letterpress > monotype adventures

monotype adventures

Created by hww3. Last updated by hww3, 12 years ago. Version #35.

October 4, 2007

Over the weekend, I made a special trip up to the shop in order to get the last remaining piece of the casting automation puzzle set up: the caster cycle input. I'm using a magnetic reed switch, and am using the action of the air bar clamp activation rod (Ian, you'll have to provide the official nomenclature) to detect a casting cycle. The attachment ended up being pretty elegant. I just need to run some tests to ensure the switch is spaced properly from the activating magnet, otherwise the cycle timing will be off. Hopefully it won't matter too much when the caster is at speed.

The air bar adapter turned out to work very well without having to make any alterations to the receiving caster, and it takes about 10 minutes to install the whole system (assuming you're not disconnecting all of the air hoses), with the only tools required being a screwdriver (and maybe a pair of pliers if your grip isn't very good). I personally don't anticipate swapping very often, as the whole motivation for me is my inadequate keyboard and accessories.

I just put some finishing touches on the caster control program this evening, and everything looks set (using the simulated interface, but the real thing should work as well.) That means I should be casting from the computer this coming weekend. I can hardly wait, and I might take a vacation day Friday to work on it.

All in all, I'm pretty pleased with the way things turned out; it's taken a bit longer than anticipated, but I think the wait was worth it. The only things left are some cosmetic bundling of the "air spaghetti" and to mount the controller and air manifold somewhere out of the way (maybe hang them from the ceiling.)

I've also been making good progress on the "e-ribbon" generator. I have a simple keyboard emulator, and I think the initial focus will be a simple application that reads a text file with caster control codes embedded in it (ala HTML) to control the ribbon generator. I really don't see the benefit in going down the WYSIWYG path, though I might create a visualization tool that allows one to see the alignment of the lines in a job. That would make it possible to detect rivers and other undesirable artifacts without having to cast and proof a job.

I've got a few photos related to all of this activity… first, a screenshot of the casting application after an "e-ribbon is loaded, and one during the simulated casting session. Additionally, a snapshot of the airbar adapter (apologies for the less than artistic photo, my father was apparently not paying close attention to what he was doing).

August 5, 2007

We spent the afternoon hooking up the new Super Caster. The motor controller box somehow got misplaced during the equipment move, leaving only a pair of control buttons on the front of the machine, so we had to make alternate arrangements. Luckily, my uncle is an industrial electrician, and he volunteered to find a suitable substitute. He came through with flying colors, and even installed everything. We also wired the pot and verified that the heating elements are working. That just leaves the water hookup and compressed air for colling the mould and matrix, respectively. Exciting times!

July 17, 2007

The caster control board prototypes have both been verified, and everything seems to work as expected. There are a few items I'd like to change, so I might do a second revision. Things such as flipping the connectors over so I can use right angle jacks; making the power connection a jack instead of a terminal block (that way I can use power supplies akin to laptop power bricks) and adding a circuit to keep the caster-side electronics powered down until a usb connection is made. Oh, and there were one or two minor changes I made to the PCB on the fly. It makes for an ugly reverse side, but it works.

I also verified that all of the valves work, but some of them seem to have restricted airflow, so I'm going to swap them out. Then it's just a matter of a day or two at the machine to get everything hooked up.

As if I didn't have enough to do, I thought about making a custom control panel with buttons that correspond directly to buttons in the GUI. It's probably not that hard to do, mostly a matter of ripping an old keyboard apart and fastening the right switches to a panel. Well, maybe that's a project for once everything's up and running, eh?

For the curious, I've added a photo of the finished board. As a side note, this board would also work well for anyone needing 31 channels of digital output… lots of physical control ideas come to mind...

June 5-12, 2007

Started adding some CasterControl notes. The PCB design has been validated and I'm able to control the valves via computer. I'm just waiting on a resistor and a few more quick disconnects and the electro-mechanical interface will be complete. Wiring up the cables and valves has proven to be a time consuming and tedious chore; each wire needs to be identified, stripped, crimped and inserted into the appropriate connector frame. I figure the cables and valves will probably take about 2-3 hours each (there are 3 for each valve bank).

I also finished a first pass at the low level IO glue, which will allow the valve bank to be controlled by a higher level computer language (pike). Initially, the interface will monitor for caster readiness by polling; eventually, I'd like to have the interface report asynchronously when a code change is required. It's probably not a necessary enhancement, but it will make things a little bit more elegant.

As soon as I get some time during the day, I'll take some photos of everything. For now, you can look at the pre-production board layout below.

May 30, 2007

A few items of note this past week:

The first pair of prototype circuit boards arrived from the fabricator. It looks like some time will be spent this weekend stuffing the parts onto the board. Hopefully, it won't be a complete failure.

Last weekend, my father and I headed out to the Brooklyn Navy Yard to pick up a pair of Monotype Supercasters from the Woodside press. They're now safely resting in the shop, and I'll try to get them sited this weekend, as well.

May 2, 2007

Over the past year or so, I've been working on a computer interface for use with the typesetting system I've been working on. Finding an off the shelf interface that meets my criteria has been somewhat difficult. There have been a few "close, but no cigar"s.

In the end, I decided to design my own. With 31/1 isolated out/inputs and a USB interface, this is a good bit more complicated than anything I've done before. The most recent spurt of activity has taken up all of my evenings for the past few weeks, and I'm pleased to report that it's almost done. The hardest part has been laying out the circuit board, but I'm 95% there, and I thought it might be fun to show off what I have so far.

Please be gentle, I'm not a professional layout technician, so I relied heavily on auto-routing, with a fair amount of cleanup afterward. I have a fairly high level of confidence that this will work, assuming that I haven't made any mistakes since moving things from the breadboard :O Once I put the finishing touches on it, I'll send it out to have a prototype made and then we'll see how close to functional the design is.

• see the board
• see a screenshot

Once it's done, I'll post some more details as well as a more in depth discussion of how the system is intended to work.

December 1, 2006

The past week or so at the shop has been busy and productive. My father and I were hard at work running new electrical circuits, including installing the new rotary phase converter which will be used to power the 3-phase motors on the English composition casters.

Sunday we had a visit from Stuart Bradley at the Railway Station Press, a fellow student from Monotype University 6. We caught up a bit and Stuart received the full tour before continuing on to Virginia.

June 11, 2006

My father and I moved the two English composition casters into the new shop area. Still remaining is the American caster, which is in a disconnected area of the building. We have to figure out how to "drag" it around to the front so we can move it in.

May 28, 2006

This past weekend, I accompanied Ian Schaefer to Charlotte, NC to pick up an English Composition Caster from Heritage Printers. While I was there, we loaded up an American Thompson type caster, some spare parts and mats for me as well. I've updated the Matrix Inventory and it's reasonably complete. There are a number of oddball matcases that came along with the equipment from Richmond, so I'll be trying to get that sorted out over the course of time as well.

January 25, 2006

I've posted an update on our attempts to get the American caster fired up.

November 26, 2005

As promised, I've started working on a page describing how to move a monotype.

November 21, 2005

A little spray of magic rust penetrator did wonders to loosen the new caster up. It looks like it will be a good machine; it appears to be rebuilt, because it has an "R" ahead of its serial number. It's a 15x17 caster, and the manufacturer's plate also says "division of Lanston Industries", which would indicate it was overhauled relatively late in the game (probably after 1954).

It has the display attachment, but not the low speed gearbox, so I'm somewhat limited in the kinds of display type I can cast. Because of this, I'm still keeping my eye out for a full fledged display caster of some sort.

I also took an inventory of all the wedges, stopbards, keybars and moulds. Next weekend I'll probably dig into the mats to see what I ended up with, as most of the matcases weren't marked in any detectable way.

It's looking like this caster will be the first to be set back up, as it's got a single phase motor and a gas pot fitted for propane. We'll take the gas tank from the gas grill and run an extension cord to the motor (it's already fitted with a plug, how nice!). My father even joked about wheeling the whole thing outside on nice spring days.

November 14, 2005

Well, it's the day after the big move (it was actually two days, but we got so little sleep, it all sort of ran together). Everything made the trip safely, except for some spilled composition mats. I'm working on a blow-by-blow for the next ATF newsletter; I'll post more details of how we actually did the move once I've caught up on sleep. In the meantime, some photosfor the unloading are available.

November 8, 2005

This weekend we're moving some more Monotype equipment. This time, it's a US 15x15 Composition caster with the Display attachment and a Keyboard (something I've desperately needed). The equipment comes to us by way of Dave Klinger in Richmond, VA. With a little luck, by Sunday it should be resting safely in the Barn.

I should have pictures up after I get home from the move.

July 13, 2005

I finally had a chance to contact Duncan Avery about the origins of my casters. The next morning, the answers were waiting in my inbox. Duncan wrote:

Caster 22441 was first supplied to A.Thom, Dublin 13 July 1948. It was then return to Monotype for allowance in 1968 and went into the Printing Dept at Salfords. Certainly I was much involved in the mid 1980s when Harold Berliner bought a great deal of matrices at that time when the Printing Dept was closed and I guess he purchased the Caster then. The machine was originally supplied fitted with Display and leading attachments which presumably are still on.

Caster 28276 first went to Advertizing Agency Services Co, New York in November 1964. They sold it to Daguerreian Era of Pawlet, Vermont in 1975 but I have no further information on it.

Well, that's certainly exciting news; I'm anxious to find out if all of the display attachment is still present. I had noticed that the set width scale was present on the earlier caster. I'll have to dig out my parts book to see what else is involved to see what I've got, and what's missing. It would certainly be easier than finding a separate sorts caster.

June 25, 2005

I started working on the software to drive the composition caster. Initally, it will consist of a "keyboard substitute" to generate an input file for the caster driver. The plan will be to get the simple keyboard application finished in the next two weeks. I'm less concerned about my hardware configuration, as Rich Hopkins is employing the same type of Clippard air valves on his "Mac Mono" system. I guess I got lucky!

June 15, 2005

I've recovered (mostly) from my experience at Monotype University 6 and have started writing down my account at monotype university 6

April 11, 2005

After a few go-arounds with Yellow Transportation, my casters were delivered this morning. Initial reports are that there was no obvious external damage to the crating. I'll have more time to look at them later this week. As a side note, I'm once again reminded that any given experience with a huge company like Yellow is entirely dependent on the competence of the person you manage to get on the phone. Luckily, I got someone good on the line at the last minute to save the day. Your mileage may vary. :)

April 2, 2005

My two monotype casters have started their journey east. They were loaded onto a Yellow Transportation truck in Grass Valley, California yesterday. They should arrive in about a week. Then we get to begin the fun of making room and moving them into their new home.

March 26, 2005

I received the IO Warrior starer kit from Code Mercenaries this past week, along with some other assorted parts that I'll be using to construct the air valve interface. Last night I spent some time soldering the IO Warrior components together. It's been a few years since I've done any sort of electronics soldering, so it took me a few connections before things were moving smoothly. After I finished putting the thing together, I plugged it in to my Powerbook, and (much to my amazement,) it was immediately detected. I started the included Prober application and within a few moments, I was controlling LEDs and reading input at will. So far, so good.

The air valves that I'm going to try are Clippard Minimatic 3 way air valves. They're relatively inexpensive (less than $25 each new) and come up often enough on eBay that I was able to get 35 of them with manifolds for under $70 after a week or two of scouting things out. They are quite small and draw less than a watt, meaning I don't need to employ any expensive relays or driver circuitry.

The IO Warrior's IO lines are TTL, which means that they're not capable of driving a load on their own. That means we have to come up with a way to turn on power to the valves based on the logic level (on or off) of the IO pins. Luckily, we don't have to supply much current to the valves (around 50 milliamps at 12 volts), so there's a readily available IC that we can use just for this purpose. We'll be using the 2803A IC, which is available from multiple sources. The 2803A is a 8 element Darlington Transistor Array that has built in clamping diodes for use with inductive loads such as solenoids. We can connect the 2803A to our 12V power supply, and connect the IO to the base of one of the Transistors. The power leads to each valve gets connected to the emitter and the common power connection. With any luck, when we apply a voltage to the base, in the form of a "1" on the output pin of the IO Warrior, current should flow to the solenoid and the valve should turn on. Not wanting to blow the multiple components up at a time, I tested each component separately, then once the theory seemed sound, I hooked everything up. The moment of truth had arrived. I connected the IO Warrior back up to the Powerbook, turned the 12V power supply on, and started up the Prober program. I sent an output "1" to the pin I had wired up and miraculously, the valve went "click"! A little bit of testing to make sure the valve could be controlled repeatedly leads me to believe this approach will work just fine. My only concern now is whether the Minimatic valves have a high enough flow to be usable. I guess we'll only know by trying, eh?

Now that the mechanical questions have been ironed out, I can start working on the interface program to convert input text to punch codes, and then from punch codes to valve commands. I'm writing a binding between the IOWarrior library and the Pike programming language. This will allow me to create a cross platform (Mac, Linux, Windows) interface and application. Using a high level language also allows me to worry about the design problems at hand, rather than waste time dealing with memory and i/o.

Once the casters arrive and are up and running, I can test to see if the air flow is sufficient, and then work on a more permanent set up (proper circuit boards and enclosures and such). Stay tuned!

March 15, 2005

While I'm waiting for my Composition Casters to be crated and shipped from California, I've busied myself with all of the preparation work required to get them casting type. I'll need to install a new power service, as well as plumbing for water and an air compressor to supply compressed air to drive the caster.

More importantly, and perhaps more problematic, is the lack of a keyboard to punch the paper tape that drives the casters. The 16x17 keyboards needed to take full advantage of the casters I've got are more scarce than the more common 15x15 or 15x17 keyboards commonly used in the US. Additionally, I'd also need keybanks, keybars and stopbars for each matrix case arrangement (MCA) I'd want to use, not to mention a source for the paper tape. All in all, it could be a significant pain and drain on my time and money to go this route. I figured that there has to be a better way. After a little thought, determined that there were basically 2 paths: come up with my own way to punch tape, or figure out how to drive the caster without punched tape.

The effort required to fabricate an alternate tape punch seemed to be a fool's errand, and I'd still need tape. plan is to initially bypass the keyboard situation altogether by hooking the comp caster up to a computer. Now, I'm not the first person to have come up with this plan, In fact, Monotype had computerized tape punching systems (part of the reason there weren't any workable keyboards available). The problem is that all of the solutions were relatively hardware specific, and rather fussy to get and keep running. I figured there had to be a reasonably low cost way to do this.

The caster is driven by paper tape which acts as a 31 position valve bank, causing air to flow through pipes that push up pins on the caster when there are holes in the tape. The typical way to replace paper is to use a bank of solenoid valves to drive the positioning pins. The way that the various techniques diverge is the way in which these valves are connected to the caster, as well as the mechanism for driving the valves. The solution I am going to explore involves the use of a device called an "IOWarrior", which is made by a German company. The IOWarrior in it's simplest form is a 32 line 125 Hz digital I/O module with a USB interface. Its advantage is that it supports multiple computing platforms (Mac, Windows, Linux) and that it's relatively inexpensive: $75 for a development board. Most other comparable solutions are several times more expensive, and don't support all 3 platforms that the IOWarrior does. I'm planning on coupling the IOWarrior to the valve bank using Darlington transistor arrays. I've got a set of Clippard Minimatic solenoid valves that I hope will do the trick (I'm not sure if they have a high enough flow rate yet). I can get everything set up and tested before the casters get here, so I'm not wasting time with it later.

Attachments

This page (letterpress/monotype adventures) contains 10 Attachments

Name	Type	Size	Updated	Updated By
caster_1_sm.jpg	image/jpeg	14 kb	Wednesday, 2 November 2005	Bill Welliver
caster_2_sm.jpg	image/jpeg	13 kb	Wednesday, 2 November 2005	Bill Welliver
monotypelogo.gif	image/gif	11 kb	Thursday, 3 November 2005	Bill Welliver
caster_3_sm.jpg	image/jpeg	2 kb	Sunday, 12 February 2006	Bill Welliver
cboardpop.jpg	image/jpeg	145 kb	Wednesday, 18 July 2007	Bill Welliver
casterboard.png	image/png	51 kb	Wednesday, 2 May 2007	Bill Welliver
casterss.png	image/png	30 kb	Wednesday, 9 May 2007	Bill Welliver
airbar.jpg	image/jpeg	114 kb	Monday, 8 October 2007	Bill Welliver
screenshot1.png	image/png	33 kb	Monday, 8 October 2007	Bill Welliver
screenshot2.png	image/png	32 kb	Monday, 8 October 2007	Bill Welliver

Posted in letterpress | RSS Feed | BackLinks