My First Post

Ever since we upgraded our website this Spring, I’ve wanted a space to write about the things that are going on at ShopBot. It’s a bit of a busy place, but also exciting. And, we are certainly hard at work developing new CNC stuff. I’ll try to make a contribution every month or so and will also recruit some commentary from others here at ShopBot. This first column is a bit on the technical side, but expect the themes to be pretty wide-ranging.

Recent suggestions on the Talk ShopBot Forum that we could do a better job of communicating actually provided the impetus to get me started here. Hopefully this webcolumn will help serve to keep everyone up to speed on some of the things going on here at ShopBot, as well as provide an opportunity to reflect on issues that are hot topics on the Forum or in the CNC business in general.

Gearing and Resolution

The question of how to improve ShopBots and the cutting they do is probably something we spend 85% of our time thinking about, so I’ll comment on some of the specific related issues that have recently come up on the Forum. The first of these is gearing and resolution. Gearing and higher resolution are good things, and assuming that the motor can keep up and the backlash is minimal, the more the better.

Early PR Belt-Box

Along with a number of ShopBotters, we also like belt-box gearing using no-stretch timing belts. Getting reduction this way has a number of advantages, one of the most important being that it dampens some of the vibration of the stepper and gives smoother motion without a lot of backlash. Our first ‘PR’ CNC tools used belt-boxes. The drive system worked very well and the photo on the left illustrates the idea. Our problem was that we had trouble reliably building these boxes at reasonable costs. We eventually decided that the Oriental Motors (OM) tapered-hob gearhead gave us motion that was nearly as good, from a gearbox that was pretty bullet-proof and at a cost that would work for the PRT. We have had practically no failures in thousands of these gearboxes over the last 7 years. The fact that many customers with 6- or 7-year-old PRTs are now considering upgraded control boxes to run these same motors is a pretty good testimony to their utility and durability. I believe that OM makes a very good product.

But, if we could get a tight, robust belt-box, we would definitely consider it. Here’s the spec: We would want a double reduction (so the pulleys did not get so big they are in the way) gear box with something between 5 and 10:1 reduction (7.2:1 is good). The box needs to be able to handle overhung and thrust loads (the latter to allow the option of helical rack and pinion). Output capability should be to about 600 RPM. Oh yes, and at a reasonable price because for our typical motor/gearhead, the gearhead cost is already well over half the cost of the unit.)

That said about gearheads, there is also an advantage to straight-drives. They are simple, straightforward, and cost-effective with few parts to wear. To a degree, resolution can be achieved with good microstepping, and the microstepping certainly does reduce/eliminate the tendency that steppers have to resonate. On our PRTalphas, we believe that for many standard woodworking applications the ~1200 steps/inch resolution provides pretty smooth motion, while benefiting from the advantages of a straight-drive, and from the significant robustness of the closed-loop drive system.

The down side of microstepping is that as the load on the motor is increased, the microstepping motion can become less and less linear, even though it sounds smooth. Aggressive cutting of hard materials can certainly benefit from higher mechanical gearing. It provides more mechanical resolution so that there is less reliance on microstepping for positioning, at the same time it provides more power so the motor is not as heavily loaded. This works for smoothness in cuts and against chatter. We did not initially offer gearheads as an option with PRTalphas, but for customers who can benefit from additional power and resolution we now offer PRTalpha tools with a 7.2:1 tapered-hob, gearheads (and similar upgrades for existing PRTalphas).

Motor Amps

The Geckodrive stepper drivers that are used in the new PRT Version 4g Control Board provide 2.5 times more resolution and significantly more speed (6-8 in/sec cut speed; 12-14 in/sec positioning speeds) than the old version of the PRT Control Board (v3). This improvement was made possible for us by the higher speed controller that we developed for the PRTalpha. Because these drivers are also capable of handling more current and voltage than the present PRT Control Box, it has been suggested that we switch to higher current motors and higher voltages. While it is certainly the case that we could get somewhat more performance with these additional changes, for the moment we aren’t doing it. Here’s why:

1. By using the existing motors at ~ 2amps we have a solution that works well for existing PRTs. The present board is a straight-forward, drop-in replacement for existing ShopBots. It requires no additional power supply or heat sinking and works well with existing motors. The simplicity of the Control Board and driver mounting allows us to provide it at a very low cost. The performance improvement is very impressive.
2. Keeping the current demands on the Geckos relatively low means we do not have to worry about heat from the Geckos or heat-sinking them. This allows a simple, low-cost, connection scheme. The Geckos snap easily in and out of the PRT 4g Board.
3. By not increasing the voltage or current requirements for the power supply, we can stick with an enclosed, commonly available, UL/CE compliant power supply.
4. We are not pushing the capabilities of the drivers. As drivers are the most vulnerable component of a stepper system, this seems prudent.
5. Additionally, over the years we have come to appreciate that the power of these PRT motors is about right for the tool and the cutting that ShopBotters are doing. The primary difference is that now, rather than cutting at 1-2 inches/sec, they will be able to cut at 4in/sec with nearly full power and at 6-8 in/sec with pretty good power.

In short, the current motors still seem to us a good, conservative, and practical choice. Yes, we could get a little better performance by beefing up the board, motors, and power supply; but it does not seem to be a good tradeoff at the moment or for the typical uses of our tools. But we’ll remain open on this one.

After-Market Control Boxes

The above topic leads into another — the potential usefulness of the after-market control boxes for PRT tools such as the one developed by Dirk Hazelenger.

Dirk was interested in exploring Geckos and other drivers to run various of his CNC projects. After talking with us about how the PRTalpha controller could be used to operate other drivers, Dirk experimented and found that the Geckodrives worked well for him (actually, he was very excited about how well the combination worked) and put together a full control box to take advantage of them. [See: www.botrods.com]

Dirk’s box is different than the Version 4g board. In concept, it is more like the Version 4 board described in the system layout on our developer’s page in offering a universal interface. More importantly, it is a direct response to the interest of several ShopBotters in using higher voltage power supplies and higher current motors. Unlike the 4g, in Dirk’s box the drivers are individually wired, mounted, and heat-sunk. This means that the box is ready to run a much wider range of motors, including the 4amp OM motors that have attracted attention. You could buy an interface board from us for this type of project, but you would have to do everything else yourself. Dirk’s board is ready to go. It seems to me to be an excellent option if you are interested in pushing capabilities to the max with your present PRT. Dirk’s board is also a great option for people looking to retrofit existing CNC tools or to add CNC capabilities to milling machines or other equipment and begin with a full motion and control system. For this latter application, because of the way the controller is sold to Dirk, he will also be able to provide a good CAD/CAM design system at a reasonable price.

In making his system market ready, Dirk has had one glitch. That was the initial choice of the Geckodrive 212 as the driver. I wanted to make it clear here that that choice was totally my error. Based on a quick look at the specs and some experience with other Geckos, it looked to me like the 212 would be the easiest way to connect to existing PRTalpha electronics and so I suggested it. It turns out that because of the way the 212 step-multiplier works the 212 is simply not suited to any kind of CNC action in which there are a lot of very short back-and-forth moves as there will be in a 3d carving file. The 202s don’t have this problem, and Dirk quickly converted to them (which is probably what he would have used originally if I hadn’t pushed the 212s). I think the control boxes are working great now, and there is some sample work elsewhere on the Forum.

There are other products as well. I know less about them so won’t comment here. However, I do want to emphasize our general support of all the variations of ShopBots that ShopBotters want to explore. There is nothing I find more interesting than visiting a ShopBot shop and see the incredibly inventive and creative use that our tools have been put to and the clever ways in which they have been modified. In many ways ShopBots are the ‘erector sets’ of robotic tools and we find all the new twists, inspiring and exciting.

Back to Cutting

Here’s an applied example of the upgrade discussion. Chris Burns, here at ShopBot, has a home shop with an old PRT. At the moment, this tool is his “Tech Ed Dept” for schooling his kids. He and son Bryan (6) recently became fascinated by 3D carving, partly stimulated by James Booth’s 3D clip-art files, but also encouraged because of how much faster 3D goes now that he is using a PRT with Gecko drives. Chris finds he can cut 3D at 8 in/sec in XY and 8 in/sec in Z. The higher resolution stepping of the Geckos helps, but so does being able to maintain a good full speed in both the XY and Z for contouring.

Here’s one of Chris’ and Bryan’s new bread boards on which they have put one of James’ carvings (cut bread on one side; have a pretty image when the other side is facing). This one took about 40 min to cut, ash was from a tree cut in his yard I think …[Bryan ran the tool and finish-sanded the part.]

Coming next …

I’ll plan to report on several development projects ongoing here at ShopBot.