-   -   WPM to BPS calculation (https://www.radiobanter.com/moderated/170948-wpm-bps-calculation.html)

 Klystron March 18th 08 02:44 PM

WPM to BPS calculation

I am trying to convert "words per minute" into "bits per second."
Bits per second, in turn, is APPROXIMATELY equal to baud, a common
measure of modem (or other means of data transmission) speed. I need to
quantify one factor: How many letters are in a "word?" If we assume that
there are 5 (five) letters to a word, my calculations look like this:

WPM = 50
LPM = WPM * 5 # letters per minute
BPM = LPM * 8 # bits per minute
BPS = BPM / 60 # bits per second
BPS = 33.33

I have assumed 8 bits to the byte, which is quite generous
considering that Morse cannot encode an 8 bit character set or, for that
matter, the full ASCII character set, which is only 7 bit.
Can anyone see any obvious errors? Is 50 words per minute really

--
Klystron

 [email protected] March 18th 08 03:51 PM

WPM to BPS calculation

Here is how Morse speed is usually calculated:

If the text is typical plain-language English, the test word "PARIS"
is used. WPM is the number of times PARIS can be sent in 1 minute,
using proper spacing between dits and dahs, letters, and words.

It turns out that the word PARIS and one word space equals exactly 50
"dit times", with a dit time being the length of time the key is
closed for a dit. (A dah is three dit times, the spaces between dits
and dahs inside a character are one dit time, the spaces between
letters are three dit times and the spaces between words are seven dit
times.)

So if the word PARIS is sent 50 times in 1 minute, that minute is
divided into 2500 dit times. Which is 41.66 bps.

The reason for the difference is that there are so many different
timing issues in Morse Code. The elements, characters and spaces are
all different lengths, with the most-common characters (like the
letter E) being the shortest.

73 de Jim, N2EY

 [email protected] March 18th 08 11:45 PM

WPM to BPS calculation

On Mar 18, 3:30 pm, Bill Horne wrote:
wrote:
Here is how Morse speed is usually calculated:

It turns out that the word PARIS and one word space equals exactly 50
"dit times", with a dit time being the length of time the key is
closed for a dit.

So if the word PARIS is sent 50 times in 1 minute, that minute is
divided into 2500 dit times. Which is 41.66 bps.

I'm sure your explanation is correct, but it leaves me confused: I know
bps baud, but they're close, and the Model 15 Teletype I used to own
operated at 45 baud. It seems illogical that Morse would be so high in
the bps count.

The difference has to do with how the coding is done. The following is
all from memory:

60 wpm Morse works out to 3000 bits per minute or 50 bits per second
using the "PARIS" formula.

Your 45 baud Model 15 Teletype was in all probability what hams called
a "60 wpm 5-level Baudot" machine. We had similar machines at the
University. (In this post I use the term "Baudot" to mean the 5-level
TTY code US hams used for many years until FCC allowed us other codes
like ASCII in the early 1980s)

"Baudot" takes 7 bits to send a character: one start bit, five data
bits, one stop bit. A space between words is a character, so to send
the word "PARIS" would take six characters including the space
character. That's only 42 bits, rather than the 50 bits that Morse
requires. Thus the difference - the Baudot machine uses 16% less bits
to send the same message. The speed difference works out to about 10%
because the Baudot stop bit was longer than the others in the machines
US hams typically used.

So you don't get the full 16% advantage that you'd expect from the raw
numbers. But since only six of the 42 bits are stop bits, the
difference is small.

To make it even more of a sporting course, the above WPM advantage of
the Baudot machine is message-dependent, same as for Morse. In Morse,
the message-dependency comes from the different characters being of
different length; a five-letter word like "TENET" takes a lot less
time to send than one like "JUICY", while in Baudot they both take the
same time to send.

But in the Baudot code the numbers and some other characters are sent
by shifting from "LTRS" to "FIGS", (letters to figures), so sending
mixed groups could take a lot of extra characters that Morse does not
require.

For example, in Morse you could just send the group "6A8G7" as 5
characters, but to send it on a Baudot machine you had to send
"figs6ltrsAfigs8ltrsGfigs7", which is 10 characters.

So the WPM are really approximations, and the BPS/baud measures took
over.

73 de Jim, N2EY

 Jim Haynes[_5_] March 18th 08 11:47 PM

WPM to BPS calculation

In article ,
Bill Horne wrote:
I'm sure your explanation is correct, but it leaves me confused: I know
bps baud, but they're close, and the Model 15 Teletype I used to own
operated at 45 baud. It seems illogical that Morse would be so high in
the bps count.

Your Model 15 Teletype at the nominal 60 wpm speed, which is actually
368 chars/minute and 45.45 baud works out like this. The character
length is 7.42 bits long (for ancient, interesting reasons I won't go
into right now) and the bit duration is 22 milliseconds. The character
duration is therefore 7.42 * 22 = 163.24 milliseconds, and that works out
to 6.12595 characters/sec = 367.55 characters/minute. To convert that
to words you have to figure 6 characters per word because the space
between words is also a character. So the speed is actually 61.26
words/minute.

Teletype speed is sometimes confusing because there are a couple of
other speeds out there. Western Union liked to use a 7.00 unit
character rather than 7.42. With 45.45 baud, or 22 ms pulses, this
gives 154 milliseconds/character, or 6.49 characters/second, 389.6
character/min and hence 65.9 words/minute. This is completely
compatible with 7.42 unit code because the baud rate is 45.45 for
both. But then there is European 50 baud Telex using a 7.5 unit
code. This is a 20 millisecond bit for a character length of 150
milliseconds, 6.67 characters/second, 402 chars/minute, 67 words
per minute. This is not compatible with the other two codes because
the baud rate is different; but if you say something like "66 wpm"
you could be talking about either scheme.

Now when you get to ASCII, the old Teletype machines transmitted 8
data bits per character and used an 11.0 unit code. This makes 100
wpm work out to 110 baud. Electronic terminals don't need 11 unit
code; they can do just fine with 10. Thus the words-per-minute is
numerically equal to the baud rate. 100 baud - 10 ms/bit -
100 ms/char - 10 chars/sec - 600 chars/min - 100 wpm.

Morse has already been explained. A Morse dot is actually two bits,
since there is the dot followed by the space that makes it distinguishable
from what comes next. A Morse dash is four bits, counting the space,
and the word space is three dot times or 6 bit times. Then the
word PARIS contains 50 bit times counting the space. So one word
per second is 50 bits per second and 60 wpm. As an aside, the
military sends a lot of encrypted 5-letter code groups, so instead
of PARIS the Signal Corps uses CODEZ as a test word more statistically
correct for their kind of traffic. And CODEZ contains 60 bits.

 Phil Kane March 23rd 08 03:03 AM

WPM to BPS calculation

On Sat, 22 Mar 2008 14:04:16 EDT, Klystron wrote:

It just seems inconsistent with the way
that so many hams have fought tooth and nail to hold onto Morse and to
hinder the move toward digital modes.

The joy of Morse is not the speed at which data is transferred but the
means of transferring. A good Morseist (mot me....) doesn't need a
computer or software to decode it.

And I know several Morseists who not only use "high speed data modes'
of those modes.

Morse is for fun.
--

73 de K2ASP - Phil Kane

From a Clearing in the Silicon Forest

Beaverton (Washington County) Oregon

e-mail: k2asp [at] arrl [dot] net

 [email protected] March 23rd 08 03:13 AM

WPM to BPS calculation

On Mar 22, 1:04�pm, Klystron wrote:
wrote:
So if the word PARIS is sent 50 times in 1 minute, that minute is
divided into 2500 dit times. Which is 41.66 bps.

� �It still seems like an awfully slow data rate.

Compared to what? And for what application, in what bandwidth?

If you have a pile of data to send, or a picture, etc., 41.66 bps is
quite slow.

But for a real-time conversation, 41.66 bps isn't all that slow. The
average person doesn't talk or type at a sustained speed much faster
than 100 wpm. 50 wpm isn't that much slower.

I have seen people
throw 14400 baud modems in the garbage because they
considered them to
be so slow as to be worthless.

11 years ago, when I first went online, it was with a 56k modem. I
gave up on dialup modems several years ago and went broadband. I don't
think anybody who has a choice is still using dialup.

But that's because the options exist, with no significant downsides. A
14400 modem uses the same phone line as a 56K modem. DSL can be run on
the same phone line and not tie it up for telephone calls.

Operating on the limited bandwidth amd high variability of the HF
amateur bands is a completely different thing.

A data rate of 42 bps is about 3 orders
of magnitude slower than that. It just seems inconsistent
with the way
that so many hams have fought tooth and nail to hold
onto Morse and to
hinder the move toward digital modes.

A lot of hams like Morse Code and use it on the air. It has a lot
of advantages. Why should they give it up?

And how has "the move toward digital modes" been hindered by hams?

73 de Jim, N2EY

 Paul W. Schleck[_3_] March 23rd 08 04:03 PM

WPM to BPS calculation

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

In Klystron writes:

wrote:

[...]
So if the word PARIS is sent 50 times in 1 minute, that minute is
divided into 2500 dit times. Which is 41.66 bps.
[...]

It still seems like an awfully slow data rate. I have seen people
throw 14400 Baud modems in the garbage because they considered them to
be so slow as to be worthless. A data rate of 42 bps is about 3 orders
of magnitude slower than that.

Many types of communications vary over many orders of magnitude of
information rate, yet are considered useful and up-to-date.

For example, the Casio WaveCeptor on my wrist:

http://www.eham.net/reviews/detail/2497

receives a ~ 1 Baud Pulse Position Modulated (PPM) signal from radio
station WWVB in Fort Collins, Colorado, which transmits on 60 kHz. It
takes about a minute to send the complete time code to synchronize my
watch. Slow? Yes. Useful? Yes, very much so, especially when
considering the coverage and reliability that can be obtained from such
a low-bandwidth, groundwave-propagated, Very Low Frequency (VLF) signal.
The watch only needs to receive the time code at most once per day,
which it does so automatically in the early hours of the morning sitting
on my desk or dresser. A faster data rate would require something other
than a VLF signal, and would not improve much on the quality or
usability of the communications. It would definitely increase the
price. Witness the much greater success in the marketplace of
WWVB-based watches versus more advanced, higher bandwidth, but much more
expensive, "Smart Personal Object Technology" (SPOT) watches:

http://www.spotstop.com

One of the most current and widely used communications technologies
among young people is cellular telephone text messaging:

http://en.wikipedia.org/wiki/Text_messaging

(sometimes also called "Short Messaging System" or SMS)

According to this recent demonstration on the Tonight Show with Jay
Leno:

the realizable data rates are comparable in order of magnitude to that
of fast Morse code that can be sent and received by human operators.
Just try telling a teenager with an SMS-capable cellular telephone that
it should be thrown in the trash because it isn't fast enough, or isn't
of sufficiently novel technology, and see his or her reaction.

To give you an amateur radio example, the Automated Position Reporting
System (APRS):

http://www.aprs.org

uses 1200 Baud AFSK packet. Faster, but still an order of magnitude
slower than technologies you imply should be thrown out. Since APRS
reports important, but compact, telemetry at periodic intervals, the
technology meets the requirements of many users utilizing VHF radios and
Terminal Node Controllers (TNC's). Again, substituting much higher data
rates would really not improve the technology or better meet the
requirements of the users which it serves.

To even give you a Morse code example, consider the simplicity and
effectiveness of the NCDXF beacons running on the HF bands:

http://www.ncdxf.org/beacons.html

A relatively low data rate On-Off Keyed (OOK) Morse Code signal is able
to quickly convey to the listener the quality of the communications
link, and required link budget, to various points around the globe. All
that is needed to be transmitted is a station identification, and the
same symbol (in this case the letter "V") sent at 10 dB power steps from
100 Watts to 0.1 Watt. Complex modulation/demodulation equipment to
achieve "orders of magnitude" faster data rates would not only not fit
on the HF bands, they would not seem to offer much improvement in the
quality of the service.

I suppose one could implement a beacon network using something like
PSK31:

http://www.psk31.com/

which might even be able to demonstrate realizable communications link
budgets below 0.1 Watt. But even that advanced digital mode would only
have data rates comparable to Morse code. Though the NCDXF beacon
network is a Morse code service, note that Morse code knowledge is
really not necessary to utilize it effectively. A synchronized time
base and a chart of which station transmits at which time would enable
very fast determination of the link budget to the beacon locations. If
you can't remember what a "V" sounds like in Morse Code (". . . _" like
the intro to Beethoven's Fifth Symphony), I suppose you could put that
on the chart as well. After all, the use of similar charts are how
pilots usually decode the Morse code identifications of aeronautical
beacons.

There are even a number of excellent software packages linked from the
NCDXF site above that could automatically monitor the signals, decode
the Morse, and record the quality of the communications paths over time.
One such package is Faros:

http://www.dxatlas.com/Faros/

one of many advanced signal processing software packages for amateur
radio that exploits the ubiquitousness of of inexpensive personal
computers with sound cards in most home ham "shacks."

Focusing simply on information rate disregards other aspects of the
communications and the channel over which it is transmitted. These
important aspects include the bandwidth and propagation characteristics
of the available channel, the complexity of the required transmitting
and receiving equipment, the amount of data that needs to be
transmitted, and how quickly and often it needs to be conveyed.

Single-attribute measuring contests may be fun, even ego-boosting to
some, but are really not very useful or impressive to those who actually
design and use practical communications systems.

It just seems inconsistent with the way
that so many hams have fought tooth and nail to hold onto Morse and to
hinder the move toward digital modes.

I'm not sure that I understand your line of reasoning here. You are
implying cause-and-effect. In other words, use and advocacy of Morse
code somehow directly contributed to the obstruction of other
technologies. Can you give direct evidence of specific examples? If
you are implying that licensing requirements obstructed the development
of advanced digital modes, that really doesn't appear to be the case.
Witness the success of Tucson Amateur Packet Radio (TAPR):

http://www.tapr.org

and the Radio Amateur Satellite Corporation:

http://www.amsat.org

which have developed or championed many promising digital technologies,
developed by amateurs with widely varying degrees of Morse code
operating skills.

Furthermore, if the only technologies that you believe should be saved
from being thrown away are those at 14.4 kBaud and up, those
technologies are only practically realizable on amateur radio bands at
high VHF and up. Such bands have been open to licensees without need of
a Morse code test for going on 17 years now. Even before then, these
bands were accessible to Technician-class amateurs since at least
shortly after World War II, with a license that only required a minimal,
5 WPM (essentially individual character-recognition) Morse code test.

If you are saying that someone *else* should have developed these
technologies (other than you, of course), and that since they haven't,
then someone *must* be to blame, well, you can't really dictate how the
world should turn out without taking an active role to help make it that
way.

--
Klystron

- --
73, Paul W. Schleck, K3FU

http://www.novia.net/~pschleck/
Finger for PGP Public Key

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (SunOS)

iD8DBQFH5epg6Pj0az779o4RAvfbAJ4kewTvCX5mqHimGwfXkK tQCusKFwCgxKPZ
ovhE2D69Thi8oiiqsv5I9X8=
=4RMi
-----END PGP SIGNATURE-----

 Steve Bonine March 23rd 08 05:21 PM

WPM to BPS calculation

Phil Kane wrote:

Morse is for fun.

Indeed, this says it all.

73, Steve KB9X

All times are GMT +1. The time now is 01:16 AM.