CNC Router Table

It’s been a cold winter.  In an effort to help moderate the temperature in my shop I decided to add another piece of equipment.  The extra mass will help smooth out large temperature swings.  What machine did I drag home this time?  That’s a bit of a long story because I found out later that the machine I brought home wasn’t exactly the machine it was advertised as.

The machine is branded as a Torcam (not Tormach!) ~ 24″ x ~ 24″ x ~ 3″  (X / Y / Z) Router table.  It is constructed out of aluminum extrusions and utilizes linear rails and ball screws.  This was the main reason I decided to purchase the machine.  Once I saw the linear rails and ball screws (and how little the machine had been used) I was sold. After loading the machine into the Sienna, (yes we did break down and buy a minivan for our family and it has been once of the best loathed decisions we ever made!) I snapped a quick picture of my purchase:

The machine did not come with a control which didn’t bother me one bit.  I had full intentions of fitting a more up to date modern control anyway.  The very very reasonable price I purchased the machine for left plenty of funds to put together a new motion controller.

But I was curious the whole time about this Torcam company.  I had never heard of Torcam before and I didn’t do any research regarding the company beforehand.  After some digging on the internet I found out that Torcam was a machine tool distributor company out of Ontario Canada who built and sold CNC machines for the educational market.  It seems they rebranded machines for sale and it is doubtful that they actually designed and built full machines but I could be completely wrong.  It appears the were a going concern in the 1990s to early 2000s and then they disappeared.  Given the timeframe of business operations (just before the internet exploded) and what appears to be a limited market for product, very little information is easily found about Torcam and their machines today.

This machine appeared to be very well designed and assembled with care. But who actually made the machine? As soon as I purchased the machine I posted a picture on Instagram and Stefan Gotteswinter immediately commented “ISEL?”.  ISEL is a German CNC machine builder who also sells various motion components.  ISEL has been in business for a very long time and according to Stefan builds good components and machines for the price.  I think most in the industry agree that ISEL stuff is built to a price point and does the job very well.

The machine does look suspiciously German and like something ISEL would manufacture so I decided to find out.  After taking a few covers off I noticed this:

It was confirmed.  This machine was made with ISEL components and I also now had an approximate date of manufacture.  All the components say made in West Germany.  That gives you a good idea when this machine was made: early 1990s.  I suspect Torcam started importing these machines and selling them.  I don’t know what control they shipped with it (did they make their own?) but the hardware was ISEL made.

I made a video and posted it on Youtube (see below).  Shortly after posting John commented on the video: “What you have is a Techno Isel router table. Originally released in the late 80’s and early 90’s I can with what was called a machine 100 MS cos controller. Back in the day a new on would be about $ 8000 or so. I have the same machine from the 80’s it ran model and prototype production 24-7 for about two years. I mothballed it for some while then had a new controller built and I still use it today. I made a mount for. 3 h.p. Ryobi router when I first got it in 1987 and it still works like a charm. I run V carve desktop and Mach 3 on it making sings and doing woodwork. My table is 52 by 52 . By the way the stepper can get warm but they seem to convey the heat well. Never any problems running it for 10 to 12 hour runs. Just stay inside the feeds and speeds. The ball screws are a big plus but keep them clean. Like yours mine had no goers on the x and y rails.”  Thanks John!!!!

The first step in getting the machine working was building a stand.  I took a Saturday morning and put together a quick wood stand.  I would have liked to have a welded or concrete stand for it but the weather didn’t permit me working outside so I settled for wood.  Maybe in the future I might make a more substantial stand.

I also took some time to make up some leveling feet that would screw onto the legs of the stand:

Once the stand was built it was time to decide upon the motion controller.  I looked at a few options like the Centroid Acorn and Mach 4 but I decided upon LinuxCNC.  Lot’s of folks are scare of Linux but let me tell you that this was a very straight forward process to get going.  I used the Mesa 5i25 and 7i76 LinuxCNC plug and go kit.  It was pricey but is a proper motion control interface that utilizes a FPGA in the 5i25.  If you purchase the plug and go kit it has the proper firmware flashed on it already that saves you from having to re-flash the Mesa board.  Even that though isn’t as hard as it sounds!

For motors and drives I used some stuff sitting around in the shop for a few years.  I purchased 3 motors and drives used a number of years back.  The motors were 60BYGH303-13 425 ounce inch dual shaft steppers that were almost a drop replacement for the small steppers that the machine came with.  The drives were knock offs of knock offs drives.  Very little information is available for the CW230 stepper driver but I did some comparing and it appears that they are copies of the Keling KL4030 drive which seems to be based on an older Leadshine or Gecko drive.  I set them up to run at 36 volts (using a linear power supply which will probably burn out) and used 1/8 micro stepping – the highest you probably should go.  I built up a panel and put all the bits inside.  Here you can see it in progress:

Setting up LinuxCNC was a simple as wiring up the Mesa interface board, installing LinuxCNC on an older computer and running the configuration wizard.  You need to be careful to enter the information into the wizard properly.  I entered everything carefully and once done I ran LinuxCNC and moved the table around.  I used stock drive timings for the KL4030 that were directly out of the wizard.

I wired up the homing switches and tried to tidy up the wiring as best I could with some cable chain and wire loom.  Now I need to mount a spindle and start cutting out parts!  I may spend some more time tuning the drives and getting the system dialed in but so far I’m very pleased.  I’m hoping to post some more information regarding setting up LinuxCNC soon so stay tuned!

I also made a video of the work.  Have a look:

 

5C Collet Chuck

A few months ago I purchased one of the popular import 5C collet chucks for my home shop.  I’ve been investigating different ways to employ a proper collet setup in the home shop for awhile.  At first I was considering going the ER collet route and purchasing, or making, an ER collet backplate for my lathe due to the large grip range of ER collets and that they are very plentiful.  From a manufacturing engineering standpoint ER collets are not considered proper work holding collet, being designed specifically for tooling, but they actually do a good job in the home shop for work holding provided you are aware of the short comings:

  1. ER collets are generally not available in square or hexagon.  This isn’t as big of deal as it may seem – many folks use ER collets and simply grip on the the edges of non round stock.
  2. ER collets require more grip length than almost all work holding collets.  This is probably the biggest downfall to using ER collets in the home shop.  Holding onto a very short part in an ER collet in most cases is asking for trouble.  Even more sketchy would be holding onto just the edges of short square or hexagonal parts in an ER collet.
  3. ER collets require relatively high tightening torques.  This isn’t a big deal with the smaller sizes, but once you get into the larger sizes (greater than ER20) it becomes a pain.  For example ER32 is recommended to be torqued at 100 foot pounds!
  4. No emergency or soft collets available.  I suppose you could make up some soft ER collets fairly quickly though.
  5. No ER pot chucks, clutch collets, step collets, oversize collets, or whatever you want to call them.

Most of the above reasons are relativity minor when comparing ER to standard work holding collets.  Many of the above downfalls of ER collets are offset, especially when you are starting out, by the fact that you can use ER collets and collet chucks for both work holding and tool holding.  ER collets also have a very large grip range – meaning you need fewer collets to cover a range of sizes.  This can save money on tooling, which can be a big deal in the home shop and was precisely why I was seriously considering using ER as I already had a some collets in the shop.  When you consider you can purchase the ER collet backplates for less than $100 or make them easily in your home shop it’s a logical choice.

But I decided to go with a standard work holding collet, mainly for reasons 2 and 5.    I chose 5C as it is by far the most popular work holding collet available.  There is a plentiful used market and new collets are inexpensive.  Soft and clutch collets are inexpensive and I can get them next day from a local tooling supplier.

There are a few options for the actual collet chuck.  Import ones are available from numerous suppliers for below $200 and this is the route I went.  I actually ordered it off Amazon Canada.    If you are looking for something of higher reputation (note generally most of the import one are actually decent) you can purchase a standard accuracy Bison ones for around $500 with a stated .0008″ TIR.  A super precision one is available for $900 with a stated .0004″ TIR.

My import 5C collet chuck has less than .0008″ TIR, which is less than the stated accuracy of the standard Bison one at less than half the cost.  It is very well finished and so far works exceptionally well.  I have ordered inexpensive $12 emergency collets for it, and also I have been using it with a custom bored 3″ pot chuck recently for a repetitive job.  With careful loading I was indicating less than .0005″ runout on this job.

A few weeks ago now I also made a video of the chuck, including some of the mounting of it on the 2 lathes in the shop.  I recommend people to get a standard backplate one and either make up your own backplate or buy one.  By mounting the chuck on a backplate it gives you an interface to adjust the TIR to zero – if the mounting system is directly manufactured into the chuck your options are probably limited to regrinding the taper in situ to improve accuracy of the chuck.

If you are a more of your make your own tools type Andy Lofquist over at Metal Lathe Accessories has an interesting 5C collet chuck kit that you can machine yourself.

 

Cross Slide Screw Support

A few weeks ago I finished a project that I had on my mind for a number of months.  There was a lot of play in the cross slide feed screw on my import bench lathe.  This showed up as backlash in the feed screw – when you grabbed the toolpost and applied force in alternate directions you could see the entire cross slide move back and forth.  Some of it was from backlash in the feed nut itself, but most of it was between the feed dial and the support casting itself.

I tried tightening up the nuts themselves tor reduce the amount of clearance – but then it bound and you couldn’t turn the feedscrew at all.  This wasn’t the best design from the get go.

The first thing I did was modeled the entire assembly up in Fusion to get a clear picture of what was going on – and to give a good starting point for the modification: The nuts aren’t shown on the end of the feedscrew but you can see where the assembly is constrained for axial movement – at the right side on a shoulder machined into the feedscrew itself and on the left side the inner bushing of the dial.  These are just 2 plain bearing surfaces – and they weren’t machined the best to begin with.  No wonder it wasn’t the best!

I thought about doing what Stefan Gotteswinter did.  If this was my main lathe I would copy what Stefan did as it is the best solution by far.  Angular contact bearings are the way to go in this situation.  Since I’m keeping this lathe around primarily for cutting metric threads (the Standard Modern now in the shop doesn’t have a metric transposition gear) I decided to scale back the project and see if I could just stuff a deep groove axial bearing and a roller thrust washer into the space without having to modify the leadscrew, or make up a new dial.

Below is what I came up with:

I incorporated a deep groove ball bearing (6900-2RS) and a 10mm needle style thrust washer on the opposite side.  This required a new housing and the old cast iron bearing support to be shortened up.  The new housing was doweled to the cast iron block for location.  2 counter bored cap screws hold the entire assembly together.  The cross slide screw required minimal rework – a shoulder had to be turned for the bearing to sit against.  I also turned down the shoulder on the screw that previously was a bearing support.

The ball bearing is preloaded using the existing nuts.  Care needs to be taken not to overload the ball bearing as deep groove ball bearings aren’t primarily designed for axial load.  In retrospect I should have flipped the positioning of the deep groove ball bearing and thrust washer around when thinking about cutting forces as the cutting tool pushes away from the work piece.  If I have problems I can always make a new bearing housing.

The cross slide now is super smooth with no backlash due to the support.  There is a bit of backlash in the screw, but I don’t get too bothered by that on a manual machine.  It is significant improvement with not too much effort or time required.

I made a video of the entire project as well:

 

 

The Future is Here: Introducing a Laser Bandsaw

A few months ago Max and I recorded a podcast where Max and I theorized on the what the home metal shop of the future would look like.  The podcast idea was inspired by the Making It podcast hosted by Jimmy Diresta, Bob Clagett and David Picciuto where they talked about what they think the future maker workshop will look like.  During the podcast Jimmy, Bob and David mentioned the idea of a laser bandsaw – something that Max and I also talked a bit about on our podcast.

A month or so after the podcast Rod Shampine reached out to me to talk shop over the phone one night.  I had a very enlightening conversation with Rod, who is an exceptionally gifted mechanical engineer with over 50 patents.  Rod also has his PhD and has worked on some exceptionally interesting projects – both on the job and for hobby.  He is an active home shop machinist as well.  Over the last few years Rod has done a significant amount of work in the 3D printing world – he is very active on Thingverse and also has did a significant amount of work on 3D printers themselves.

In the conversation Rod told me he had acquired all the hardware to put together a laser bandsaw prototype.  At first I thought he was making a joke, but he went into specifics about safety, power supplies and the actual laser itself.  Rod certainly had a workable design flushed out – one that both had us very excited.

In December Rod sent me an email that he had finished his laser bandsaw.  It could only really cut paper and balsa wood but to my knowledge it is the first working prototype of such a device.  Obviously a laser bandsaw is exceptionally hazardous – particularly to your eyes.  Rod pointed this out numerous times.  But with proper eye protection and proper design a device could be made to work.

Shortly afterward Rod posted his working prototype on Youtube:

So thanks to Rod the future is now here.  I’m watching with interest to see where this all goes.

$200 Shenzhen DRO (JCS900-2AE)

About 6 months ago I purchased a digital readout off eBay for the Rong-Fu mill drill. Originally I had planned to purchase either iGaging scales or standard import calipers and utilizing a tablet based DRO. Once I started looking at prices however I was shocked to find that for less than the price of either the iGaging scales or the import calipers I could have a full blown 2 axis DRO complete with proper glass scales.

I went ahead and ordered the scales off a eBay seller. It was a typical Chinese eBay seller that sells everything from DROs for machine tools to various useless cell phone and house gadgets. The total for the order was around $200 USD plus about $30 for shipping. I then communicated the scale lengths I needed via email. In about 2 days I had a shipping confirmation including a tracking number.

I was excepting to wait about 3-4 weeks for the shipment to arrive, typical of most stuff ordered from China. I was shocked at the end of the week when I received an email from DHL that my shipment was to arrive on the following Monday – about a week for the entire process! Sure enough Monday afternoon a DHL driver dropped off the 2 boxes.
The one box contained the DRO – a JingCE JCS900-2AE 2 axis DRO unit. The other box contained the 2 glass scales both of proper length. Also included was a large amount of hardware, mounting brackets and associated items you would need to install the DRO.

I spent the next few days thinking about how to mount the scales. The X axis was easy – I decided to mount it to the front of the table using the T slot already present. I thought about mounting it to the back of the table but I didn’t want to loose any Y axis travel. The Glass scales are rather bulky – something to note if you are considering installing them on a smaller mill like the X2 mini mill. The Y axis was a bit of a different story – there really isn’t anything to fasten the scales to. I decided to make up a bracket to hold the Y axis scale. That took a fair bit of work to do.

Y Axis Scale (behind the fabricated bracket)

X Axis Scale

After mounting the scales and trying out the DRO I also fitted a inexpensive import digital caliper to the quill to get a .001″ resolution readout for Z depth. This also took a few hours to do properly. 2 brackets were made out of aluminum to hold each end of the caliper. I modified the caliper using a Dremel tool. I drilled mounting holes using a standard off the shelf masonry drill bit – a poor man’s way of drilling hardened steel. High speed steel usually won’t touch hardened calipers.

Hard Drilling Using a Masonry Bit

After using the DRO for 6 months I can say that it is a very good unit. I haven’t had any issues. As far as accuracy and repeatability is considered, I really don’t have the proper measuring tools to qualify the DRO but I will say that I tested it using a dial indicator over the travel of the table. At each point where I tested the DRO it corresponded to the dial indicator – within at least .0005″ (as best as my judgment permits). I also ran the table up against a hard stop several times to test the repeatability and each reading was easily within .0005″. I probably should do a proper gauge R and R study on it, but just with the general testing I’ve done it’s easily within .001″. And to be honest doing work closer than .001″ on a Rong Fu mill drill is unreasonable.

I filmed and edited a number of videos showing the install and finally a video review of the DRO.  The first video shows the hard part: the installation of the Y axis scale.

The second video shows the installation of the X axis scale and also the mounting of a digital caliper on the quill.

The final video is me talking about the DRO itself and contains much of what is written here.

If you are interested in reading the manual, I scanned a copy of it and it is available here.

One thing to note is that you will get little to no support with the DRO. To me this isn’t a big deal at all when you consider the price. The next closest DRO in price in the North American market is approaching 4 times the cost. And the unit looks suspicious like this unit. If something breaks I am willing to try and fix the unit myself or simply replace it.
If I had a high end knee mill I would probably buy a Mitutoyo DRO and be done with it. But putting a Mitutoyo DRO on inexpensive import mills is a bit like putting lipstick on a pig.

It was a $200 well spent. Having a DRO on a milling machine is exceptionally handy. I won’t say it is a necessity, but it greatly improves your efficiency – especially on larger mills or making larger parts. Time will tell how durable the unit is but I think it is an excellent addition to a home metal shop.

Fixing a Mill Drill Stand

If you listen to the podcast you already know that I purchased a Rong Fu Mill Drill.  While some people have issues with the round column, the mill drill is a significant step up in machine capacity and machining performance when compared to the X2 mini mill.

For those unfamiliar, the Rong Fu Mill Drill looks to be a heavily modified drill press.  There are several size variations, but most utilize a R8 tapered spindle with provisions for a draw bar.  The dovetail table has a relatively large travel of about 450 mm (over 17″) and 200 mm (just under 8″).

While there is much debate on the origins of the Taiwanese mill drills that started showing up in the 1970s, the most probable explanation is they are simply rather crude copies of the Fehlmann mill drill machines.  Fehlmann is a Swiss machine tool manufacturer and they still build a number of mill drill machines, although I suspect you if have to ask the price you cannot afford them.  Besides the very similar appearance, the main reason I think the Rong Fu mill drills are copies of the Fehlmann is primarily because of the tapered gibs on the Rong Fu table.  Fehlmann being a Swiss machine tool company in and of itself is another telling reason why they were copied.

Round column mills are not just limited to two companies.  Emco also  manufactured several round column mills around the same time as Rong Fu started.  A German company  also manufactured a nice home shop mill drill branded as Ixion around the same time or slightly before the Rong Fus started flooding the home shop market.

The Rong Fu mill drill I purchased came with the typical flimsy tuna can stand that is oh so common on import machine tools.  I did not purchase the machine new and the previous owner was selling the stand with it, otherwise I would have passed on the stand and just built one.  Initially I was going to weld up a new stand out of 2×2 steel tubing, but then I thought could I just dump a bunch of concrete in the bottom and kill 2 birds with one stone; adding weight and rigidity?  That and I find concrete a very useful engineering material in the home shop from previous antics.

And that’s exactly what I did.  For less than $75 and one day’s home shop work, which is less than what the material alone would have cost for a tubing stand, I now have a rigid machine tool stand.

I’ll be posting further on the mill drill as I use it, but so far it has been a great addition to the shop.

Schaublin 102 is Making Chips!

Although I’ve talked about it with Max on the podcast, I’ve never announced on the blog that I picked up a made in Switzerland Schaublin 102.  102 is the turning radius in millimeters (about 4 inches).  The lathe was in pieces, but in very workable condition.  I dragged it home and it sat for a few months until I found the time to get to working on it.

This week I finally managed to get the 102 making chips.  It took some work mostly in the drive area.  I didn’t have access to the proper voltage to drive the existing motor so I decided to replace the original Schaublin motor with new 3/4 HP Baldor that I picked up a year ago for $50.  I also wanted to keep the mechanical variable speed drive working.  I could have got the old motor rewound, and I might do that some day, but the $800 that I was quoted was a bit rich.

After modelling up the existing motor in Fusion I designed up a pulley to fit the Baldor, spacers to place the new motor in the same location as the old one and a motor mounting plate.  I used old school methods to make up a plate to mount the VFD and associated electrical components.

I made a montage type video of all the work:

I was pretty happy with how it turned out.  Here is an animated gif showing a test cut that I did with the lathe:

Aside from making up the required parts, I spent a fair bit of time cleaning out the bed, cross slide, tailstock and the interesting air – oil lubricating unit for the spindle bearings.  I also have several hours fishing the air – oil lubrication lines back into place in the headstock.

I’ve never used a plain turning lathe before, and quite frankly until I did I thought they were a bit of a joke.  In the past no carriage or leadscrew caused me to immediately write off plain turning lathes as useless machines.  Nothing could be further from the truth.  In fact, as I used my import lathe more (which has a carriage and leadscrew) I realized that I do 80% of my work without such features.  If you have the chance to pickup a plain turning lathe in good condition, jump at it!   Many people devalue such machines and as such you can sometimes get a very good deal on a lathe that is exceptionally capable – and a joy to use.

Next up is a proper toolpost for the lathe, a backing plate for a Buck 6 jaw chuck I picked up, and probably a faceplate.  That is unless I manage to pick this stuff up used somewhere.  I really don’t count on that happening though.  Parts and associated tooling for Schaublins usually demand high prices.