Get the teleprinter up and running

[amcewen]

So, Western Approaches have a bunch of teleprinters from when they were operational in the Second World War. They've lent us (might well be long-term loan or donation if we want one) to see if we can get it up and running and talking to something across the Internet... It's a Creed Model 7B:

@magman2112 has bought a book about them with a bunch of details in it - chapter 2 of that is also available online, and there's a Post Office manual for it

[amcewen]

@kayharvey and I have investigating it and making a start on cleaning today.

I think we've got most of the cleaning done - the keys are drying, and need their rollers oiling before they're reassembled. Next steps we think are...

• Reassemble it all
• Check the springs, particularly on the receiver "print head" and the one on end transmitter rail which is currently unhooked
• Check over the wiring with a meter (there's a wiring diagram at the end of Chapter 2 in the book)
• Possibly clean the carriage (we've only worked on the transmitter and receiver units today)
• Work out how we power it - the motor is marked as DC, but bits of the manual imply that you can maybe run them on AC and Dean at Western Approaches said they'd just stuck a plug on one a while back.
• It feels like it'll make sense to run it through @magman2112's mains isolator unit, so finding out how that works would be handy
• Fire it up and see what happens when it's running!
• Find a new ribbon for it

[pkharvey]

Looking for clues anywhere they turn up...

RTTY The Easy Way has some more clues about powering the motor on page 4 (and an OK overview in general) but couldn't see how it matches with our motor marked "13.↵100-110V. D.C.↵160-250V. D.C."... could it be selectable, with a link or something else?

However there are several different voltage ratings for the motors; the more common types are:

• 230V AC: Connect direct to the mains.
• 160/220V DC: Wire for 220V (adjustment Link is usually on the rear side of the motor) and feed from the mains via a 1000PIV 1A bridge rectifier.
• 160V DC:Use half wave rectified mains or suitable transformer and rectifier (a mains to 350-0-350V transformer in reverse will do).
• 110V DC:Requires an appropriate transformer, rectifier, etc.
• 24V DC: Obtain or construct a 24V 4A DC supply - regulation is not critical, but the on-Load voltage should be between 23V and 26V.

The motor shown in the picture of a 7B on this page is marked slightly differently with "№13↵100/110 V DC↵OR↵160/250 V DC" and hints the motor can be "configured":

Although some 7B teleprinters had synchronous AC motors, most had DC motors that can be configured for different voltages, and governors that can be adjusted for 50 or 45.45 Baud operation.

Found here was a clue about there being an actual power supply box for the 80 V DC signaling:

The Creed 7B was used on both point to point links and terminal points on the home defence network. It has no paper tape facilities making it unsuitable for message centres. The mains power unit, a large box normally sited under the desk provides the 80 volts DC signalling. The motor has an integral speed governor. This makes the Creed 7 series teleprinter less reliant on good quality mains power and in a war situation suitable for use with a standby generator.

Now just found this Teleprinter No. 7B instructions for major overhaul and repair, there is some more detail on our dc motor №13 on pdf page 42 (paragraphs 183, 188 and 189 might be useful)

Finally, a 7-page P.O. engineering instructions — motor and governor unit also has info covering dc motor №13.

Clue in info on D.C. power supply for teleprinters Nos 3 and 7 from a.c. and d.c. e.l. mains mentions a Rectifier No 22

Pulse power supply for teletype from the 1940s (with luminous mercury thyratrons!) Not our model, but I remember seeing this one, I think I saw it before with @amcewen Teletype Model 19 Part 4: Thyratron Power Supply

The REC-30 power supply delivered 900 mA at 120V DC, enough to power 15 teletypes.

Some discussion on interfacing the 80-0-80v signalling.

Bombarded with information.

[Sean-anotherone]

Based on it's age, apart from the paper button inlays this might benefit from a dousing in paraffin, or a good session in an engine degreaser.

Once working, it would be pretty cool to have it printing messages out (arduino driven perhaps) when a button is pressed ?

[Sean-anotherone]

https://www.radiomuseum.org/r/creed_teleprinter_7b.html

http://www.samhallas.co.uk/repository/telegraph/ep_telegraphy_4_1.pdf

http://www.samhallas.co.uk/repository/telegraphy.htm

[Sean-anotherone]

suggests motor can be AC sync or DC 22-25V, probably best not plug it into mains until this is checked

[pkharvey]

A YouTube playlist of 20 videos by Jerry Walker - a series on a Creed 7B Restoration

I noticed his motor is the 24V DC version, unlike ours.

[amcewen]

That YouTube playlist is an excellent find @kayharvey! Having watched some of it I think we should maybe do a bit more disassembly and checking of the receiver mechanism (and maybe the transmitter, but I've not got to that video yet... ;-) before we fire it up.

@Sean-anotherone, there seem to be an assortment of motors for them, but ours has "100-110V. D.C. 160-250V. D.C." written on the side of it in big letters. @kayharvey and I were thinking that rectifying the mains with a suitable bridge rectifier (I think I've got one knocking about still, if I can find it) is probably the easiest option. Looking at @kayharvey's notes above, we should do a bit of checking for configuration options on the motor too first.

[amcewen]

Watching this video it seems we can run the data at +/-20V rather than +/-80V (or possibly even +/-15V, which might let us produce it from a bench power supply?)

[9600]

@kayharvey and I were thinking that rectifying the mains with a suitable bridge rectifier

Maybe with a transformer in between for added safety? I was thinking to run our Creed motor this way, with a 110V step-down and bridge rectifier.

Also possibly of interest:

https://github.com/fablab-wue/piTelex/wiki/Example_RPi

ISTR reading that while lower signalling voltages may pull the solenoid in, it may not do so fast enough and result in issues.

We should get one of our Creed machines up and running too, then set up a Liverpool-Halifax teleprinter link :)

[pkharvey]

I brought in today from storage for a future session, a 1.5 kVA isolation transformer, a KBL407G 700 V 4 A bridge rectifier, and a FB2506 600 V 25 A bridge rectifier just in case or for good luck.

[amcewen]

(this comment by @kayharvey using @amcewen 's laptop and github session)

I learned in this Telex FAQ the difference between usage of the terms 'teleprinter' and 'telex'. The explanation makes sense though I have read some articles that appear to use 'teleprinter'/'telex' interchangeably.

• A Telex machine is a teleprinter used on the telex network.
• A Telex message is the communication transmitted/received on the telex network.
• "A Telex" usually means a message sent on the telex network.
• A Teleprinter is a device for transmitting and receiving the typed word using International Telegraph Alphabet No.2 code designed for start-stop telegraph systems.
• A Teletype is a teleprinter manufactured by the Teletype Corporation.
• A Creed is a teleprinter manufactured by the ITT-Creed Company.

The PiTelex link from @9600 might be a good solution. The text says it's to connect teletypes (with FSG) to the i-Telex network over internet. FSG is fernschaltgerät ("Remote Switch"). Might need some investigation if our model can be used or adapted.

And you (must have)* a so called "fernschaltgerät" (FSG), e piece of old fashioned switchbox with the TW39 protocol, you may find some still on ebay.de here above some examples.

Idea earlier for the 'receiving/printing' side to use a 24 V DC single polarity power supply with a H-bridge driver such as the ST/TI L293D, datasheets say up to 36 V at 600 mA (1.2 A peak) and the 'D' type has integrated output clamp diodes for inductive transient suppression.

Could computer-generate the Baudot to drive the inputs, but might be good to use some isolators, so that after our efforts, our isolation transformer isn't bypassed?

For powering the DC motor, rectified mains will give 325 V peak! I still don't know how to interpret that №13 motor, is it the first range or the second range printed? We can only try by setting the variac to the lowest range first. If it runs too slow, it's the higher range. Running it slow might be good for problem solving if something's not working right. I have a giant variac at home that can be used if the one here can't be made to work, though my one is missing the cover that goes over the live terminals, plus needs a plug and socket on leads to go onto its terminals. @amcewen mentioned it was described in one of the videos (in the Jerry Walker series) that lowering the motor DC voltage has the effect of increasing its speed, so we might not be able to carefully turn up the variac to go from lower to higher speeds before the governor starts to limit it.

For the keyboard transmitting side, the interesting pages in the Teleprinter Handbook 2nd ed (Hobbs et al) are on pages 2.11 and 2.13. It looks as if the mark and space contacts, switching between mark and space could be detected by sampling continuity (and as we saw today, we think may default to an 'idle' mark or space while in transmit mode when there's no latched key. We can verify this next time with a beeping multimeter and turning the keyboard slowly by hand. This could be sampled at a higher rate and debounced when we get to the stage of "transmitting" the typed message into an ESP32 or whatever we end up using.

Servicing done today, an extension spring was replaced under the cams on the transmitter side. The combination bars were degreased and re-oiled with 3 in one oil and reassembled.

[Sean-anotherone]

If it's useful I have a variac-box a bit like the one Mike brought in, its fully working.  My dad used to use it to "gently run up" valve-based radar sets.  It goes up to about 280V.  Variac output is hooked up to a 13A socket. Sean

On Saturday, 27 February 2021, 21:46:38 GMT, Adrian McEwen <notifications@github.com> wrote:  

(this comment by @kayharvey using @amcewen 's laptop and github session)

I learned in this Telex FAQ the difference between usage of the terms 'teleprinter' and 'telex'. The explanation makes sense though I have read some articles that appear to use 'teleprinter'/'telex' interchangeably.

• A Telex machine is a teleprinter used on the telex network.
• A Telex message is the communication transmitted/received on the telex network.
• "A Telex" usually means a message sent on the telex network.
• A Teleprinter is a device for transmitting and receiving the typed word using International Telegraph Alphabet No.2 code designed for start-stop telegraph systems.
• A Teletype is a teleprinter manufactured by the Teletype Corporation.
• A Creed is a teleprinter manufactured by the ITT-Creed Company.

The PiTelex link from @9600 might be a good solution. The text says it's to connect teletypes (with FSG) to the i-Telex network over internet. FSG is fernschaltgerät ("Remote Switch"). Might need some investigation if our model can be used or adapted.

And you (must have)* a so called "fernschaltgerät" (FSG), e piece of old fashioned switchbox with the TW39 protocol, you may find some still on ebay.de here above some examples.

Idea earlier for the 'receiving/printing' side to use a 24 V DC single polarity power supply with a H-bridge driver such as the ST/TI L293D, datasheets say up to 36 V at 600 mA (1.2 A peak) and the 'D' type has integrated output clamp diodes for inductive transient suppression.

Could computer-generate the Baudot to drive the inputs, but might be good to use some isolators, so that after our efforts, our isolation transformer isn't bypassed?

For powering the DC motor, rectified mains will give 325 V peak! I still don't know how to interpret that №13 motor, is it the first range or the second range printed? We can only try by setting the variac to the lowest range first. If it runs too slow, it's the higher range. Running it slow might be good for problem solving if something's not working right. I have a giant variac at home that can be used if the one here can't be made to work, though my one is missing the cover that goes over the live terminals, plus needs a plug and socket on leads to go onto its terminals. @amcewen mentioned it was described in one of the videos (in the Jerry Walker series) that lowering the motor DC voltage has the effect of increasing its speed, so we might not be able to carefully turn up the variac to go from lower to higher speeds before the governor starts to limit it.

For the keyboard transmitting side, the interesting pages in the Teleprinter Handbook 2nd ed (Hobbs et al) are on pages 2.11 and 2.13. It looks as if the mark and space contacts, switching between mark and space could be detected by sampling continuity (and as we saw today, we think may default to an 'idle' mark or space while in transmit mode when there's no latched key. We can verify this next time with a beeping multimeter and turning the keyboard slowly by hand. This could be sampled at a higher rate and debounced when we get to the stage of "transmitting" the typed message into an ESP32 or whatever we end up using.

Servicing done today, an extension spring was replaced under the cams on the transmitter side. The combination bars were degreased and re-oiled with 3 in one oil and reassembled.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[amcewen]

Yesterday @kayharvey found a list of the Baudot codes which will be handy for working out whether a peg or a gap in the transmitter is a mark or a space, so we can start talking about them properly :-D

[magman2112]

I’ve not been contributing much to this project, but have been watching developments with interest.

One thought I had regarding the motor supply was to daisy chain the lab power supplies we have in the space to build I higher voltage supply. I know we have at least 2 supplies, possibly 3. These are typically 0-30V supplies and have the benefit of metering and current limits. With 2 supplies, we can make a 90V supply (The Farnell one is dual channel). With 3 supplies we can make a 120-150V supply. This is obviously just for testing, once we know what supply it actually needs, we can make something more permanent. For a more permanent supply, CPC sell relatively inexpensive toroidal transformers up to 2x55V secondaries - after rectification this would be about 150V DC, so this is readily achievable - we just need to know the current capacity needed.

On the serial interface side, I recently picked up a Tektronics 832 Serial Datacom tester off eBay. It appears that this will output 5 bit serial data (though I doubt this will be baudot). It also states that it outputs up to 25V on the serial port. This unit is now in DoES on the members rack, it needs a bit of investigation, but may give us a quick fix for testing purposes. Again, longer term would likely be a Raspberry Pi with a serial adaptor optimised for the right signal voltage and possibly a set of prepared messages and proximity triggers/timers.

[amcewen]

We've been slacking a bit on updating things here, but are keeping a video diary of progress - https://www.youtube.com/watch?v=3YZY0edw8Co&list=PLx0MNQgBzO2XejhD4XcjIWMq6_-TW47H-

@kayharvey and I worked out a few weeks back that we can run the receiver solenoid at anything above +/-6V, which should make signalling easier than finding +/-80V somewhere :grin: and last Saturday @kayharvey, @magman2112 and I daisy-chained four bench power supplies to get the 100-110V DC the motor needs, and tried powering it up for the first time!

[Sean-anotherone]

I wonder if a local rewind firm would be able to work their magic on that motor and make it 24V.  As the governor is centrifugal, it would not need to be particularly accurately worked out, probably just 110/24 x #turns on each pole. Or just hollow out the motor case and fit an r/c car motor in there, probably got about the same amount of power! Sean On Thursday, 15 April 2021, 12:39:16 BST, Adrian McEwen @.***> wrote:

We've been slacking a bit on updating things here, but are keeping a video diary of progress - https://www.youtube.com/watch?v=3YZY0edw8Co&list=PLx0MNQgBzO2XejhD4XcjIWMq6_-TW47H-

@kayharvey and I worked out a few weeks back that we can run the receiver solenoid at anything above +/-6V, which should make signalling easier than finding +/-80V somewhere 😁 and last Saturday @kayharvey, @magman2112 and I daisy-chained four bench power supplies to get the 100-110V DC the motor needs, and tried powering it up for the first time!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[magman2112]

I’ve managed to find a possible inexpensive partial solution for the Power Supply, a DC-DC converter from eBay.

https://www.ebay.co.uk/itm/900W-15A-DC-DC-Boost-Power-Converter-8-60V-to-10-120V-CC-CV-Step-Up-Module-/384098970206?_trksid=p2349624.m46890.l49286

This will still need another mains power supply, but a PC or Laptop PSU, may be sufficient, we will just have to try it and see. With a bit of luck, this will arrive before our next workshop on the teleprinter.

I also picked up another DC-DC converter from eBay that may help with the signalling, but this will likely take a bit longer to be delivered.

https://www.ebay.co.uk/itm/DC-DC-Step-Up-Down-Boost-Buck-Voltage-Converter-Module-45-390V-45-390V-/323884986234?_trksid=p2349624.m46890.l49286

[magman2112]

Last Sunday, we tried the DC-DC convertor with a mains powered 48V power supply without a great deal of success. at anything above about 80V, the motor would start ramping up and down in speed, which appeared to be the voltage fluctuating on the DC-DC convertor under the load of this motor.

We tried inserting a capacitor bank on the output of the DC-DC convertor, which made a small difference, we now got the teletype motor to work above 90V, but it wasn't always reliable. I suspect that the DC-DC convertor is more specified for an amplifier rather than a high inductive load, such as the teletype motor.

We have subsequently decided to go back to a linear transformer based power supply, for which I have built a shopping list from CPC for the majority of the parts (I already have the Bridge Rectifier and IEC mains socket in stock).

Teletype PSU Parts 28APR2021.xlsx

@amcewen has said that he will obtain these parts from CPC, so that I can get the PSU built for our next time we work on the Teletype.

I have also identified some possible power connectors for the Teletype-PSU connection, but these are for an up to 8mm cable diameter, which is OK for mains flex, but may be a little tight with the old flex on the teletype at present - it is worth a try though.

[Sean-anotherone]

It may also be the impact of the centrifugal governor intermittently loading a psu designed for steady load .  An optical servo governor might work better.  Linear psu far simpler and quicker though. Sent from Yahoo Mail on Android

On Wed, 28 Apr 2021 at 15:09, Mike @.***> wrote:

Last Sunday, we tried the DC-DC convertor with a mains powered 48V power supply without a great deal of success. at anything above about 80V, the motor would start ramping up and down in speed, which appeared to be the voltage fluctuating on the DC-DC convertor under the load of this motor.

We tried inserting a capacitor bank on the output of the DC-DC convertor, which made a small difference, we now got the teletype motor to work above 90V, but it wasn't always reliable. I suspect that the DC-DC convertor is more specified for an amplifier rather than a high inductive load, such as the teletype motor.

We have subsequent decided to go back to a linear transformer based power supply, for which I have built a shopping list from CPC for the majority of the parts (I already have the Bridge Rectifier and IEC mains socket in stock).

Teletype PSU Parts 28APR2021.xlsx

@amcewen has said that he will obtain these parts from CPC, so that I can get the PSU built for our next time we work on the Teletype.

I have also identified some possible power connectors for the Teletype-PSU connection, but these are for an up to 8mm cable diameter, which is OK for mains flex, but may be a little tight with the old flex on the teletype at present - it is worth a try though.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[amcewen]

@magman2112 Thanks for that. Parts now ordered.

[amcewen]

Making a note of the data cable wiring here for somewhere to lodge it... Taken from page 46 of the overhaul and repair manual (not to be confused with the different wiring diagram on page 2.30 of the (physical) Teleprinter Handbook)

Colour	Description
Slate	Alarm and bell
Green	Transfer contacts: Mark
Orange	Transfer contacts: Space
Blue	Transfer contacts: Common. (Also connected to Send/Receive switch: Send, via a 4K resistor)
Black	Send/Receive switch: Receive
Green/Red	Send/Receive switch: Common
White	One side of the electro magnet
Brown	Electro magnet common
Red	Other side of the electro magnet

To test it today we wired it up to the loopback circuit Fig 12.2b on page 12.2 of the Teleprinter Handbook.

We were running with +/- 15V, initially with current limits, but that wasn't needed once we got the wiring set up correctly.

• +ve from right-hand power supply to Orange (NOT BROWN)
• -ve and +ve of the two channels of the power supply are connected together as 0V, and then connected to red
• -ve output of the left-hand power supply is connected to Green
• Blue and White are linked together to provide loopback.

To show mark/space we wired two LEDs back-to-back in series between blue and white, so we'd get them to illuminate alternately whether there was a mark or a space.

[amcewen]

Turns out the address had gotten mangled (it lost the street details) in the CPC order, so it got returned to them. I've placed the order again - the caps are back-ordered so they'll follow on next week when they get them in stock.

[amcewen]

I've ordered a new ribbon for it too. It's a half-inch ribbon, which seems pretty standard, although the spindle seems to be 3/4" which is less common. The only one with a bigger spindle hole was going to ship from the US and be too expensive, so I think we'll need to re-wind the new ribbon onto the existing spools.

[Sean-anotherone]

Or just soak the old one in ink and wind it from end to end a couple of times? Sean

On Friday, 14 May 2021, 15:40:59 BST, Adrian McEwen ***@***.***> wrote:  

I've ordered a new ribbon for it too. It's a half-inch ribbon, which seems pretty standard, although the spindle seems to be 3/4" which is less common. The only one with a bigger spindle hole was going to ship from the US and be too expensive, so I think we'll need to re-wind the new ribbon onto the existing spools.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[amcewen]

The power supply parts have arrived (apart from the caps, which are back-ordered). They're in the CPC box next to the teleprinter.

[amcewen]

The ribbon has arrived. I've left it next to the teleprinter

[amcewen]

We've been remiss in updating things here, but there have been plenty of developments (see the latest videos on the restoration playlist for full details).

• Ribbon is fitted
• Loopback tested, and it needs at least +/-30V on the signalling to get clear characters
• The power supply has been built up
• We've got initial code to send data to it, but haven't confirmed if the timings are right yet

[amcewen]

@kayharvey has built up a circuit with a high voltage h-bridge, and wired that to D1 and D6 on a Lolin ESP8266 NodeMCUv3-clone. Checking the pinouts, D1 is Arduino GPIO 5 and D6 is GPIO 12.

[amcewen]

This week we were looking at fixing the figs/letters switching, and just got as far as dismantling almost all of the teleprinter tin order to get to the combination head unit. Next will be to dismantle and clean that.

The screws are all in the storage box:

1. Control lever unit
2. Ribbon holder
3. Motor
4. Combination head unit
5. Motor drive washer
6. Receiving cam unit
7. Automatic starter
8. Small driveshaft cover
9. Longer driveshaft cover
10. Typehead bearing and bracket

[magman2112]

@kayharvey and I have been working out the connections from the microcontroller to the teletype, so we thought it worth capturing the connections on GitHub as well.

Teletype Connections Transfer Contact Mark (Green) -> -12V Power Supply (PSU 12) Transfer Contact Space (Orange) -> +12V Power Supply (PSU 12) Transfer Contact Common (Blue) -> Controller Keyboard In One Side of Electromagnet (Red) -> Controller H-Bridge Out 1 Other Side of Electromagnet (White) -> Controller H-Bridge Out 2 Send/Receive Switch Receive (Black) -> +12V Power Supply (PSU 12) Send/Receive Switch Common (Green/Red) -> Controller Send/Receive Switch In

Additional Controller Connections +80V (PSU 80) - H-Bridge Supply (Vs) +5V (PSU 5) - ESP8266 (5V) +12V (PSU 12) - H-Bridge Supply (Vcc) Ground for all 3 PSU’s (PSU 5, PSU 12 and PSU 80)

[pkharvey]

Logic board (ESP8266/MC1489/PWD13F60) pin assignments (needed for programming the ESP8266):

Function	Pin on new board	Pin on old board
LIN1/HIN2 (PWD13F60)	D7/GPIO13	D6/GPIO12
HIN1/LIN2 (PWD13F60)	D6/GPIO12	D1/GPIO5
KB IN (MC1489)	D2/GPIO4	n/a
S/R IN (MC1489)	D1/GPIO5	n/a

Illustration of physical layout of the new board:

[pkharvey]

Today's efforts were in wiring the control box (with ESP/H-bridge/+-12/5/80V) and the black box (25-way D-sub to teleprinter spade terminal block) and trying it out.

It was suspected that the connectors were at fault as we could disconnect/reconnect the 5/12 V connector repeatedly and find sometimes the ESP runs (flashes the LED, sending the Baudot) and sometimes it doesn't.

We were powering the control box from a 12 V 1 A simple power supply.

It was also suspected there could be a ground loop (as a guess) somewhere as we noticed the LED on the 80V power module blinking and making a hissing sound (suspected to be coming from the 80 V module) with the USB plugged into the laptop.

On removing the laptop USB, the led on the 80 V module went stable, hissing stopped and the ESP would sometimes run the code, send the Baudot and (weakly) actuate the solenoid.

Trying again on the larger bench power supplies (DoES Farnell and "Aeternum" 10 A power supply), both showed that even with the current limit maxed, the current limit LEDs still lit while H-bridge was switching, but it did drive the solenoid much better, however the ESP still only occasionally ran its code when power was applied - either not doing anything at all, or printing in full its Baudot "hello world" string.

Capacitors were added to input (1000 uF) and output (100 uF) of 80 V with not much effect. 470 uF on 5V output also had no effect on ESP startup.

80 V was dialled down to 50 V (and kept there), with the effect of quieter solenoid. Hissing was slightly quieter also.

[pkharvey]

Final try was to power-cycle the ESP with a power-only USB cable to the laptop with the 5 V from the module pulled out. This caused the ESP to power up and run every time.

[Sean-anotherone]

As each  link in the chain seems to work independently, perhaps opt isolators (with a suitably rated transistor each) are your friend? Or. more in keeping with the vintage, relays? Sean

On Saturday, 11 September 2021, 20:30:22 BST, kayharvey ***@***.***> wrote:  

Final try was to power-cycle the ESP with a power-only USB cable to the laptop with the 5 V from the module pulled out. This caused the ESP to power up and run every time.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

[magman2112]

Whilst working to straighten the roll holder on the print carriage, I managed to break a rod and a screw which attach the roll holder to the carriage. On further investigation and additional disassembly, I found that some of the other parts for this mechanism are quite fragile as well. The replacement parts needed are 1 rod with screw threads on each end along with 2 threaded shafts with a portion of the shaft turned to a smaller diameter (one of these is not broken, but feels a little fragile)

I have stored the miscellaneous screws bolts and washers in position 15 of the parts box and the the broken rod in the teleprinter box.

I will generate some CAD drawings for the respective parts and will ask @sean if this is something he can possibly make for us.

[Sean-anotherone]

I'll have a look certainly, but my lathe won't cut threads and I only have metric dies, so might need help from someone with a stash of older tooling. Sean

On Saturday, 15 January 2022, 16:06:02 GMT, Mike Gorman ***@***.***> wrote:  

Whilst working to straighten the roll holder on the print carriage, I managed to break a rod and a screw which attach the roll holder to the carriage. On further investigation and additional disassembly, I found that some of the other parts for this mechanism are quite fragile as well. The replacement parts needed are 1 rod with screw threads on each end along with 2 threaded shafts with a portion of the shaft turned to a smaller diameter.

I have stored the miscellaneous screws bolts and washers in position 15 of the parts box and the the broken rod in the teleprinter box.

I will generate some CAD drawings for the respective parts and will ask @sean if this is something he can possibly make for us.

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you were mentioned.Message ID: @.***>

[amcewen]

@magman2112 has found more manuals on the samhallas.co.uk site, in particular C5007 which details the typehead adjustments!

That has far better diagrams, so we're looking through the adjustment process for that.

• Clearance Y seems within bounds, maybe a little too close. @pkharvey noted it doesn't seem 100% parallel to the bellcranks, but not terrible.
• Clearance Z seems good.

In manual C1005 there's also a good description of how the latching mechanism operates, which seems to confirm @paulgeering's thoughts on the mechanism, but it hasn't gotten us any closer to getting it to work.

Going to see if we can get another teleprinter to compare it with.