The weather was very nice so I disassembled the drum sander, took it outside, and slathered blue paint all avo. I painted all exposed particle board surfaces because I want to make sure it does not swell and crumble if I accidentally spill a few drops of water on int in my workshop. I did not do a very neat job but from a distance, it does not look so bad.

I did not paint the tilting table because the paint there would be a hindrance and not a boon.

I also improved the inner geometry of the dust cover by gluing in two pieces of polystyrene and grinding them on the drum until there was just a small gap.

The P40 abrasive with velcro finally arrived and it is scary stuff.

Just putting it on mangled my fingers so much I had trouble logging into my phone via fingerprint recognition for a few days. I forgot to put in the safety screw on the right side as you can see in the picture. It did not appear to be a big deal, just folding the paper into the cut was enough and it run without a hitch for a few hours before I noticed the missing screw.

I was baffled at first that even with 40 grit it took me two hours to flatten four boards, i.e. with the same speed as with 80 grit. The machine was stalling all the time whenever I tried to be a bit more aggressive with it. And after two hours I realized that the leather belt stretched a bit and did not have enough tension to power the drum effectively. I knew that the belt would stretch with time so it is a tad embarrassing that it took me two hours to realize it had done so already. After I corrected the tension (by making a jig for precise cutting of the belt at an angle and shortening it), I breezed through the remaining six boards in about an hour and a half, i.e. twice as fast.

After that, I cut these boards into strips with my table saw.

I took some of these and I glued an end-grain cutting board of the biggest size  I intend to make – ca 30×50 cm – and I am currently in the process of flattening it.

This one will not be for sale since it is a learning piece. I will use it personally to see how it performs but it has some gaps and thus it will probably serve mostly as a pretty background for photographing knives.

I learned that I will have to take more time and care with the glue-up to save time when flattening the glued piece. Jatoba is extremely hard in all directions but trying to sand or plane the endgrain is truly a penance. I spent three hours today doing it and I am nowhere near finished. At least that is an assurance that if I make and sell cutting boards from this wood, they will last.

When I figure out the ins and outs of making the cutting boards I will post about it. As far as the drum sander goes, it is finished and functional. All that remains is to figure out where to store it when it is not in use

 

My brother and my sister-in-law were busy as bees and between the three of us, we managed to do all the necessary work re-painting the kitchen in just two days. On Wednesday I had to do some little cleanups and shopping and on Thursday I could continue to work on my drum sander. So I did.

After the first successful test run of the drum sander, I glued up a 50×1500 mm leather belt to drive it. For better precision, I bought an already-cut leather strip (a prefabricate for leatherworkers to make belts), cut both ends at an angle of 45°, and glued it up with special extra-strong elastic cement for shoes. I was thinking of reinforcing the seam with sewing, but so far it does not appear necessary – the glued joint can withstand all the force my hands can exert, which is significantly more than the spanning spring on the belt sander can do. I run into a minor setback here – I made the driving wheel on the drum sander with a groove as I would for running it with a rubber belt. That was a mistake, the leather jumped the low fence quite easily and when I tried to make the fence higher, it jumped the spanning wheel and the motor wheel. The best solution (which, luckily, did not take long to find out and implement) was to remove the fence on one side and to make the cylinder wider. The leather belt is not precise and needs some space to oscillate from side to side.

After this, It was running quite well. The crowned spanning wheel does keep the belt on track and on the motor wheel, and on the driving wheel of the drum sander, it moves a bit side-to-side when running. With that problem solved, I could move on to dust collection. Pushing the wood with one hand and vacuuming the dust with the other is neither an easy nor elegant solution.

The dust collection does not need to be especially sturdy. Its main two functions are to stop sanded-off particles flying all around the shop and to prevent me from accidentally grinding off my fingers on the drum. Thus I opted for the easiest and cheapest material to use I have available – cardboard. And it is held in place by four M6 fly screws that go into threads cut directly into beach wood.

I cut a piece of pine board to put across the drum and I glued on it fences from cardboard that slot into the grooves I pre-cut in the two pillars holding the drum. To prevent the edges from fraying, I glued strips of sturdy paper over them. To attach the vacuum cleaner I used a reduction I saved from my old hand-held circular saw when it broke.

As far as inner geometry goes, I tried to glue in two strips of paper to reduce the space between the drum and the walls. I tested it with these and without them and I think that they did help with the dust collection, but they were not very stable and had issues (dust collected behind them). I am currently in the process of finding a better and more permanent solution to getting the inside of the housing to conform to the drum better. For now, it works, albeit it can be improved.

Initially, I intended to glue the belts on the drum but rsmith’s link in the comments under the first article gave me the idea of using velcro. If I can change the belts reasonably fast, I can use the drum sander not only for material removal and flattening of surfaces but also for getting a reasonable surface finish on completed cutting boards. And I also could have a separate set of belts to flatten metals, if I need to. So I went and ordered some velcro and a set of belts. The velcro was so far the most expensive part of this whole assembly- about 30,-€! I hope it was worth it. I glued it in a spiral around the drum and I wound up a strip of abrasive around it in the same direction. On the trailing edge (left side of the drum) the velcro suffices to hold the paper in place. On the leading edge, I had to make a groove, bend the paper into it and secure it with a screw.

The next day I made a few optimizations. I improved how the whole assembly is held near the belt sander by inserting two M10 hammer-in nuts into the table. I made one of the two supporting legs adjustable in length so I could adjust it to the uneven ground of my workshop. I made two long M8 fly screws and I used them together with two hammer-in M8 nuts to attach the legs in a secure yet easy to disassemble way. And finally, I made a specialized push tool to shove the pieces through the sander without endangering my fingers and without exerting myself needlessly.

It is made from an old furniture leg and two offcuts from an old spruce board. The table height is just below my navel so I can easily brace this thing against my belly and gently but firmly push with my body weight.

And with all that done, it was Friday evening and I went to bed late. Yesterday, I ran it for about 6 hours to test it thoroughly. In those 6 hours, I managed to flatten twelve jatoba boards approximately 15×30 cm. I think I could do the job faster if I used a coarser abrasive, but unfortunately, 40 grit was not in stock and the 60 grit I ordered had not arrived yet. Thus I had to do with 80 grit. I ordered the 40 grit finally yesterday and the 60 grit should arrive on Monday. I still have some unflattened glued-up boards and I will probably wait for those to see how much faster I can do the same work.

I do consider the test to be a success. The machine ran the 6 hours (three 2-hour segments) without a single problem. The boards are not tutti flatti perfetti, but they are flaterooni accepti. As in within a few tenths of a mm over the 30 cm. That is enough for a good and practically invisible glued joint.

I learned a few things while doing this. The food-safe PVA glue does not gum up the abrasive but it does create small resin-like chunks that can attach to the wood and/or the table and make the board run unevenly. So I have to brush those off, especially at the beginning when it grinds off the glue runoffs. I also learned how to incrementally lift the table for best results and I got a few more ideas for improvements and optimizations.

That was significantly easier and quicker than I thought it would be. It is not completely finished yet and I will write about the finishing touches. However, as of now, I have a fully functional drum sander. I made it in about three weeks and so far it cost me less than 200,-€, most of which was the velcro and the abrasive belts.

And while I am waiting for the coarser abrasives to arrive, I will paint it.

With the drum sander being essentially functional, I could start fixing it in its working place and truing the drum. Both issues were simple but not easy, so it took me a whole day and some more to do it properly.

The first problem to solve was how to fix the drum sander besides the belt sander so it could be powered by it. Here has shown my first serious mis-measurement. I had to trim a sliver of the belt sander base because I could not position the drum sander close enough for the wheels to align properly. Then I stood in front of the problem of how to fix the drum sander in place so it does not slide around.

From the start, I wanted to do it with two M10 bolts at the back but my first attempt was a bust. You can see the results of that first attempt, those two yooge holes in the back of the base. The problem was not that it was not stable or somesuch. The problem was that the bolts would be in the way if I wanted to sand something longer than about 40 cm. What followed was about an hour of serious thinking and faffing around in the workshop and when I started to work on another solution, I accidentally found two metal angle irons with pre-drilled holes.

I undid all the work I did in the last hour, screwed these two onto the back of the belt sander base, and cut the 10 mm holes open. That way I could slide them onto the M10 rods inserted into holes in my workbench and fix them in place with winged nuts. I also had some play to slightly re-position the belt sander to align the wheels properly before tightening the nuts.

With the back of the belt sander being fixed to the table, I could run it but I also added two legs on the front just in case. They are not strictly necessary – in fact, one of them is a tad short and is not even touching the ground properly and I will have to add a screw extension – but I feel better with them.

With the belt sander standing, I could run it for real. I used an old trizact belt to power it. That works for now as an impromptu measure, I will make a proper sturdy leather belt to do the job. For now, I stood in front of it for a couple hours running a piece of particle board with 60-grid sandpaper glued to it to true the drum, holding the board in one hand and vacuuming the particles with the other.

Here you can see the drum being sanded down to being nearly, but not entirely, concentric. There are still visible gouges made by the table saw. I had to run the sandpaper under it quite a few times after each lifting of the table, and I only could lift the table about ¼ of a mm at a time. I only ran it at 20% speed at first because the drum was so unbalanced and the whole assembly vibrated violently. After a while, I could increase the speed to 40% as the cylinder became more and more concentric and thus better balanced. That is also the top speed at which I intend to use the machine since higher speeds would burn the wood and destroy the abrasives (I built the machine so the surface speed of the drum is approximately the same as the surface speed of the belt so I can transfer experience from one to the other).

Late in the evening truing the drum was finished and it was flush with the table across the whole length to within a tenth of a mm. I left a bit of a gap here so it can be seen.

With the drum being concentric and true, I could finally fix some abrasive to it. For a test, I used an old torn 40-grit belt. First I wound it around the cylinder, fixed it with rubber bands, and trimmed the edges to flush (these scissors are used for trimming abrasives and nothing else, in case you are wondering). After that, I cut grooves into the sides of the cylinders, tucked a bit of the belt in it, and fixed it in place with a screw.

It worked reasonably well as a temporary solution and I learned a few things when running it for a few hours with some testing pieces of wood. The leading edge held up fine but the trailing edge eventually tore off. It seems that at the trailing edge, the screw is completely unnecessary and a bit of double-sided adhesive tape would suffice plenty. That is good to know, but I am already looking for a better way to do it in the final version. I am ordering some velcro and abrasives with velcro.

And that’s it for now. The 40 grit did hog off material quite nicely. I flattened a seriously twisted and bent piece of jatoba that I definitively would not be able to flatten manually, ever. Although manually feeding the stock through the machine is still hard work (for me) – I did get some workouts on my back and my pecs that way. Still an order of magnitude less work than trying to flatten one of the hardest woods on the planet by hand.

There is still a lot of work to be done and I will do it and write about it. Optimize the assembly/disassembly, optimize the attachment of abrasives, optimize the legs, make dust collection attachment, make a proper and safe push tool to feed the wood through, and finally, paint job. Only I do not know when I will do all this it because I have to take about a week’s pause from this work. I have to paint the kitchen now. I hate it, but it needs to be done and it has to be done now because now my brother can come by to help with things that need four hands.

I took Sunday off and on Monday I had other things to do but yesterday and today I could work on the drum sander some more and I progressed significantly. First I finished and attached the propulsion wheel. I started that right at the beginning but it progressed slowly because I glued it from several layers of old plywood and I had to wait for the glue to dry a few times. Just before the two-day break, I soaked the surface with epoxy resin to make it harder and on Tuesday I turned and sanded it until it was passably round. I will probably give it another layer of epoxy once everything is finished, I will use the epoxy on other parts too to improve their surface hardness and strength.

The wheel is attached to the axis with an M4 screw going through a thread drilled and cut into the axis. I hope it is strong enough and won’t get sheared off in work because I do not have any better idea how to do it. I need to be able to eventually remove this wheel in case a ball bearing gets busted and needs replacing. I am also leaving myself a bit of the axe poking out so I can power the thing with my drill in case powering it with the belt grinder does not work.

On the right side, I added a fence with a shelf to lift the base 55 mm; on the left side, I simply added a 55 prism. Then I screwed two spruce boards on it to reinforce the structure. Though I screwed them on with the wrong grain direction of the core but it seems to hold strongly enough. Should a problem occur, I will reinforce it later or replace it with properly oriented boards.

The base under the propelling wheel is unsupported. This has a reason – I need to be able to slide this side over the base of the belt grinder because I need to align this wheel with the motor and the spanning wheels of the belt grinder. It should not be under too much stress but I did reinforce it with steel under the wheel where there will be most of it. Again, I might add some more support later, for now, I think this suffices.

That was Tuesday done. I attached it improvised to the belt grinder and it worked, albeit it was loud as hell. I think that can’t be helped, the ball bearings are not perfectly aligned and the drum is not concentric yet and I can only solve the second of these two issues.

Today I made the tilting table to adjust the sanded thickness. And here I must thank to rsmith for pointing me to a page with useful information about how to make a very simple one.

I started by cutting-out a piece from the lifted base. This was the plan all along, to use this space for some sort of mechanism to lift the table. On the front, I added a 55 x 20 mm beech board across. Beech is reasonably strong hardwood, it should be able to support the table. In the middle of the board is inserted a 24 mm long M8 nut with a sufficiently long rod through it. To operate the screw I tightened on it a hex nut and a winged nut against each other.

I decided to make the table from a 15 mm particle board because the piece I had had a very smooth surface and it is straight. However, particle boards are not exactly known for their strength, so I had to reinforce it. I reinforced the rear end with stripes of beech wood for the hinges to have better and stronger anchor points. I put it under weight and went for lunch while it was setting

After lunch, I continued reinforcing it. On the front end, I added an offcut of angle iron. That also provides a hard surface for the screw to lift the table.

For longitudinal support, I screwed on two beech strips. I forgot the exact measurements, somewhere around 25×15 mm or somesuch. I tested it and it seemed to be sturdy enough against bending, but not against twisting due to the table being supported by a single point on the front. It should not be a huge problem if the sander is loaded in the center as it should be, nevertheless, I decided to reinforce it further with three offcuts from the same particle board intended for firewood.

One continuous piece would be better and I had a suitable one – the one that I cut out of the base in the morning – but I completely forgot about it until these were screwed on and I did not want to go through the hassle of trimming it and drilling all the holes again. I think it is strong enough and if not, I can add more support later. I say that a lot.

And this is where I finished today. I think the work is progressing nicely. I have a tilting table with a range of about 35 mm, which is plenty for my purposes. It takes about 80 half turns to lift from zero to full. That is enough sensitivity to sand off only tiny layers of material when needed. The whole assembly so far is sufficiently light and compact that I can easily lift it and put it under the table next to the table saw when not in use. I am reluctant to be optimistic but I do feel ever so slightly positive about this project.