Some performance questions on Desktop

[Tomlin]

Hi everyone

New guy here to the forum, thanks for your kindness and patience.

I have a few questions that I've not seen answers to yet and would like to pose here. First a little background:

I'm currently using a Probotix V90 and while this machine has served me well for the past 2.5 years I am ready to move up to something much more rigid and that can handle a sizable increase in feed rate and function more like a small-production machine. Because this machine is made of MDF, PVC and other non-metal construction, I get a lot of gantry sag issues in the middle of the machine that causes issues when milling out acrylic products that I make. The V90 doesnt run well for me over 28 IPM, so I need something much faster. Naturally, I have two main concerns with whatever new machine I go with:

1) A very sturdy/rigid machine (especially flat table and well-designed gantry)

2) Much faster feed rates in .125 and .25" cast acrylic.

Question 1)- I see per the Desktop specs, the machine is capable of cut speeds at 4" per second (I calculate to 240 IPM). I also see per some charts that I can run between 75-100 IPM using .125" bits in cast acrylic.

My main point here is, Would this be too much to expect from the Shopbot Desktop machine? In other words, specs on the "label" are one thing but as I cost out production requirements, set expectations, etc. I would like to have an honest feel as to if 75-100 IPM would be way faster than what is "safe" for this desktop machine.

Question 2)- I see that this machine has the T-slot table, which I think we can safely say is going to be very flat.

Is it reasonable to expect to be able to get a truly flat surface by attaching an mdf spoil board to the T slot and surfacing? This is important because the parts we make have a critical thickness that gets pretty thin and we have about a 10-20 percent rejection rate on parts due to gantry issues.


Other thoughts:

I'm super interested in the laser engraver option- any news on this upgrade?

I know that from testimonies the Desktop is a great machine but as you can tell, I have been working with a hobby level machine for so long now that I am a bit anxious about making sure that those quirks are not repurchased in a new machine.


Thanks!

[bleeth]

Eric:

Those speeds are realistic, but can vary depending on bit selection, quality and sharpness as well as the particular Acrylic being cut. Cast cuts better than extruded and using poly-carbonate (Lexan) may also require adjustment. Smaller bits may require multiple passes and/or ramp-in toolpathing strategies.

Many use an MDF spoilboard. IMHO for the work you are doing an issue with MDF can come in due to its changing shape from moisture absorbtion and evaporation. You may be better off with something more stable (Like PVC) although this will obviously dull your bit bottoms faster on cut through machining.

Contact SB direct for updates on laser engraving. It may be worth your while to run up to Durham and have some test cutting done if there is no-one with a desktop set up for precise Z height cutting (That would be experience and spoilboard choices). You will need to co-ordinate with them to ensure that your design file is translatable to SB conmtrol software correctly.

[Tomlin]

Dave,

Thanks, that is very appreciated. I would love to hear other's input as well.

I do realize that it depends greatly on RPM, Cast vs. Extruded (we only use Cast btw) and bit types, but on the whole, I guess I've just been riding a bicycle for so long with the V90 that it's hard to even imagine being able to cut a part that currently takes 2 hours to mill at 22 IPM at only 1/3 of that time. It's exhilarating contemplating how much more fun and less time consuming the machine time would take.

Would a valid test be to calculate per guidelines what is "possible" and then put a bit in and just start pushing the feed rate up until we develop chatter or a broken bit?!? :D

thanks again!

[bleeth]

Definitely a good way to go. Each project has so many variables that it always makes sense to approach your limits slowly. You may find that your own comfort zone is reached before the machines is! Also remember that ramping for changes in direction reduces actual rate of cut and this becomes more true when your toolpath has lots of corners close to each other, as in clearing a small pocket and/or 3-d work. Even with experience on other machines it makes sense to be prepared to spend some time learning the operational idiosynchrocies of your new machine and being prepared to trash a few parts. Too many folks have gotten a new machine and had a job deadline so close that they set themselves up for problems.
Post in the "looking for shopbotter" section and you may find a desktop owner willing to do a demo for you closer than you may think.

[Brady Watson]

The Desktop will rival the V90 in every category. It is a real hunk of beef - probably every bit of 100 pounds. It is hard to predict just how fast you can cut without knowing the shape of your parts. Naturally a square will cut faster than a big convoluted French curve, like the ones a draftsman would have in his tool kit. It's just like driving your car...You're just not going to do 0-100-0 in the 7-11 parking lot. Acceleration and deceleration can play a large role in throughput. In general, you should have no problem cutting up to 90 IPM with a single flute tool, and up to around 180 IPM with a 2 or 3 flute tool. The DT is a LOT stiffer than the V90, and it is made to run production all day long.

Hold down is an important factor in how fast you can cut as well. I would recommend making a vacuum table to hold your parts while they are being cut, and possibly even a part-specific vacuum jig (Google BradyVac (https://www.google.com/#hl=en&safe=off&tbo=d&output=search&sclient=psy-ab&q=bradyvac&oq=bradyvac&gs_l=hp.3..0i13j0i13i30l3.147.1412.0.2589.8.8.0.0. 0.0.128.664.6j2.8.0.les%3B..0.0...1c.1.2.hp.t6gG4G 1kCiI&pbx=1&bav=on.2,or.r_gc.r_pw.r_qf.&bvm=bv.41867550,d.dmQ&fp=876a991549296b91&biw=1366&bih=639)) to make hold down a no-brainer. If parts lift, you will have to run at a slower speed, especially running an upcut geometry tool made to keep plastic chips from welding onto the parts.

Yes, you will want to fasten a piece of MDF (I'd recommend Medex moisture-resistant MDF) to your bed so that you can maintain its flatness with the tool. Since flatness is critical, I would go so far as to glue the MDF to the t-track to eliminate the chance that there could be a void or at the very least, use thin double sided 'mounting tape' like that used in the vinyl sign industry (check HarborSales.net). Mechanical fasteners will cause inconsistencies right out of the gate - avoid them - flatness is king.

-B

[bleeth]

Hey Brady:

I have to strongly disagree with you on one point.

Most draftsmen today don't even know what a set of French Curves is! They are as foreign to today's draftsmen as a slide rule is to the younger batch of "engineers".

[Tomlin]

Mechanical fasteners will cause inconsistencies right out of the gate - avoid them - flatness is king.

-B

Such great post in addition to all the others- thanks again.

I have to say, your comment on mechanical fasteners is something I was glad to read, although it's too bad because you feel sometimes that it's the best way. Currently we use a double sided tape to hold parts and it works well other than when you cut thru to the spoil board and thru that tape (bit gunk). We use screws to hold the MDF spoil board down to an MDF bed, and then on top of that the V90 gantry is MFD too, so with the Bosch Colt router on the end, it's started sagging over time and we have a bowl-shaped bed that is at it's lowest about .0625" where the edges are level :-(

I'm torn between the CAMaster Stinger 1 and the SB Desktop. The possibility of having an add-on "turnkey" laser engraver is extremely enticing, even at the expense of the lesser weight/ aluminum vs. steel and the smaller size. I need both a quality CNC and a laser engraver so this makes it very nice to consider that add-on in the future if it pans out for SB to offer it.

[Brady Watson]

My advice would be to call SB to find out about the laser engraver. Only they can give you up to date info.

I have a saying I tell customers when I consult with them, "Drywall screws are bad news!" - and screws in general are problematic on a CNC. This is especially true when it comes to your machine bed. Everything moves - expands & contracts with even the smallest change in temperature and humidity. The best advice I could give is to replace the bottom board with a good quality plywood like MDO or birch ply, and then GLUE the MDF/Medex spoilboard on top. That will give you the most stable combo and eliminate dips in the center of your table...although I think a lot of this has to do with the V90s light chassis. This would not be a problem on the SB Desktop, but I would still adhere the spoilboard to the AL bed.

If you HAVE to use screws, I recommend the Kreg pockethole screws...Lowes has them. They don't break like drywall screws and they pre-drill the hole. You have to remember that wood moves - even homogenous laminations/formulas like MDF. If you screw around the perimeter, the boards are locked together at these screw 'pinch points' - and the MDF will 'gather' - causing it to bow like a rainbow in between fasteners...then air, dirt, chips and moisture gets in between layers and it just gets worse as time goes on. When you glue them, they move together a lot better, and a lot more consistently.

If the laser engraving head is not available, consider a stand-alone laser engraver...Personally I would choose a stand alone setup anyway because after buying my 1st SB with plasma cutter on one Z and 5hp spindle on the other, I realized that some things don't mix well...and swapping out heads and setups can be inconvenient when you have to do production. I personally would not want a laser head anywhere near my spindle...dust, debris & optics don't mix well.

-B

[shilala]

I'm torn between the CAMaster Stinger 1 and the SB Desktop.
I just bought a PRStandard 48 Buddy last Spring, and Camaster Stinger II or III were the only other choices left after reading volumes and removing candidates.
I ultimately chose Shopbot because of the endless rave reviews, extensive online resources, and the people at Shopbot.
When I bought my machine, I was literally talked out of buying some upgrades that didn't make sense. They knew that because they actually asked what I was going to do, and what I wanted to achieve. I got nothing remotely like that speaking to Camaster, they seemed kind of annoyed that I called.

Here's a simple little thing that's meant the world to me...
I had trouble with a little $15 USB hub that came with my machine. I called Shopbot, and had to leave a message for Frank. I turned machines back on, so I missed his call that came 10 minutes later.
A few hours later, I missed his call again. He left a message that said "I didn't want to hold things up, so I shipped that adapter you needed, etc. Call me back and let's make sure you don't need anything else...". Mind ya, I hadn't even spoken with him yet and he had already looked up my info and shipped the part.
At 9:00 that night, he called again. Yes, 9:00. From home, I imagine. Over a cheap little part that he'd already shipped. It was just bugging him something awful that he may not have covered my problem 100%. I missed that call too. He called me again the next morning and we talked.

Now, I had read these sort of lunatic's ravings about Shopbot prior to ordering from them. I wasn't so sure I believed them, so I called. Ever since that moment I picked up the phone, they have done nothing but fall all over themselves to make sure I'm the happiest customer they ever had.
When these guys say "call and ask", they mean it. You'll be immediately speaking to friends.

So far as my machine goes, I could not possibly be happier. Just today my brother-in-law stopped by and when looking at the machine said "damn, is that thing built nice!!!"

Don't be torn anymore, brother. I can't say if you'd be happy with the Stinger, but I can't possibly imagine you not being pleased with Shopbot.

[tomwillis]

Eric - I integrated a small 40w Universal on my 2nd Z and the whole system runs off WinCNC software. It was very expensive to integrate - because of a steep learning curve and hiring techs to integrate. I agree with Brady - that keeping a router/spindle sep from the laser is better. I usually drape the laser with a garbage bag and put a small plastic baggie around the laser nozzle - when using the router on the other Z. The biggest obstacle I've found, however, is vibration. I think many of the quality laser tables are belt driven which give faster output and accuracy. Mine works great for what I use it for - so it probably depends upon the precision of your parts.
Tom

[dana_swift]

Brady & Dave on "French Curves". Draftsmen used to keep a drawer full of greenish yellow curves, today all of them have been replaced with Bezier curves (or NURBS based on the same math.) Pierre Bezier was most definitely a frenchman! So today we still use French curves.. just different ones. Now they are just created graphically for each situation instead of having the old green ones that never quite fit the need.

Personally I will take modern Bezier curves over any plastic template any day. Especially because its easy to tell the Shopbot to follow a Bezier that was drawn in PartWorks, and very difficult to tell it to follow a piece of green plastic!

http://en.wikipedia.org/wiki/Bezier_curves
(Check out the animations at the bottom to see how they work, and how software converts the control points into tool motion!)

Most people never use the node editing mode in PartWorks to discover that many things are done with Bezier curves whether they knew it or not.

Cool stuff!

Oh- I still have a deci-trig-log-log K&E slide rule from when I started engineering school. I have my fathers also!

Sigh-

D

[JackM]

Dana... I have been lurking and researching my Boomer-Sooner butt off for a couple months now and I have to say, you are one of the more compelling characters on the forum boards. Partly because you are from T-Town but mostly because you have a Renaissance Man cachet. Your posts are always insightful and interesting. Hope to meet you and several other of the "usual suspects" from the SB forum.

[Brady Watson]

Dana,
I must be showing my age...I had a French curve in HS drafting class; two actually. One was that opaque snake vomit green color and the 'real nice' one was clear acrylic. I don't think either of them got very much use in drafting, but sure were fun for random doodling and strange geometric constructions ;)

Thanks for the Wiki link - gonna check that out.



Most people never use the node editing mode in PartWorks to discover that many things are done with Bezier curves whether they knew it or not.


Absolutely. I am amazed at how many users never do any node editing. When I was first starting out, I would trace things out with arcs...and that never worked out as smoothly and perfectly as beziers do. I think it was watching James Booth @ an ArtCAM demonstration when I really 'got it' and now rarely use arcs for vector shape construction, unless that section of the part truly is an arc or arc section.

Mastering node editing is the key to all of these programs, especially if one plans on progressing to 3D. You must master 2D drawing before you can do any real work in 3D. Otherwise, even creating a good cross section for rail sweeps etc will be frustrating.

-B

[dana_swift]

Jack- thank you for the kind words. Meeting the shopbot community is one of the great things about having a shopbot. Go to the various shopbot events, you will be stunned at what people are making with their bots. Inspirational to say the least!

Brady- Bezier curves are fascinating. They are easy to explain to someone just wanting to draw a specific curve and not worry about the math behind. Yet most people never hear an explanation of what the 4 control points do and how to manipulate them to get exactly what is needed, so they seem confusing and remain unused. Tragic.

Hope you enjoyed the animations.

D

[bleeth]

As a marine carpenter/shipwright, learning early on about "fair" curves made working with beziers in software a natural. I still have my box of acrylic ships curves and do get them out for sketching. Many of the desks we build for our projects are curved but even when they are somewhat serpentine in shape the architect always draws the original design as a series of a few connected arcs and shoves the centerpoint around to achieve a close tangent instead of using beziers. This is probably due to making it easier to lay out for framing, etc., although from the state of so many of the architectural plans I see I wonder how they justify their fee and for the framers on many of my projects I will cut the shape out in some cheap mdf or garbage ply and screw it to the floor so they can plumb up from it.
Interesting sidenote: My automatic spellchecker tells me that bezier is not in its dictionary!!

[GDGeorge]

Dana,
I must be showing my age...I had a French curve in HS drafting class; two actually. One was that opaque snake vomit green color and the 'real nice' one was clear acrylic. I don't think either of them got very much use in drafting, but sure were fun for random doodling and strange geometric constructions ;)
-B

So... If I take one of my old french curves and tape it to the work surface on my Desktop, I could set up the digitizer and...

Cheers,
J

[Tomlin]

Ok, so to bring this thread back on topic I first want to say "Thanks!" to all who gave such a great insight into the product.

I'm thinking that I read somewhere that although you can get the 2.25 hp router elsewhere for less money, that Shopbot actually modifies it for this machine?

[bleeth]

The router is not modifed, but the springs in the z, the mounting plate, and configuration in the controller as well as appropriate wiring and connectors is in their price. (You can buy a crate engine for less than a whole one properly installed)

[Tomlin]

Hi Dave

Thanks for the reply. Sorry, but what is a crate engine?

[hh_woodworking]

Crate engines are replacement engines purchased new and complete, shipped in wooden crates. Can be complete or missing existing fuel system,alternator, water pump or other accessories.

[bleeth]

I have one in my truck now (Ford 5.7). Starter, Oilpan, distributer, intake and exhaust manifolds, AC etc. had to be stripped from the old one and installed on the new one.

[Tomlin]

Ah, I got you now. Good illustration there!