Possible Shark 3D CNC Router

[Gerry Grzadzinski]

I'm not sure if it's a typo or Ted knows something I don't, but his $200 number is not accurate based on what I know.

I believe that Ted is saying he can change two extrusions and turn his 4x4 into a 4x8, for about $200 for the two extrusions.

My experience is the weak point is one of the following.

1. Open loop stepper motors.

Millions of machines of all types use open loop steppers every day with 100% reliability. They are not a weak point, if used properly. Steppers have a finite amount of torque. If your machine needs more than they can provide, bad things happen. Run them within their limits, and they'll provide trouble free service for a very long time.

The other issue I have is the machine doesn't do what I'd like when going around curves or making turns. On a straight line, I can comfortably machine a lot of materials at 200+ipm at decent depth. When it's time to turn or go on a curve, it can get hairy.

You never actually say what issues your having here, but it's apparently a control software issue. There are several Mach3 settings that can probably address your issues. And there are several alternatives to Mach3 that work as good or better, if Mach3 won't do what you need. You're not locked in to Mach3 if you buy a CNC Router Parts machine.

[Ross Moshinsky]

I believe that Ted is saying he can change two extrusions and turn his 4x4 into a 4x8, for about $200 for the two extrusions.

Which isn't accurate as far as I know. Unless I missed something he'll need to lengthen two pieces of extrusion, add several cross beams, extend rail system and extend the rack. All of which add up to more than $200.

Millions of machines of all types use open loop steppers every day with 100% reliability. They are not a weak point, if used properly. Steppers have a finite amount of torque. If your machine needs more than they can provide, bad things happen. Run them within their limits, and they'll provide trouble free service for a very long time.

Millions of machines do use open loop steppers but millions also used closed loop. With the machine talking to the control system and the control system talking back, you're going to get better results. I'm well aware you know more about CNC machines than me but this is a universal truth.

You never actually say what issues your having here, but it's apparently a control software issue. There are several Mach3 settings that can probably address your issues. And there are several alternatives to Mach3 that work as good or better, if Mach3 won't do what you need. You're not locked in to Mach3 if you buy a CNC Router Parts machine.

The issue is the machine looses accuracy and can jump a tooth. As I said, my machine could be the contributing factor BUT you're talking about a home made, bolt together machine, with an imprecise rail system. Mach3 is an EXCELLENT value but I've yet to find a way to cut in a straight line at 200ipm and then have any control of the speed in the curves. It will slow down, but it's constantly fighting to get back up to 200ipm, which can be problematic. It wouldn't doubt me one bit, there is a software solution to my problem but I haven't found it.

My point is that when you make an extrusion router, it's normally coupled with Mach3, and it's really up to you to dial in everything and Mach3 may or may not be able to do what you want. ShopBot for example has their own control software that's dialed in for their machines. I doubt they allow you to dial in those settings but theoretically, they've already done it for you. If your machine isn't behaving, you can rule out software and go to hardware.

[Gerry Grzadzinski]

With the machine talking to the control system and the control system talking back, you're going to get better results.

I disagree. If an open loop machine is not losing steps, then a closed loop stepper machine will not give better results. You may be able to get better performance, but also at a significant increase in cost.
Recently, however, lower cost chinese closed loop steppers have started to appear, but I've yet to see any DIY systems using them, and I'm not sold that they'll actually improve performance.

The issue is the machine looses accuracy and can jump a tooth.

OK, that's definitely a mechanical issue, and not a Mach3 issue.

Mach3 is an EXCELLENT value but I've yet to find a way to cut in a straight line at 200ipm and then have any control of the speed in the curves. It will slow down, but it's constantly fighting to get back up to 200ipm, which can be problematic.

Mach3 *should* be cutting the curves at the same speed. It always has for me.
It sounds to me like your *curves* are actually a series of straight segments, and mach3 is trying to maintain a constant velocity through these sections. Messing with the CV settings should make a difference.
If your curves are G2 and G3 moves, then it should be running at a steady feedrate through the curves.

ShopBot for example has their own control software that's dialed in for their machines. I doubt they allow you to dial in those settings......

Actually, this page discusses how to "tune" your ShopBot for different types of cutting. It's not necessarily "dialed in" from the factory.
http://www.shopbotblog.com/2008/03/a...m-performance/

[Roy Harding]

... ShopBot for example has their own control software that's dialed in for their machines. I doubt they allow you to dial in those settings but theoretically, they've already done it for you. If your machine isn't behaving, you can rule out software and go to hardware.

Shopbot ABSOLUTEY allows one to dial in those settings. It's impossible for a factory to set rigid settings that will work for all users. The correct settings depend upon what I'm cutting (species of wood, bronze, aluminum, plastic, etc), and how I'm cutting it (straight parts cutting, v-carving, relief ("3D") carving, etc, etc). In my own case, I have a variety of settings files which are fine-tuned to what I'm specifically doing. Gerry linked to an excellent article which explains the basics of this process for Shopbot users - one assumes that the principles discussed in that article apply to all control software, including Mach 3.

All successful control software (which would include Shopbot and Mach 3) HAS to enable the user to change these settings, or the machine would quickly become useless to the majority of owners.

[Ted Reischl]

I'm not sure if it's a typo or Ted knows something I don't, but his $200 number is not accurate based on what I know.

The reality is extrusion routers work reasonably well. The bolt together nature of the machine is rarely the weak point. My experience is the weak point is one of the following.

1. Open loop stepper motors.
2. Low tolerance rail systems.
3. Under developed R&D resulting in various levels of design issues.

For example, I have an early 4x8 CNC Routerparts machine. I have two "major complaints". The first being the machine is simply too small by 1-2". To mill the spoilboard, you're right on the edge of the rail system. Making it 2" wider and longer would make life much easier and allow not only milling of the spoilboard to not be a nerve wracking event but also allow someone to throw a 4x8 sheet on the table and cut it from edge to edge. This may or may not have been resolved in newer models. The other issue I have is the machine doesn't do what I'd like when going around curves or making turns. On a straight line, I can comfortably machine a lot of materials at 200+ipm at decent depth. When it's time to turn or go on a curve, it can get hairy. I'll admit, my machine probably isn't the most tuned in and tightest it could be, but I haven't figured out a way for the machine to slow itself down enough where I'm comfortable running at the higher speed. In the end, I've found going deeper and slower works for me better, but there are some jobs I feel I should be able to run faster and simply can't as it's not worth the risk. I should also mention, this could be down to Mach3's ability to tune in those changes, but that's part of what you're buying into when you go with an extrusion machine.

The alteration to my machine only involved raising the Z axis about 6 inches overall. So I purchased two extrusions that were 10" and replaced the 4" ones.

I did not buy a complete kit from CNC RouterParts. Being a bit on the cheap side I hunted around the internet and found the extrusions for the ways at a substantial savings. Since I was doing that, I made sure I sized everything to give me enough travel to overshoot the spoil board. The actual travel on my machine is 39 X 51, but I refer to it as a 3' X 4'.

Do I think that a welded machine is more rigid than a bolted machine? Positively. No doubt in my mechanical mind. However, most hobby guys are not equipped to do welding, nor would they be happy cutting the machine apart to make alterations.

The rest of your comments about feeds has mostly to do with how Mach 3 is set up, and also how much the gantry weighs. Some of the guys focus on rigidity until they wind up with a grossly overweight gantry and then wonder what happened to their speed.

Here is a video of my machine cutting at 800 IPM, about 500 IPM on curved surfaces:

I apologize for the lousy focus. . . cell phone cameras, grrrrrr. And no, the film was not speeded up at all.

https://www.youtube.com/watch?v=Pt8n43_YRjI

I do not cut at those speeds, scares me a bit.

Did a few other things when I built the machine, instead of using aluminum plates to connect major components, I used steel for stiffness. I did not use cheezy L shaped brackets to connect things either, I tapped the end of the extrusions so I could use 5/16 bolts.

Just for the record, I would never state that an assembled (erector set machine) will be as rigid as a commercial machine. But for a hobbyist wanting to cut at 250 IPM? Not a problem.

[Ted Reischl]

Which isn't accurate as far as I know. Unless I missed something he'll need to lengthen two pieces of extrusion, add several cross beams, extend rail system and extend the rack. All of which add up to more than $200.

What "crossbeams"? The rest is correct, but I have done the research and it is about $200.

The issue is the machine looses accuracy and can jump a tooth. As I said, my machine could be the contributing factor BUT you're talking about a home made, bolt together machine, with an imprecise rail system. Mach3 is an EXCELLENT value but I've yet to find a way to cut in a straight line at 200ipm and then have any control of the speed in the curves. It will slow down, but it's constantly fighting to get back up to 200ipm, which can be problematic. It wouldn't doubt me one bit, there is a software solution to my problem but I haven't found it.

The machine does not "loose accuracy and can jump a tooth". It looses accuracy WHEN it jumps a tooth. Mine doesn't. "Imprecise rail system"? Ever look up the tolerances on CRS plate? Then buy some and find out it is way more accurate than the specs? A piece of CRS plate is extremely straight and the thickness is consistent.

If you are cutting straight lines and then expect the machine to maintain 200 IPM in tighter curvy details you need to understand the physics of motion a bit better. Formula 1 race cars slow down for a reason in the curves, it is called G forces. Those G forces get too great and the car flies off the curve. Same thing on a machine.

I posted this on another part of this thread, but if you would like to see a bolted together, non rigid, inaccurate ways machine cutting at 800 and then slowing to about 500 on curves, have a look, BTW, that is a .25 endmill hanging way out of the collet:

https://www.youtube.com/watch?v=Pt8n43_YRjI

[Jim Gunter]

I want to thank everyone for your input. I decided not to get the Shark 3D on Craigslist. Cannot say I know exactly what I will end up getting but it will be larger than the one for sale. Going to save up a while longer and hopefully get a Stinger, Shopbot or another in that category.
In my quest for the best fit for me, I am going to start another thread and ask about the software that is needed to make the machine operate. Proprietary, open access, easy, cumbersome, $$ etc.
Thanks Again!

[Art Mann]

I have only run my CNC router for a few hundred hours but I was a manufacturing engineering team leader for an OEM automotive electronics plant for a few years. I had responsibility for the purchase, operation and maintenace of millions of dollars of off the shelf to full custom robotic manufacturing equipment. We bought machines with both open loop steppers and closed loop servo motors, depending on the application. The idea that stepper motor motion control is inherently inaccurate or unreliable in a small CNC router is simply wrong. There must be 10's of thousands of stepper motor CNC routers across the country churning out high quality work all day every day. I know for a fact that similar machines are cranking out millions of electronic assemblies daily.

I talked to two different engineers/technicians at CNC router Parts on a few occasions. I was impressed with their designs, for serious hobby or light commercial work. Their kits are substantially better than the low cost turn key machines you can buy on the internet. Of course, the quality of a kit built CNC router is highly dependent on the person doing the building. I didn't build one because I didn't want to waste a month of time assembling metal rather than cutting wood. They now offer a couple of smaller machines which are more complete and can be assembled in days.

[Mike Heidrick]

Art, a cncrp build can be hours not weeks. You will need a base and controller but me and a buddy built a 4x4 one in less than nine hours and it was jogging at that point. Had only a few pieces of extra hardware left. Very nice kit.

[Dick Jensen]

Howdy, I'm new to this website but need to add my insight.
I bought a Shark HD 2.0 from Rockler when I retired 1 1/2 years ago. It does everything I want and does it well. The company is very responsive to any questions or problems - which is a BIG plus. They also make a scaled down model called the Piranha which I also bought recently. It is limited more in size but is just as accurate as the larger Shark.
I did a test drive of about 5 different CAD/CAM programs before settling on Aspire from Vectric. I concluded that there are 2 different approaches to designing in 3D. The majority of programs are, understandably, written by "engineering" types. But I don't think that way. So I chose Aspire because it handles designing the way an artist thinks. Their included tutorials are very understandable and well worth the time you spend on them. This software is amazing.
Since I retired to Utah (which still baffles me) I have created a new market in doing carvings of the Mormon temples.
I do think $3,000 is high for second hand.

Hello CNC community. I am a home hobby wood worker and have been making furniture from Oak rough boards for some time now. Mostly for home and the kids. Bedroom sets kitchen cabinets, living room furniture etc.
I plan to retire from my paying job in a couple years and thought about adding CNC routing to my shop. Some sign making, special touches to furniture, relief carving and who knows what else.
I see there is a used Shark 3D for sale at $3,000 locally. Would this be a good machine to get into?
I believe in buying good equipment once (if possible) but do not plan to go into high volume production.
Thanks in advance for your insite.
Jim

[jim cox]

I bought the CNC shark about five years ago from a Rockler store, I really was a novice as far as how a CNC really worked. I knew what they did and their capacity in a workshop, but how to run it, I was lost. The shark may not be the best one out there as far a the bed, (they have become better now) but as far as downloading pictures of anything you wanted to add to a sign or sign making, the program is very easy. This is worth the money to me. If you feel you need to do mill something more than 29 inches wide, you might want to look at something else, but as far as the learning curve, the quality of the machine, its a good choice. one thing that i did do was narrow my choices down to three, I called each companies support number to see how long it took for them to respond to a simple question, one company it took over a week, the second, three days and the shark, I had someone on the phone right away and they took the time to answer my question. think about this when you are in the middle of a project that a customer is waiting on. Customer service is a huge factor.

[Jerome Stanek]

Shopbot has a new desktop called the The Shopbot Desktop Max with a real spindle

[Dave Martin88]

My experience with a Shark HD machine is much like Dick Jenson's. I've had mine for about 2 years and has done almost everything I've asked of it. They have a good following of users and supporters so easy to get help if needed. I also use Vectric programs, started with VCarve and then upgraded to Aspire. I've done a bit of 3D work and the Shark handles it well. Assuming the Shark for sale is an HD model and has some decent add ons, the $3K isn't a bad starting point but would expect it to sell for closer to $2500 as long as it includes at least some software like VCarve? Software can be expensive so if nothing is included, think the price is really high.

I do have a horse in this race since I own a Shark and getting ready to put it up for sale- I want a much larger cutting surface so have ordered a Camaster Stinger II 4x4. If I had the room, I'd keep the Shark as it is fully functional but just not enough room for both. For carving signs and basic 3D, the Shark does well, maybe not as fast as the big machines but I'm retired so can take the time! The software supports a "tiling" function that lets you index your projects to cut larger than the machine bed along one axis.

Jim, good luck on finding a router. Can't argue with the advice to buy the second machine you want first but the cost difference is significant- I decided to go with a more entry level machine to find out if I was going to like working with a CNC, see if there was a market for products in my area, and also to start learning because I personally think the learning curve is a bit steep if you plan to do your own 3D design work, etc. Obviously, prior CAD experience changes that. The Shark has done all this for me but I'm sure a lot of the other machines that people have mentioned would do as well or better. I had no interest in a DIY CNC but for many, that's part of the fun. I wouldn't build my own table saw so didn't want to build my own CNC. Enjoy the hunt.