[Ron Dunn]

The Background

I recently added myself to the growing list of customers for the Microcarve machines. I bought an MV3 - a fixed gantry machine with MDF sides and table, an aluminium frame, and an oilite bearing/unsupported rod guide system.

microcarve MV3

I currently own a Probotix Fireball V90. During this thread I'll make a lot of comparisons between the V90 and the MV3. This should be interesting, as the V90 was designed by the same person as the MV3 - John, otherwise known in these forums as Microcarve. I bought the V90 before John started making the Microcarve machines.

Woodworking is my hobby, computer software is my profession, and CNC offered the chance to combine both. My original requirements were a machine that would help me build some toys for my children (the justification), be available as a complete kit from one supplier, have good quality instructions, and a strong support base. The Probotix V90 met all these criteria.

FireBall V90 CNC Router -

While I do have a good range of woodworking tools, I had no metalworking capability. I'm also not confident in my ability to do precise tool-quality work in wood. My objective was to make things, not machines. These factors meant I wasn't interested in building a machine from plans. Also, living in Australia means that I don't have the wide range of cheap DIY products available in the USA. Even including shipping, the V90 kit cost me far less than a set of parts to build a similarly sized machine would have cost me locally.

My machine came, I set it up, and I spent the next few months writing firmware to drive it from a USB port. After three iterations of hardware and software design I had something working to my satisfaction, and I started on my original plan of toy making.

As often happens, you do something for yourself and others want it as well. My wife's friends started seeing the toys I was making, and began to ask if I'd make some for them as well. I got more ambitious, designed a wider range of accessories, and quickly realised that one, small CNC machine wasn't going to keep up with my plans.

Then I flip-flopped for ages. I looked at Tormach with lustful eyes, I looked at Shopbot, but as much as I liked these machines I had two problems. Firstly, one machine can only make one cut at a time, no matter how good it is. Secondly, I'd have to sell a LOT of toys before making a profit on my time and expense. I needed an additional machine, and it had to be cheaper.

Then I looked at the Sherline and Taig machines. Again. Something keeps drawing me back to these pretty little machines, but I also see a lot of advice about not using them for wood, and their performance envelope was a little slow.

Did I want another V90? Yes and no. I had a great learning experience with the V90, but I'd also found some limits in its design that I felt were affecting my productivity. I'll discuss these limits in future posts in this thread.

Then John came out with the MV2/3 range. The fixed gantry style suited most of my work, and addressed the issues I had with the V90. I loved the aluminium frame, and saw it as a progression in design from the V90 and his subsequent A4 machines. I watched his progress, saw good reports from customers, and decided to place an order.

This thread will be about the experience of buying an MV3, from ordering to production. Because I think they address a similar market, and because I think they are both good machines, I'm also going to contrast the V90 and the MV3 at each stage of the experience. I hope this comparison will help others to choose between these machines.

You might notice the "and" in the thread title. I originally wrote "vs", but these are both good machines and that seemed unnecessarily competitive. I think there is a place for both machines, and I don't plan to get rid of my V90 any time soon.

I'll make future posts on the following subjects:

* Ordering and Delivery
* Setup and Alignment
* Performance
* Support
* Conclusion

Posts will be made every few days - I've got two machines to run and a full-time job to satisfy at the same time

I hope people with similar requirements to mine will enjoy the comparison.

Ron.

[crumbaker]

looking forward to this, was wondering why the microcarve is cheaper initially if it's the better machine.

[Ron Dunn]

Ordering and Delivery

In this post I'll discuss the experience of buying a machine from each vendor.

Let's go back to my original criteria. I wanted a CNC machine that was available as a complete kit from one supplier, with good quality instructions, and a strong support base.


Complete Kit

I doubt that I would have bought the MV3 as my first machine. The MV3 is a machine kit only, and doesn't have any options for motors, controllers, power supplies, etc. This was important to me for my first machine, as I didn't want any finger pointing between vendors. I wanted a one-stop solution, where everything was assembled, tested and supported by the one company.

Probotix does a great job of handling a first machine order. Their web site offers a good/better/best option for most items, which lets the buyer match capability to budget.

Microcarve will recommend components to complete the machine. In fact, their recommendations range from HobbyCNC for those inclined to DIY to save a few dollars, to the same Probotix solution that I originally purchased for my V90. The only thing missing is a one-purchase option.

Actually, for Australian readers, that separation of purchases could work in your favour. My original V90-everything purchase came in one shipment that was over the tax-free threshold. That delayed my shipment and increased my cost while I paid the import tax. Separating the order into two purchases from different sources would have avoided that problem.

I recommend that Microcarve thinks about creating a detailed set of purchase instructions that would tell first time buyers exactly what to buy, where, and how it all fits together.


Instructions

Speaking of instructions, Probotix also has great documentation for all components on their web site. I was able to review the instructions for assembly, wiring, etc. before I placed my order. The assembly instructions for the V90 were particularly clear and precise.

John hasn't yet put full instructions for the MV3 on his site, although a fellow user has created a great web site with pictures and instructions that fills the gap. It just isn't easy to find if you're not already part of the machine community. Here I'd recommend that Microcarve fill this gap on their site, to help potential customers get a feeling for how the machine fits together.

I'll deal more with this subject in the next post in this thread.


Support

Microcarve (John) has a strong presence in CNCZone, and that is how I first came to learn of his machines. Probotix occasionally post here, but there are also a few users (self included) who respond to posts about the V90 machine.

Both suppliers use Yahoo Groups as their support forum. For this error in judgement I'd like to stake them both on an anthill, pour honey over their genitals, and giggle like mad as the bull-ants (fire-ants for you Yanks) punish them severely. Yahoo Groups is the forum you use when you have contempt for your customers, or when your last technology update was in 1994.

OK, now that I've vented, the content of both support forums is both relevant and helpful. Probotix questions tend to get answered a little faster, but I think that is because there are more users of V90 machines at the moment.

John provides most of the answers in the Microcarve group. His answers are easy to read and highly relevant. Actually, I find I learn a lot just by reading John's posts, as he often explains *why* he has designed or made a machine in a particular way.

Len from Probotix used to participate in the V90 group, but lately I don't see him as much. Probotix bought out a new machine last year - the Comet - and I wonder if that took a lot of his time and attention away from the support side of the business.

In each case I had pre-sale and post-sale questions. All questions were handled quickly and accurately from both groups. I was particularly pleased by the post-sale questions. In each case there was a small issue on receipt of the packages. In the V90 case it was a missing grub screw, for the MV3 it was a small point of shipping damage. Each vendor was far more worried about fixing the problem than I was ... I'd say they both take pride in a high degree of customer service.


Shipping

Living in Australia, I'm always nervous about ordering items from small suppliers in the United States. That caution drives me to watch suppliers and customers for a while, to see if there is a pattern of slow- or non-delivery. If so, I don't order.

Both Probotix and Microcarve have great reputations with their customer bases. In both cases my goods were shipped within the quoted timeframes, I had shipping tickets from reputable transport providers, and I was able to track the progress of my orders through shipping and customs.


Comparison

So, who has the better ordering and shipping experience?

Probotix has an edge for the first-time buyer, offering a complete, everything-included kit option. That wouldn't be so important to someone who knew what they were doing.

Microcarve doesn't offer anything more than the machine, but they have an edge that I haven't discussed to date - being willing to make small customisations to the machine if requested by a customer. In my case, for example, John made me a custom 43mm spindle mount to hold a Kress 1050. Probotix doesn't offer that service.

I'm going to call this one a tie.

[Fish4Fun]

Ron,

You have chosen to applaud one of my favorite people in the world, Microcrave (John). He is not so much a "company" as a guy who loves to build CNC machines and does it really well. He is a great guy, and I don't say that about many people. He honestly doesn't care about "promoting", he cares A LOT about making sure people are happy with his machines. He views "sending them off to their homes" much like a parent might about attending a child's wedding. As an A4 owner, I can assure you that I could not be happier; I can see you have had a similar experience:-)

John lives only a few miles from me, but I doubt seriously that gives me any advantage over people like you who live 12,000 miles from him (except shipping costs, lol). He is straight-up, honest to a fault and has un-heard of customer support. I don't want to say ANYTHING BAD about anyone else, I just cannot imagine dealing with anyone better than John.

The worst thing I could say about John is that he is a CNC junky; that is, he is always fixated on his next machine build and how he might make it better than the last. I was planning a DIY build (had been planning it for over 20 years) when I ran across his "My Newset Desktop Machine" Thread. I blame him for losing a day of work (but not "missing" ;-) ) because when I started reading the thread it was 04:00; I ordered my A4 @ ~18:00 after spending the whole day reading (job description "owner" has occasional advantages). I own a retail fishing tackle business and have owned/operated it for several decades and I am here to tell you that most fishing tackle MFGs are "very in tune" with their clientele, but I have never encountered ANYONE more passionate about their products than John. As a business owner I pride myself on customer service, but I am humbled by John's passionate dedication and integrity.

Thank you for starting a thread that is obviously about praising John (even if you are praising other people as well), there is no one I can think of who deserves the accolades more. I do not want to diminish anyone else's products, but I do want to stand up and say "Thank You" to John. My A4 is: "All that and a bag of chips".

Now...who was that other company you mentioned......? hehe

Fish

[Ron Dunn]

Setup and Alignment


The big event! The boxes arrive.

Before we get into setup, one comment on boxes and packaging. Probotix' packaging is definitely better than Microcarve's. In the interests of saving customers money, Microcarve has gone for a lighter, standardised packing solution. Personally, I'd prefer to see this as an option, but with a much stronger parcel with better internal contents control as a standard. I had some minor cosmetic damage due to postal abuse, nothing serious and nothing lost, but this is an area of risk that could easily be addressed.

The contents for both products are well organised. In each case I had everything that I ordered, so I proceeded to setup.

As I mentioned in my last post, Probotix posts very clear assembly instructions on their web site. Given that this was my first CNC machine, the process of assembly was incredibly smooth. The only part that gave me any real difficulty was the assembly of the side tubes and screws to the front and back. This required some manual dexterity, which isn't exactly my strong point. From start to finish, assembly of the frame took me around two hours. If I had to repeat the process, I think that time would have halved.

Microcarve doesn't post assembly instructions, but great guidance is available from another customer:

Assembling the microCarve A4 CNC

These instructions are for the earlier A4 machine, but directly translate to the MV3. Assembly took less than an hour, although this is possibly a misleading comparison as for my second machine I had a better understanding of how a CNC machine should fit together.

Immediately, differences between the V90 and the MV3 become apparent. Stating the obvious, the V90 is a moving gantry machine, while the MV3 is a fixed gantry/moving table design. The MV3 has fewer structures to assemble, as the gantry does not need to be built. Also, the one-piece (MDF) side frames of the MV3 make assembly of this structure a breeze. The only point where dexterity is required is to slot the three cross members into their receiving pockets, but following Saundby's instructions makes it easy.

The MV3 is definitely a stronger machine. Apart from the PVC frame tubes, the rest of the V90 is made of MDF. The downside of this material becomes evident when the tensioning rods that pull the frame into alignment are tightened. Even using small tools and light pressure, it is very easy to deform some frame members. This problem just doesn't exist in the MV3. The solid sides and aluminium cross-bars are very strong and rigid.

Finally, the MV3 is vastly more simple to align. Partly due to the design (moving gantry) and partly to the variability in materials, squaring the V90 can be an exercise in frustration. The advantages of a fixed gantry are apparent here, as the only point that didn't immediately square itself on assembly was the pivot of the gantry axis around its rails, which affected the squareness of the Z-axis to the table. A light pressure on the rail frame as the bolts were tightened quickly fixed this problem.

This is the fundamental advantage of the MV3 over the V90. The gantry on the V90 can flex/move with just finger pressure. Add a router to the machine and you can see the axis sag forwards. That problem just doesn't exist in the MV3.

I did find some negative points about the MV3 in comparison to the V90.

* Using a hex-head bolt to hold the spindle mount to the z-axis was an exercise in tedium. The allen key was JUST long enough to reach the head, and I skinned my knuckles a couple of times on the spindle mount. A cheese-head screw would (to my mind) be a much more user-friendly solution.

* The lead screws seem to be a slightly inconsistent diameter. Installing the screws means sliding one set of bearings along the length of the screw as it assembles, and those bearings occasionally jammed (lightly) on the thread. I didn't notice that problem on the V90.

* The MV3 uses a Delrin drive nut embedded in the the table (Y-axis) and the Z-carrier (X-axis). I ordered the anti-backlash option with my V90, and these nuts are screw mounted to the relevant parts of the machine. While I've got a couple of spare Delrin nuts, I'm scratching my head over the manner in which I'll have to eventually replace them. The V90 is much more obvious.

So how would I sum up the assembly experience?

* The V90 is better packed
* The V90 instructions are better

* The MV3 is easier to assemble
* The rigidity and alignment of the MV3 is far superior to the V90

This last point is the critical one for me, and validates the reason that I chose to go for the MV3 as my second machine. Its rigidity is just so much better that I'm sure I am going to spend less time on alignment and tuning issues, and more on using the machine for production.

Next post ... using the machines.

[nio101]

Hello Ron,

Thanks for the posts, they are very interesting!
Looking forward to read the other parts...

[Ron Dunn]

Coming soon

I wrote my own controller (ie, Mach3 replacement) and it didn't cope well with the change to a heavier spindle. I was stalling occasionally, and losing steps.

I should have that fixed up by tomorrow, and expect to be cutting great parts this weekend with another set of review comments to follow.

[Ron Dunn]

Things have slowed to a crawl in the Dunn shed.

My controller sucks. Well, it probably isn't quite as bad as I'm saying, but my acceleration/deceleration strategy simply isn't good enough to cope with my new machine.

I can cut straight lines very well. Early experience shows that the MV3 is much quieter in operation than the V90, which I attribute to the heavier construction of the machine. Where I fall apart is in cutting arcs, which I believed I've traced to a software design problem.

I'm taking a two-part approach to solving the problem. As an insurance policy I've just ordered a KFlop board, which intrigues me with its programmability. At the same time, I'm going to redesign my own controller software to better manage mid-arc direction changes.

In the meantime I've put everything back onto the V90 so that I can continue to make parts while I continue my foolish quest for a DIY controller

[dorz]

cant just use mach3?

[Ron Dunn]

Good question, dorz.

The simple answer is that I'm too pig-headed to use Mach3.

When I got interested in CNC, the first thing that struck me was how UGLY is most CNC software. It is as if the developers have a rare form of autism, where they give 100% to the engine and 0% to the user interface. At the time, Mach3 was the perfect example, with graphics that looked like they came from Web 0.1 and absolutely no attention to detail in the screen layout.

The second thing I noticed is that most CNC software guys are mired in the past. Parallel ports, for example ... they died so long ago that the corpse doesn't even smell any more. They're often using development tools that are years out of date. They don't support cross platform users.

Software (well, data, actually) is my profession, CNC woodworking is my hobby, so I decided to put the two together and write my own controller. Mainly just to prove it could be done, but also to create something that I actually liked using. Short story - as discussed above - it just isn't good enough, with a severe problem in mid-arc acceleration.

If I was a sensible person, rather than the pig-headed person I mentioned earlier, I'd have simply bought a product like Mach3 and put up with it. I've spent hundreds of dollars on hardware and software, and more hours than I care to value writing code. I just don't feel happy going down that path.

As it turns out I've rushed into the wrong solution. The KFlop I mentioned earlier isn't suitable for use with the Probotix drivers, and frankly, I didn't do enough investigating before choosing the product. The maker has a generous return policy, so I'll write the cost of two-way international shipping off as a nice little lesson in patience and diligence.

Now I'm back on the horns of a dilemma.

Do I go back to my own code? I've spend the last couple of hours strategising solutions, all of which will take me weeks to complete with the limited coding hours I have available.

Do I buy a controller? I still don't like Mach3, but maybe Mach3 plus Smooth Stepper, with a third-party screen, will keep me working in the short term.

I'm also looking at the Planet-CNC USB controller. At least the software UI doesn't make me want to pluck out my eyes, but the controller itself seems to be something of an unknown quantity. I don't want to rush into another KFlop-style mistake of insufficient evaluation.

*sigh* ... decisions ...

[milehigh]

I'm also looking at the Planet-CNC USB controller. At least the software UI doesn't make me want to pluck out my eyes, but the controller itself seems to be something of an unknown quantity.

You could always try the DIY version to see if it's suited for you. When I get my MV3 set up the way I want it, I will probably give the DIY Planet-CNC a shot myself.

I also agree on everything you said in terms of software. I've been wondering why CNC software seems to resemble software I ran in Windows 98. Just take a look at websites. Compare a page with modern design and one that is ancient, and you'll see what I'm talking about. I wish CNC stuff would follow in terms of design.

[microcarve]

Now I'm back on the horns of a dilemma.

Do I go back to my own code? I've spend the last couple of hours strategising solutions, all of which will take me weeks to complete with the limited coding hours I have available.

Do I buy a controller? I still don't like Mach3, but maybe Mach3 plus Smooth Stepper, with a third-party screen, will keep me working in the short term.

I'm also looking at the Planet-CNC USB controller. At least the software UI doesn't make me want to pluck out my eyes, but the controller itself seems to be something of an unknown quantity. I don't want to rush into another KFlop-style mistake of insufficient evaluation.

*sigh* ... decisions ...

I could be wrong....and probably am wrong....

(Said so, right up front!)

But the way I see it is there's going to be a wire running to a computer
whether it's an old desktop....(free and reliable)...or a $10K new age
laptop.

Whether the wire is a USB or PP...doesn't matter to me. I have to avoid
tripping over it either way.

I run an old P4 XP machine that's maybe 9-10 years old. It has 256MB
memory. No networks...no background junk running....(which is easy
with XP). It runs 24/7 until a power glitch resets the computer....then
it comes right back up. I never notice it most times.

I've had 3 laptops in the last 6 years. 1 really nice $$ one. None
of them is reliable enough for me to leave running for a few days at
a time. 2 don't even work anymore....I left them running for weeks.

Either way the computer had to be in a place it wouldn't get moved to
be used for the larger machines I ran. So the desktop size wasn't so
much of a space hog that the difference was worth considering.

Laptops have been huge headaches for many, many people for a long
time. Some just won't work at all. The more advanced they get the
more problematic many of them can be.

Still, there's a physical connection between a machine and whatever
computer runs it, so the loooong proven, nearly bulletproof, old crappy
out of date desktop makes way more sense to me. A machine will only
cut anything so fast and I've never seen a thing that a laptop would
cause to be any better.


Mach3 can be customized and that's what I did. I didn't like it for the
first 4-5 times I tried it....I ran CNCPro on an old DOS machine. Customizing
is a PITA for anyone without some programming knowledge, but a sheer
breeze for anyone who has some. The screens can look any way someone
may want. It's the inner workings of Mach3 that causes it to be left
running indefinitely on my old junk machines.

Why leave it running? There used to be a debate...(never was any
definitive answer)...about shutting down and constantly restarting
computers. I always just left mine running. I hate waiting for boot times.
Habit. But it's very long proven to be fine here....

IMHO...a computer that runs any machine should be dedicated to running
that machine. Making changes to even the file system, changes the way
the computer works and acts. That's the reason Windows has to be
constantly re-installed. It gets junked up and isn't reliable anymore....
only fix is re-install at some point.

But the dedicated computer does nothing but run the cnc machine and
mine has worked great for years.

I did have a total HD failure once. I had backed up a copy of my Mach3
xml file, so I was back in business within a few hours.

I'd never trust a wireless connection to any computer for running a
cnc machine, so 'm stuck with USB or PP. It's a wire to a computer.
I don't see any benefit to the computer being a laptop. Only potential
problems.

Mach3 ain't pretty, but it's not too hard to modify.....
And, you can rely on it. It runs huge industrial machinery worldwide
on a daily basis.

As a side note....Fish (A4 owner)...bought the extra stuff that runs
the KFlop and he LOVES it. He's gotten some kind of Huge benefits
in performance from it. I wish I had the time to try one myself.


John

[ger21]

The KFlop I mentioned earlier isn't suitable for use with the Probotix drivers

Can you tell us why it isn't suitable. The Kflop outputs step and direction, so why won't they work together?

[UUU]

Software (well, data, actually) is my profession, CNC is my hobby too.

But I bought Mach 3 and am happy with it warts and all. I spend enough time at work honing and bulletproofing user interfaces without starting all over when I get home!

[nio101]

What about EMC2?
AXIS UI seems to be more pleasant than Mach3's UI...
And if you can create a better UI, it'll be a great contribution to this free community-developped product!

The planet-cnc USB package seems to be nice... but you will always be more limited using a micro-controller than when your whole PC is used to interface to your CNC...

[Ron Dunn]

Can you tell us why it isn't suitable. The Kflop outputs step and direction, so why won't they work together?

Hi Ger, the KFlop has a maximum 3.78us pulse width, the Probotix drivers require 5us. Tom Kerekes advised me that they were incompatible.

Ron.

[ger21]

Thanks.

[Ron Dunn]

What about EMC2?

nio101, isn't EMC restricted to the parallel port? It would be interesting to learn otherwise.

Then again, I have a different issue with EMC being run on Linux. I work for Microsoft, and I'd prefer not to use a competing platform. I'd like to use the stuff that pays my salary if I can

This raises a point that John touched on in his thoughtful post. Frankly, many of the comments about Windows are mythology. I use Windows7 day in, day out, on devices ranging from tablets to netbooks to notebooks. I present and demonstrate to customers and audiences every day. Not once have I ever had it "freeze", although I certainly agree that Windows 7 is not designed to be a real-time operating system.

In my own controller (I know ... the failure) I worked around any potential problem through buffering at the USB device. The control software buffers steps at the control hardware, which are released on a timed interval. I've never had a problem in this aspect of the controller, the buffer has never run dry in mid-stream, and any jitter in the pulse rate was due to my calculations on the controller, not the host netbook.

I have thought about using old PCs with parallel ports, but it just doesn't sit with my vision of how I'd like to work. I'm dreaming of a tablet PC on an articulated arm, driving everything through a touch-screen interface. That means USB only.

As my wife pointed out last night, "it is just a hobby!" ... I hope she's telling the friends that are waiting on toys the same message

[nio101]

nio101, isn't EMC restricted to the parallel port? It would be interesting to learn otherwise.

Yes, it is, I guess. But PCI parallel port cards are cheap and USB is not mandatory for me.

Then again, I have a different issue with EMC being run on Linux. I work for Microsoft, and I'd prefer not to use a competing platform. I'd like to use the stuff that pays my salary if I can

OK, I can understand that! (^_^)

[microcarve]

I have thought about using old PCs with parallel ports, but it just doesn't sit with my vision of how I'd like to work. I'm dreaming of a tablet PC on an articulated arm, driving everything through a touch-screen interface. That means USB only.

Ahh..that's the reason I never got very interested in USB stuff as well....

I'm seeing it more as built like the machining centers, with everything all
integrated into one unit.

I have tons of designs, but so far, they'd all have to be DIY because
I surely couldn't ship one cost effectively.

I have this cool little stand alone controller that plugs directly into a
Hobbycnc board and an SD card plugs into it. Completely portable
machine....

But the project got sidelined years ago due to my input not being as
well as it should have been. I regret not being a better beta tester.

I also ran a few mini itx boards for awhile to work towards integrating
those. But, I must have gotten on board too early with those....both
failed within 6 months. I noticed they'd been discontinued when I went
to check on warranties. I never got a chance to try any more of them since.

My laptops got dust in the keyboard. Power supply failed on one. Too
much trouble. An old desktop can be fixed in a very short time.


John