[Jasen Betts]

And that's basically Manchester 2 encoding.

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

Not really, manchester is basically "clock xor data",
or for differential manchester "clock xor m(data)"
where m is a memory function

John's scheme needs a bit more bandwidth, and isn't balanced
but decoding is simpler, johns scheme is basically
"ramp < data" where ramp is a ramp signal synchronised to the clock
that starts below data=0 and ends above data=1

 

[Bill Sloman]

Not really, manchester is basically "clock xor data",
or for differential manchester "clock xor m(data)"
where m is a memory function

John's scheme needs a bit more bandwidth, and isn't balanced
but decoding is simpler,  johns scheme is basically
"ramp < data" where ramp is a ramp signal synchronised to the clock
that starts below data=0 and ends above data=1

Except that the simplest way of decoding Manchester 2 pretty much uses
John's scheme. He re-invented the constant cross-section roller when
he should have re-invented the wheel.

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

 

[Allan Herriman]

I don't see how they can be equidistant and not at 256Mhz

For a design with a single clock, that clock must run at 256MHz to get
the required 3.9ns resolution.
A designer using an FPGA would probably choose this method.

Now consider a design with a 16MHz clock feeding a delay line with
sixteen 3.9ns taps. A multiplexer can select any of the taps, which
gives the required 3.9ns resolution. The highest frequency in the design
is only 16MHz though.
A designer of an ASIC would probably choose this method.

An ASIC allows the delays of the taps to be matched reasonably well. A
phase detector can compare the input and 16-tap delayed clocks and this
can be used to trim the total delay to be exactly 1 period of the input
clock.


For more information, try searching for Delay Locked Loop.

(As an aside, if I was doing this in an FPGA I would use one of the
onboard 5GHz transceivers to get 200 ps PWM resolution.)

Regards,
Allan

 

[Bill Sloman]

No, I had in mind long/short pulses, not a constant bit rate.
Long/shorts can convey a clock and data on a single wire, and receive
separation takes just an RC. Manchester takes a bit more work to
receive.

Sure. But there are a lot of single wire serial communication
protocols, and re-inventing Morse code isn't exactly something to
boast about.

 

[Jon Kirwan]

For a design with a single clock, that clock must run at 256MHz to get
the required 3.9ns resolution.
A designer using an FPGA would probably choose this method.

Now consider a design with a 16MHz clock feeding a delay line with
sixteen 3.9ns taps. A multiplexer can select any of the taps, which
gives the required 3.9ns resolution. The highest frequency in the design
is only 16MHz though.

But the MSP430 Timer D will lock tightly at a wide range of
oscillator frequencies, with up to 25MHz inputs but down, so
far as I can now tell, a few kHz. That is a huge dynamic
range of operation and fixed delay taps won't cut that
mustard.

Bill's comment about varying the current to adjust delays
(twisted ring osc, that he mentioned) might also be applied
(given my ignorance) to adjust delay taps -- but I simply
cannot buy the idea that it could do so, and do so accurately
enough, over such a wide operating range. It just sounds
insane to imagine.

Jon

 

[Bill Sloman]

But the MSP430 Timer D will lock tightly at a wide range of
oscillator frequencies, with up to 25MHz inputs but down, so
far as I can now tell, a few kHz. That is a huge dynamic
range of operation and fixed delay taps won't cut that
mustard.

Bill'scomment about varying the current to adjust delays
(twisted ring osc, that he mentioned) might also be applied
(given my ignorance) to adjust delay taps -- but I simply
cannot buy the idea that it could do so, and do so accurately
enough, over such a wide operating range. It just sounds
insane to imagine.

A ten thousand to one range - 25MHZ to 2.5kHz - would imply changing
the tail currents from 1mA to to 100nA.

That does happen to be one thing that you can do with well-matched
bipolar transistors on a single integrated circuit - 60mV off the base-
to-emitter voltage cuts the current by a factor of ten, and four
decades takes 240mV. The text-book examples show plots that cover a
rather wider current range.

<snip>

 

[Bill Sloman]

So find a (cheap!) 1-wire dpot that solves Joerg's problem.

Cheap parts aren't really my thing. I approve of them in principle,
and use them when I can, but throughout my career I designed in a lot
more parts with tightly specified high performances that weren't all
that cheap.

Quit being bitchy and do something useful.

One of your typically courteous and constructive contributions.

 

[[email protected]]

But the MSP430 Timer D will lock tightly at a wide range of
oscillator frequencies, with up to 25MHz inputs but down, so
far as I can now tell, a few kHz. That is a huge dynamic
range of operation and fixed delay taps won't cut that
mustard.

Bill's comment about varying the current to adjust delays
(twisted ring osc, that he mentioned) might also be applied
(given my ignorance) to adjust delay taps -- but I simply
cannot buy the idea that it could do so, and do so accurately
enough, over such a wide operating range. It just sounds
insane to imagine.

I just looked at the manual and it says 8MHz-16MHz for x16, 8MHz-25MHz
for x8

that an the involved describtion of how to change frequencies, I think
point
to a delay line

something like a long delayline, first pick the right tap to get a
full period
then pick the taps get close to x16 or x8, adjust as you go

afair the DCM in a spartan3 FPGA does phase shift etc., with a
delayline with 45ps
per tap

-Lasse

 

[josephkk]

But the MSP430 Timer D will lock tightly at a wide range of
oscillator frequencies, with up to 25MHz inputs but down, so
far as I can now tell, a few kHz. That is a huge dynamic
range of operation and fixed delay taps won't cut that
mustard.

Bill's comment about varying the current to adjust delays
(twisted ring osc, that he mentioned) might also be applied
(given my ignorance) to adjust delay taps -- but I simply
cannot buy the idea that it could do so, and do so accurately
enough, over such a wide operating range. It just sounds
insane to imagine.

Jon

I see your point. I'll bet that the fancy subcycle resolution does not
work below a 1 MHz basis PWM counter clock. It is a rather special mode
and undoubtedly has constraints.

 

[Jasen Betts]

No, I had in mind long/short pulses, not a constant bit rate.
Long/shorts can convey a clock and data on a single wire, and receive
separation takes just an RC. Manchester takes a bit more work to
receive.

Long/short is how 1-wire works, and probably several other things.

Sinclair Research was using long/short mark followed by long/short
space to record data on audio cassette in 1982. Mark length was one
bit, space length the next, (thus, the bit-rate was data-dependant,
the average was about 1300 bits per second obvoiously a pattern like
10101010 woulde result in a DC bias, this didn't seem to cause any
problems.)

 

[[email protected]]

Or some logic in the corner of a FPGA.

Not when John was an undergrad. It took a lot of relays, then. ;-)

 

[Bill Sloman]

On Tue, 28 Aug 2012 07:21:27 -0700, John Larkin

[snip]

I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.

"Invented?"  Bwahahahahaha!  That's like Gore claiming he invented the
Internet.  Your "invented" method can be found in ancient textbooks.

                                      ...Jim Thompson

I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.

It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

That's exactly what you were at the university to learn about, but you
went off and re-invented the wheel instead. You seem to have done just
enough academic work to pass the exams, rather than learning a bit
about the subject you had enrolled to study.

We sold hundreds
of supervisory control systems using the Manchester-with-CRC comm
protocols I defined. I designed all the control, i/o, modems, power
supplies, and color graphic display controllers (quite a novelty at
the time, when a mediocre Conrac color monitor cost $2400.) I wrote
the PDP-11 RTOS, too. All that was fun, pioneering stuff for an
undergraduate.

Re-inventing the wheel is pioneering?

I'm a systems/circuit designer, which isn't so much
about inventing pieces as it is about putting pieces together the best
way, a step or two up the abstraction stack from semiconductor
physics.

But it does help if you have some kind of deeper understanding of the
pieces you are putting together.

Got a link to the Manchester prior art that you refer to?

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

 

[Jeroen]

The 0.75T one-shot works in an FPGA, too.


Well, 74-series TTL. Ironically, it's fairly recently that relays
started getting really good, small/fast/reliable/cheap enough to use
in signal circuits instead of CMOS multiplexers.

I must have been looking the other way. What relays are that?

Jeroen Belleman

 

[Jeroen]

On Tue, 28 Aug 2012 07:21:27 -0700, John Larkin

[snip]

I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.

"Invented?" Bwahahahahaha! That's like Gore claiming he invented the
Internet. Your "invented" method can be found in ancient textbooks.

...Jim Thompson

I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.

It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

Oh nonsense! Some things are so simple that reinventing them costs
less than looking them up. OK, so it's been done before. So what?

Jeroen Belleman

 

[Joerg]

On Wed, 29 Aug 2012 09:04:23 -0700, Jim Thompson

On Tue, 28 Aug 2012 07:21:27 -0700, John Larkin
[snip]
I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.
"Invented?" Bwahahahahaha! That's like Gore claiming he invented the
Internet. Your "invented" method can be found in ancient textbooks.
...Jim Thompson
I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.

It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

Bill, think back: 35-40 years ago there was next to nothing in search
functionality, other than a massive card registry. You had to plow
through zillions of publications because the categorization in the paper
card registry was mediocre at best. To top it off you had to order whole
stacks from other libraries, only to find out that it wasn't what you
thought in there.

This produced some sad results. The worst was a Ph.D candidate at an ivy
league place who, years into his work, found out that someone in a far
away country had already researched out the very same topic. It had been
published in a foreign language that nobody at that institute spoke. In
Germany this pretty much meant "game's over", regardless of reason.

Oh nonsense! Some things are so simple that reinventing them costs
less than looking them up.

Absolutamente. That's how I also see it. You only have to watch the
patent area because that can cause things to hit the fan. People patent
the most mundane stuff these days and examiner let it pass. Which is why
I have an IP disclaimer in my standard agreement that is not negotiable.

OK, so it's been done before. So what?

Happens all the time. Almost 10 years ago I was very unhappy with the
gable diverters for metal roofs. So our roofer and I designed a new
type, on the hood of his Chevy truck. It performs much better than the
stock stuff. A few month ago I saw the exact same design from a vendor
in Europe. Couldn't believe it. But ... so what?

 

[John S]

On Aug 29, 6:30 pm, John Larkin
On Wed, 29 Aug 2012 09:04:23 -0700, Jim Thompson

On Tue, 28 Aug 2012 07:21:27 -0700, John Larkin
[snip]
I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.
"Invented?" Bwahahahahaha! That's like Gore claiming he invented the
Internet. Your "invented" method can be found in ancient textbooks.
...Jim Thompson
I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.
It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

Bill, think back: 35-40 years ago there was next to nothing in search
functionality, other than a massive card registry. You had to plow
through zillions of publications because the categorization in the paper
card registry was mediocre at best. To top it off you had to order whole
stacks from other libraries, only to find out that it wasn't what you
thought in there.

This produced some sad results. The worst was a Ph.D candidate at an ivy
league place who, years into his work, found out that someone in a far
away country had already researched out the very same topic. It had been
published in a foreign language that nobody at that institute spoke. In
Germany this pretty much meant "game's over", regardless of reason.

Oh nonsense! Some things are so simple that reinventing them costs
less than looking them up.

Absolutamente. That's how I also see it. You only have to watch the
patent area because that can cause things to hit the fan. People patent
the most mundane stuff these days and examiner let it pass. Which is why
I have an IP disclaimer in my standard agreement that is not negotiable.

OK, so it's been done before. So what?

Happens all the time. Almost 10 years ago I was very unhappy with the
gable diverters for metal roofs. So our roofer and I designed a new
type, on the hood of his Chevy truck. It performs much better than the
stock stuff. A few month ago I saw the exact same design from a vendor
in Europe. Couldn't believe it. But ... so what?

Well said Joerg. Thank you.

 

[Joerg]

On Wed, 29 Aug 2012 09:04:23 -0700, Jim Thompson

On Tue, 28 Aug 2012 07:21:27 -0700, John Larkin
[snip]
I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.
"Invented?" Bwahahahahaha! That's like Gore claiming he invented the
Internet. Your "invented" method can be found in ancient textbooks.
...Jim Thompson
I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.

It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

That's exactly what you were at the university to learn about, but you
went off and re-invented the wheel instead. You seem to have done just
enough academic work to pass the exams, rather than learning a bit
about the subject you had enrolled to study.

I actually hadn't heard the term "Manchester" when I designed the
supervisory stuff.

*WHAT* ?!

You hadn't heard of Manchester United?



[...]

 

[Bill Sloman]

21:27 -0700, John Larkin
[snip]
I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.
"Invented?"  Bwahahahahaha!  That's like Gore claiming he invented the
Internet.  Your "invented" method can be found in ancient textbooks..
                                       ...Jim Thompson
I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.

It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.

Oh nonsense! Some things are so simple that reinventing them costs
less than looking them up. OK, so it's been done before. So what?

And some things look that simple, and aren't. When you look them up
you find out about the less obvious gotcha's.

You can save money by skipping the literature search, but every now
and again it costs you a lot more than you save, sometimes just in
paying royalties to the guy who invented and patented before
inspiration struck you.

 

[Jeroen Belleman]

My scope makes occasional clicking noises as I change the vertical gain.

I assume that's because somebody decided that relays worked better
than whatever else they tried.

Well, yes! If you compare ON resistance, control signal feedthrough,
and OFF isolation with your typical MOSFET switch, relays win with
ease. If you also consider lifetime, switching speed, size and
parasitic L and C, the picture is not so rosy, alas.

Jeroen Belleman

 

[Bill Sloman]

On 2012-08-29 19:21,BillSlomanwrote:
On Aug 29, 6:30 pm, John Larkin
21:27 -0700, John Larkin
[snip]
I first invented a Manchester decoder (and encoder, and a supervisory
control system that sold megabucks) when I was an EE undergrad.
Deriving the receive clock is easy: a transition detector followed by
a one-shot of 0.75 times the bit period.
"Invented?"  Bwahahahahaha!  That's like Gore claiming he invented the
Internet.  Your "invented" method can be found in ancient textbooks.
                                       ...Jim Thompson
I didn't say I invented it first, just that I invented it. Before the
internet, it was hard to access prior art. It's not shocking that lots
of people independently discovered the same circuits.
It's a bit depressing when an undergraduate with free access to a
university library declares that he couldn't access prior art.
Oh nonsense! Some things are so simple that reinventing them costs
less than looking them up. OK, so it's been done before. So what?

And some things look that simple, and aren't. When you look them up
you find out about the less obvious gotcha's.

More often, you find a circuit in RSI or somewhere and discover that
it's fundamantally unreliable: beta dependent, bad corner cases, stuff
like that. RSI is better read for amusement than guidance.

You are telling me this? I've now published five critical comments in
RSI, pointing out where their refereeing process had failed - in 1972,
1996, 1999, 2004 and 2009 - and they rejected similar comments I
submitted in 1980 and 1995.

Nobody is saying that every academic paper is to be relied on, but
rejecting the whole of the scientific literature because some of it is
flawed rather misses the point. We make progress because some of the
papers that get published reflect dramatic steps forward.

Larsen N T 1968 Rev. Sci. Instrum. 39 1–12

http://rsi.aip.org/resource/1/rsinak/v39/i1/p1_s1?isAuthorized=no

is difficult to fault. If you'd read that more for amusement than for
guidance you'd have missed quite a bit

You make up stuff like that.

What makes you think that?

s.e.d. is populated by a lot of lurkers
who don't actually design stuff but pontificate on how it should be
done.

Perhaps. I'm not actually designing much stuff at the moment, but I've
done enough design in my time to be able to talk about the process.

You certainly pontificate about stuff that you have designed - not all
that expertly, though do you seem to be tolerably good at eventually
finding and correcting your mistakes - and on any number of other
subjects where you are much less well-informed. You would yourself
seem to be a prominent part of that population.