View Single Post
  #29   Report Post  
Old November 18th 17, 05:20 AM posted to uk.radio.amateur,rec.radio.amateur.homebrew
rickman rickman is offline
external usenet poster
 
First recorded activity by RadioBanter: Nov 2012
Posts: 989
Default Morse Key Contacts?

Jerry Stuckle wrote on 11/17/2017 11:00 PM:
On 11/17/2017 6:57 PM, rickman wrote:
Jerry Stuckle wrote on 11/17/2017 3:23 PM:
On 11/17/2017 3:37 AM, rickman wrote:
Jerry Stuckle wrote on 11/16/2017 10:34 PM:
On 11/16/2017 9:17 PM, rickman wrote:
Jerry Stuckle wrote on 11/16/2017 9:02 PM:
On 11/16/2017 7:28 PM, rickman wrote:
Gareth's Downstairs Computer wrote on 11/16/2017 4:01 PM:
On 16/11/2017 20:04, rickman wrote:
Gareth's Downstairs Computer wrote on 11/16/2017 12:55 PM:
On 16/11/2017 17:07, rickman wrote:
Bob Wilson wrote on 11/7/2017 9:47 PM:
On 11/4/2017 5:42 AM, Gareth's Kitchen Komputer wrote:
On Mon, 21 Dec 2015 23:33:17 +0000, gareth wrote:

For those who make their own Morse Keys, what do you use for the
contacts,
for
I have found the phosphor bronze pillar for the dot contact
on my
ersatz
Vibroplex
to be very noisy and scratchy?

I'm fairly sure that I asked this question before, but it has
been a
habit
of never put off till
tomorrow what you can put off till the day after.

I chose phosphor bronze because in the days of making one's own
electronic
organs, phosphor bronze was suggested as a suitable keying
matreial
for
the
keyboards.






If a question's worth asking, Gareth...







I have not made lots of keys, but when I do I use contacts from
old
relays.
I still have some relays I bought as WWII surplus when I was in my
teens,
back in the 1950's. They have strange coil parameters, weird
mounting
arrangements, etc., but nice little silver contacts.
Bob W, WA9D

Why use a mechanical switch? There are magnetic hall effect
switches
around. Or a mercury wetted switch can be operated by a magnet.
Are the
magnetic switches too sloppy?


Perhaps you confuse mercury wetted with reed relays?

I am thinking of reed relays, so I guess I used the wrong name, but
aren't
they wetted with mercury to prevent corrosion building up?


In the closed atmosphere of the glass tube, there should not be any
corrosion, I think. ISTR some gas in there, but don't know about
mercury wetting.

Call it what you want, corrosion, oxidation, just plain pitting. The
point is a solid surface will wear from the arcing that happens when
contacts break connections (which also happens on initiating connection
because of switch bounce). Mercury doesn't have this problem as it
is a
liquid and reforms the layer every time it is "pitted".


Mercury also does not make a "clean" break. It "puddles" as the
contacts
are pulled apart due to surface tension. It leads to irregular break
timing. This isn't a problem at low switching speeds as you have in
regular
switches and relays, but can be at higher rates as in CW. Additionally a
magnetic field for switching accentuates this problem. Reed relays are
good
for things like security system contacts, but not a CW key.

I haven't timed such switches. What is the timing precision requirement
for CW? Does optical satisfy it?


Never tried optical but as long as you can get a clean make/break of the
beam it should work. The problem as i see it would be that clean
make/break; the beam will fade in and out. Probably not as bad as a reed
switch - but for high speed CW you need clean operation.
That's why many keys use brass contacts or similar; the have a clean
make/break. And even if they get pitted they are soft enough to burnish
rather easily and thick enough to last for years.

So no numbers? Let's try to make some.

According to Tom - W4BQF, "All high speed code (above about 55 wpm) is
sent with a keyboard/keyer or a computer keyboard". So we can use 55 WPM
as an upper limit for using a manual key.

Dot length = 1.2 seconds / WPM = 1.2 / 55 ~= 22 ms

Switch bounce time for many switches is some single digit ms. So I'm not
sure what "clean" make/break means. The bounce time for reed switches is
about the same as other switches I found.


No, it is not bounce time - "clean" is how quickly it makes/breaks. Brass
contacts are virtually instantaneous. However, reed switches operate via a
magnetic field. This field is not a simple make/break - it gets stronger
and weaker depending on the distance from the magnet. IOW, brass contacts
are digital but the magnet field is analog. The actual switching does not
necessarily always occur at the same magnetic field strength. Additionally,
switching off to on requires a stronger magnetic field than the release.
The two combined can result in varying width pulses at high speed. Optical
is also analog and can have similar problems.

And before you dismiss all of this - why don't you see any paddles with reed
switches or optical? If these methods are so good, why aren't they in the
high-end paddles/keys?


Whether or not reed switches are used in keys is not related to your
understanding of electronics.


It is. That is the context in which they were brought up. But I know you'd
rather change the context than admit you are wrong.


There is nothing going on here except that you are mistaken about the
functioning of switches. How about you address the issues rather than
diverting the discussion.


Reed switches are *not* analog any more than mechanical switches in that
they don't vary the connection continuously. They are still metallic
switches and make or break when the metals touch. The fact that the
magnetic field varies continuously is no different from the pressure from
the paddle varying the position of the mechanical switch contact
"continuously". I found no information indicating significant variations
in the action of reed switches. The variation in pull-in and release
field strength gives a small amount of hysteresis which is desirable in
any switch. It is not enough to distort the key times from one character
to the next.


Reed switches are not - but the magnet field which triggers them are. And
there is a huge difference between the magnetic field and finger pressure on
a paddle. That's exactly why reed switches are not used on paddles.


"There is a huge difference"... but you can't say what that difference is.
The motion of the key paddle is analog which results in an analog
displacement of the mechanical switch until contact is made or in the case
of a reed switch, results in the analog change in field strength until the
threshold is reached where the switch pieces are pulled together. All
analog until the switch pieces make contact which results in a discrete
output change.

The real difference is in the bounce time. A mercury wetted reed switch has
no bounce time, contact is singular and certain. A mechanical switch has a
variable resistance until the point of contact stops moving and settles
down. That's the bounce time.


Likewise optical is only optical in the movement of the paddle and the
signal seen by the detector. Once the signal passes through a threshold
detector with hysteresis the signal is purely digital. The strong suit of
optical is that it eliminates all mechanical issues of wear and failure.


The same is true for optical - except that optical can also be affected by
ambient light, making the switch even less reliable.


Lol. It is easy enough to exclude ambient light.


I did a bit of reading about reed switches yesterday and they have much
less bounce time than other switches and the mercury wetted types have
virtually no bounce time. In fact while looking for info on typical
bounce times one of the pages I found showed a rather elaborate circuit to
debounce the two mechanical switches in a keyer. If the switches were so
instantaneous, why would they need a debounce circuit?


Once again it isn't the bounce that causes the problem.


Bounce is a problem that makes the timing of the switch closure uncertain
and must be eliminated. Compensating for the uncertainly can't be done. So
what is the problem in mechanical switches if not bounce? If a switch
bounces for 5 or 10 ms, that is a significant portion of time for a 22 ms dot.


Your analysis above shows a complete lack of understanding of how switches
in general work and not just reed switches. Try reading about switches a
bit. You will quickly find that metallic switches do not make or break
cleanly with an "instantaneous" connection. Just use google or any other
tool to find some info on this and read it.


Nope, my analysis is right on. And it is exactly why reed and optical
switching is not used in keys and paddles.


You actually haven't analyzed anything. You just keep repeating the same
unsupported conclusions ignoring the way mechanical switches operate.

--

Rick C

Viewed the eclipse at Wintercrest Farms,
on the centerline of totality since 1998