>>From: ard_at_p850ug1.demon.co.uk
>>
>>>
>>> Any Teletype ASR33 experts out there ? I have a couple
>>> of problems with a teletype that recently came my way.
>>
>>I believe the ASR33 manauls are available on a web site somewhere. If
>>you don't have them as printed books, then you want to get these files
>>and print them. They make life a lot easier. OK, I did a complete
>>strip-down, clean, oil, reassemble and adjust job on an ASR33 without the
>>manuals (and with never having seen the manuals), but then I tend to do
>>crazy things like that...
>>
>>>
>>> i) When I type a character from the keyboard in local
>>> mode, something else gets printed. What gets printed
>>> is not consistently the same character. I've watched
>>> the selector mechanism while I hit rubout (all marking
>>> codebars) and I can see that the codebars are not
>>> being consistently selected betwee key presses:
>>> sometimes I get 5 of them, sometimes 6 and
>>> occasionally all 8. The high order bits seem to be
>>> worst affected by this inconsistency. However if I
>>> hold down the repeat key and select a character, the
>>> right codebars are being selected. I punched a number
>>> of characters onto tape to verify this and they seem
>>> to be punching OK. This suggests to me that this isn't
>>
>>If the character _always_ punches correctly, then the right codebars are
>>being selected. So I assume this is not the case.
>Hi
>I think you need to make this point clearer. When a character
>is punched, the same character is miss printed on the platen?
>This is how I read his original post.
Even using the REPT key to workaround the codebar issue, I wasn't able to
connect the carriage problem conclusively to either a rotational or a
vertical problem. I think at this point I'm going to take the carriage off
again and try a more thorough dismantle/clean/oil. Once I resolve the
codebar issue it will be easier to diagnose the carriage problem with a
clean mechanism.
>This tells me it is something sticking in the linkage that goes
>between the code bars and the print head( or he has it installed
>wrong but that doesn't explain why it starts working with repeats ).
>I would remove the print assembly and put it in some solvent
>while moving the levers. Then clean it off/dry it and re-oil
>it. You most likely have some dry oil that is sluggish. Getting
>oil into the right place will take a bit.
>In the service( military ), we used to clean things in a bath
>of water and heavy detergent. This was done in an ultrasonic
>cleaner. We'd then rinse and bake it at about 150F for a few
>hours. Then Re-oil/grease, place on test bench and fine adjust.
>I suspect that a good dish washing liquid would work.
I remember using an arklone bath at a place I worked years ago for cleaning
circuit boards. Probably taken a few years off my live expectancy breathing
that stuff in.
>
>Basically, there are 3 possible problem areas :
>
>1) The keyboard contacts are not closing correctly, so the parallel data
>output of the keyboard is incorrect.
>
>2) The distributor disk (rear right of the typing unit) is not correctly
>serialising the data from the keyboard. Unlikely, but possible.
>
>3) The selector unit (rear left of the typing unit) is malfunctioning,
>and is not correclty responding the receiving magnet.
>
I can see the armature 'stutter' when I punch the rubout key - by stutter, I
mean that its movement looks slightly different between keypresses. So I'm
inclined to think that the magnet is just responding to what it is receiving
which is incorrect. I cleaned the distributor and reseated the brushes,
readjusted the brush holder to the alignment mark per the adjustment
procedure in the manual. So I'm back to option 1) - the keyboard - as the
source of the problem. One thing I noticed last night was that the keyboard
cover is broken. There are four round tabs, one at each corner, that fit
into corresponding holes on the end plates. These are broken, so the cover
can rock from side to side, pivoting on those center projections. I wonder
if that could cause some sort of alignment problem within the keyboard ?
Thanks for the suggestions - I will get to it again tonight and see what
else I can find out. I hate the idea of opening that keyboard up...
>I would check (3) first. It may need to be removed, dismantled, cleaned,
>and re-lubricated. It's not uncommon for parts to stick on old, gummy,
>lubricant.
>
>> an electrical problem since whatever code the keyboard
>> is sending is eventually being sensed correctly.
>>
>> ii) In addition to i) above, even if I do get all the
>> codebars moving as required, the wrong characters are
>> being printed. I removed the carriage, cleaned and
>> oiled it and replaced it but this hasn't helped.
>
>How did you clean it? I have found that soaking compete assemblies in
>solvent doesn't help much. You really have to take all the parts apart,
>clean them, and then re-assemble them. If yoy've done this, then there
>are a lot of sdjustment you'd have to set up.
>
>A character decoding problem must be in the carriage. There are 4 parts
>to the decoder mechanism :
>
>2 bits -> one of 4 levels of lift of the type cylinder
>1 bit -) rotate cylinder left or right
>2 bits -> one of 4 angles (pairs of characters) of rotation of the cylinder
>1 bit -> an extra 1 character angle rotate (it moves the selector bars
>for the previous rotation selection slightly...).
>
>Figure out which mechanism is malfunctioning by comparing the position of
>the character you get with the position of the character you should be
>getting. Then work out what's not operating properly.
>
>-tony
>
>
Received on Fri Nov 15 2002 - 20:35:31 GMT
This archive was generated by hypermail 2.3.0
: Fri Oct 10 2014 - 23:35:27 BST