Hi Nicolas,

Many thanks for your thoughtful reply and I have checked those parts of the machine, all the springs are in place. I have checked the operation of the selector mechanism a number of times and it seems fine. Sending the various bits, either as a single bit, or combinations of bits, generates the correct mark-space positioning of the code bars. I was puzzled for a while about the common bar, but it became clear that there are in fact two 'common' arrangements. The first one is bits-1 and 2 which is the common code bar that comes from the selector mechanism and this is used for the vertical positioning of the type box. There is also another 'common' for bits-4 and 5 which are part of the horizontal positioning of the type box. Bit-3 selects which side of the neutral alley is used as you have pointed out.

In the meantime I have found there is a spring missing on the bell crank that links the code bar for bit-4 to part of the horizontal shift mechanism. It seems that the missing spring causes the horizontal shift to miss a column of characters depending upon the direction of travel which is what I was finding when sending the character pairs. In that part of the machine there are three small springs which restore the 'plates' (I'm not sure what else to call them...) for bit-4, bit-5 and the combined bit-4-5 combinations which determines which column of characters in the type box is used. Bit-3 selects which side of the neutral alley is used as you have pointed out. It's a very clever system for sure. Nick England sent me a link to a useful document (you may have already seen it):

https://www.navy-radio.com/manuals/tty/jh/ttyman3/TeletypeManuals/model28/Instructional/28_Printer.pdf

and pages 21 and 22 show the relevant functions for the type box positioning which helped me to locate where the spring is missing. It's a tiny little spring and I hope I can get some to replace the set of three... They are Teletype part number 45027 which I found from the assembly diagrams in one of the Teletype manuals. Of course there may still be other problems once the springs are replaced and I'll have to deal with that situation if it arises...

I hope your restoration work is going well! Thanks again for your suggestions.

Regards

Dale

On Tue, 14 Oct 2025 at 02:07, John Doe <jdoe10449@gmail.com> wrote:

Hi Dale, how ya goin'? Greetings from Canada.

I'm new to the group too and I've been restoring an M28 ASR for the last 6 months or so. The mechanism is still fresh in my mind and I'd like to take a crack at helping you with your machine.

Good work testing and gathering the data to share. I can't claim to be an expert but something didn't sit right with me in your message. Your quote here:
"The problem seems to be with bit 4 going from mark to space if bit 5 is mark. There is no problem with bit 4 = space (...) The state of bits 1,2 and 3 don't seem to matter."
It's my understanding that the only difference between H-T, Y-Z, P-L, Q-W, is Bit-3. Please see the attached "Dales_Machine_1.PNG" for a map. Both columns are fixed by bits 4 & 5 - the left column (HYPQ) if bit 3 is on or the right (TZLW) if bit 3 is off - and the rows are selected by the 4 possible combinations of bits 1 & 2.
If you and the others in the group can agree that Bit-3 is the problem, let's start with checking the selector mechanism (I know you checked it already). See attached "Dales_Machine_2.PNG". Let's hope it's just a lack of lubrication or a missing spring. Repeated information: Check that all springs are present on the push levers and intermediate arms and verify that movement of the mechanism is snappy and free of sticking.

Amendment: You may need to check the springs on the selecting levers that ride the cam. I failed to highlight them in the attachment because you will probably need to remove the typing unit to be able to see them; they're underneath. It's not a difficult job - just 4 screws to remove and the typing unit should pull straight up. Get a sharp 10" long flat head screwdriver - preferably hollow ground and press down hard on the selector side to prevent the driver from slipping.

Hope this helps.

Nicholas Gromak