind of significant down side. 2. Base dimensions- I set up my base layout over a year and a half ago but it's not too late to change it but I'm having a hard time finding raw dimensions which is what I need since my base is a different design. Right now my shafts center to center are 35 11/16" on each pair. I'm re-drilling the holes anyway so I can adjust them to 36" exactly if needed but it might mean another hour or two of measuring and head scratching =P. I also need another dimension before I drill. Right now, between the tips of my shafts on center are 19" (see attached image) when measuring from one gearbox to the nearest gearbox in the next pair. Is that correct? How critical is this dimension? Just to explain why I'm complicating things my gearboxes have significantly different dimensions than yours and look to have longer shafts as well. Also my base is a flat platform I'm drilling holes in rather than a frame. Thanks for all the help! Attachments:
|
|
|
Post by GA-Dawg on Aug 9, 2016 15:03:46 GMT
Let me start off by saying that I started this build wanting the most motion possible as that is what I "thought" was needed for a good experience. So the dimension of the machine where designed to accommodated a +30 degree angle. Which is still what you would most likely want if you plan on using a screen setup. I have switched over to a Oculus (I just finished reading your thread and see your going to do that as well) and I would say the amount of angle is not as critical for this setup. So that means the rods you have should be fine. With a Oculus the thing that is needed is FAST movement to make sure you stay in sync.
On the dimensions. I would first need to ask what size is your upper platform? and do you have that made yet? The side spacing your asking about is related to the upper platform size.
Cheers GA Dawg
|
|
|
Post by Trip Rodriguez on Aug 9, 2016 21:31:00 GMT
@ga-Dawg Thanks for the reply!
I did a bunch of digging around and reading last night to try to find all the info I was using almost 2 years ago when I started this and found that what I was working off of was Fabi's dimensions.
My distance center to center on shafts in each motor pair is 190mm (hence why it was not exactly 36". My distance between pairs (diagonal) I gave was an error, it's actually 450 mm I think (I'll check in a few minutes and edit this).
The upper platform is not yet started so that is flexible. I was going to use Fabi's measurements, but did see that Thanos and some others recommended going down to 50mm between rods at the top from 100mm so I'll do that I think. Does that mean I should change other dimensions at all to match?
Making an image with my dimensions now, posting it SOON™!
|
|
|
Post by GA-Dawg on Aug 9, 2016 22:30:02 GMT
I would recommend that you finalize the upper platform first. The upper platform mounting side is parallel to the rods and the rods are parallel to the motors. So the "19" side is somewhat depended on that upper platform.
|
|
|
Post by Trip Rodriguez on Aug 9, 2016 23:28:51 GMT
Here are my dimensions, looking for advice before I drill the holes in my floor to mount the gearboxes. I'll be out of town the next couple days so you won't hear from me until Saturday night or Sunday! Hoping to be ready to drill this weekend!
|
|
|
Post by Trip Rodriguez on Aug 10, 2016 0:28:21 GMT
I've seen now where several people have said that with VR go for high speed and that means less travel. I had my torque arms made with four different settings so for short travel i can use a shorter torque arm setting and gain mechanical advantage for more speed at the cost of less travel.
If you look at my drawing you will see my torque arms can be set at 100, 130, 160, or 190 mm's. Should I go all the way to the 100 mm setting or keep to middle ground? How much effect does this have on my other dimensions (because I might want to switch later to a different torque arm length)?
|
|
|
Post by Trip Rodriguez on Aug 10, 2016 9:34:13 GMT
@ga-dawg Can you tell me how you limited the travel for use with the Rift? What I really want to know right now is if you are using only a portion of the total travel of each actuator at all times, or if under some circumstances (heave, sway, surge) you are still using full or near full travel on the actuators.
I'm considering mechanically reducing travel to gain torque since my sim is basically intended 100% for VR.
More torque should mean faster movement, better for VR.
|
|
|
Post by GA-Dawg on Aug 12, 2016 12:11:24 GMT
So just a quick update. Things are going OK. I got thing tuned in xsimultor (not x-sim). Lag is just about gone but I not 100% happy with YAW. I think a upgrade is in the future. I did create a nice E-brake setup out of a hammer. It's a cheap solution, but is has a good feel to it. I will post the details just in case anyone is interested.
|
|
|
Post by paulg100 on Aug 12, 2016 19:19:48 GMT
HI GA
Did you look at this in BFF ?
"PitchStart=127 The mid stroke positions of all actuators in the position feedback 0-255 scale. Example; if your feedback potentiometer voltage output is 0.5 to 4.5V when the actuator moves over it's full working range, assuming your drive hardware uses a max possible 0-5V voltage range to detect actuator position then .. FALim1 = (4.5 - 0.5) / 5 * 255 = 204, and the min and max positions on the 255 scale are 0.5 / 5 * 255 = 25 and 4.5 / 5 * 255 = 229 And the actuator mid position is, PitchStart = 25 + 204/2 = 127"
these are the pots you used? the more expensive ball bearing type (I have one to test also)
file:///D:/Motion%20Simulation/6DOF/BI%206153V1A180l.5.pdf
The volts are 0.25 to 4.75 not 0.5 to 4.5 with is the pitch start default of 127
Does this make a difference to our home position?? or just affects the failim range I think so not related to the homing offset issue.
|
|
|
Post by GA-Dawg on Aug 15, 2016 15:27:26 GMT
Hi Paul,
Home position did not move so it just effective the range.
|
|
|
Post by GA-Dawg on Aug 15, 2016 19:00:49 GMT
Hi All, So I got everything working fairly well and thought I would share my thoughts. First motion is subjective so this is really just what I think and not much fact to back it up. Well I finally had a chance to get some real seat time in (4.5 continues hours on Project Cars this weekend ). Motion was driven by xsimulator and I only used my Oculus. Problems. 1. There was a problem with xsimulator and yaw. When the value hit +/- 180 degree the platform would flip to the opposite direction. So I had to turn of yaw. I need to research to see if others have this problem. 2. Still having a problem getting the accl ramp down on VFD. Why is this important. The closer the ramp is to 0 the closer the motion matches what the eye is seeing. So for now I have had to give up so movement to have things "feel" in sync 3. Center of gravity (COG). Some motion driver software allow you to programmatically adjust for COG but not all, and after play for awhile you really start to notice when COG is not right. 4. This has nothing to do with the sim platform but still a problem. The Oculus will not hold the image (sinking into the seat). Other have found that it's due to roll so I will follow what they have done. 5. Butt Kickers....Need to hook them up.
What's next.
I have decided to redesign the upper platform out of aluminium and adjust for COG. I am hoping that the lighter weight will also help with the accl ramp. If that doesent work than I will have to think about switching out the gear box. I originally chose the 80:1 as I read the ideal RPM was around 20. With my setup I would never be over 22.5. With a 60:1 I would get closer to 30 RPM and it seams most are able to get there accl ramp down. For the record my accl ramp is at .18 right now.
Cheers GA Dawg
|
|
|
Post by paulg100 on Aug 15, 2016 21:02:38 GMT
"1. There was a problem with xsimulator and yaw. When the value hit +/- 180 degree the platform would flip to the opposite direction. So I had to turn of yaw. I need to research to see if others have this problem."
yes this is a common issue with many of the plugins. some have an alternative option allocated to extra dof 1/2 to allow it to top out at 180 and wash out back to center but it depends on how the plugin has been made.
its not a bug. Can be fixed by changing the plugin but learning how to do that needs a degree of programming know how from what i can see.
simtools 2 will have a proper 6dof math plug in with adjustable COG apparently. Also a simvibe like plugin for transducers as well. allot to look forward to.
|
|
|
Post by GA-Dawg on Aug 22, 2016 14:58:01 GMT
Hi All, So I have been focused on my accl ramp thinking it was causing my sync issue. Well as it turns out it was my setup. I was running two computers with a crossover cable. I decided to upgrade my main computer to handle the Oculus better and in doing so I decided to use just one computer with a multi video card/monitor setup. After getting the new computer in place I also switch back to BFF as it handles the 6DOF better than the current public version of xsimulator (ver 1x). I was able to increase the varies settings which I had to originally decrease, and the sync issues never came back... . One point my accl ramp is currently at .19 and I thing the only way I can reduce that is by lightening up the upper platform (remake it out of aluminum). But I not sure if there is really a need to do so at this time. So I will go back and update the build parts for this thread to reflect where I ended up. Hope others fine this thread helpful Cheers, GA Dawg
|
|
|
Post by Trip Rodriguez on Oct 23, 2016 20:52:37 GMT
Thanks for that update @ga-Dawg, this is helpful. Now I'm wondering if I made a mistake going to the 130mm crank arms and putting my gearboxes closer together to try to get more ACL speed as it seems that turned out not to be your issue.
I was planning to use a separate computer to run the motion also, good to learn that is not the way to go.
|
|