Making a Palm Press

Printing lithographs and monoprints without a press
Nik Semenoff

Concept behind the Palm Press

Rubbing the back of a sheet of paper pressed to an inked surface has been used for centuries, but the Japanese have carried the technique to great heights in their woodcut prints by using a baren made from a bamboo leaf over a braided core. Wooden spoons and other items have been used by contemporary block printers to print their editions, but the Japanese baren is still the preferred choice of printing small woodcuts. It has been duplicated in plastic and assembly of metal balls, but the traditional bamboo baren cannot be beaten for delicacy in printing. Unfortunately this tool cannot produce enough pressure for any printing any media other than woodcuts. The Japanese technique of using thin waterleaf paper and waterbased ink allows for a softer gentler application of pressure to give complete control of transfer of ink to paper.

Because of the high cost of good traditional bamboo barens, some Japanese printmakers started to experiment with steel bearing balls assembled in the holes of metal speaker grills. Plans for making such simple devices are available on the internet as it is now impossible to know who indeed produced the first such ball barens. A number of names have been mentioned and there were even attempts to motorize the action by attaching the ball assembly to an oscillating platform driven by a small electric motor. Some Japanese craftsmen have tried to improve the ball baren by stacking two balls in each cell, hoping to get a rolling action in the bottom one that contacts the paper. The ones I have seen are all individually hand made and very expensive, but are preferred to print heavier passages than what the bamboo units will handle.

One of these ball barens was shown to me with the hope it could print waterless lithographs, as the artists did not have a press and not the space for one. After some tests, I came to the conclusion there were far too many balls, which greatly reduced the pressure needed for printing lithographs. I started research into making a more effective unit by reducing the numbers of balls, using low friction materials to ensure that the balls revolve, and designed a handle that allowed the printer to put greater pressure on the unit. Initial tests were promising, but when the concept was tried in a high school art class, the pressure was found to not be sufficient for making consistent good prints. By not being careful and applying more force to one side of the unit would produce circular lines in the image from the action of printing. A different approach had to be taken. On the suggestions of my daughter, who was teaching the art class, some sort of roller might produce better results.

Using a roller system

My first attempt was to use a wide small diameter roller made of rubber, but the need to turn the unit while applying pressure made it difficult to operate. Other versions of even small diameter rollers, but made of metal were no better. Finally I decided to try a collection of small ball bearings mounted on a metal shaft. These work as rollers or wheels to transfer pressure to the paper. Smooth revolving of the bearings and their individual rotation, allowed me to get over the problem of turning and applying greater pressure to the print. The next problem that arose was the stress on the wrist to balance the handle over the bearing rollers as this would lead to a form of carpal tunnel syndrome. By staggering the rollers on two separate shafts, this stress was greatly reduced.


Ball bearings come in both metric and imperial measurements, but metric bearings are more common in international commerce and made in larger numbers, making them much less expensive. Since the pressure any human can apply to the bearings is very little and few revolutions are required, one does not need the more expensive units. While metric bearings are considerably less costly, the metric rod used for making the shafts is not readily available in North America and very expensive. The only metric rod available in Canada is drill rod steel, supposedly a quality product, but I have found variations in the diameters of similar rods. This has caused problems in mounting bearings on the shafts.

A handle for the palm press

While injection molding with suitable plastic would make a very nice unit, the cost of making the molds is extremely expensive. For the small amount of units I might sell, I decided to use hardwood for the handle portion and five-ply Baltic birch plywood for the side panels that support the bearing shafts. This concept has proven to be quite acceptable and gives printers a comfortable unit for printing. Wood screws are used to put together the unit, after the bearings and shafts have been assembled. To get the placement of the predrilled holes into the exact position to allow for the adjustment of the shafts, I built a special drill press that drills two holes at the same time. The same drill press is used to put the holes into the plywood side panels.

Assembling the bearings on the shafts

After trying one unit that had a total of nine bearings, five on one shaft and four on the other, I came to the conclusion that a lesser number of seven would work as well. Depending on the width of the bearings, the rod for shafts is cut to accommodate four bearings, plus twice the thickness of the plywood. I add a small amount to make sure the shaft has a bit of space between the plywood and the end bearings. Since two different sets of assembled bearings are required, one has to layout the spacing on each of the shafts so that the one set of bearings will cover the space in between the bearings on the other shaft. I use four bearings on the back shaft and three on the front one; this means there will be a space approximately the width of a bearing between those on the back shaft. I have found suitable plastic pipe that I cut for the spacers.

Bearing inner bore diameters are kept to very close tolerances and are usually mounted on shafts that fit the hole very tightly. The normal procedure is to heat the bearing until it expands enough to slip on the shaft; on cooling, the bearing will not move on the shaft but spin within the assembly as planned. Since there is more than one bearing mounted on the shafts, the procedure gets more complicated. Even heating the bearings does not assure the bearings will just slip on easily, so an arbor press has to be used to arrange them on the shaft. They have to be positioned very accurately as once they cool, they cannot be moved. The spacers help, but since they are made of plastic, they can be compressed and reduce the spacing between any two bearings. A temporary removable metal gauge placed against the newly mounted bearing, will keep the spacing between it and the next bearing, as the bearings are positioned.

Modifying an arbor press

Arbor presses from China are available for a reasonable price. These come with a standard anvil that has four U shaped cutouts to support work pieces. A more accurate method is needed that will apply pressure in a direct line so the bearing does not try to go on at a slight angle. Because of the very close tolerances between bearing and shafts, this can prevent the assembly of the unit. I have built a special anvil that has a hole that is just slightly bigger than the shaft and in perfect alignment with the rack on the press. This gives support to the bearing and ensures the shaft is pressed into the bearing hole without binding. As the bearings are mounted, it will allow the rod to go into the anvil hole as the bearing slides farther on the shaft.

Assembling the four bearing shaft

Start by mounting the first bearing away from one end -- about the thickness of the plywood and the thin spacer. From the other end, now slip on a spacer the width of your bearings and put the next bearing over the hole, then press the shaft into it and through the anvil. The bearings are quite hot and have to be handled with special pliers. Continue putting on spacers and bearings alternatively until all four bearings are in place. If your calculations are right, there should be the same length of rod sticking out of both ends of the assembled unit. This should be the thickness of the plywood end panels and the thin spacer to keep the bearings away from the plywood.

Assembling the three bearing shaft

Start by placing the first bearing away from one end the distance to the first space on the four bearing shaft. Using the same procedure as before, mount the other two bearings. When placed next to the four bearing shaft, this unit should cover the three spaces between the other bearings. Use spacers just long enough to leave the ends uncovered to the same thickness as the plywood panels. I have found spacers necessary, as some of the metric rod is not quite as thick, allowing the bearings to slide sideways on the shaft. This was quite surprising and very frustrating, as it is supposed to be quality drill rod steel product meant for precision work.

Making the handle

Hardwood of a suitable width to support the bearings and give you a comfortable hold is cut to individual lengths. This will be just a shade longer than the shafts. Leaving a small amount to act as a stop along the topside, cut away the thickness of the plywood at each end. There are two holes at each end to screw on the side panels, which must be very accurately placed in relation to one another and the holes that will be put into the side panels. While I decided to build a special drill press with two spindles, there are methods of using a jig on a standard drill press. The holes are predrilled deep enough to take a wood screw without splitting the handle. Forming the handle to be more ergonomically comfortable is an option, but will require extra care in construction.

Making the side panels

I have found that the handle should be at an angle to the printing plate - about 17 degrees higher towards the front. Five-ply Baltic birch plywood is the best to use, as it is much stronger than common three ply made in North America. Layout the end panels so that the bearings just stick out at the bottom about an1/8th of an inch, and about the same amount front and back. The backset of bearings should just clear the handle, leaving enough space above the front set for fingers. Cut the panel as a pair and somehow temporary glue them together. Temporary hot gun glue is a possible solution. I use the same double spindle drill press to drill holes to fit the handle and also the larger holes to take the shafts, but standard wood working procedures can be used. All these holes must be perfectly aligned on both panels, and fit the holes on the handle as well. Use Forstner or lipped brad-pointed drill bits for the shaft holes so there is no tear out. The top set of holes should be a little bigger than the diameter of the screws being used to assemble the panels to the handle. After sanding and painting (optional), glue a thin piece of dense urethane foam (about 1/16th inch thick) on the top edge of the panels. Trim foam to the width of the plywood. This will act as a "spring" when assembly or adjustment has to be made on the alignment of the two shafts.

Assembling the palm press

Screw one side panel to the handle, but don't completely tighten the screws. Put the four bearing shaft into the back hole and the three bearing assembly in front. Put the other side panel into place and lightly set the screws. To make sure that the two shafts are absolutely parallel to the printing surface, I use an accurate machinists granite surface plate whose surface is flat to .0001 of an inch. By pressing hard against this surface, I tighten the four top screws. I try to rock the unit to make sure it is steady, then loosen and retighten the screws if required. If the wood handle should twist due to change in humidity, this would not let all seven bearing to come in contact with the paper, leaving unprinted areas. This "spring" feature allows the unit to be adjusted.

Testing the unit

To make sure that all the bearings are perfectly aligned, I use the following method. I roll up a small plate with black ink and place a sheet of smooth laser paper on it. By rolling just partially across the back of the sheet once and removing it, I can see if all the rollers apply equal pressure. Depending on the spacing between the bearings, there may be a thin white line -- or none at all. This is not too important as the press is rolled over the sheet many times in all directions, obliterating any unprinted lines. If there is a twist in the unit, one end of the shaft will produce less pressure at anyone of the four corners. This can be adjusted out by loosening the screws and pressing down before tightening them again.

Printing with the palm press

This unit will turn out quite acceptable small prints on smooth rag papers, but not intended to replace a large lithographic press. Since the pressure is limited when compared to the traditional transfer press, I have found that ink should be modified to increase its' bulk and tack. Under a traditional press, there would be tremendous ink squeeze and darkening of image with this modified ink. A palm press will allow it to be pushed into the grain of the paper and produce a good print.

To make a "press", I use a sheet of Mylar that is bigger than the plate and paper. I attach one end of the Mylar to a flat hard surface with wide masking tape and place the inked plate and paper on top of it - under the plastic. By restricting the other end of the Mylar with my hand to make sure there is no shifting, I roll over the entire surface with fairly heavy pressure. One can test to see if ink transfer is even and complete by carefully lifting one end of the paper, while making sure the rest of it stays in contact with the inked plate. If more pressure is required, those areas can be given special attention. If even greater pressure is needed somewhere, tilt the unit so that either only the four or three rollers are making contact. This will increase the pressure over a much lesser area.

Large flats are a printing problem even with traditional litho presses. Many lithographers use screen-printing for their flats to make sure there is no salt-and-pepper areas. The palm press is even less efficient for printing flats, but if a punched-hole registration system is used, adding more ink to an area is one solution to the problem. Smooth rag papers produce good results, as will some dampened varieties. Due to the slight shifting of the paper from the roller action, I suggest punched-hole registration for color work.

Making offset images

I have found that the palm press and some simple materials, plus registration pins can be used to edition small sized offset prints.  The use of some sort of tight and accurate registration system is necessary and the common punch pin method is perfect.  Since the use of the palm press is usually on 8 x 10 and a little larger plates, any heavy duty three hole paper punch can be used for punching aluminium plates.  Since the registration pins one buys for graphic use are ¼ inch in diameter, make sure the punch you acquire is the same dimension.  The latest punches seem to have larger holes, so maybe you will have to go around second hand stores for the right size.   You should have two sets of registration pins as this makes setup easier. If only one pair is available, then you can tape down one end of the offset blanket instead.

There are a number of possible materials to use as the “blanket”. For the palm press it should be soft and not very thick, but firm enough not to ripple or go out of shape.  One suggestion is the clear smooth vinyl covering used for table tops. Mylar either frosted or clear is another alternative even though it is not soft.  I have found some place mats for the table with just the right resiliency and firmness to make good offset material.  You will need a smooth hard surface to make up the press system. It should sufficiently larger than your plates to you can roll off the edges with comfort.   The blanket should be longer than the plate by a couple of inches to allow it to be folded back from the surface.  First punch a set of holes at each end to take the registration pins. Use the blanket to position the pins on the flat surface by taping the pins down with masking tape. You are now ready to prepare the plate.

The drawing is put on the plate without having to mirror the image since the offset method reverses it.  The plate should have a set of hole punched at one end for registration purposes.  Process the plate as usual and prepare the ink on a slab nearby. You can use a small brayer since the image size will likely be small. Roll up the plate on a siliconed subplate of some sort to help keep the area clean.  Place the plate over one set of pins and place one end of the blanket over the same set, while the other end fits over the set.  The blanket has to be only lifted from the plate set as the anchored set keep things in register.

Roll over the blanket with the palm press until the image has transferred to it. If clear material is used, this helps in judging the rolling.  Lift the blanket from the plate end and remove the plate, replacing it with a sheet of paper with corresponding punched holes.  The first print will be weak as the ink has to first build up on the blanket before it splits evenly between paper and blanket.  Because pinpoint registration is possible with offset, you can rollup the plate again and double print the sheet of paper.  This concept can be used to increase the ink coverage just as I have seen done with large flat bed offset presses by printmakers as registration is perfect.

Modifying ink

Thick bulky ink is the answer for efficient printing with these units. I presume waterless lithography is the process being used, as it is the most practical in this case. Ink is modified according to my suggestions for using my waterless process, but it should be made a lot more viscous than used with a regular press. Epoxy hardener and magnesium carbonate are the best materials to use for thickening Van Son Rubberbase Plus ink. Cleanup with water by using my suggested cleaning solution will keep the studio less toxic.

Another print media for high schools

The cost of lithographic and etching presses has prevented these media being taught in high schools. Only screen-printing and relief seem to be the one being used today. By using the less toxic waterless process and water cleanup of ink, lithography can now be presented as an alternative printmaking media. The recycling of aluminium plates according to my method reduces the cost of material within the budgets of most school systems. It is being looked at seriously by the art teachers within the school system in this area and has been taught at one high school this past year - more art teacher are likely to follow.


The palm press is not a replacement for a well-built litho press, but can produce acceptable prints on any smooth hard surface like a Formica table. I would not recommend a large edition of multicolor prints in a large format, as that is a major undertaking even for the professional printer with a press. It will allow printmaking students who graduate and can't afford a press or have access to one at a cooperative workshop, to still make prints. I have known many graduates who have given up on printmaking because of lack of equipment.

Teaching high school students the fundamentals of lithography by this method will introduce them to how easily editions can be made, since relief and screen-printing is much slower in practice. Instead of rag paper, inexpensive smooth laser copy paper produces good prints, but is limited to letter size. The more advanced students could be allowed larger sheets of better paper as an incentive.

The palm press may look like a very simple device and not expected to produce good prints. The making of a good unit takes great care as there is precision in the placement of the bearings as rollers, which is necessary for efficient printing. It will not replace a transfer press, but in the meantime, printmakers can continue to produce art.

Artists can use it to produce small editions of one or multicolor lithographs. There are a number of smooth rag papers available that produce very good results, rivaling prints done on a proper press. Commercial printers have always known that smooth clay filled papers gave the best reproductions of lithographs and relief photo-engravings. Rag papers for artists come in smooth, plate and calendared varieties, with the calendared preferred for this technique. A soft waterleaf paper like Arches 88 also gives good results. I have used dampened BFK with acceptable results, but like the hard calendared papers better. Rising, Crescent and Somerset are the companies that produced suitable rag papers that can be used with the palm press.

Considering the various suggestions I have seen for simpler printmaking methods, the use of the palm press and waterless litho produces more professional prints in my opinion.  I have produced small editions for research that are practically indistinguishable from the plates printed on a press

Using this innovation

My practice is not to patent any of my developments, but make them available to the printmaking community free of charge. By putting this innovation into public domain, I, nor anyone else can now patent the palm press. My daughter Sasha disclosed the palm press at the Southern Graphic Council conference in Miami in March of 2000. She demonstrated how it could be used in a high school art classes as another printmaking media. By using water-soluble ink and recycling of plates, waterless lithography's more direct approach to image making can have a major impact in teaching art.

Updated March 2003