Gryphon CNC extrusion

discussion of design changes / improvements / suggestions
joelcrawford
Posts: 8
Joined: Sat Oct 25, 2014 1:57 pm

Re: Gryphon CNC extrusion

Post by joelcrawford » Tue Apr 14, 2015 9:16 pm

Custom belts on custom systems maybe all hunky-dory for hobbyists that have time to construct the various odds and ends, machine parts, but I can tell you that once you add nema-23s and a trim router to the shapeoko 2 the 5mil (or whatever they are) belts won't even come close to cutting it. Or expand the system even with the stock motors to 1000m and the distance really has issues. R&P gearing is what 90% of production machines use. And let me be perfectly honest here, the idea of being out a few thousandths here or there really doesn't bother most people that use a machine for volume. Perhaps if you were prototyping, milling, or making intricate puzzle pieces that might be an issue, but since the work we do is much like a cabinet shop the machine is well within specs.

If I had the money I'd buy a SB Alpha because it would do everything I need and a bit faster. But when you're running 60 80 120 IPM with .4" cuts on walnut on a machine that's sub 3K, that's pretty darn impressive.

I know that y'all are diehard CNC hobbyists, but there is a much larger group of people that have some cool ideas and they think they can make some of them. They don't want to kit together, or source parts, and certainly they don't want to experiment. They just want to make something - That's where this product falls in. Just like the Shopbots - They just "do".

Gadgetman!
Posts: 1061
Joined: Sun Aug 25, 2013 12:44 pm
Location: Norway
Contact:

Re: Gryphon CNC extrusion

Post by Gadgetman! » Tue Apr 14, 2015 10:30 pm

I have a 1000x1000 ShapeOko, the first generation with 'tinfoil-thickness' (OK, they're about 2mm thick steel)motor mounting plates on the ends of the X-axis, and with my NEMA17, 6mm wide belts and a putzy 300W Chinese spindle, I get better than .5mm every bl**dy time.

the NEMA17, the pulleys, the belts, even the Makerslide rails and V-wheels are all stock items that can be sourced from multiple sources.

By the time I have replaced the steel with 6mm 6082 aluminium and he belts with 9mm wide belts, I expect the accuracy to improve quite a bit.
I expect that to happen at about the same time as those aluminium racks of yours have worn down... Which alloy is it, anyway?
(Get the distance between the rack and the pinion just slightly wrong, say 0.1mm, and all bets are off)

The fact is that your Y-rails is a horrid waste of material since the core area doesn't add any rigidity to the design, just weight.
There's no mention what the racks are made of, really(we just assume that it's aluminium, and hopefully of the same exact alloy as the beam itself as different alloys may have different temperature coefficients)


I notice that there are 6 V-wheels on the side plates... Notice how they aren't placed in pairs, but are slightly offset from each other?
Did no one tell them about bending forces?

If it works for you, how nice.
A pity there's so many others with a rather different experience.

Here's a good one:
https://www.youtube.com/watch?v=5d4LCG1QcLs
Feel free to watch the whole thing, or just skip to around 27minutes and listen to why it skips while tying to lift the spindle...

;-)
Weird guy...
Shapeoko 2014F: 1000mm X/Y, 300W Quiet Cut spindle, Arduino Uno/G-shield and GRBL 0.9i
15x30 drag chains, custom spindle mount, 9mm belts, 8mm endplates, 6mm motor plates.

WillAdams
Posts: 8542
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Gryphon CNC extrusion

Post by WillAdams » Tue Apr 14, 2015 10:35 pm

Right, but this is a forum for discussing a particular belt driven machine, and CNC in the context of that machine. If you wish to discuss the CNC Mogul / Gryphon specifically, you’ll need to find another forum.

That said, one thing I’ve been curious about is a description of the different options for linear motion and of their relative merits and tradeoffs.

Further discussion of such would be welcome.
Shapeoko 3XL #0006 w/ Carbide Compact Router w/0.125″ and ¼″ Carbide 3D precision collets
Nomad 883 Pro #596 (bamboo)

joelcrawford
Posts: 8
Joined: Sat Oct 25, 2014 1:57 pm

Re: Gryphon CNC extrusion

Post by joelcrawford » Wed Apr 15, 2015 12:41 pm

Gadgetman! wrote:I have a 1000x1000 ShapeOko, the first generation with 'tinfoil-thickness' (OK, they're about 2mm thick steel)motor mounting plates on the ends of the X-axis, and with my NEMA17, 6mm wide belts and a putzy 300W Chinese spindle, I get better than .5mm every bl**dy time.

the NEMA17, the pulleys, the belts, even the Makerslide rails and V-wheels are all stock items that can be sourced from multiple sources.

By the time I have replaced the steel with 6mm 6082 aluminium and he belts with 9mm wide belts, I expect the accuracy to improve quite a bit.
I expect that to happen at about the same time as those aluminium racks of yours have worn down... Which alloy is it, anyway?
(Get the distance between the rack and the pinion just slightly wrong, say 0.1mm, and all bets are off)

The fact is that your Y-rails is a horrid waste of material since the core area doesn't add any rigidity to the design, just weight.
There's no mention what the racks are made of, really(we just assume that it's aluminium, and hopefully of the same exact alloy as the beam itself as different alloys may have different temperature coefficients)


;-)
My Y rails are Maker Slide which is the same as Shakeoko 2. If you're talking about the X rails, then you're wrong. I could stand on a piece of 4ft, which I couldn't do on maker slide. Listen you need to look at how production machines work. There is a huge disconnect between the home hobbits and how a real machines are built - Both for space and money reasons. I don't know of a production class machine that uses belts. I know of only a handful that use screwdrive for the X - I don't know ANY that use belt - It's just a bad way of doing it when you have machines flying around at 100-200ipm with 1" cuts. I have no doubt that my rack and pinions will wear out over time, but they haven't yet, there's no loss in accuracy yet and we are over 1000 hours. If they fail by 3 or 4 thousands of hours the machine will have generated 50-100x it's cost so buying a few rails isn't a big deal.

As for the video, I've seen it. But I have first hand experience here.

I own a Gryphon and I own a shapeoko with upgraded everything (seriously everything) and I've run shopbots and I just wanted to clear up some of the things that people mistakenly said without any direct knowledge of the machine or how production machines vary from desktop machines.

In the end I need numbers. I need a production level machine. I need parts per hour and a machine that's not broken all the time. Seriously, we run it 10 hours a day. So far so good.

Hans
Posts: 290
Joined: Tue Dec 25, 2012 5:19 am
Location: Connecticut, USA

Re: Gryphon CNC extrusion

Post by Hans » Wed Apr 15, 2015 2:46 pm

I think it's a silly argument that the beefier rails are inefficient. It's a simple fact that they're several times stiffer than Makerslide. It's also a simple fact that the S1 and S2 are light-duty machines. There's no reason to take offense, it's a simple fact: for cabinet making, the drive system of a stock S1 or S2 is too weak and too flexible for high-feed woodworking. You can't buy a Shapeoko that fits a full sheet out of the box, and a lot of people don't want to mess with customizing everything. At the same time, I wouldn't argue over "production class" because it depends on what you're producing. For light applications like engraving and small parts the Shapeoko is a great machine. For cutting 1" plywood, I doubt even a Gryphon is very competitive if it can't push a .5" tool at 800 ipm (~200 lb cut). Even a ShopBot will only do 240 ipm with a .25" 2-flute endmill .25" deep, that's a 40-pound cut. People find success on a stock S2 at more like 4 pounds cutting force.

Joel, may I ask what tools/WOC/DOC/feedrate/rpm are typical for your production work? I want to calculate what sort of cutting forces you've found usable on your machine.
Shapeoko #1,123

Gadgetman!
Posts: 1061
Joined: Sun Aug 25, 2013 12:44 pm
Location: Norway
Contact:

Re: Gryphon CNC extrusion

Post by Gadgetman! » Wed Apr 15, 2015 3:39 pm

No one questions that profile's ability to hold a person's weight...
Still doesn't excuse the horrid design with lots of 'useless' aluminium in the middle doing nothing but adding weight to the design.
(And cost... Not just in the material, but the die for extruding it must have been horrendously expensive... sealed cavities are difficult to extrude)

Doesn't the Gryphon have two of those beams, one above the other?
Should it really need that?
Or could it be that the beam doesn't resist twisting force significanly better than a pair of Makerslides bolted together?
(You need unbroken flat skin on the outside to add that kind of stiffness. you know, beams like cvoinescu's MegaBeam, or the beam used on the SO3)

So, your machine may do its job, but that's more in spite of than because of the engineering choices made.
Weird guy...
Shapeoko 2014F: 1000mm X/Y, 300W Quiet Cut spindle, Arduino Uno/G-shield and GRBL 0.9i
15x30 drag chains, custom spindle mount, 9mm belts, 8mm endplates, 6mm motor plates.

LouisV
Posts: 316
Joined: Fri Dec 13, 2013 3:41 am

Re: Gryphon CNC extrusion

Post by LouisV » Wed Apr 15, 2015 6:38 pm

Just my two cents but CNC Mogul and Gryphon must be owned by the same guy or at the very least both companies are working together. I say this because their machines are identical other than being different colors. CNC Mogul is not an open source machine, so it begs the question how the other company is able to sell an identical machine if they're completely separate from one another with no affiliation. Just an observation :geek:

Secondly, belt drive systems are not to be underestimated. Comparing a rack and pinion system to 6mm wide 2mm pitch belts is one thing, comparing them to 9-15mm wide 3mm pitch belts is another. Through my testing I've reached very high feedrates with great results. I love rack and pinion systems, (take a look at CNC Router Parts), but if they're not designed right or they skimp out on using strong materials then I question if its really worth it.

joelcrawford
Posts: 8
Joined: Sat Oct 25, 2014 1:57 pm

Re: Gryphon CNC extrusion

Post by joelcrawford » Wed Apr 15, 2015 6:42 pm

Hans wrote: Joel, may I ask what tools/WOC/DOC/feedrate/rpm are typical for your production work? I want to calculate what sort of cutting forces you've found usable on your machine.
Of course the speed is dependent on the material. On hard maple with a .25 uncut we make .3"-.4" passes around 90-110ipm. On walnut we take it down to 80-90 or so. On baltic birch ply we're running a compression bit and we need to dig in for the first pass usually at .4+ and that drops to about 60ipm.

Upcut bits get run the fastest, downcuts and compression bits get about a 20-25% knock on speed because of the waste interference. 1/8" and 3/16th bits run much slower to allow for failure stress. v-carve bits I run in the 160-200. Plastic or alum thin stock gets run at 200ipm on single flute upcuts single pass.

I'm thinking I could get away with 120ipm minimums on the hardwoods when I double the HP on the spindle. That's the biggest drawback is that the machine can handle the forces but the router is the point of fail for anything more high speed. Rapid rate we have set to about 300imp.

The steppers are big enough that when you forget where you are standing the side plate assemblies will knock you over. I've skipped the steppers but only doing stupid things, I've never skipped them because of standard cut - long before that happens I'm killing the router and need to back off. Perhaps I'll find that point with the spindle upgrade.

joelcrawford
Posts: 8
Joined: Sat Oct 25, 2014 1:57 pm

Re: Gryphon CNC extrusion

Post by joelcrawford » Wed Apr 15, 2015 6:55 pm

Gadgetman! wrote:Doesn't the Gryphon have two of those beams, one above the other?
Should it really need that?
Or could it be that the beam doesn't resist twisting force significanly better than a pair of Makerslides bolted together?
Actually I'd rather like it to have 3. If that was the case it would support a 5hp or better spindle. The worst design element (imho) of the shapeoko 2 was the use of double *unattached* rails at the X. So much twist it was unusable with any real speed or weight or any combination of the two. The gryphon system fixes that on two critical levels:
a. it's a hunk of rail
b. it's two of them

That was a huge selling point for me coming from a supercharged S2.

Seriously, look at some of other kits or some of the Chinese import CNCs. X rails are often times 6" x 6" alum or steel beams. I'm not an engineer but I'm going to guess that the main issues for design for a CNC is all about the X and how the gantry mates to it and the performance of those two pieces. Again, having fiddled with these machines now for some time, that's the money shot - the synergy between those parts and the performance that you can get out of it.

joelcrawford
Posts: 8
Joined: Sat Oct 25, 2014 1:57 pm

Re: Gryphon CNC extrusion

Post by joelcrawford » Wed Apr 15, 2015 7:12 pm

LouisV wrote:Just my two cents but CNC Mogul and Gryphon must be owned by the same guy or at the very least both companies are working together. I say this because their machines are identical other than being different colors. CNC Mogul is not an open source machine, so it begs the question how the other company is able to sell an identical machine if they're completely separate from one another with no affiliation. Just an observation :geek:

Secondly, belt drive systems are not to be underestimated. Comparing a rack and pinion system to 6mm wide 2mm pitch belts is one thing, comparing them to 9-15mm wide 3mm pitch belts is another. Through my testing I've reached very high feedrates with great results. I love rack and pinion systems, (take a look at CNC Router Parts), but if they're not designed right or they skimp out on using strong materials then I question if its really worth it.
They are two separate companies is all I know. I believe that have some interaction when it comes to having the parts made (which I believe is all made by local suppliers with the exception of some bearings and odds and ends) but I'm not paid by either and I only know Phil the owner of Gryphon Systems thru phone calls asking about this or that. (I know you're thinking this guy must work for them to be so passionate - but y'all made me this way with the guess work in the thread every time I googled gryphon and saw it. It was a hatchet job, seriously. So I figured I'd add some factual input.)

Belt drive is fine for short runs. And your right r&p systems do need to be tight and I think there's a weird partnership with the aluminum and steel gearing in that regard, but again as is with 1000s of hours it's still painful to run your fingers down the rail because they are sharp.

The thing is that even the system you are working on with kickstarter, LouisV, if you built it to the same size as I've got, while I'm sure it's a fantastic machine as is, I would have to wonder where the money went because it would be more expensive. Even without having matched the electronics or steppers, it would be hundreds of dollars more. And that's fine, you have a nice hobby machine and that's your niche (btw, nice job on funding).

See? This is the point I'm making. I researched this for months and made 15 calls overseas just to make sure I didn't miss some CNC out there. The Gryphon CNC is a massively amazing deal for the money. I don't know anyone with the same benefits (or even different ones) that comes close for the feature to dollar ratio.
Is it perfect? Nope. But it exceeds my expectations at that level price.

Post Reply