Question IBT-2 or Sabretooth + Arduino Uno verses JRK data transfer speeds

The old saying goes, "bigger isn't always better". Based on this, how does a controller setup like IBT-2's or sabretooths (2 x 32) paired with an Arduino Uno Rev3 compare in data transfer, bit rates etc for more acurate motion to using a JRK G2 18v27 for example.

Yes, I need something that will cope with high amp spikes, but the data transfer must also be an important consideration yes? It has been said that the JRK has much better transfer speeds then the Ards. Don't know if this still holds true with the newer written codes for the ards and IBT-2's for example but i guess bit rates are bit rates.

Would the difference be that noticable with a well tuned rig and Sim Tools settings?

Been looking around for a while now on the forum for comparisons, but it's a bit of a grey area to a novice electronics tinkerer. Any enlightenment would be most welcome before I commit the coin.

Cheers.

I'd be curious to throw the Kangaroo in the mix as well (in place of the Arduino that is). I previously had my rig set up with some JRKs as they seemed like the cleaner config, but recently swapped to the Sabertooth/Kangaroo combo as I've been thinking about converting to 6 DOF and you can run two motors with a Sabertooth for the same cost as one motor with a JRK.

I guess from a very basic standpoint, the Arduino is talking serial to the Sabertooth (or the computer is talking serial to the Kangaroo), whereas the JRK is directly USB to a computer. I know USB is ultimately just an implementation of serial (universal serial bus after all), but I'm also not sure what is really going on within the board of the JRK. Maybe it's still doing serial translation to something else on the board it just has a built in converter to allow easy USB hook up.

All that said, it is wildly noticeable how much faster my Kangaroos talk to the computer when set from the default 9600 baud to 115200. Setting downloads and uploads go from a few seconds to under a second. However, I'm not sure how much that really matters in the long run when it comes to a sim. Taking the big picture approach, your more realistic limitation is going to be the mechanical inertia of your system. Say you have to send 16 bits of data to your motion controller from your computer. At 9600 baud that's ~0.0017 seconds. At 115200 baud that's 0.000139 seconds. Yeah, that's way faster, but are your motors capable of reacting and moving your rig in 1 thousandth of a second to begin with? I believe you also set a data transmission internal in the settings (default is 20 ms I think). As long as you can transmit everything you want to send in that timeframe, I would assume it would be a moot point. Though I guess it depends what kind of motion you are trying to sim as well. Smooth pitch and roll like a plane is likely less reliant on rapid direction changes compared to rally driving or something along those lines.

@Map63Vette thanks for that insight. I don't know much about the kangaroo controllers. I live in Australia and I certainly know what they taste like, as I've had a many a kangaroo steak! I just installed a sabretooth 2 x 32 into my setup. It's taking over from two IBT-2's. It will control my two front vertical (pitch and roll motors). My two rear horizontal (Yaw/Traction loss and surge motors) are staying on IBT-2's.

It will be interesting to see if there is much of a change in data speeds between the sabretooth and the former IBT-2's. Hopefully it'll be noticeable and I'll report back how the "seat of my pants" feeling is with the sabretooth.

I'm seeking snappy realistic motion. My Sim has a large range of motion for a worm drive motor setup, . (a link to a recent demo of the new added surge axis) but I'm after a realistic feel from the rig over an arcade ride. I still have much to learn and tweak in the sim tools axis DOF's Vs Tuning centre inputs. It's all a bit of a journey at the end of the day.

Again @Map63Vette thanks for the input.

I know I've read in the past where people have seen nice fidelity benefits when switching to JRKs, but not sure how old those posts are and if code has been updated over the years to close that gap. I've had fairly little true seat time in even my own sim as I found my motors weren't really the best fit (or maybe my pivot point if my latest measurements are true). I'm mostly just going off of theory and what you could extrapolate from pure numbers at least, so take it with a grain of salt.

I know one of the other main reasons I was interested in the kangaroos specifically was their built in PID auto tuning. The PID values are really the heart of a motion controller as they are going to be what determines how a motor gets to it's target location (or if it even does). That's not to say that their tuning algorithm is faultless, but I figured it probably had a better shot than me randomly throwing numbers into the JRK settings until graphs looked okay. Hand tuning is probably still good to get those last little bits sorted, but my hope was the auto tune would get 90%+ of the way there. There are some good guides on this site already for setting those values that I read first and would advise looking over regardless. Always good to understand what is going on under the hood, even if you aren't manually turning the wrenches so to speak.

@Map63Vette I'm hearing you on tuning PID. Was a steep learning curve trying to get my head around PID, but I found some reasonably straight forward literature on it. I still feel like I'm experimenting for the most part with it. Just making minor adjustments here and there, testing and checking for jitters and heat. Not sure how else to know what are optimum settings for any given controller/vs application.

Give me a spanner and an impact drive any day!

Yeah, the struggle I had with PID is that it never quite feels like there is one right answer. Maybe I just don't have enough seat time tuning either, but it certainly seems like multiple different combinations of PID terms can get you a similar feel. I'm sure higher fidelity rigs with better parts probably make that more noticeable though.

I really need to fire up Assetto Corsa again now that I've got my Sabertooth setup going for a better one to one comparison as that's what I've had more time with on the JRKs. I started playing with flight sims just before I got the Sabertooths, which don't tend to use the motors quite the same way. I know previously I would have one motor that would drop out every couple of minutes because the heat switch on the brushes would trip (wiper motors). When I swapped to the Sabertooths I couldn't get one motor to work only to find out I had actually cooked it to the point of breaking it after removing the heat switch. In the couple of flights I've done since the change my motors don't seem nearly as hot though, so thinking the PID settings probably have something to do with that.

Yes. I've noticed significant changes to motor temps (I also use heavy duty truck wiper motors) when making changes to PID settings and PWM settings. There are important things like running the correct Fpwm frequency with certain motors. I was running my IBT-2's s at 35khz for a long time (that was the default setting when SMC3 loaded) not realising they are only rated to 25khz.

Found that out by accident on a different forum looking for something completely different! So much technical stuff to navigate. It's hard enough knowing the right way to ask something when searching this stuff, let alone finding a straight answer. I basically try and read up on all the spec stuff now on what I am using in the rig, to make sure I'm not cocking something up via ignorance.

But I'm pretty slow on the uptake.

Yeah, I know what you mean. I've had the idea to build a 6 DOF for a while now and keep second guessing myself every time I read a new thread. It's also tough because I think we tend to use components in a way that they weren't really originally designed for, so it's always a bit "seat of the pants" and hard to find specs on. Regular DC motors aren't really meant for holding applications to begin with, but they've been used to great effect on sim rigs for a while now.

I know I only finally dipped my toe in years ago when I got a pile of them for free from work as that was the biggest expense I wasn't ready to throw money at. I cut up an old treadmill for metal tubing for my rig as well, so in the long run about the only parts I actually bought to build my first rig were the motor controllers and power supplies.

It's just wild reading threads too because people all over the world are building stuff, which makes it even harder to copy designs as sourcing parts is so different. My problem has been finding that goldilocks motor that's not too big and not too small for my design, but also readily available in the US. I'll find one I like, but they it's built in the UK or Australia or something and costs more to ship to the US than the thing is worth, so it's back to the drawing board.

Yes I've noticed that. Even the motors I use which are really great for a box standard worm drive wiper motors (they are all metal worm and gear) are now out of stock here in Oz. I have found similar motors on buy China and Alibaba, but they're always minimum quantities to buy. Cheap, but you'd need to buy 100 units at a time before they'd sell them to you.

Bugger that. I get all sorts of crazy ideas to build better and bigger, but the truth is, the rig I've built is kick ass and It'll take me a long time to justify in my head the hrs of build time, research and coin to build a 6DOF. Now I've got mine to a 4DOF I'm pretty happy. Just got it all painted real pretty and have been shooting a video of reassembling it.

I'll post it in time lapse when I'm finished. I'm a sucker for old school muscle, so I've painted it up like a late 60's early 70's muscle car. like my car in my avatar.

I think at the moment I'm really just looking for a project or something to do as I've been really bored around the house. My dad also got into flight sims and my wife gave him some money toward flight lessons for Christmas, so I've kind of got the bug to do it with him and thought a better home rig might help me practice for the real thing. It's probably a waste of money as I've barely even used my current 2 DOF, but we'll see where I end up. I bought my first parts to set up a test actuator control system the other day. Figured that was cheap enough and would give me something to do for a bit. Was thinking I can also buy parts over time instead of all at once so the cost doesn't seem so bad, lol.

I had seen the car in your avatar and wondered. I'm a big fan of old muscle as well and have an old Dart with a modern engine in it as my other tinker toy. Just too cold outside right now to spend my money on it instead, haha.

@Map63Vette a Dart with a modern engine. That'd be interesting. Seen a few of those projects here as well (newer engines in older cars). Here's the latest video of my now completed 4DOF platform.

Neat! That's a really clever idea on how to get surge on a more 2/3 DOF platform.

I was fortunate in my original design concept, a single spine mid frame. It lent itself nicely to adding a rail system to carry the top frame.