Collecting dust

For a while now I got by with using a small cyclonic separator in the form of a DustDeputy clone. This is built to sit between the vacuum cleaner and the nozzle hose separating dust from the airflow and collecting into a container. A plastic bucket in this case. It is simple yet very efficient. Even collecting fine plaster and fibreglass sanding dust without clogging the filter on the shopvac. Meaning I can sand in the house without making a mess or messing up the shopvac.

It has some limitations though. Shopvacs are not intended to run continuously. The extra bucket fitted with the cyclonic separator makes for a bit of a clumsy setup and is a little tippy until it has enough in it to weigh it down. I like the general portability and I will keep it for doing small jobs elsewhere. However I am considering stepping up the design and making it more permanent.

Most basic/smaller commercial cyclonic dust collectors are separate units having to be plumbed into an existing dust collector. Other actually come as separate units. They are also often 2hp and upwards meaning that they put a decent load on the home workshop power supply before even running another tool. The really basic units are just a blower and a dust filter bag with a motor in the 1hp range displacing around 500-750cfm which is more than adequate. The 100mm ports are also easily suited to ducting for the home workshop. This forms the basic unit for a home setup than can run continuously. Steel drums are readily available for ~$5 and will make a great bin for collecting into. Which leaves the cyclone… a basic sheetmetal construction and a few lids to make it all fit together in a modular fashion so it is easy to service or clean. The blowers can be a little noisy so one consideration is to build it into a soundproof enclosure, outside the workshop. This will free up space and make it easy to get to in order to empty it making plumbing a 100mm tube through the wall the only remaining obstacle.

Dipping my toes in electronics again

A few months ago I picked up a Vertex BS-0 dividing head during a relocation sale. It was heavily marked down as it was missing the faceplate, drive centre, dividing plates other than the #1 fitted and the tailstock. The spindle nose appears to be threaded for 1.125" x 12tpi Whitworth similar to Myford lathes which makes obtaining a faceplate or making one fairly simple. Spindle taper is MT2 and my existing rotary table tailstock may suffice for the dividing head.

This leaves the dividing plates. Which happens to be the problem facing me making my first set of gears for a bead roller and needing one of the other plates for the size I’m contemplating. I could change the pitch and diameter to suit the plate I have. However I will just run into this problem next time I need to build something and then there may not be room to adjust the design.

A replacement set is available. They are not cheap either and in thinking about it the age old method of indexing can be very tedious and easy to make a mistake. This is one area where technology is still catching up. Luckily many people out there have caught on to the simplicity of indexing and motion control using servos and basic electronics. There are not many commercial solutions and the better ones are only available internationally. All requiring some retrofitting and modification to suit the particular device. Research revealed many published solutions based on common PIC and Arduino stepper controllers.

The basics are this:

·         Stepper motors generally have something like 200 steps per revolution

·         The BS-0 dividing head worm gear feeds 4 revolutions per degree of spindle rotation

·         This allows for 800 steps per degree or 288000 steps over 360 degrees

This is a fairly high resolution for indexing and even mathematical rounding shouldn’t have too much of an impact for diameter that would fit the dividing head or under the spindle of most milling machines. Even with the stepper directly coupled to the feed drive on the indexing head. As it turns out the electronic solutions are not only simpler and more process fault tolerant but make indexing as simple as entering the number of divisions and stepping through them one by one. Building the control into a pendant style enclosure would also allow it to be swapped between rotary table and the indexing head. This would also allow the steppers to be integrated as a 4^th axis for later CNC conversion.

Now to come back to my Tom Senior mill power feed. I had a choice between stepper and servo with servo being a better choice for constant feed and simpler speed control. However consider the possibilities of synchronising the indexing head feed with a stepper controlled table feed and being able to cut helix or spirals!

Going sideways with a horisontal

Decided I have tripped over the old mill for the last time. So after months of sitting around in pieces I started cleaning and cleaning and cleaning… and more cleaning. The old swarf tray had been patched and welded and bumped and bent so much that it looked a little tired. It was also the one thing that looked a little out of place given all the nice round curves on this machine. So I welded up a new one and used the old to transfer all the holes I planned to re-use. The base also had a section used as a sump for flood coolant and as I don’t use full flood coolant in the home workshop decided to leave out the drain holes. This also reduced the machines footprint by a lot.

After that there was a lot more cleaning and taking everything apart for cleaning. Stripping paint and painting and slowly reassembling. Cleaning the spindle revealed that the original was a little battered and worn and bearings needed replacement. As luck would have it I found a spare spindle on a local trade site which I bought some time ago. Sourced new spindle bearings. Installed new oilers. Replaced some thrust bearings and straightened the knee screw. Welded up some castings. Ordered new v-belts. Rewired a new single phase motor with a magnetic contactor switch and finally all up and running

Here is a photo before fitting the motor:

With the gib adjusted in the centre the table binds a little with it reaching the end of travel to the right. That was expected given most work is done on the right hand side of the table feeding to the right. I will need to source a surface plate, long enough straight edge and some marking fluid so I can scrape it in.

One thing I have not addressed is the table feed mechanism. This mill was heavily modified to use a sprocket chain to power the feed gearbox. Some parts were since lost or broken and what was there was horribly cobbled up. One sprocket was crudely welded to the end of a drive shaft which I ended parting off on the lathe. The drive worm that feeds the worm gear on the leadscrew is missing along with part of the telescoping drive shaft and the bracket allowing for it to be disengaged. Research revealed the specifications for a replacement worm though the remaining missing parts complicate the rebuild. At this point the machine is not original and later models replaced this mechanism with a servo powered solution allowing ease of change in direction and finer speed control. I am tempted to give the table feed an independent drive motor. This allows the feed to be disengaged and the feed motor turned off when not needed, reducing wear. However it requires most missing parts to be built.

Alternative is a more compact stepper or servo driven solution fitted directly to the table with a basic control panel off the side of the machine. This may end up being more elegant without detracting from the machines appearance. One benefit of the original setup was that it was driven off the spindle making it in tune with the spindle speed. That means feed would adjust with the spindle allowing for 3 basic adjustments on that speed. An independent drive mechanism would not offer that without adding some feedback from the spindle. Which would mean a little guesswork to get the feed ration dialled in. However this can be achieved by adding a sensor to the spindle and using a stepper driver to control step speed in relation to spindle speed. This all warrants a little thinking. In the meantime the machine can be operated manually.