Saturday, August 25, 2018

Work on model 14, 15 and 19 teletype parts

MODEL 14 TD RESTORATION

Marc is in contact with a fellow YouTuber who does CNC work on gears to see if he can make a replacement for the motor pinion gear I would need to hit the standard 60 wpm transmission rate (about 6 1/8 rotation per second). This is achieved by taking the 1800 rpm synchronous motor, putting on a 9 tooth pinion gear and coupling that to a 44 tooth gear on the main shaft. 

I am also pursuing a possible source of another 14 TD which might have the right gears in place, but it will be a week or so before I know if this will pan out. 

We worked to get the wiring of Marc's TD correctly installed. Once that was completed and verified, he continued to install the labels, covers and other parts to finish this up. Next week he will run a test to transmit the contents of some test tapes to one of the model 15 printer units.

MODEL 15 PRINTER RESTORATION

My model 15 teletype has a bent typebar for the letter Z which results in the letter printing faintly in the superscript area above the line of the other typed characters. This is due to a deformation of the shape of that typebar, which sits on the outside edge of all the typebars on the carriage. 


Z typebar prints too high compared to other letters
I can see from the rest position of the typebar that it is bent out of position. The typebar shape is complex, with several bends the ensure the typeface strikes right in the middle of the ribbon opening, flat and at the proper height. To fix the misalignment, I need to adjust several of the bends. I am studying and making a detailed plan for the bending necessary.


Leftmost (Z) typebar out of position
MODEL 19 TELETYPE DESK RESTORATION

Ken and I began to install all the wiring, receptacles and switches into the desk today. Fortunately, the Teletype company drew up a wiring diagram, which is distinct from the more usual schematic. The wiring diagram shows each part in place with the actual wiring that is connected. This includes the jumpers that connect different terminals on the terminal blocks.

It goes so far as to identify the wiring bundles that are laced together with string - in one long group of wires there are several bundles. Each bundle runs to one or two receptacles and I can distinguish them by the number of wires hooked together. A circuit continuity tester gave me resolution to the individual wire in each bundle.

By the end of the day, I had two fuses and five receptacles fully wired, additionally I had four phone jacks wired but not fully verified. Ken had three switches on the front of the desk wired and the power lines installed through the flexible armored cable and


Finished wiring receptacles and fuses while the signal jacks at right are nearing completion
Power switches and one armored conduit installed, more to do on right
The wires from all these receptacles, fuses and switches run to five terminal blocks that connect together the 110VAC, 120VDC, send, receive and test lines. There are quite a few jumpers that run from one part of the terminal blocks to another part. Many of these are loose and have to be replaced as part of the wiring. I still have to install the metal stands that the five terminal blocks attach to, then complete the wiring.


Terminal blocks being wired and tested, but not yet mounted in place
MODEL 14 TYPING REPERFORATOR

Mike Albaugh brought over his typing reperforator. This is a unit that will take the incoming characters from the signal lines and punch them on 11/16" paper tape, using a chadless method that leaves the hole with a hanging chad.

In today's terms, because of elections, we think of chads as things that hang attached to the main paper, but the original terminology calls the completely removed circle of paper a chad and that chad falls into a waste chute. If the hole remains partially attached, it is called chadless by Teletype.

The reason that the hole is only perforated partially is to leave the paper in place since characters will be typed on the tape. If there were holes, the characters would be obscured by the holes. This way, you can read the typing on the paper tape but the pins in the Transmitter/Distributor will push the flap away and accurately detect where a partial hole was punched.

This unit has a selector unit and typebars, much like the model 15 printers we restored, but it is quite a bit less complex in many ways. No function box since actions like tab, carrier return or line feed don't make any sense when performed on paper tape. Instead a special character is typed that represents each of those functions. Thus, this machine only needs type bars to print all the letters, figures and control function characters.

When Mike got it wired up, he found that the selector mechanism was too gunked up to perform properly when we tested it. The motor and solenoids were moved and it was dunked in Simple Green to remove old lube. At the end of the day he began to remove, reassemble and lubricate parts of the typing and punching mechanism. It will take at least another session to get this working properly. 

Tuesday, August 14, 2018

Successful trip to RTTY Electronics for teletype parts

TELETYPE RESTORATION PARTS SAFARI

On Monday five of us journeyed to visit Paul Cemburas, Mr. RTTY, of RTTY Electronics, who maintains an enormous stock of teletype parts, machines and related goods. We had previously sent him a list of parts that Marc wanted, a list that Bob wanted and a list that I wanted.

My parts included missing items from my model 15, a spare part I wanted for my model ASR 33, and new gears for my model 14 TD to convert it to 60 wpm operation. Marc had a similar list including thumbscrews, the glass panel for the cover and the hand crank to advance the paper manually on the model 15. Bob Rosenbloom has a longer list because of the large number of machines he has to restore.

In spite of an enormous collection of parts, each of us had requests which could not be fulfilled. For example, both Marc and I were interested in buying the crank that is used to manually space paper up out of the machine, but they are very rare. We saw one in person and admired it, but none was for sale.

Converting my model 14 TD requires a matched pair of gears, one on the main shaft and a smaller pinion gear on the motor shaft. Paul had a partially machined gear for the main shaft but no pinion gear. Once I drill mounting holes in the gear I could install it, but not until I find a pinion to match.

Marc did have a pleasant surprise, finding that he could add the motor stop optional feature to his model 19 system. The model 15 he owns has this feature, allowing a user to start up the motors of the teletype, send a message and then power down the motor by sending the Figures-H character.

It consists of many dozen brackets, levers, screws, spring and other parts, installed on both the baseplate and the printer modules. Marc decided to check availability of the parts, using the parts catalog to generate a long list of part numbers. Amazingly, Paul had every single one of the parts in stock.

I was unable to pick up the stock line feed upstop for my ASR 33, nor the model 15 parts I was seeking. None of them were critical - two rubber guides that position the cover over the baseplate, a new drawstrap for carrier return, and the unobtainium hand crank.

Bob did pick up quite a few parts that he requested, although some parts such as the cover plates for the front of the model 33 teletypes were unavailable. Overall, we were all satisfied with our trip. Plus we saw many rare and unusual teletype machines.



Saturday, August 11, 2018

Progress on model 14 Transmitter Distributor restoration and a motor investigation

WORK ON MODEL 14 TRANSMITTER/DISTRIBUTORS

I disassembled my model 14 and oiled everything on the main shaft in preparation for trying out a potential new gear set. The gear I received from RTTY Electronics needed some holes drilled but was otherwise cut with the necessary number of teeth to gear the synchronous motor down to the 60 wpm rate we wanted.

Marc had some gear sets he received with his teletype, but none of the motor pinions fit my gear properly. It meshes perfectly with the 80166 gear on his TD and that is the part I still need to make this unit work at the intended rate, rather than the faster 75 wpm rate it was geared to produce. 

The clutch on my TD is a later type, a drum clutch, unlike the friction clutches on most model 14 TDs including Marc's unit. This has shoes that expand under spring pressure to rub against the inside of the drum when the clutch is engaged, then they retract when the clutch is disengaged after one rotation of the shaft. 

After lubricating all the parts I tried to reassemble it. There is some trick I have to learn to compress the shoes in order to fit them up inside the drum; I wasn't successful on Friday in figuring out the needed technique. 

I partially reassembled my TD and set it aside for when I have a new pinion gear, at which time I can also complete reassembly of the drum clutch. If I don't find or make a suitable pinion gear for 60 wpm, then I will need to reassemble the old 75 wpm gear set and make use of an alternate method to slow down the TD. 

I have a number of methods that could be used to make my model 14 TD produce teletype serial streams at the 60 wpm rate:

  • buy or make the pinion to mate with my new main shaft gear to operation at 60 wpm
  • use a variable frequency drive to slow the frequency of the mains down and slow the motor
  • build a box that will accept and buffer data at 75 wpm and drive a 60 ma current loop at 60 wpm

Marc worked on his model 14 TD. It had been disassembled to send the base and covers out to be powder coated. With the coated parts on hand, he began assembly of the unit. Most parts went together easily and the wiring was obvious. The wires to the main contact plate, however, don't match the schematics or other TD units, which was where he stopped at the end of the day. 

Ed had bought a synchronous teletype motor to disassemble, as he was curious to see its construction including the motor start winding and its centrifugal switch. He pulled it apart although it was not obvious that the bearing on one end would pull the centrifugal switch mounting bracket off with it. This stripped out the two small screws that hold the switch to the end plate.

Ed picked up a replacement plate and screws from RTTY Electronics. While there, he also picked up a new set of brushes for my TD. The brushes slide on the brass rings on a disc, making contact to route the various pinhole microswitch contacts from the paper tape reader to the serial output line. 

The brush that rides on the segmented ring, the one that breaks a rotation into a start bit, five data bits, and 1.42 stop bits, was worn down quite a bit by the action of the slots between segments. The other ring is continuous and its brush was much less worn. 

My Model 14 paper tape reader is the wrong speed, but we make continuing restoration process on the Teletypes

TELETYPE MODEL 14 TD RESTORATION

I was given a transmitter distributor, which is a paper tape reader for the 11/16" 5 bit baudot tape used with older teletypes. At our work session with teletypes we worked on two of these units, one that came with Marc's model 19 system and one that I was given. As well, Marc made some adjustments to his model 15 teletype to correct two flaws we had discovered.

The model 15 has a feature that will stop the motor based on a character sent to it (motor stop is the H key in Figures mode), then restart the motor if new data begins arriving. One function of this feature is to flip the switch from SEND mode to RECEIVE ONLY mode during a stop. 

We had found that it was bouncing the switch nearly out of SEND mode on every typed character, and from time to time it would move it all the way to RECEIVE. This was frustrating to users trying to type messages on the keyboard as it would suddenly stop sending their keystrokes. 

Marc did some adjusting and it now works perfectly. The lever doesn't bounce at all except when seeing an H character (H key in Letters mode), where it moves just a bit but never forces the lever to RECEIVE ONLY. When a stop is sent (H key in Figures mode) it will flip the lever to RECEIVE and stop the teletype motor. Worked correctly every time. 

The other flaw we had discovered was with the autorepeat feature for the spacebar on his keyboard. This is designed so that a normal tap on the space bar produces exactly one space character, but if you press down hard on the bar and hold it, a string of spaces are emitted until you release the bar.

On this keyboard, it would autorepeat on any tap, so that single spaces were usually turned into 2 or 3 in spite of how briefly the typist hit the bar. Marc adjusted the mechanism until it was performing perfectly, providing single spaces unless you intentionally pushed down the bar and held it. 

Marc's TD (Transmitter Distributor) is disassembled, since he had the cover and baseplate powdercoated along with the rest of the model 19 external metal surfaces. He spent the time cleaning up dirt and corrosion from his parts, adjusting it properly and cleaning all the contacts to remove old oxidation. Someday soon he will begin reassembly of the TD. 

My TD needed cleaning and relubrication before I could test it. I spent time freeing up mechanisms, scrubbing off dirt and cleaning most of the corrosion on metal surfaces. I then cleaned the contacts and inspected the unit carefully. It appeared ready for a test, but we had to provide connectivity first.

The model 14 TD has a set of metal contact fingers on the back which are intended to slide into a baseplate such as the desk from Marc's model 19 system. We worked out the required wiring to provide 110V AC to the TD and hooked up the current loop for the output signal. 

I crimped ring terminals on some wires including a phone jack cable for the data output and a household plug cable for delivering power. These were hooked to the appropriate fingers of the contact mechanism. 110V is provided to two fingers to power the AC motor and to two other fingers to power solenoids that control function. The phone jack hooks to two fingers for the line common and MARK output side.

The TD is designed for both the normal NEUTRAL signaling method and POLAR signaling. NEUTRAL signaling delivers a 60ma 120V current for MARK bit values and no current to indicate SPACE. POLAR signaling delivers current in two different directions, one for MARK and the other for NEUTRAL. We wire our TDs for the NEUTRAL method we are using, but could use them to drive a POLAR line with a bit of rewiring. 

I fired up the TD and verified that the motor spins properly, a tape advances through the reader properly until it either runs out (sensed by one microswitch) or the tape becomes too taut and pulls up a safety switch bar. The result on the signal output line seemed correct as I pushed or released each of the five pins that sense/read the paper tape holes.

It was time to hook this in a current loop and see if the tape we read will be properly typed on one of the model 15 printers. Marc had produced a few test tapes with his model 19 keyboard, which has a paper tape punch called a perforator that transfers keystrokes into tape characters. I put in the first tape, holding a one line message, and transmitted it.

The results on the printer were gibberish. Having a suspicion that this might be a speed mismatch, we hooked up an oscilloscope to the line and observed the duration of the bits being produced. Our normal 60 word per minute units have a bit cell time of 22 ms and the stop bit, being 1.5 bits long, is 33ms. The scope showed a stop bit that was only 26 ms long.

We produced another tape, this one with a string of R characters, which would allow us to easily see and measure data bits. That confirmed that the bit cell was a bit more than 17ms, reflecting a transmission rate of 75 words per minute. The steady string of R characters produced gibberish on the printer, but a repeating pattern that further proved this was a speed mismatch problem.

I will need to replace the motor pinion and main shaft gears with a set designed for 60 wpm operation. We plan to visit RTTY Electronics to look around and buy spare parts on Monday, so that would be a good time to pick up the new gears I need. 

There is the complication that model 14 TDs had three different sets of gears for 60 wpm from a 60Hz AC synchronous motor. The motor pinion is always the same, but the main shift gear is one of three different part numbers. I am not sure which I should be buying, but hopefully I can spot the correct one when I look at them on Monday. 

We put the huge thyratron based model 30 power supply into service powering our teletype current loops as well as the perforator solenoids on the model 19 keyboard. Ken and Ed took lots of measurements of parts and watched the operation to understand this beast more thoroughly. 

Monday, August 6, 2018

VCF West exhibition this past weekend was a success

VINTAGE COMPUTER FESTIVAL EXHIBITION AFTERMATH

The exhibit at VCF West became one ASR 33 working and a broken one open on the table for inspection, two model 15 wired together to allow attendees to chat, and my model 15 printing the Reuters news feed. Near the end, we had a partially working model 28 join the other machines, as one of the team at our booth bought it from the show consignment room.

One of two model 15 teletypes linked together for chatting
My temporary fix for the line feed on the ASR 33 was erratic, but I had received a replacement bracket from a memory of the teletype oriented Greenkeys mailing list. I swapped that part into the machine, made a few adjustments and the line feed problem was fixed for the remainder of the show.

My print cylinder on the ASR 33 became more and more clogged with ink over the course of the show. I will need to remove it, soak it and get it cleaned up. This is the first of the after-show tasks on my list.

ASR 33 working with Altair 8800 clone
I noticed from reviewing the news stories on my model 15 that the letter Z was only partially printing and raised like a superscript above the print line. I looked closely and could see that the typebar for that letter was bent out of position a bit. Fixing this is the second after-show task on my list.

Model 15 printing Reuters news feed
My ASR 33 tends to print the first two columns crowded together, which I believe is due to a misadjustment for the carrier return mechanism. Adjusting this properly will be the third after-show task on the growing list.

I was lucky to have quite a few friends who worked the booth with me, engaging with show attendees and showing off the teletypes. In addition to the demonstrations I had planned, we added the game ZORK and had a few people trying out their BASIC skills on the Altair-duino and ASR 33. 

Friday, August 3, 2018

Wrapping up and transporting teletypes to CHM for the Vintage Computer Festival West

VINTAGE COMPUTER FESTIVAL EXHIBITION PREPARATION

Last night I performed my first tests of the new circuit and modification that would allow the SBC 6120 (PDP 8 clone) drive the ASR 33 teletype using the Reader Run Control feature which throttles the paper tape reader to protect against overruns. 

The test did not work properly. Some portions are working, but the small relay board was not turning on and off in sync with the Data Ready bit of the Harris 6402 UART chip on the PDP 8 clone. I picked off that signal and brought it out to a 74HC04 inverter chip, then routed it to an Arduino type relay board that will switch the 12V to the teletype control feature.

I realized that I had powered the new circuit and relay board from a separate power supply, but had no tie of the ground between the SBC 6120 and the inverter chip circuit. Can't respond without a complete circuit for electron flow. The fix would be to power the relay board and inverter from the 5V supply on the SBC 6120.

I am not sure if I will have time to make the changes and test again. Likely this will be done after the end of VCF West. I still have to write up some support material for my booth helpers and begin to collect and pack everything for transport this afternoon. 


Thursday, August 2, 2018

Readied everything for Vintage Computer Festival West, transporting to the site tomorrow

VINTAGE COMPUTER FESTIVAL EXHIBITION PREPARATION

One of my planned demonstrations was a Hangman program but it won't work in 4K Basic on the Altair 8800 clone, both due to limited room for program statements and use of syntax that didn't exist in 4K Basic. 

TicTacToe and Prime Number Generator both worked fine. I made another copy of the paper tapes for safety and then switched over to the PDP-8 clone. I booted OS/8 and ran a FOCAL program to play Towers of Hanoi. I also listed the directory for ideas for other possible demonstrations. 

I did notice that the ASR 33 letter O and numeral 0 are both filled in with ink when they print. I believe that gunk has built up in the type cylinder for those. I will use Q-Tip swabs and alcohol, plus tools, to try to clean out the letters.

To support my potential modification allowing the SBC 6120 to read paper tape, I began wiring up the inverter board and hooking it to a relay module. Construction was on a PCB built like a breadboard socket, letting me easily hook up the unused inputs to VCC or ground, add a capacitor across the power rails, and wire the input and output of the one gate I will use.

At this point, I realized that my relay modules allow me to wire up the outputs of the relay to either NC or NO points, thus I don't need the complexity of an inverter. I still decided to use it since this will buffer the UART pin and minimize any impact I might have on its operation.

Next steps to make this work are changes to the SBC 6120 and the teletype. A wire has to be tacked onto a pin of the UART on the main board and routed down to my quickie board as the input. The relay lines are connected to the Reader Run Control board wires, to operate the reed relay on that control board.

Finally, I had to wire up a 120VAC switch to manage the shorting of the output of the Reader Run Control board - it must be shorted for most uses, but opened up to allow the board to work for PDP-8 style systems like the SBC 6120. I found an extension core with a remote switch and used that to wire across the Reader Run Control outputs.

After tacking on the wire to the SBC 6120 UART at pin 19, leading to my newly built board, I cobbled together a test setup. I am using a triple power supply since I need 3.3V for the Altairduino to power the MAX3232 level converter, 5V for the board and relay I build for the SBC 6120, and 12V driven through the relay to operate the Reader Run Control reed relay in the ASR 33.

I could hear the 12V close the reed relay in the teletype, but the circuit I built wasn't working. Occasionally it would start buzzing the small relay board rapidly but it did it when no I/O was occuring through the serial port. Something is wrong and I don't have enough time to debug it.

As a result, I will abandon the attempt to read paper tapes with the PDP-8 clone. I have adequate demonstrations on hand without it. I am therefore done with changes.

I updated the scripts and included pictures of key switches and settings. Multiple copies were printed out and brought along. Next I worked on some signage to place on the table near my exhibit. I have six stands to hold a letter sized page, I just needed content for each.

I picked up the model 15 teletype this morning and brought it over to my place for staging and testing. I chose to test the model 15 while it sat in the back of my car to minimize moving the units around; this is both for my back and the sake of the equipment. I ran a power line out and set up the deskside PC, control box, teletype and prepared for testing.

The intended behavior is for the PC based program to provide a short menu on the teletype, then allow me to print news and weather stories from Reuters and the National Weather Service using their RSS feeds. The program BaudotRSS written by John Nagle is what drives the machines.

The program is seemingly working properly, however the keyboard is encoding incorrectly. I need to type the letters N and W to trigger printouts, but get V and M instead. I pulled the keyboard and manually tested connectivity to the various microswitches for the desired letters.

Offline I am getting the right encoding reliably, but the keyboard is erratic in working. I suspect it is tarnish whose cumulative effect provides too much resistance for the Nagle designed interface box to handle properly. I first swapped to my second box, to see if there is any difference using that one.

The second interface box decodes the keyboard identically - sadly that is incorrectly. I could try to burnish, deoxidize and otherwise clean up all the contacts on my keyboard, but fortunately I discovered that running the BaudotRSS program configured for a keyboardless teletype will simply type out the news feed, item by item.

This is the behavior I want, thus I am set for VCF without having to fix whatever is going wrong with the keyboard. I moved the teletype back a bit in the car so that I could shut the rear and mentally checked that item off. Model 15 demonstrations are ready to go.

I expect that I can set up a secondary configuration file with no news feed defined, also keyboardless, which should print out the weather forecast. I will work this out and test it later. 

Wednesday, August 1, 2018

ASR 33 repaired and ready for show; testing all demos to be ready for exhibition

VINTAGE COMPUTER FESTIVAL EXHIBITION PREPARATION

With my line feed and other issues fixed, I ran through the Altairduino (Altair 8800 clone) demonstrations to verify they all work. I then ran the non-tape based SBC 6120 (PDP-8 clone) demonstrations to validate them.

If I complete the reader control modification to the SBC6120 and teletype, I will test a paper tape based demo for that machine after the modification is working.

MODEL ASR 33 TELETYPE REPAIRS

I shaved down the line feed repair part I made from a rubber deck shoe sole and put it onto the teletype. It works perfectly! Linefeed is back in operation and I could turn my attention to the oddball behavior associated with a toasting electronics smell.

I did some differential diagnosis and discovered that the problems all stem from the paper tape reader. The reader engages to try to read a character, even if the reader control is set to Stop or Free, pushing up the pins and generating bits based on the hole pattern detected by the pins.

With no tape in the reader, the pins are all up and this is detected as a character 11111111 which is a rubout. If I put in a blank section of tape, the pins are all down and it is detected as an idle character 00000000. The reader pushes the pins up when the teletype is plugged in, even without turning the power switch to either Local or Line.

Whether I flip the switch to Start or Stop, the reader sits partway through a cycle with the pins pushing up. With the switch in Start, if I hit any key on the keyboard to trip the distributor, that also advances the paper tape. Thus, the paper tape reader is engaging when it shouldn't with power switch OFF, it doesn't trip the distributor clutch to complete its cycle, and this ORs in bits with valid keyboard encodings.

Since the paper tape reader is an important part of my planned demonstrations, I had to get this fixed.  The operational description in the model 33 documentation was a good starting point, where I could look over how the mechanisms related to paper tape reading work.

From that background, the symptoms led me right to the failing area. There is a microswitch that is closed when the paper tape reader trips the distributor clutch to encode the character sensed by the reader. The microswitch then energizes a big solenoid that drives the pins up through holes in the paper tape and after release will move the tape forward one position.

To have the pins sticking up and the holes jamming in data bits, the big solenoid had to be on permanently. It switched on when the plug was inserted into the wall outlet, without turning on the teletype to Local or Line. That meant the switch was closed. The last symptom, that the paper tape reader didn't advance until keyboard keys were pressed, showed that the clutch trip coil was not tripping the distributor clutch.

I looked at the clutch trip coil and related microswitch. The armature of the trip coil was out of position, so it would not move to trip the clutch. Further, it was allowing the microswitch to close. I moved it back into position and tested the machine again.

Everything now works! The toasty smell was heated resistors in the power supply, since the big solenoid in the reader is intended to be powered intermittently, not steadily as occurred with this defect. No more heat or toasty smell now. 

Preparing for VCF West exhibition, disaster strikes on ASR 33, work continues on model 15

VINTAGE COMPUTER FESTIVAL EXHIBITION PREPARATION

As soon as I arrived home from a trip to visit family on the east coast, I worked on the modification to allow the SBC 6120 (PDP 8 clone) to drive the Reader Run Control board on the ASR 33 teletype, allowing the software (e.g. OS/8 and its apps) to start and stop the paper tape reader, avoiding overruns.

I want the inverse of the state of the Data Ready pin of the Harris 6402 UART chip to drive the relay of the Reader Run Control board. The board in the teletype is designed to operate a reed relay when the control line is active with 12V (RS 232 levels). The chip, of course, uses TTL logic levels of 0 or +5V.

My board has to have an inverter to reverse the sense of the pin - since I want the motor to run only when data is NOT ready - and then something to convert to 12V and drive enough current for the teletype board.

I could use a MAX3232 board to convert the TTL level to RS232 voltage, but I am not certain that it can handle the current demands of the reed relay.

The other and simpler choice it to use a small relay board to switch a 12V supply  The relay board doesn't demand much, thus the inverter chip will easily drive it. The only complication for my exhibit is that I need a 12V supply in addition to the 3.3V supply that handles the Altairduino connection. 

The other major work I needed was a replacement for the disintegrated plastic piece in the line feed mechanism of the ASR 33. Apparently this piece which spaces a lever below a stop tab has a very narrow range of heights that will work properly.

The proper part is devilishly hard to spot on the parts catalog, especially because the plastic portion is not an independently ordered unit. Therefore, the metal bracket comes with the plastic in place, under part number 187078 for those who are also trying to locate this.

Too short and the line feed does not occur. Too high and it produces spurious line feeds. I have to get it to about 1mm of the correct size to have the teletype work properly. I decided to use some acrylic and machine it with a Dremel tool. I can start with it a bit high and grind off a bit at a time until it works perfectly.

That proved much harder to execute than in my imagination, particularly because the acrylic melts when being cut, drilled or ground. I had to abandon that and try with other methods, such as my original bit of shoe sole trimmed down a bit or with sets of rubber grommets. The sets are a bit too high, but a single grommet is too low.

While I was trying to work on the line feed issue, I began to smell a toasty electrical smell from the CCU (power supply area on the right of the teletype). Operation began to degrade. When switched to Local mode, typed input isn't being decoded correctly, instead every keystroke acts like a Rubout.

When in Line mode with the Altairduino driving it, print occurs normally and correctly, but my keyboard responses are being corrupted. I thought it would be useful to attempt to read a paper tape in either local or line mode, so I installed one of my demo tapes.

The paper tape reader didn't run when switched to Start. It did appear to be blocking the keyboard input, as if the paper tape reader was jamming in bits on top of the codes coming from my keyboard.

This is fairly disastrous if I can't find and fix the problems before Friday, as the ASR 33 is the center focus of the entire exhibit. I will need to think about the symptoms and study the documentation in the hope that I can narrow this down enough to locate the failure point.

I will need to pick up my model 15 teletype and get it over her, so that everything is staged for an easy move over to the museum on Friday when we set up for VCF. Not sure when I will have time to do that. 

MODEL 15 TELETYPE RESTORATIONS

While I was away, Marc and Ken continued to work on the two model 15 and one model 19 teletypes. Marc's model 19 printer is typing clearly and accurately. My model 15 printer types clearly and accurately. Marc's other model 15 printer is typing with mis-selections and it appears that it needed 70ma not 60ma to work.

Marc has built a box to hook together two teletypes and supply 60ma of 120V DC, letting him type on the keyboard of one unit and print on the other unit. It allows him to test the machines. He discovered that his other model 15 had the selector magnets incompetently repaired, with one of the two electromagnets wired in reverse. He fixed this and all three printers now work perfectly.

Work remains on the model 19 reassembly and on the two model 14 Transmitter Distributor boxes (paper tape reader). This won't take long to complete. Since the desk of the model 19 is back from the powercoating and looking great, we do have to replace all the wiring inside and start assembling the entire model 19 system.