5 Oct
2026
5 Oct
'26
8:26 p.m.
My ADM-3A Dumb Terminal has an aggravating tendency to key bounce. It's
not the highest quality keyboard I've ever encountered, and the
circuitry appears to have (at best) 2-key rollover.
Currently I'm using it to talk to my PiDP-8/I at either 110 baud (to
simulate the Teletype experience) or 1200 baud.
Has someone already made improvements to the key scanning circuitry, or
do we just live with it ("that's what they all do")?
I haven't used one regularly since 1981 (senior year in college) 😉 but
don't remember it being quite this annoying.
thanks.
5 Oct
5 Oct
9:27 p.m.
I did elecctronic maintenance in a data centre which included working with a
small number of ADM-3As in the early 1980s.
I do not recall any complaints about key bounce and I know darn well I would
have heard. Those would likely be mechanical key switches so they may have
oxidised over their long career (but all 58 keys? That does not seem
likely.) Cleaning the contacts with a aerosol contact cleaner may be
possible or it may be possible to take them apart and clean them.
After 40 or more years on the job, some of the capacitors may be failing -
that is the most common time related component failure . You might look
hard at the schematic focusing especially on electroytic capacitors. If the
power supply becomes noisy due to sagging filter capacitors or the bypass
capacitors are failing and no longer cleaning up the noise on the power
busses then that might contribute to your problem. When diagnosing an
electronic problem, you are wise to start by being suspicious of the power
supply and the power distribution busses. As far as I recall, the ADM3A
would be mostly SSI TTL circuitry and those ICs would require proper noise
suppression.
The only problem I distinctly recall with an ADM3A was when some academic
spilled rum and coke into the keyboard. I can't recall if I ever fixed it,
I suspect that I scrapped it.
--
Richard Loken VE6BSV : "...underneath those tuques we wear,
Athabasca, Alberta Canada : our heads are naked!"
** rlloken(a)telus.net ** : - Arthur Black
6 Oct
6 Oct
11:49 a.m.
I'll check the p/s, but since everything else works correctly, I suspect
dirty key contacts and a poor debouncing circuit.
Pressing (for example) the "2" key, sometimes it outputs "22",
sometimes "2v"... the debounce circuit should take care of the first,
and the second has falsely set bits (0x32 vs 0x76).
Some of the other keys have either or both errors too - will have to log
the errors associated with each key and see if I can find it in the logic.
On 10/5/26 23:27, Richard Loken wrote:
I did elecctronic maintenance in a data centre which
included working with a
small number of ADM-3As in the early 1980s.
I do not recall any complaints about key bounce and I know darn well I would
have heard. Those would likely be mechanical key switches so they may have
oxidised over their long career (but all 58 keys? That does not seem
likely.) Cleaning the contacts with a aerosol contact cleaner may be
possible or it may be possible to take them apart and clean them.
After 40 or more years on the job, some of the capacitors may be failing -
that is the most common time related component failure . You might look
hard at the schematic focusing especially on electroytic capacitors. If the
power supply becomes noisy due to sagging filter capacitors or the bypass
capacitors are failing and no longer cleaning up the noise on the power
busses then that might contribute to your problem. When diagnosing an
electronic problem, you are wise to start by being suspicious of the power
supply and the power distribution busses. As far as I recall, the ADM3A
would be mostly SSI TTL circuitry and those ICs would require proper noise
suppression.
The only problem I distinctly recall with an ADM3A was when some academic
spilled rum and coke into the keyboard. I can't recall if I ever fixed it,
I suspect that I scrapped it.
--
Richard Loken VE6BSV : "...underneath those tuques we wear,
Athabasca, Alberta Canada : our heads are naked!"
** rlloken(a)telus.net ** : - Arthur Black
12:24 p.m.
Does it get better, or WORSE, with extended use?
Sometimes, oxidized or dirty contacts clean themselves a little with more
use, and are especially bad with non-use.
40? years ago, I was handed a "dead" TRS80 keyboard. After repeatedly
beating on one key, it gave a hit! And, would work a small part of the
time. I tried with others, and determined that extended excessively heavy
use seemed to re-enable keys, at least sporadially!
So, I dug out a Rochester Dynatyper Keyboard Actuator. It was like the
KGS80, a box of solenoids intended to be set on top of a typewriter
keyboard to turn a typewriter into a cheap printer (back when "cheap"
computer printers were a thousand dollars). Those would "print"
relatively reliably at about 8 characters per second. There also existed
the Escon (Walnut Creek) Model E-P, which was the same idea but Selectric
specific, and mounted inside the Selectric, and pulled the keys down; that
could almost handle the 14.8 characters per second maximum of Selectrics.
I set the Rochester on top of the keyboard, and sent enormous text files
to "print". After a while some of the keys were working! By the next
morning, they all worked!
Before trying that, consider that if the problem is mechanical, it might
help those keys, or it might wear them out, and finish killing them.
--
Grumpy Ol' Fred cisin(a)xenosoft.com
On Tue, 6 Oct 2026, Charles Morris via cctalk wrote:
> I'll check the p/s, but since everything else works correctly, I suspect
> dirty key contacts and a poor debouncing circuit.
> Pressing (for example) the "2" key, sometimes it outputs "22",
> sometimes "2v"... the debounce circuit should take care of the first,
> and the second has falsely set bits (0x32 vs 0x76).
> Some of the other keys have either or both errors too - will have to log
> the errors associated with each key and see if I can find it in the logic.
12:43 p.m.
It does get better with repetition, but never goes completely away.
The contacts are almost dry-circuit (4.7K pullup + one TTL input) so
will only clean up mechanically.
I'll try pulling a couple of key caps and see if I can carefully clean
the contacts.
On 10/6/26 14:24, Fred Cisin via cctalk wrote:
Does it get better, or WORSE, with extended use?
Sometimes, oxidized or dirty contacts clean themselves a little with more
use, and are especially bad with non-use.
40? years ago, I was handed a "dead" TRS80 keyboard. After repeatedly
beating on one key, it gave a hit! And, would work a small part of the
time. I tried with others, and determined that extended excessively heavy
use seemed to re-enable keys, at least sporadially!
So, I dug out a Rochester Dynatyper Keyboard Actuator. It was like the
KGS80, a box of solenoids intended to be set on top of a typewriter
keyboard to turn a typewriter into a cheap printer (back when "cheap"
computer printers were a thousand dollars). Those would "print"
relatively reliably at about 8 characters per second. There also existed
the Escon (Walnut Creek) Model E-P, which was the same idea but Selectric
specific, and mounted inside the Selectric, and pulled the keys down; that
could almost handle the 14.8 characters per second maximum of Selectrics.
I set the Rochester on top of the keyboard, and sent enormous text files
to "print". After a while some of the keys were working! By the next
morning, they all worked!
Before trying that, consider that if the problem is mechanical, it might
help those keys, or it might wear them out, and finish killing them.
--
Grumpy Ol' Fred cisin(a)xenosoft.com
On Tue, 6 Oct 2026, Charles Morris via cctalk wrote:
> I'll check the p/s, but since everything else works correctly, I suspect
> dirty key contacts and a poor debouncing circuit.
> Pressing (for example) the "2" key, sometimes it outputs "22",
> sometimes "2v"... the debounce circuit should take care of the first,
> and the second has falsely set bits (0x32 vs 0x76).
> Some of the other keys have either or both errors too - will have to log
> the errors associated with each key and see if I can find it in the logic.
7 Oct
7 Oct
6:17 a.m.
From my personal experience with such type of contacts I have in my
ADM-3A I favor DeoxIT. They make a great syringe version of their
product. The version for pots works the best for me as it adds some
lubrication to the contacts without compromising the conductivity. It
will also be the least aggressive mechanical cleaning as the product
does the work for you as you work the key pressing. Hope this helps.
On 10/6/26 3:43 PM, Charles Morris via cctalk wrote:
It does get better with repetition, but never goes
completely away.
The contacts are almost dry-circuit (4.7K pullup + one TTL input) so
will only clean up mechanically.
I'll try pulling a couple of key caps and see if I can carefully clean
the contacts.
On 10/6/26 14:24, Fred Cisin via cctalk wrote:
> Does it get better, or WORSE, with extended use?
>
> Sometimes, oxidized or dirty contacts clean themselves a little with more
> use, and are especially bad with non-use.
>
>
> 40? years ago, I was handed a "dead" TRS80 keyboard. After repeatedly
> beating on one key, it gave a hit! And, would work a small part of the
> time. I tried with others, and determined that extended excessively heavy
> use seemed to re-enable keys, at least sporadially!
>
> So, I dug out a Rochester Dynatyper Keyboard Actuator. It was like the
> KGS80, a box of solenoids intended to be set on top of a typewriter
> keyboard to turn a typewriter into a cheap printer (back when "cheap"
> computer printers were a thousand dollars). Those would "print"
> relatively reliably at about 8 characters per second. There also existed
> the Escon (Walnut Creek) Model E-P, which was the same idea but Selectric
> specific, and mounted inside the Selectric, and pulled the keys down; that
> could almost handle the 14.8 characters per second maximum of Selectrics.
>
> I set the Rochester on top of the keyboard, and sent enormous text files
> to "print". After a while some of the keys were working! By the next
> morning, they all worked!
>
>
> Before trying that, consider that if the problem is mechanical, it might
> help those keys, or it might wear them out, and finish killing them.
>
> --
> Grumpy Ol' Fred cisin(a)xenosoft.com
>
>
> On Tue, 6 Oct 2026, Charles Morris via cctalk wrote:
>
>> I'll check the p/s, but since everything else works correctly, I suspect
>> dirty key contacts and a poor debouncing circuit.
>> Pressing (for example) the "2" key, sometimes it outputs
"22",
>> sometimes "2v"... the debounce circuit should take care of the first,
>> and the second has falsely set bits (0x32 vs 0x76).
>> Some of the other keys have either or both errors too - will have to log
>> the errors associated with each key and see if I can find it in the logic.
12 Oct
12 Oct
9:58 a.m.
Update: Each time I had a key that bounced, I pulled the key cap off and
sprayed a very small amount of DeOxit D5 on the contacts.
After actuating it about 20 times, no more trouble :)
The other issue turned out to be self-inflicted! As I mentioned, the
keyboard has (at most) 2-key rollover and I type fast.
When hooked up to simulated TSS/8 at 110 baud for the TTY experience, it
was easy for me to overrun and cause a misfire.
A real 33 ASR (I have two) won't let the typist go too fast because the
next key is physically blocked. At 1200 baud I can type normally,
although I still have to pay attention to fully depress and release each
key.
20
days inactive
26
days old
6 comments
4 participants
participants (4)
-
Charles Morris -
Fred Cisin -
Richard Loken -
Wayne Foux