Dedicated thread to discuss improvements/fixes relating to Chinese 30X0 desktop mill/routers
My modded machine...
Attachment 292432
I purchased a 3020T machine from "fleabay" about a year ago and being the type of person who cant help but mod everything, I have made some significant changes to the machine that I would like to share with the community that will hopefully help others who are needing to get the most out of their 30X0 machines.
I originally bought the machine mainly for the purpose of cutting carbon fiber panels (1mm-3mm thickness on average). When I received the machine I was unfortunately one of the many who bought from a seller who was incompetent at packaging as well as a false advertiser. The machine arrived with shipping damages as well as USED parts! Fortunately it worked so I decided rather than going through the hassle of returning and having to wait months for another machine I'd try and make a deal with the seller to get a partial refund. I ended up getting refunded $175.00 from the $500.00 I paid as well as some new parts (deck panels, spindle motor and a few other small parts). So I only has $325.00 into it and decided I'll invest some time and a little more $ to improve things.
The first upgrades I did were the common ones, shielded wires and limit/home switches. Unfortunately I made a big mistake when finishing up the install of the shielded wires. I had the shield of the spindle wires temporarily twisted around some wire with the other end twisted around a screw connected to the frame of the machine (which was also connected to the the shield of the parallel port cable) to test that everything was working correctly. It worked fine but when I went back to solder the wire to the shield I didn't think to TURN THE POWER OFF ON EVERYTHING FIRST!!! I went a little too far with the solder and melted through the casing of one of the spindle power wires and shorted 50v to the shield! It instantly fried my computer and break out board as well as damaging one of the stepper drives. Lesson learned...
Moving on...I was on a pretty tight budget and after getting a new PC I needed a new stepper driver. I must have read through 2 or 3 hundred posts regarding the TB6560 based stepper drives and their common woes. Unfortunately my budget only allowed for one of these sub-par drivers so here's what I did. I scoured ebay and found a "new" version that looked promising with respect to some of the design flaws know on previous models. I purchased the driver from seller wfyb but I just looked and they don't seem to have them available still, however I found the same model in this listing...
Controller Board for Engraving Machine CNC TB6560 4 Axis Stepper Motor Driver | eBay
Here's mine...
Attachment 292434
The biggest differences in this model are better diodes on the outputs, larger and better placed capacitors on the the outputs and opto-isolated inputs. I did add a small 5v fan to the voltage regulator heat sinks on the side of the driver. I have been running this driver for over 6 months now and NEVER LOST ANY STEPS, which is common in older designs. This driver is 1/16th micro step capable and that's the setting I use. I also have it set via the dip switches for fast decay and the highest amp output, 3.5A/channel. I'm running Mach3 at 35k kernel speed vs the usual 25k...this works better in my tests (your results may vary depending on your PC, stepper motors etc...). I do however limit rapid speeds to 1200mm/min, going much faster may loose steps but for this small machine 1200mm/min is plenty fast enough...after all average cutting speeds range from 200mm/min to 800mm/min depending on the material your cutting and a host of other factors. Anyway, this setup has yielded VERY consistent results for me, I don't lose steps and the machine will return to a given point after running tens of thousands of lines of code without any measurable discrimination via a dial indicator.
After just a few weeks of running the machine with the new driver I started wanting to do some milling of aluminum parts. Immediately I discovered 2 main weak points that needed to be addressed. The weak DC spindle motor and the flex of the X-axis allowing the spindle to move upward. I resolved the spindle issue easily with this...
Attachment 292436
You can see the original motor on the right, I briefly tried the cheap harbor freight motor in the middle but had way too many issues with the collet not being accurate so I spent the $100.00 for a 1.25hp Makita router (on the left) and IT"S THE BEST MONEY IV'E SPENT ON THIS MACHINE PERIOD! THe collet is very accurate, it has smooth variable speed control and the best part LOAD DETECTION, meaning it will maintain consistent rpm under load, and that is a HUGE deal. RPM range is from 10,000-30,000, perfect for my projects. To fit the larger diameter motor body (65mm) of the Makita router I had to machine a new holder bracket. I used two 3/4in MDF pieces stacked to make the new motor holder and primed and painted them black. I also used some steel bands left over from an exhaust pipe mod I did on my car. Don't use the cheapo thin galvanized bands, they, make thicker steel ones that are much better. In addition I needed longer M6 bolts and some stand-offs (spacers) to prevent the bolt heads from contacting the spindle motor body. The bolt pattern is just barely wide enough to fit the 65mm diameter Makita motor.
Attachment 292438
To address the other main issue, the flex in the X-axis, things were a lot more involved... I happened to have two old printers laying around that had some decent linear rods (one 8mm and one 9mm) with brass bushings. The brass bushings and high rigidity of the rods are what makes this mod viable. The bushings slide very smooth and don't have any play at all. At the time I was still using AutoCAD and LazyCAM (Now I use Fusion 360 fro CAD CAM, simply AWSOME) so I made a simple 2d drawing in CAD of a bracket setup I thought would work well. The design takes advantage of leverage to strengthen the X-axis, by locating the two new support rails where I have them they have a great deal of leverage to hold everything in place allowing the use of small diameter light weight rods. Again I used 3/4in MDF for a many reasons, vibration dampening, fairly light weight, low cost, dimensional stability, easy to machine and plenty strong for the job. The design includes two brackets located on the top of the Z-axis (primed and painted black), two main brackets on either end of the X-axis (primed and painted with Duplicolor Chrome) and 4 clamping blocks mounted to the main brackets holding the new X-axis support rails. The Duplicolor Chrome paint really makes it look like it's made of aluminum, even in person, everyone that sees it thinks it's metal including fellow machinists...LOL. The way the clamping blocks are designed is crucial to the setup, it allows you to install the rods and jog the X-axis back and forth to ensure exact alignment, then you pre-drill and screw the blocks in place. I used some angle aluminum to fab the mounts for the rail guides to attach to the brackets I made on top of the Z-axis. See pics for details of mounting screw locations...
Attachment 292412 Attachment 292440
Attachment 292442 Attachment 292410
Attachment 292414 Attachment 292416
Attachment 292420 Attachment 292418
After these mods I figured it would be a good idea to replace the linear bearings and Y-axis rods. I bought LM12UU's for the X-axis and Z-axis and LM16UU's for the Y-axis. For the Y-axis rods I found an Ebay listing for 1 meter 16mm dia. hardened and chromed shaft that was pre-drilled with a 5mm through hole, perfect for tapping the M6 threads needed to mount them on the machine. I used a chop saw with a fiber disc to cut the shaft to the 16in length required, tapped the threads and tossed the old stainless shafts.
Attachment 292426 Attachment 292428
As you may have noticed, I have my machine mounted on a rolling cart with the electronics underneath and a desktop and monitor conveniently next to the machine. I used some black foam core sheets to enclose the sides of the cart to keep the chips away from the CPU, power supply and stepper driver board. I also cut some well placed vent slots in the sides to allow good air flow so everything runs cool even after hours of continuous cutting.
Attachment 292422 Attachment 292424
As I mentioned earlier I have switched from AutoCAD + LazyCAM to Autodesk's Fusion360 for CAD and CAM. I can't stress enough the tremendous benefits of using Fusion 360. Not only does it save me time but it also uses much more intuitive tool paths leading to less tool wear, machine stress and more accurate parts.
With all the mods I've made to my machine the capabilities in terms of accuracy, cutting speeds, chip load and materials that can be cut have increased dramatically. It's now able to effectively mill cast iron, mild steel, aluminum alloys and much more without chatter and to accuracy within +/- 0.01mm or 0.0004in. Of course it won't take large depth cuts like a big mill but it does very well with aluminum @ DOC of 1-1.5mm @ 380mm/min for rough cuts and a finish pass of 0.5mm DOC @ 280mm/min. I use a 3 flute 1/4in high helix end mill from Suncoast tools for aluminum, it makes a quick job of clearing chips and leaves a surprisingly nice finish.
I hope this helps some folks and I'll post pics of some things I've made with this machine...