Metal Type: Home | Library | Forum | Free Ads | Store

Venturing into matrix engraving...

Started by Jason, August 05, 2022, 04:17:53 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Printle: A Printing Word Game from Metal Type


What follows are the fumbling first attempts of a rookie, so any feedback or suggestions are very welcome. I'll do my best to add to this thread as my attempts move (hopefully) closer to success, but at least these early notes might encourage others not to be put off by early failures, because I'm certainly not going to be stopped by the fact that I don't really know what I'm doing!


With all of the pieces now set up (cutter-grinder, fitting machine, pantograph, etc.), I started last week to get ready to cut my first mat. Everything I'm doing is largely flying blind. While I've read extensively on these things, have learned from folks at the ATF conferences, have watched all of Rich Kegler's footage of Jim Rimmer again and again, I'm still learning mainly by trial and error, and getting to know the machines as I go. So, these experiments today were a dismal failure, but it's all part of the process.

My two problems leading to this point were:
1. How to make reliable and usable patterns for the pantograph.
2. How to reliably finish cutting tips after the tool is ground to a zero point.

A few weeks ago I bought a fancy laser cutter, and after playing with it I've come to the following process. I prepare my artwork in Adobe Illustrator, including an alignment "corner" to position the cutter on the mat, and eventually to do the final positioning on the fitting machine.

[Note: I've just today re-read Ed Rayher's little pamphlet on cutting mats for his Cherokee type which discusses using a high-dot for this purpose, a practice I'll now adopt, as I realize now that head- and side-bearing is the more important reference point to mark, rather than the bottom of the artwork.]

I then scale the art (for the 14pt ornament I'm cutting I fit it into a 14pt square, scaling the actual glyph to allow for beard so I don't have to dress the type, and then scaled up x15), and laser cut it out of 1.5mm acrylic. I experimented with engraving into 3mm acrylic, but the laser cut marks on the floor of the engraving were quite prominent. The 1.5mm pattern gives me a very clean and accurate edge. This pattern is then laminated to a matching piece of 3mm acrylic, which provides a good stable base, and a perfectly smooth floor to the pattern.

I then, late last week, got my 1-ft bar of 3/8" x 2" half-hard 360 brass ($80), which I cut to .5" pieces on the bandsaw (switching the short piece at the end to a milling vice and cutting the last few pieces on the drill press with a jeweler's saw on an arbor, the same way I chamfer flat mats).

These rough blanks then went through the fitting machine to true up all four sides. I ended up with 22 pieces out of my 12" bar. Eventually, I'll mill all of these on the fitting machine to get them all to the exact same thickness, but I didn't want to go to far until I tested the entire process.

I then ground 5 cutters to a zero point. This is the real trouble piece of the puzzle. Jim dragged his cutters across a stone to try to get the right sized tips, but this is obviously very crude. I'm also doing it for the first time, so I don't have high expectations at this point. I'm also not sure if I should drag with one of the "corners" of the cutter against the stone, or one of the "flats." I opted for the corner. I marked one of the four corners with a Sharpy, then held the rod with that mark facing the stone (in my case, a diamond plate from Lee Valley). I then held the rod at about 45-degrees, and dragged for about 2 inches with light pressure. Then did another rod for about an inch. And a third one for about 3/8". This actually worked reasonably well. Under the ATF microscope, I *think* I ended up with one at .009", one at .005", and one at .003". Not the 10/5/2 I was shooting for, but close enough for a first try.

In the second image you can see the diamond plate at the front used to drag the tools on to create the tips.

I then made a jig for the pantograph cutting table, which allows me to position one of the mat blanks against one corner, with the mat held in place with a lug.

From here, it was really sloppy. Following Jim's process shown in Rich's footage, I inserted the .009" cutter into the panto and lowered the cutter until it was just touching the mat surface. I then lifted the quick release (I think it lifts it about .050) and lowered the arm to .015 for the first cut. I put in a .100" follower, lowered the quick release, and did the first "counter" cut. This seemed to work, but, man, I had very prominent swirl patterns on the floor of the engraving.

I then lowered the cutter to .030 and did my second cut with the same cutter/follower combo, this time cutting only the actual character shape.

Then I swapped in the .006 cutter, and switched to a .075 follower, and did the third cut to .045.

Finally, I swapped in the .003 cutter, a .036 follower, and did the final cut to .053

I then, VERY STUPIDLY, wheeled the panto arm DOWN instead of UP and stabbed the mat, leaving a deep diamond-shaped stab wound in the floor of the mat. Ugh!

Using an ATF depth gauge, I checked the mat only to discover it wasn't deep enough (it was two steps deeper on the gauge than a .050 standard). Darn, but oh well, it's my first try. Obviously, I wasn't careful enough setting the cutter against the face of the mat blank, and so my .053 depth setting was in reality more like .048

The .050 standard.

My not-deep-enough mat attempt.

I'm going to try the entire process again next week and will report back here.


Second day of early attempts was a significant step forward. After re-grinding the cutting tools and this time dragging them on the stone with one of the "flats" against the stone, I didn't even worry much about the specific tip size. I dragged one of them 2", one 1", and one 1/2", checked the tips to see that they were relatively clean grinds, and got to work.

This time I also added a drop of oil to the matrix blanks for each cut, and blew out the mat between cuts (here's hoping my positioning jig is reasonably accurate).

I also got in a lot closer to the mat to check to get the cutter juuuuuuuust touching the mat prior to setting the drop for each cut. First cut was to .015" with the larger cutter and a .125 follower. Jumping ahead, I now realize that I should be doing both the .015" and .030" removing the entire image area (not just the image) so that the counter is .030" deep. Today, I only cut the counter to .015", which is practically nothing.

After the first two cuts I switched to the middle cutter and a .075" follower, and cut just the image to .047".

Then blew out the mat, put it back in with a drop of oil, swapped in the smallest cutter and a .036" follower, and went for the final cut to .053 (going a bit deeper to be sure I wasn't short like last time).

Taking out the mat it was, as hoped, deeper than necessary, so I trimmed it on the fitting machine to close to perfect depth.

Yes, that's a broken corner on the bottom of mat, but early trials is what broken mat blanks are for!

I then did some adjustments to the shims in my foundry mat holder, and tinkered to get it positioned over the mould.

Casting a few test sorts to get the image aligned (it's a 14pt ornament, but I had the 18pt mould in so just went with it), the first few sorts revealed something very strange: there were about 5 different heights to the image when cast. Huh?

I then took the mat out and looked at it under the microscope, and, sure enough, the floor level was all over the place. Did I forget to cut some areas with the final cut? I don't see how, as I was pretty thorough. Then I took a good look at the spindle in the pantograph and, sure enough, the main component was loose in the arm, which would have allowed the spindle to move up and down during engraving. Yeesh.

So, apart came the spindle to make sure everything was snugged up.

I decided to take a chance and put the mat back in (the final cutter was still in the spindle, along with the .036" follower), and did another cut to .053 (remember, I trimmed off a couple of thou before the test casts).

After blowing out the mat I put it under the microscope and, this time, the floor was pretty level (still plenty of swirl, and some variation in level, but my cutter tips are still very crude).

I then trimmed the face again until I was getting a good reading on the ATF depth gauge.

.050" standard.

The newly trimmed mat.

Then, back into the caster. Now, as some of you will know from reading my other recent posts to this forum, I'm getting really terrible type out of the caster right now (I'm still diagnosing that), but all I needed was one good cast to see what the image looked like. And, well, not terrible!

Next test, I'll cut the full counter to .030", and then just the image for the last two cuts. I'll also try to refine my cutter tips a bit more. I'm still not worrying about head or side-bearing, I'm just throwing it down on the mat to see if it'll cast. Still a LOT to learn and practice, but a big step forward from the first attempt.


Good question. Not even going to try until I sort out my caster issues and get some good type. However, the sort measured good with a type-high gauge. I'm still thinking my caster issues are pump/piston, so I'm going to try all three of my small-type pumps to see if that solves the problem. Only one piston, unfortunately.


No time to get the caster going today, but given the progress engraving, I cut three new patterns today on the laser cutter using Ed Rayher's "high dot" for positioning on the mat and then on the fitting machine. Photos show all three patterns cut from 1.5mm acrylic, with the 3mm blank bases below, onto which the pattern will be laminated (they have 3-M sticky film on the back). Then a couple of progressive close-ups, showing the high dot, and the edges of the pattern.

These are all Primula ornaments, cut to engrave 14pt mats at 15:1.


I'm going to start another topic specifically on finishing cutting tools, but after doing some more diagnostic maintenance on my Really Awful Type problem, the caster was spitting out nice, clean, solid, and shiny type yesterday, although I only cast a couple of 6" lines.

The goal was to see if I could get a full 6" line cast from my very first matrix, and, at last, some good type came out.

That's not a great photo, but if you look closely at the mat you'll see where I, once again, stabbed the mat with the finishing cutter: it's between 7 and 8 o'clock on the mat below the engraved image. If you also look closely at the type you'll see a tiny cone sticking up at the same position, right on the edge of the type. (Kevin, you'll also notice the flash I've mentioned on both sides of the face, but in this case it's because the face of the mat was still pretty rough and needed a rub on some emery cloth.)

I still haven't had a chance to print these sorts (that's tomorrow), but I did cut another mat today after coming up with a better (but still very crude) way to finish my cutters (see the new thread on that topic). Tomorrow I'll trial cast the new mat, which I think will have a better face, and will pull some proofs of type from both the first and second mats.

By the way, these are Primula ornaments, which I love. My plan is to cut about a dozen of them in a couple of sizes, but let's not get ahead of ourselves.



Another bit of the equation that I've put into the sequence is marking the matrix blanks for positioning the "high dot" prior to engraving. I'm using a side-bearing of 8pts, and a head bearing of 24pts. This is relatively arbitrary for the moment, but it puts the image in a safe place on the mat that works well in my mat holder.

I'm lucky to have a good set of point-size standards, shown below:

I'm using these to score a light cut into my mat blanks for both the side- and head-bearings. I then use the intersection of these lines to position the high-point engraving, just inside that intersection (cut with a fine cutter to a shallow depth). Following Ed Rayher's lead, I'll then use that dot, positioned in the cross-hair of my fitting machine's microscope, to trim the side and head of the mat so that the middle of the high dot is exactly 8pts from the left edge and 24pts from the top edge of the mat, theoretically (!) placing the image in perfect alignment with other mats produced using the same specs/process.

Val Lucas

Question about milling machine/fitting machine- you and Ed seem to have the magical machines that will finish the brass mats to exactsize and position easily. I started looking at milling machines to see if there's one that could perform the same functions- I have been cutting and filing and finishing the flat mats to size by hand, and I am terrible at it. However, my eyes glazed over with the possibilities for milling machines (small ones, of course...) do you have any insight into what I could look for to get a milling machine that could do what the fitting machine does?

Also, thanks for posting the pattern process- great way to avoid having the slope of polymer plates, which has been Jim W's method but gave me some trouble with fine lines.


Your sample castings look a bit rounded at the corner. Do you know if that is a casting problem, or the matrix not having quite sharp corners at the edge of the face (or perhaps an artifact of the photo of the type)?

Have you looked at the tip of your cutter in the microscope after cutting the mat to see how it has worn?

I noticed in your other thread of making the cutters that you seem to be using about a 45 degree angle to cut the tip. This seems to me to be a rather steep angle and might contribute to rapid wear of the cutting edge. This angle is the clearance angle (which avoids the tip of the cutter behind the edge from rubbing on the work) and for most cutters (lathe tools, endmills, twist drills, etc.) it is usually more in the 5-10 degree range. Also if you are not grinding the tip square to one of the sides of the cutter the steeper angle will produce more swirl marks.


Val, Ed and I (and Patrick Goosens) are very lucky to have these ATF fitting machines, which are amazingly accurate and very sturdy machines which are remarkably easy to use. I don't know anything about commercial milling machines, but I would think there has to be a good substitute out there. One of the reasons I'm making foundry-style mats is to avoid the difficulty of milling flat mats. Jim Rimmer used a Ludlow SuperSurfacer to mill/true-up the front and back of his flat mats, and this could also be done on the fitting machine, but I just don't like running anything past a spinning blade assembly when it's being held on an edge that's only 1/8" thick. With foundry-style mats there's much more depth, and so milling the four sides is far more stable and safe. Yes, it's a lot of extra (expensive) metal, but I like the stability of the process. I'm also drawn to the history of foundry-style mats.

Kevin, thanks for your feedback. I'll have a close look at the finishing cutter I used yesterday to see how it's worn. My plan today is to grind three new cutters to a more accurate finish, so that I can really focus in on the tip/follower ratio and reduce the weight of the next mat, and with the new cutters I'll greatly reduce the angle of my finishing grinds.

I'm curious, though, about your comment on grinding to one side of the tool. Ed uses what I assume is the ATF method, which is finishing a screwdriver tip (along the middle of the tip), which makes a lot of sense to me, as the cutting edge turns on a centre axis. Grinding to one side would spin the cutting edge in circles length-wise, essentially cutting circles, and this seems far more likely to produce swirl marks. You're really good with the technical side of these things, could you expand on this a bit for me?

In my graphic below it shows how I'm imagining the cutting operation. On the left, with a screwdriver tip, the cutting edge is spinning like a helicopter blade, which should create a smoother floor to the mat. On the right, the cutting edge turns in circles, and this seems far more likely to create swirl patterns.


Jason, the central chisel tip should be able to give a good finish, as long as it is perfectly level. If one end of the chisel tip cuts a bit deeper than the other end it will produce swirl marks that would be pretty much unavoidable. (the chisel tip does, however, explain the steep angle you use to shape the cutter tip)

You are correct that the geometry I suggested will cut a doughnut shape, but as you move the cutter around all the doughnuts will overlap leaving a flat surface, provided again that the cutting lip is level.

In either case the level lip will cut a small flat-bottom patch, and moving the cutter will overlap these to create a smooth overall surface. If the lip is not level it will cut a small cone-shaped patch (either point up or point down depending on where the lip is highest) and it will be very difficult to get an overall flat surface because you effectively have a sharp-point cutter.

The geometry I suggested provides much more back support and less negative rake for the cutting edge, but it is not center-cutting so it cannot be used for a straight plunge cut. It is also more difficult to produce because you need a second cut to ensure there is end clearance just behind the cutting edge at the midpoint of the main pyramid face; here the cutting edge would be moving almost parallel to itself making a shear cut.

I suspect that making a macroscopic model of the shape and seeing how it "cuts" as you turn it in your hand and move it around might be very instructive. So far the best materials I can think of to do this are all vegetables: potatoes, carrots, rutabagas... They are all easy to carve and unlike modeling clay, they preserve their cut shape.


Thanks for all of this Kevin, I always appreciate your very technical discussion on these things, even if I fumble to grasp everything you're saying.

Given my crude methods to finish the tips, I'm not sure there's much I can do but make due with what I'm producing. I've refined the process a bit, switching to a very fine Arkansas stone held on my glider table, and the jig I made to hold the cutting tool *should* keep it level to produce a level edge to the tip, but as this is all hand-work, there's no real precision.

I also discovered another problem I wasn't aware of, which is that the rod holder of my grinder was out of true, so the rod wasn't sharpening to a central point (the point was slightly off axis). Thus, when spinning in the pantograph, it was cutting a circle, rather than a point (if you follow me). This would certainly explain why the mats I've cut so far have all come out "bolder" than I want.

I discovered that the rod-hodler assembly on my grinder has a collar around it, with three set screws. Putting a gauge on the rod and fussing with the three set screws, I managed to get the rod to turn with zero movement off-axis. I also checked my panto to discover that there's a bit of fuss to my collet, so I have to be quite careful when changing my cutters to be sure they're in straight (like putting a small bit into a drill). I'm now putting my gauge on the tool once installed and turning it to be sure it's in straight with no off-axis "wobble" as it turns.

I also verrrrrrry cautiously made a few new tools, making sure the rods were true in the grinder, and then hardly dragging at all on the stone (tiny drag on one side, and half of that on the otherside to get the screwdriver tip). I managed to get nice clean .002 and .003 cutters, which I'll use for the finishing cuts of my next trials.

Quick Reply

Please leave this box empty:
Type the letters shown in the picture
Listen to the letters / Request another image

Type the letters shown in the picture:

Shortcuts: ALT+S post or ALT+P preview

Printers' Tales - Over 30 stories from the pre-digital age. Buy now on Amazon/Apple Books

☛ Don't miss our illustrated newsletters. Click here to see examples and subscribe. ☚