G27 belt drive mod

[eKretz]

Ah, I'm looking at the pics on my phone, the black part of your bearing looked like a rubber seal. Those are known as flanged bearings, sort of like pillow block bearings. Being difficult to turn then should be caused by the heavy grease they're packed with - that's why I recommended washing out the bearings with acetone or something so you can get it out of there. Cleaning it out and replacing it with a light synthetic oil will really make a big difference.

 

[sesselpupser]

that's why I recommended washing out the bearings with acetone or something so you can get it out of there. Cleaning it out and replacing it with a light synthetic oil will really make a big difference.

I'm not sure it's possible to clean them out, I'm reading the various datasheets (specifically the ones about seals and lubricant) and everything implies that they're properly sealed units that fail for two reasons - mechanical wear and the lubrication drying up - at which point the entire bearing has to be replaced; there's nothing here about removing the seals to re-grease them or anything. There is a grease, um, hole in the bearing but no obvious way to get the old grease out, the seals themselves are apparently interference fit on the... Inner part (I think the extent of my bearing knowledge is showing now, right?) and trying to pry them off isn't getting me anywhere. Do you know if it's possible to open this type of bearing up?

Looking at this datasheet pretty much confirms these bearings, sold to me as 25mm bearings, are actually 1". That means the steering shaft will be off-center by 0.4mm, but it's not a high-speed shaft so that shouldn't matter too much, I hope. Also the pipe fitting to connect the shaft to the wheel just arrived, that was sold as 'chrome' but it's just really rough chromed plastic... It feels tough enough, though, so I guess we'll see how that works out. Still waiting on the shaft coupler, I actually did manage to find 3.175 to 3.175mm couplers on Ebay last night so if this 3.175 to 4mm one is good enough I'll order everything else.

 

[eKretz]

Hmm, you mention seals but you said they didn't have rubber seals. Can you clarify? If you only have steel covers on the bearings those are shields, not seals, so it's possible to wash out the grease by injecting a solvent through the small gap between the bearing race and the shield, then working the bearing around by hand, injecting some more working it around etc. until you have the grease cleaned out. Usually 3 or 4 iterations does the trick. Then you inject some light synthetic oil and work/rotate the bearing a bit - again 3 or 4 times should do it. Only rotate the bearing by hand, and do not use compressed air to spin the bearing when there's no lubrication or you can damage it. I use an old syringe to inject acetone and lubricant into the shielded bearings on Fanatec wheels.

I would not recommend attempting to remove the pressed in shield, they are very thin and fragile and are almost always destroyed upon removal. In machine repair work the bearings mostly either get automatically replaced or new shields are bought and installed if they are very expensive bearings that are still in good shape and only need to be cleaned and relubricated.

Being off center that amount might add a considerable amount of drag or it might be fine, depends on how rigid your setup is.

 

[eKretz]

OK I just downloaded the data sheet for that bearing. Looks like it does have rubber seals. That means you won't be able to relubricate it. These are self-aligning bearings, so your minor misalignment you mentioned should be fine. The grease fitting on the bearing is only for the self-aligning feature, it won't get grease or oil into the actual ball bearing itself.

 

[sesselpupser]

OK I just downloaded the data sheet for that bearing. Looks like it does have rubber seals. That means you won't be able to relubricate it. These are self-aligning bearings, so your minor misalignment you mentioned should be fine. The grease fitting on the bearing is only for the self-aligning feature, it won't get grease or oil into the actual ball bearing itself.

Aw. Well thanks for taking a look at that for me, I really appreciate it, and when I said 'seal' before I meant 'shield' because the black bit in the photo is metal. Moot point now anyway, of course! I think I'll get the 25mm pipe, assemble what I can of the steering shaft and see what it's like turning the wheel with both bearings in place. If they're no good I might see if I can return them because they're not metric like the site said they were, I'm not happy about that but I can work around it if they'll do the job. The site only has a 7-day returns window, though, but they're the ones who mixed up the UCF205-16 (imperial measurements) and UCF205E (metric), not me!

 

[sesselpupser]

The pipe arrived, my laughably unreliable, super cheap digital caliper reckons it's 25mm - it usually measures about .2mm over and it said 25.35, I'd expect at least 25.7mm if it was a 1" pipe. Anyway, it would appear the 25mm bearings are metric after all, but the flange/block/housing/whatever isn't. Anyway, after hammering one of the bearings on with a piece of wood, I realised that was probably a really bad idea so I tried to hammer it off - no dice, of course. Then I remembered I had a set of dumbbells that had a 1" hole, so I, being a clever girl*, took a 2.5kg weight and used it like a hammer by sliding it down the pipe over and over until the bearing came off again. So there's absolutely no play between any of the bearings and their shafts which is great, but I'm still waiting for the 3.175mm to 4mm shaft coupling before I order more stuff so I can check that it'll take enough torque without slipping. I mean, it should be fine, but I don't want to order another £100 of really specific, non-returnable components only to find that it isn't fine.




*Jurassic Park reference, of course

 

[sesselpupser]

I just got the 3.175mm to 4mm shaft coupler, of course I wasn't able to put much torque on it through either the motor or the 4mm shaft but when I get home later I'll try clamping the shaft and powering the motor to see if it slips when I counter-rotate the motor beyond the stall torque.

Apart from that, I've made an early attempt at connecting the 25mm shaft to the wheel, it seems to be alright but the cheap pipe mount I got appears to taper so when I tighten the grub screw it goes a bit sideways if I don't put the shaft through a hole in the mount to keep it pointing the right way. That just means the mount has to extend further back down the shaft than I originally thought but that's ok, I'll just make it a two or maybe three layer thing to make sure it's all straight and there's no play in anything.

So now I just have to order all the other parts and wait for them to come in, I imagine the 3.175mm to 3.175mm shaft coupler will take a week or two so I'll order that right now. Or not, because my Paypal account is acting up. I'll do it later I guess.

 

[eKretz]

If you want to give yourself a little cheap insurance on the coupler just put some flats on the motor shafts.

 

[sesselpupser]

If you want to give yourself a little cheap insurance on the coupler just put some flats on the motor shafts.

I could, but I'm trying to keep all the original components unmodified if possible so I could just reassemble it if I ever need to. I think it'll be ok though, there's not going to be that much torque.

 

[eKretz]

I could, but I'm trying to keep all the original components unmodified if possible so I could just reassemble it if I ever need to. I think it'll be ok though, there's not going to be that much torque.

You ought to be able to reassemble it still as long as you only make it a very small flat.

 

[sesselpupser]

Well, I tested it and I'm fairly confident it won't budge, but if it does then I guess I'll do something about it.

I've just ordered:
- 3.175mm to 3.175mm shaft coupler
- A load of wire
- The belts
- Ten each of PHR-2, 5, 6, 7 and 9 connectors (I hope the ones on the PCB are PHR! If not I guess I'll just remove the plastic shrouds, the pitch is 2mm for sure though)
- 200 PH-series crimp pins - about 150 more than I need, but the PHR connectors come in packs of ten and the pins in packs of 100, so...
- Four ferrite beads identical in size to the ones already in the wheel.

I also added some hall effect sensors (Allegro A1302s, the same as Leo Bodnar sells but for about a third of the price) to get free shipping from CPC, I'm going to see about converting the pedals to hall effect once the belt mod is done.

Also I just realised I was going to add 25mm deep groove bearings to the belts and wire order. Oops. I better get on the phone!

 

[tubedriver]

I have no idea, I'd have to have a look inside a new G27! No matter what, though, I'm quite sure the new G27 will work the same way in that it counts a predefined number of pulses and anything over that number will stop it from calibrating, so the theory is the same.

Speaking of which, I tried my G27 without the rotation limiter in it yesterday - it doesn't like that at all. The calibration routine does just turn the wheel one way until it stops, turns it back the other way until it stops and then divides the number of pulses between the two as a centre point. Without the limiter it just keeps turning, but if you grab the wheel it'll stop and turn back the other way again. I experimented with trying to get it to turn further than 900 degrees but if you do that the calibration fails, the shifter LED just blinks constantly and it doesn't know what to do. So, yeah, I do need a mechanism to limit the rotation, I guess I'll have to think about how to do that. The easy way would be a ball screw like Frex have in their SimWheel:

That also allows for on-the-fly lock configuration which is nice (you can see the adjustable rubber stoppers on long threaded rods). But that would push the cost up above £200, so... If I can't keep the cost below that I'll call the whole thing off!


Edit: The cheaper option is to copy what Logitech have done; a rack and pinion with blocks on the rack and shaft to stop it from rotating at either end. Pretty simple, I just hope it's quiet!

Some thoughts on calibration....

Just following the logic here - if one can cause the wheel to calibrate by grabbing the wheel then the calibration routine isn't limiting the rotation just based on pulse counts, it's looking for the stops. The calibration becomes the full count between stops divided by two to find the centre. It fails without stops because of a counter overflow.

So you should be able to configure your wheel for more rotation if you wish, based only on the mechanical stops, by using an encoder with fewer slots. I understand the 30-slot encoder wheel will fail on a 60-slot G27, but some smaller change may work.

Having said that, given that the lock-to-lock can be set by entering the button codes, it must be counting as well.

Good luck with your mechanicals, look forward to see what you come up with.

Cheers!

 

[sesselpupser]

@tubedriver I think you're right but I can't make my own encoder wheel right now so it's sadly academic at the moment. I do think it would be trivial - unless the G27 has a minimum pulse rate before it fails to calibrate, though I don't know why it would be engineered that way - to use a custom 50-slot encoder and stops limiting the wheel to 1080 degrees of rotation.

Another method would be to separate the encoder wheel onto its own shaft away from the motors and adjust the ratio of that shaft so that you get more rotation but the same number of pulses but I think that would be too expensive to be worthwhile.

I do really want to find another way of limiting the rotation, something that's either adjustable physically or through programming. I'm pretty stuck on that, though, so I'm relying on the backup plan for now.

 

[sesselpupser]

Ok, so here's where I'm at now:

I've got all the parts except the material to make the case with, the pulleys and the rack and pinion for the rotation stops, I think it's probably wise to wait for pay day again so I'll probably order those next week or the week after but they'll arrive pretty quick. I got a straight router bit for my Dremel and used it to make a better wheel-to-shaft mount prototype, it works well but I need to spend a little more time on getting that right; I could really do with a scroll saw, actually.

I got the 25mm deep groove bearings and they're much lower friction than the pillow block bearings I have (if that's what they're called, I don't know), in fact they're even lower friction than the bearings I pulled out of my old DFP so I'll work those into the build instead. I managed to order a load of the wrong size connectors for the replacement G27 internal wiring - my cheapo digital caliper reported the pitch as 2mm but it's more like 2.5, I'm an idiot for not checking but I think I'll probably just carefully lift the header shrouds off (I can't find the exact type of locking header it is anyway) and use normal 0.1" female headers instead, I did something similar with my DFP which seemed to work fine. At least I got the wire dimensions and ferrite cores right, eh? I did completely forget the heatshrink, though!

So it's still a couple of weeks until I'll start building the thing properly, even then I think it'll take a while because I have college work to be doing and I think I'm going away in May for a bit; I've decided I want to do my second degree (electrical and electronic engineering) in Berlin so I'm going to go have a better look around before I apply. I will of course keep the thread up to date whenever I do anything big.

 

[eKretz]

Your bearings are called flange bearings actually, they're just similar to pillow block bearings.

 

[sesselpupser]

Well, it's not much but I made this:

So that's the wheel mounted on the shaft. It's not pretty but the whole thing cost about £15, I may 3D print a cleaner alternative later on. There's not really anything else I can do without the pulleys so I guess I'll have to just wait until I can order them, then I can test my shaft centre point distance calculations and start making the mounts so I can then go on to assemble everything.

 

[sesselpupser]

So I'm still waiting for pulleys, the place I'm getting them from missed an email I sent them about two weeks ago and were too slow replying to the last one I sent them to get it done before Easter. The guy said he'd process the order today but I haven't heard from him yet.

PUBLIC HOLIDAYS. Am I right or what?

Edit: Pulleys ordered, that wasn't cheap!

 

[sesselpupser]

Well, I finally got the pulleys, the rack and pinion, but I've apparently made a miscalculation somewhere - the rack is only long enough to allow something like 1.9 turns! I had originally thought that the rack I was buying would have excess teeth so I could trim it down (since it's 20cm long). So now I'm trying to think of ways to lock the wheel rotation that don't involve converting the shaft rotation into linear motion, which is tricky. At least the pulleys are good, though.

 

[tubedriver]

Glad to hear you are making progress!
Any chance you could add a compound gear between the pinion and rack? Like this, just replace the red gear with your rack..A 2:1 ratio on B would give you just short of 4 turns with the rack you have, assuming A=C.

Another option, which doesn't involve linear motion, is to drive a simple single reduction gear (say 4:1) and install a pin in the output gear face and have it hit a stop. One shortcoming of this would be the lack of adjustability.

Cheers,

 

[Christian_C]

This is really cool, good luck completing it

I have a suggestion though, don't try to put more intermediate shafts in to get a different ratio for the encoder, gears don't mate perfectly so the more gears you put inbetween the wheel shaft and the encoder the more play will build up in the system.

I don't actually know what a G27 looks like inside or exactly how you're planning to do this but if possible i'd be trying to drive the encoder with only one gear reduction from the main shaft if that was at all possible.

 

[sesselpupser]

Glad to hear you are making progress!
Any chance you could add a compound gear between the pinion and rack? Like this, just replace the red gear with your rack..A 2:1 ratio on B would give you just short of 4 turns with the rack you have, assuming A=C.
View attachment 351767
Another option, which doesn't involve linear motion, is to drive a simple single reduction gear (say 4:1) and install a pin in the output gear face and have it hit a stop. One shortcoming of this would be the lack of adjustability.

Cheers,

The intermediate gear between the rack and pinion is what I'm probably going to explore next, I could probably even do away with the rack entirely if I could reduce 900 degrees down to something more like 45 (I haven't tried to work out if that's easily doable yet though) because then I could just have a big lever moving the blocks around. As for having a pin in a gear, I'm not too sure about that since it has to handle a lot of torque, you know how it feels physically impossible to turn a G27 beyond 900 degrees? It needs that level of firmness to the stop, hence why I want to brake the shaft the wheel is on directly.

This is really cool, good luck completing it

I have a suggestion though, don't try to put more intermediate shafts in to get a different ratio for the encoder, gears don't mate perfectly so the more gears you put inbetween the wheel shaft and the encoder the more play will build up in the system.

I don't actually know what a G27 looks like inside or exactly how you're planning to do this but if possible i'd be trying to drive the encoder with only one gear reduction from the main shaft if that was at all possible.

Yeah, that was really just a suggestion, in my design the encoder is going to remain fixed to the rear shaft of the rear motor because it's cheaper and easier that way. I would've used a timing belt and pulleys rather than gears, though, because they don't have the meshing problem you described.

Edit: oops, I forgot to say thanks for your kind words and advice!


Last night I modified my biscuit cutting rig to drill out a blind 47mm hole to make a bearing block for one of the shaft bearings. It's made (predictably) of MDF and it was an extremely tight fit so I'm kind of regretting - well, no, massively regretting - testing the fit by just dropping the bearing in because there's no way I can get it out without destroying the block, but at least it worked! Now I have to make an identically tight one for the other bearing... Also I found an old veneered chipboard CD rack which has one nicely square edge so that's going to be cut up and made into the prototype. I think I have enough material to make a few mistakes, so that's good.


Edit: So I've spent a few hours working out how I can do the rotation limiting with gears and I can't really figure out a solution that's cheap, reliable and quiet... Then I thought - what about wire? What if I used the 25mm shaft as a spool so that when the shaft turns one way, one cable is taut and the other slack then when you turn it the other way, the opposite happens. I haven't worked out the specifics yet - how to fix the cable at either end, how to make sure it doesn't tangle around itself or anything like that - but it seems like a potential option.

 

[sesselpupser]

Ok, so I got the steel rope but I hadn't anticipated how springy it would be, it's effectively just a coil spring when wound around a practical radius (if I could make the radius twice the size of the pulley on the same shaft it would be ok, though, but I can't) so it loads the shaft quite significantly even after half a turn. That'll mean the motors will work harder so it'll accelerate both their ageing and the MOSFETs', so I'm not too keen on that idea.

I'll try some alternatives (1m of 3mm steel rope was £1, so it's not a big loss at all, I'll try thinner steel rope, fibre rope and bailing twine because why not, I guess) but I had a different idea I'd like to try - a bi-directional solenoid that does the same job the gears would do, but with less mechanical complexity. I figured I can probably drive it with not much more than MOSFETs, an IR LED, IR phototransistor and some logic ICs, but sourcing the solenoid itself might be tricky. It's just a very vague idea right now but it might work.

Apart from that, I've been drilling the bearing blocks, realising I haven't got the appropriate screws yet, putting the drilling on hold and otherwise preparing the chassis for assembly. I need to get a hacksaw from somewhere to trim the various shafts to size and screws (which I've just ordered) to mount the bearing blocks, then I'll be able to test the mechanism's torque characteristics, correct any misalignment and so on. A basic test fit yesterday looked pretty good but there's only so much you can do by eye.

So yeah, TL, DR: when the screws arrive I should have something interesting to share.

 

[tubedriver]

Ok, so I got the steel rope but I hadn't anticipated how springy it would be, it's effectively just a coil spring when wound around a practical radius (if I could make the radius twice the size of the pulley on the same shaft it would be ok, though, but I can't) so it loads the shaft quite significantly even after half a turn. That'll mean the motors will work harder so it'll accelerate both their ageing and the MOSFETs', so I'm not too keen on that idea.

I'll try some alternatives (1m of 3mm steel rope was £1, so it's not a big loss at all, I'll try thinner steel rope, fibre rope and bailing twine because why not, I guess) but I had a different idea I'd like to try - a bi-directional solenoid that does the same job the gears would do, but with less mechanical complexity. I figured I can probably drive it with not much more than MOSFETs, an IR LED, IR phototransistor and some logic ICs, but sourcing the solenoid itself might be tricky. It's just a very vague idea right now but it might work.

Apart from that, I've been drilling the bearing blocks, realising I haven't got the appropriate screws yet, putting the drilling on hold and otherwise preparing the chassis for assembly. I need to get a hacksaw from somewhere to trim the various shafts to size and screws (which I've just ordered) to mount the bearing blocks, then I'll be able to test the mechanism's torque characteristics, correct any misalignment and so on. A basic test fit yesterday looked pretty good but there's only so much you can do by eye.

So yeah, TL, DR: when the screws arrive I should have something interesting to share.

Perhaps use a length of bicycle chain over two idlers, connected to the steel rope on the shaft? You don't even need sprockets, it will run over a plain wheel. You can bolt through the chain for the stops.

Here's a pic of the design that made me think of that...

This is not a feedback design, but wraps the chains around the wheel shaft (not what I was suggesting by the way, more the opposite) to pull on the steel wire to provide spring centering. It also has a beautiful implementation of a rack with adjustable stops.

Cheers!

 

[sesselpupser]

@tubedriver that's a really nice design but I can't work out how I'd attach the chain to the shaft, from the photo it looks like... I can't work out what that looks like, actually, which is the problem!

I've been busy lately trying to finish my college work in time, but I've got a few minutes while I sit in a coffee shop waiting to meet someone so here's what I've been doing lately:

- Pricing up a laser cut Delrin chassis, it's expensive! Like, £100 to do it with 10mm Delrin, probably double that for 20mm (which is what I'd prefer). That would make this mod as expensive as the G27 was in the first place - I don't think I want to own a £450 G27!
- Looking at Makerbeam and Openbeam - I think they'll be too flimsy but I might order some Openbeam bits to test, I have actually ordered a beam of each but they haven't arrived yet.
- I ordered a shorter short timing belt so I can put the motor shaft between the other two for compactness, having the motors above the steering shaft makes it a bit too tall for my liking.
- Looking at different motor configurations. So far I think I can do sequential motors driving one shaft, parallel motors each with their own pulley and belt (expensive, probably less efficient and noisier, I think), a motor at each end of the shaft with opposite polarity... Sequential motors is probably the cheapest and easiest, though, but probably the longest option.
- Staring at all the wood, screws and bits I've already prepared and not actually doing anything with it... I know it has to be precisely aligned to work properly but because my tools are awful (you just can't cut straight edges with a £15 jigsaw!) so it's just really daunting, hence why I'm looking at getting stuff laser cut. Maybe I'll start with an MDF version.

Oh and I've just ordered some 2mm steel rope and some bailing twine, pretty much just because why not, I don't think they're going to really work though. I'm now looking at solenoids but that's going to mean working out how much 'spare' current the G27's power supply has and all that kind of stuff. Meh.

 

[sesselpupser]

So, I... Uh... I've done some thinking.

First of all, I'm going to reassemble my G27 as soon as I've finished writing this post, not because I've given up, but because a) I want to be sure I still can, i.e. that I haven't lost anything, and b) I want to play, dammit!

Secondly, I took the day off of college today and I've spent some time thinking about the project.

I got the beams I ordered and I think Openbeam (15mm^2 aluminium profile with a 3mm slot) is pretty nice, Makerbeam may even be sufficient, I'm not sure though (it's 10mm^2). I don't think I'll ever actually find out without building a chassis with both, so, uh, that might get expensive!

Then I was looking at my measurements. First of all, the belt length calculator I used to spec the first belts I ordered was clearly wrong but I chose to ignore it at the time, the new shorter short belt I ordered arrived today as well, though, and it made me realise just how much more compact the whole thing could be if I shortened the other belt too - it could be about a third narrower than it is right now. But that's working on the assumption I keep the fairly long motor arrangement of having them in sequence, sharing a shaft...

If I have the motors in parallel, I could reduce the depth of the entire thing by... Oh, just over half! And even then, if I get a bit more inventive with the pulley placement I could reduce it by a further 12mm.

In addition to this, going with a beam chassis means two things:
1. I only need two 3mm sheets of whatever material laser cut with 3-4 large holes and maybe 12-16 smaller holes (for screws), instead of at least three significantly larger sheets which would then need further work done by yours truly with cheapo tools - so the likelihood of expensive errors is much, much smaller.
2. The chassis becomes modular, so I could add things like a phone holder (for use as a virtual dashboard), button panel or maybe even a real dashboard or SLI-M/Pro holder later on with minimal cost and potentially no modification at all to existing parts, that's very appealing.

So now I'm trying to psyche myself up to get on with the Sketchup drawing. Unfortunately I still haven't solved the rotation limiter problem so I am kind of just assuming whatever it is, it won't add any bulk, which probably isn't wise - I'm kind of betting on finding a rope or string that's strong but also not even slightly elastic or springy when coiled. Yeah, that's spectacularly dumb, isn't it? But still, modular chassis > non-modular chassis in this case.

 

[sesselpupser]

I've drawn what I think is a pretty good attempt at a more refined prototype:



The initial dimensions of the all-MDF model were 321 wide, 187 deep and 176 tall, now they're 246.4 wide, 156.4mm deep and 126.4 tall - that's a 240x150x120mm Openbeam chassis with 3.2mm MDF covering it, hence the ...6.4mm measurements. I obviously don't know how big the JLVRH wheel is but it looks pretty big, I think my mod might actually be smaller than the stock G27 (only very slightly though, if it is) but they've definitely got the upper hand with regards to getting the wheel sitting lower, particularly when tilted - mine is at 0 degrees. Tilting it 15 degrees like the stock wheel is going to lift the wheel up quite high, I think. Anyway, it's going to cost an additional ~£100 so once again I'm playing the pay day waiting game. I need:

- More new belts... I definitely measured once and cut twice, here. In my defence the calculator I used the first time was wrong by a long way (I'm starting to think it was for 10mm pitch rather than 5) and the decision to change to Openbeam meant the third belt I bought was only a tiny bit too long to get the whole thing to fit within the width of a 210mm beam...
- 4x M3 standoffs plus hardware (£2.27)
- Openbeam corner cubes - a bag of 12 (I only need 8 but I can't buy them individually) is £12.35
- 6x 90mm Openbeam (£6.90)
- 6x 210mm Openbeam (£12.66)
- 4x 120mm Openbeam (£5.52)
- 16x 90 degree corner brackets (£7.68)
- Another Synchroflex T5x10mm 10-tooth timing pulley, with the 4mm custom bore and M3 grub screw from the shop it's £9.32
- Another 13x5mm 4mm bearing
- About a million M3x6mm screws, six of which should be flanged (for the motor)
- two M6 screws and nylon locking nuts
- A sheet of 3.2mm MDF; I'm specifying MDF solely because I can only find Delrin in 3mm, but Openbeam's slots are 3.2mm so it - I suppose - would rattle around. The good thing is the MDF is cheap - £1.14 for a third more MDF than I need - and thin so hopefully getting it laser cut will be comparatively cheap, too. Annoyingly I can't tell if it's strong enough without just making the thing first. (£1.14)
- Laser cutting. I don't know how much this will cost so I'm just hoping I'll barely scrape by the £50 minimum most properly-equipped shops use. I mean, it's all fairly simple stuff - rectangles with several small holes cut in them - and thin, commonly-used material, so I'm just going to assume £50. I'll probably be horribly surprised.

There are, however, two problems:

1. Calculating the centre line spacing between the two motor pulleys and the 60 tooth pulley they're belted to is going to be difficult, and a lot depends on getting that right. I can see why the JLVRH mod kept it simple so the horizontal lengths of belt were parallel and everything was at 90 degrees; I've got the motor pulleys closer together than that so the belt is sloped, that makes the calculation way, way harder, but I couldn't do that because the diameter of the motors was going to add something like 40mm to the height, that may not sound like a big deal but not only does that mean the thing is, well, taller, it also means I have to buy slightly longer beams, get nine piece of MDF cut 4mm taller and the wheel would be 2cm higher than it would otherwise be, and I happen to prefer having the wheel quite low. Anyway, I think I'm going to have to see if I can get a test model cut before I commit to cutting the design as I've approximated it and I'm definitely going to have to buy the pulley before the new belts and test it with some string or something before I make any final decisions there.

2. The whole rotation limiter problem is still very much a problem! I got the 2mm steel wire in the post today, it's a lot better than the 3mm was but it's still very, very springy so I can't use that either. The bailing twine, although it looks super ghetto, is actually a pretty good option; I tested it earlier (admittedly turning the 25mm shaft by hand rather than with the wheel, since I reassembled my G27) and it's pretty tough, but of course it will snap under enough stress. It's not elastic at all, though, I tested some last night by hanging about 15kg off a 1m loop, measured the length of the loop again after five hours and it hadn't stretched even slightly. I'll test it more before I accept it as a solution but it's looking pretty good, believe it or not. In the meantime I'll look for something a little less... Agricultural. I'd still hugely prefer a big, physical block, but I don't want to increase the bulk of the wheel specifically for that reason.


More screenshots of the approximated internal dimensions:


That's the Openbeam chassis with MDF slot inserts. You can see the 25mm shaft at the front, the green components are shaft couplers, the motors are peeking around the corner there too.



The orange parts hold the bearings in place (you'll see it better in the next image), the yellow parts hold the bearings so they don't fall out (the two combined basically work like a bearing block). I'm going to glue these bits together but I've got screw holes to help with that and to add some extra reinforcement.



The blue parts are bearings, so you can see how they'll be held up and enclosed by the combination of the yellow and orange parts.


The red parts are the pulleys, of course the purple ones are the shafts. I didn't draw those motors and, actually, they're not the right size but that's not important, I've drawn the mounting holes for them from the datasheet for the G25's motors, then I added an extra mounting hole because G27 motors have three.

Oh yeah, the blue part in the first image is just illustrating the awful MDF wheel mount I made, I really don't know what I'm going to do about that... Either get one 3D printed, work a bit harder on milling my own out of MDF or maybe design a layered one to be laser cut then screwed and glued together afterwards... I don't know!

 

[tubedriver]

I've drawn what I think is a pretty good attempt at a more refined prototype:

View attachment 371094

The initial dimensions of the all-MDF model were 321 wide, 187 deep and 176 tall, now they're 246.4 wide, 156.4mm deep and 126.4 tall - that's a 240x150x120mm Openbeam chassis with 3.2mm MDF covering it, hence the ...6.4mm measurements. I obviously don't know how big the JLVRH wheel is but it looks pretty big, I think my mod might actually be smaller than the stock G27 (only very slightly though, if it is) but they've definitely got the upper hand with regards to getting the wheel sitting lower, particularly when tilted - mine is at 0 degrees. Tilting it 15 degrees like the stock wheel is going to lift the wheel up quite high, I think. Anyway, it's going to cost an additional ~£100 so once again I'm playing the pay day waiting game. I need:

- More new belts... I definitely measured once and cut twice, here. In my defence the calculator I used the first time was wrong by a long way (I'm starting to think it was for 10mm pitch rather than 5) and the decision to change to Openbeam meant the third belt I bought was only a tiny bit too long to get the whole thing to fit within the width of a 210mm beam...
- 4x M3 standoffs plus hardware (£2.27)
- Openbeam corner cubes - a bag of 12 (I only need 8 but I can't buy them individually) is £12.35
- 6x 90mm Openbeam (£6.90)
- 6x 210mm Openbeam (£12.66)
- 4x 120mm Openbeam (£5.52)
- 16x 90 degree corner brackets (£7.68)
- Another Synchroflex T5x10mm 10-tooth timing pulley, with the 4mm custom bore and M3 grub screw from the shop it's £9.32
- Another 13x5mm 4mm bearing
- About a million M3x6mm screws, six of which should be flanged (for the motor)
- two M6 screws and nylon locking nuts
- A sheet of 3.2mm MDF; I'm specifying MDF solely because I can only find Delrin in 3mm, but Openbeam's slots are 3.2mm so it - I suppose - would rattle around. The good thing is the MDF is cheap - £1.14 for a third more MDF than I need - and thin so hopefully getting it laser cut will be comparatively cheap, too. Annoyingly I can't tell if it's strong enough without just making the thing first. (£1.14)
- Laser cutting. I don't know how much this will cost so I'm just hoping I'll barely scrape by the £50 minimum most properly-equipped shops use. I mean, it's all fairly simple stuff - rectangles with several small holes cut in them - and thin, commonly-used material, so I'm just going to assume £50. I'll probably be horribly surprised.

There are, however, two problems:

1. Calculating the centre line spacing between the two motor pulleys and the 60 tooth pulley they're belted to is going to be difficult, and a lot depends on getting that right. I can see why the JLVRH mod kept it simple so the horizontal lengths of belt were parallel and everything was at 90 degrees; I've got the motor pulleys closer together than that so the belt is sloped, that makes the calculation way, way harder, but I couldn't do that because the diameter of the motors was going to add something like 40mm to the height, that may not sound like a big deal but not only does that mean the thing is, well, taller, it also means I have to buy slightly longer beams, get nine piece of MDF cut 4mm taller and the wheel would be 2cm higher than it would otherwise be, and I happen to prefer having the wheel quite low. Anyway, I think I'm going to have to see if I can get a test model cut before I commit to cutting the design as I've approximated it and I'm definitely going to have to buy the pulley before the new belts and test it with some string or something before I make any final decisions there.

2. The whole rotation limiter problem is still very much a problem! I got the 2mm steel wire in the post today, it's a lot better than the 3mm was but it's still very, very springy so I can't use that either. The bailing twine, although it looks super ghetto, is actually a pretty good option; I tested it earlier (admittedly turning the 25mm shaft by hand rather than with the wheel, since I reassembled my G27) and it's pretty tough, but of course it will snap under enough stress. It's not elastic at all, though, I tested some last night by hanging about 15kg off a 1m loop, measured the length of the loop again after five hours and it hadn't stretched even slightly. I'll test it more before I accept it as a solution but it's looking pretty good, believe it or not. In the meantime I'll look for something a little less... Agricultural. I'd still hugely prefer a big, physical block, but I don't want to increase the bulk of the wheel specifically for that reason.


More screenshots of the approximated internal dimensions:

View attachment 371102
That's the Openbeam chassis with MDF slot inserts. You can see the 25mm shaft at the front, the green components are shaft couplers, the motors are peeking around the corner there too.

View attachment 371097

The orange parts hold the bearings in place (you'll see it better in the next image), the yellow parts hold the bearings so they don't fall out (the two combined basically work like a bearing block). I'm going to glue these bits together but I've got screw holes to help with that and to add some extra reinforcement.

View attachment 371099

The blue parts are bearings, so you can see how they'll be held up and enclosed by the combination of the yellow and orange parts.

View attachment 371100
The red parts are the pulleys, of course the purple ones are the shafts. I didn't draw those motors and, actually, they're not the right size but that's not important, I've drawn the mounting holes for them from the datasheet for the G25's motors, then I added an extra mounting hole because G27 motors have three.

Oh yeah, the blue part in the first image is just illustrating the awful MDF wheel mount I made, I really don't know what I'm going to do about that... Either get one 3D printed, work a bit harder on milling my own out of MDF or maybe design a layered one to be laser cut then screwed and glued together afterwards... I don't know!

Wow, Now I can see what you are up to! Your design looks very promising.

A couple of thoughts....

You might be able to solve the issue with the motor belt length being so critical by making the plates supporting the motors and their front bearings separate from the other plates supporting the wheel shaft etc, and make them adjustable so you can set the tension on the drive belts.

The plastic cord used for line trimmers (Weed-Eaters, etc) is readily available and is quite flexible. .080" or .060" might work as a bailing twine substitute. Not sure how much it will stretch over time, but it is very strong.

MDF has very little strength when loaded in any way that tends to pull the layers apart, as will happen when you pull on the rim of your wheel. Unless you have bolts going all the way though to a flange, I would be cautious about using MDF for this, other than as a prototype.

Looks great, good luck!
Cheers,

 

[sesselpupser]

You might be able to solve the issue with the motor belt length being so critical by making the plates supporting the motors and their front bearings separate from the other plates supporting the wheel shaft etc, and make them adjustable so you can set the tension on the drive belts.

I had thought of doing this for simplicity, but for reliability I'd prefer to get the measurement right in the first place so it's all one solid piece, I think I'll just have to test it until I'm convinced I've got it right! Then, when the laser-cut MDF prototype confirms that it works ok, I'll get it cut out of something tougher when the MDF model fails.

The plastic cord used for line trimmers (Weed-Eaters, etc) is readily available and is quite flexible. .080" or .060" might work as a bailing twine substitute. Not sure how much it will stretch over time, but it is very strong.

Good call, I hadn't even thought of using that stuff. I'll get some to try out, I might even already have some lying around but I can't remember seeing any.

MDF has very little strength when loaded in any way that tends to pull the layers apart, as will happen when you pull on the rim of your wheel. Unless you have bolts going all the way though to a flange, I would be cautious about using MDF for this, other than as a prototype.

Yeah, I'm concerned about the longevity of MDF but this is just a prototype for now, ideally I'd use Delrin but it's pretty expensive - £20.46 for 500x500x3mm (and I'm not even sure that's enough, I haven't checked) vs. £1.15 for 300x900x3.2mm MDF. Also, the slots in Openbeam profile are 3.2mm wide and MDF was the only material I could find that comes in that width; I could no doubt shim 3mm sheets of something so they sit tight in the slots but for a prototype I think MDF is ok, in theory anyway. Without being a mechanical engineer I don't think I can work out the kind of forces the MDF plates will take or how much they can take without making them and using them until they fail, so I guess I just have to do it and see what happens!

However, that said, I could get two more 90mm pieces of profile and two of 60mm, add them in at the front to make a 60x60mm opening for the front 25mm bearing (which has an outer diameter of 47mm), that should make the MDF harder to damage. It's only an additional £2-3 or so, so why not.

Thanks!


Edit: I am (well, was) half asleep but thinking through this project, about what more I could do with it... Here's what I'm thinking - I'm writing this here mainly because I don't have a notepad to hand!:

- I'm designing this around a modular aluminium structure, but I've designed the outer case to cover the whole thing up... That's dumb. I think I'll add indentations in the outer case panels so that parts of the aluminium are bare for future additions like button panels, phone/tablet/display holders and so on.

- There should probably be an internal USB hub, mainly so that I don't have to run cables from the wheel to the PC when I do add external devices.

- I should add cable routing holes in the rearmost bearing mount since the PCB will be in there, just for tidiness. Maybe I should even mount the PCB backwards so that the connectors are pointing away from all the moving parts, but that would limit the ambient air getting to the MOSFETs so maybe not. Unless I add vents in the rear of the case and put a small 5V fan in there to blow air directly onto the FETs?

- Somewhat related, maybe I could use one Arduino as a button interface, a USB switcher and a temperature sensor/fan controller.

- I should definitely earth the aluminium...

- Right now I have my wheel plugged in to a home-made USB switch so I can change from PC, disconnected or PS3 whenever I want, I think I'll integrate it into the case and have two USB sockets on the outside, probably USB-B.

- I could recess the power/USB/pedal/shifter connectors by making a panel to sit in the left rear slot so that they're not mounted directly to the outer case. That would look quite a lot nicer and more professional (even though it's still just going to be a pile of MDF on an aluminium skeleton).

- If I do want to rake the wheel back like the stock wheel, how would I do that? Do I even want to? I'm not sure...

- I wonder if I should re-case the shifter... Later.

 

[Mayaman]

Neema, forget string, find some fine steel wire for this. Keep it up and don't give up. Its a great learning experience.

 

[sesselpupser]

Neema, forget string, find some fine steel wire for this. Keep it up and don't give up. Its a great learning experience.

The thing is I have no idea where to buy this stuff from, how to know how easily it'll coil around a 25mm shaft or a spool of up to 90mm, how much it'll act like a coil spring when coiled up, what the tensile strength will be like... So I have to keep buying wire which could get expensive. I've just found some wire that looked good (except for the fact that it looks like the kind of wire you'd make springs out of which is exactly what I don't want!) but it'll be £6 for one metre, so I'm not too happy to pay that. £1-2 for a metre of rope I can't use is acceptable, like the 2mm and 3mm steel rope lengths I bought, but spending £6 when it might be a waste... Not so easy to justify.

I'm trying to find a similar application so I can find a supplier that sells to that industry - like, for instance, how the steel rope I bought was the kind of stuff you'd hang chandeliers with - but I'm not having much luck thinking of a more common use case than mine!


Edit: I've revised my drawing:


Cutouts to make the profile accessible for bolting on accessories.


Reinforcing the 25mm bearing mounts with extra profile.


How the front part of the two-piece bearing block front face will look with the reinforcement.


And the other part that has been deepened from 3.2 to 6mm. I thought this was probably wise in part because the 25mm bearings are 12mm deep, so they weren't all that well supported before. I can't make it much deeper than this, though, but i think this is adequate. In truth I thought 3.2mm was adequate too, but... Yeah.

I've looked into raking the wheel backward 15 degrees like the stock wheel but because the wheel sits higher in the body of this mod (because of the clearance required for the 60-tooth pulley on the motor shaft), it would be almost 25mm higher than the stock wheel when raked, which sounds like it would be noticeable, therefore I think I'll concentrate on getting the wheel assembled first and worry about tilting it later.