Command Control Standard Status Update
Stan R. Ames
was presented last week in Columbus. What follows are my
reactions to the weeks events and a summary of what I have
recommended that the committee do next. A distribution list is
being formed so that people can follow this effort more closely.
Additional help is solicited. If you are interested in being on
the committee or just want to follow the discussions please
Email me your desire and address. This is especially important
if you have a EE bent or have strong connections to one of the
scale specific societies such as NTrack. Summaries will
continue to be posted on any significant events.
In general the clinics were will attended and well received. We
went many hours into the evening, well past our allotted time.
Suggested improvements included more bits for functions and
some attention paid to backward compatibility to existing
command control systems.
Rutger Friberg provided an inspiring clinic that demonstrated
what one can do with digital control. Rutger brought many of
his latest implementations that left many in the audience
amazed.
There was some negative comments made outside the clinics
that implied that this was a Marklin driven effort (which is not
true) and several were incensed that Marklin was giving a
clinic. A slur of the National Marklin Railroad Association was
passed around. It should be pointed out that all manufactures
that desired to present a clinic were given space. Clinics by
Badger and CTC16000 come to mind of other manufacture
clinics that were presented.
It is unfortunate that no progress was made in soliciting help
from some of the CTC vendors. Attitudes such as "standards
inhibit creativity" and "my system works the way I want it to
and if you want to base a standard on it fine but I am not open
to any changes" were expressed.
I would say that the results were a cautious wait and see
attitude, which considering the amount of information we were
providing was somewhat of a success.
Some additional observations
At the Marklin booth their was much interest and curiosity but
not much strong interest in buying.
At the CTC 16000 booth the serious model railroaders
converged and spent many hours looking at the latest CTC
16000 command station.
The point of this is that good system implementations compete.
Talk of standards at lower levels is interesting but is not a
substitute for good system implementations. The Marklin
command control stations do not currently fill the needs for
many of the larger layouts (who happen to be the biggest and
most vocal users of the current command control systems) and
therefore until command stations are marketed that have
features that compete with the latest systems that do not
adhere to the standard, and more US Model Railroaders become
users, the standards process is not likely to be totally accepted.
What's Next
We are in the process of setting some goals for the next year.
Following is a strawman. It is ambitious but I think achievable.
Any comments on this would be greatly appreciated.
1) Fix the technical issues in the standard, the radio
interference problem is the most important
2) Generate a second packet format that has more speed
control, addresses and better functions.
3) Achieve some form of CTC backwards compatibility.
4) Provide Dick Andrews a complete standard and
recommended practices proposal that can be presented to the
NMRA board.
5) Work to have at least 2 US Manufactures have product to
demonstrate at the 1993 convention
6) Encourage Lenz to complete and offer for sale his US style
command station (that would provide 4 different manufactures
demonstrating compatibility at the 1993 convention)
7) Generate a simple build it yourself command station.
8) Demonstrate the standard at one layout on the 1993 layout
tour.
9) Demonstrate multiple manufactures equipment and creative
examples of what you can do on a small Module layout at the
1993 convention
10) Complete the second source for the chip design.
11) Generate one article for a Trade Magazine.
12) Obtain support from the various gauge specific societies
and the European MRR society.
13) Prepare a clinic that shows some of the creative things you
can do with the standard. It is important that this clinic be
manufacture independent. Have committee members present
this clinic at regional and gauge specific conventions over the
next year.
Stan Ames
Elmer Beachley
last week, I'd like to add a few comments to Stan's article:
I attended the Thursday night clinic and greatly enjoyed Rutger Friberg's
presentation. Clearly some of his "toys" exceed technically any of the
products on display at the Great Train Show. I'm refering particularly to
his on-board analog sound devices and some of the animated figures in his
models (all digitally controlled, of course.)
From a technical perspective, several of his slides toward the end of the
talk included schematic diagrams of the on-board decoders, with the Lenz
chip as just a box, though. Rutger has written a book completely documenting
the Marklin system and is looking for a publisher so apparently we will be
seeing some much needed technical info on this subject.
It did seem strange though that not one of the few Marklin dealers at the
Show had ANY digital equipment for sale. (Excuse me if this is wrong but
I went through the show very thoroughly.) Here is a proposed "standard"
at a national show and no one is even selling parts!
Now I like the Marklin system. It appears to me to be the most technically
advanced, though I've not extensively studied them all. But I think that
before many will pay attention to these products, much as to be done to
provide technical and basic how-to information in easy to obtain fashion.
This does appear to be coming.
The second issue is cost. In Europe, where this system has been around
since the mid-eighties and where sales are apparently "high", the little
decoders cost the equivalent of $100 US (they cost ~$170 here). Even
with a open system, multiple sources and competition, will the demand
for this type of stuff every drive the prices low enough to meet the goal
of a $50 decoder?8!`0*hLf3f3f3f3f3f3f3f3f3f3f3f3? Hard to say.
Anyway, congratulations to Stan for putting together a very enjoyable
presentation. I learned quite a bit.
Elmer Beachley e...@unix.cis.pitt.edu e...@cs.pitt.edu
Dennis H Lippert
>passed around. It should be pointed out that all manufactures
>that desired to present a clinic were given space. Clinics by
>Badger and CTC16000 come to mind of other manufacture
>clinics that were presented.
Unfortunately, I could not attend the convention... but the way I see it,
this is basically a waste of time. All you're doing is showing the
different versions on the market right now, and each mfr. is going to
point out their advantages (i.e. OnBoard using a DC carrier... errant
locos just run away, rather than suffer damage on an AC-carrier system).
>At the Marklin booth their was much interest and curiosity but
>not much strong interest in buying.
>
>At the CTC 16000 booth the serious model railroaders
>converged and spent many hours looking at the latest CTC
>16000 command station.
Neither of these facts surprises me in the least... I've always avoided
Marklin's booth myself (the system looks strange to me, and the last I
saw it, it was too expensive for the market). Meanwhile the CTC 16000
(a beast I've never heard of... not been following this thread till now)
sounds like something I'd look into out of curiosity. The name implies
(possibly wrongly) a compatibility with CTC-16 and it's derivatives
(CTC-32, I believe, and the Mann-Made system).
>3) Achieve some form of CTC backwards compatibility.
With what? Just CTC-16? What about Onboard, DynaTrol, and the others?
>5) Work to have at least 2 US Manufactures have product to
>demonstrate at the 1993 convention
What manufacturers? Knowing this industry the way I do, you have to
go to a Marklin-sized company to find anything more than a few-man
operation. You want a small one-product company to invest in this
new system (which competes against their old system, IMHO) "just to
see if it'll fly"? I don't think *I'd* take that kind of risk...
especially since the system will be open to me to make fairly easily once
the standard is adopted.
>7) Generate a simple build it yourself command station.
Absolutely necessary.
>13) Prepare a clinic that shows some of the creative things you
>can do with the standard. It is important that this clinic be
>manufacture independent. Have committee members present
>this clinic at regional and gauge specific conventions over the
>next year.
IMHO, the standard isn't strict enough to be called a standard if
it's loose enough to "be creative".
As I said before, I've stayed away from this discussion, mainly because
I will not be in the market for a system for at least a few years, and
because I had no idea that things were this far along. The current
manufacturers have invested a lot of time and money in their products.
This shows up in some systems. The only way I can see a standard is
if it's backward-compatible with almost anything on the market... and
I think that's a technical impossibility. My second preference is for
a CTC-16 compatible system.... because it's the one the do-it-yourselfers
have been using for many years (14 of 'em, in fact).
-Den -owner of no command-control, user of many-
--
===============================================================================
Dennis Lippert - mac...@pitt.edu or mac...@vm2.cis.pitt.edu
The "Mac Maniac" operator at the Hillman Research Lab - Univ. of Pittsburgh
"All comments are mine, you'd look like a fool if *you* used them"
Michael Greene
I too was one of the attendees at the presentations in Columbus. Here's
a few of my comments/observations:
1) I attended the:
CTC 16000 clinic on Tues evening
Marklin Digital all day clinic on Wednesday
NMRA Command Control Standard Update on Thursday evening
2) Many thanks to Stan, Rutger, Tom Catherall (Marklin), Keith
and the others who were involved in preparation of & delivery
of the various clinics. It was no small undertaking to get
all the demos, etc put together. And, in my view, it takes
"educational" sessions like these (regardless of the bent to
a particular manufacturer) to help attendees like myself
develop a sense of the wide range of issues involved
in defining such a standard. And it will take more clinics
like these, plus magazine articles, videos, etc to build a
comfort level.
3) Poor scheduling (probably unavoidable) meant that if I wanted to hear
the CTC 16000 clinic, I could not attend the NMRA Statndard update
clinic before the Marklin Digital clinic. This was unfortunate,
but thanks to Stan's posting of his paper before leaving for
Columbus, I was able to read it before going to Columbus.
4) A standard in this space is clearly essential if we are to leverage
the buying volume to drive prices down and have the interoperability
we desire between various manufacturer's equipment.
5) We need to think about presenting command control in different
scenarios - trying to cover the breadth of applications
of command control from small personal layouts all the way
through large club layouts in a single session may make
it diffcult for people to understand and see pros and cons.
And it certainly makes discussions "entertaining" and confusing
given that different people are looking from different
perspectives. Perhaps having different sessions and demonstrations
for each of these areas in the future would be helpful.
6) We need to investigate other media for widespread delivery of
command control "educational" materials. For example
perhaps a short video on the subject, some of the issues,
and some demos. This might be a way to get more people up
to speed. Or perhaps a small handbook called "Introduction
to Command Control in Model Railroading" that would survey the
field and help readers understand the subject.
7) On the technical side, I'm pleased to see the definition of a second
packet format on the proposed list of goals for next year. While I
suspect the proposed packet format would meet requirements in the
short term, it left me with the feeling that it was destined to
become the "compaibility mode" or "lowest common denominator"
format. In my opinion it just doesn't allow enough room
for expansion & creativity on the part of manufacturers and users.
As a N scale modeler, I was pleased by the significant amount of attention
given to the area of command control in Columbus. However I was disappointed
by the high prices of the products available in the US at this time for
use in N scale. And the incompatible products. But this can be fixed!
Thanks Stan et. al. for your persistence in this area.
stewart l dejournett
>saw it, it was too expensive for the market). Meanwhile the CTC 16000
>(a beast I've never heard of... not been following this thread till now)
>sounds like something I'd look into out of curiosity. The name implies
>(possibly wrongly) a compatibility with CTC-16 and it's derivatives
>(CTC-32, I believe, and the Mann-Made system).
>a CTC-16 compatible system.... because it's the one the do-it-yourselfers
>have been using for many years (14 of 'em, in fact).
The CTC-16000 is Keith Gutierrez's (father of CTC-16) new system. The new
system retains compatibility with old receiver's a plus in my book but allows
more channels and does some error correcting to give better performance on
dirty track.
stewart
Bruce Perens
writes:
> with a open system, multiple sources and competition, will the demand
> for this type of stuff every drive the prices low enough to meet the
> goal of a $50 decoder?
Gosh, I don't see why not. Let's look at the hardware we are talking about
here:
First, a simple embedded-controller microprocessor. Say a 4-bit controller
with 512 bytes to 1K internal PROM. These cost about $2 in quantity.
Then a few bytes of non-volatile RAM for the ID and throttle curve data.
Either EAROM, or battery-backed RAM. Say $2 for the RAM and $2 for the
battery.
Then we have a zener diode (for the maintenance-mode-detect feature) and
the power supply rectifier and filter capacitor. Say $4 .
Then there's the power transistor to drive the motor (we'll use the
microprocessor to pulse-modulate it, so it won't run hot). $1 .
And we need a double-sided PC board. $5.
$1 for copying the instructions, and you have a $14 parts cost for a kit.
Mark it up to $20 to cover shipping and other expenses, and any model
club could distribute them.
Now, do this on a commercial scale. Use surface-mount components and
a microprocessor with built-in EAROM, and it's going to be cheaper than
that. Pot the whole thing to protect it from prying fingers, and you
should still be able to make a profit at $30.
Now, if you can find me a microprocessor with built-in EAROM that fits
in an 18-pin DIP or smaller, I can get this to fit in a Z-gauge tender.
I can do all of the software work and most of the hardware. Anyone want
to help?
Bruce Perens
Bruce Hoover,Armstrong
> In article <46...@blue.cis.pitt.edu.UUCP> eb...@pitt.edu (Elmer Beachley)
> writes:
>
>> with a open system, multiple sources and competition, will the demand
>> for this type of stuff every drive the prices low enough to meet the
>> goal of a $50 decoder?
>
> Gosh, I don't see why not. Let's look at the hardware we are talking about
> here:
>
Yeah, but if you add up the component price of the Marklin system, you
sure don't get $1500 or whatever it takes to get started with that
either. :-)
Bruce Hoover
Bill Brown
>In article <46...@blue.cis.pitt.edu.UUCP> eb...@pitt.edu (Elmer Beachley)
>writes:
>
>> with a open system, multiple sources and competition, will the demand
>> for this type of stuff every drive the prices low enough to meet the
>> goal of a $50 decoder?
>
>Gosh, I don't see why not. Let's look at the hardware we are talking about
>here:
>
>First, a simple embedded-controller microprocessor. Say a 4-bit controller
>with 512 bytes to 1K internal PROM. These cost about $2 in quantity.
I think an 8-bit, like one of the Motorola 68HC05 series might be a better
fit - if it had an a/d one might be able to "close the loop" on the motor.
A bit more spendy though - maybe $5 in qty. An a/d might also be nice
for detecting maintenance mode.
Your also going to need a little bit of hardware in the front end - need
to be sure that noise doesn't get mistaken for a (real) transition.
>Then a few bytes of non-volatile RAM for the ID and throttle curve data.
>Either EAROM, or battery-backed RAM. Say $2 for the RAM and $2 for the
>battery.
There are 1-chip micros with NVRAM and eprom on-board, including several
68HC05xxx flavors.
>Then we have a zener diode (for the maintenance-mode-detect feature) and
>the power supply rectifier and filter capacitor. Say $4 .
>
>Then there's the power transistor to drive the motor (we'll use the
>microprocessor to pulse-modulate it, so it won't run hot). $1 .
One obvious omission is that with only 1 transistor you've no way to
reverse the motor. A motor-driver chip (on the order of $3 or so)
is a more reasonable choice. In any case, it takes the equivilent of
several transistors and a couple of diodes. If one wanted to be fancy
and do an analog drive it would add a few bucks.
>
>And we need a double-sided PC board. $5.
Probably reasonable, but to get this price is gonna require a fairly
large order. Ties up capital.
>
>$1 for copying the instructions, and you have a $14 parts cost for a kit.
>Mark it up to $20 to cover shipping and other expenses, and any model
>club could distribute them.
>
>Now, do this on a commercial scale. Use surface-mount components and
>a microprocessor with built-in EAROM, and it's going to be cheaper than
>that. Pot the whole thing to protect it from prying fingers, and you
>should still be able to make a profit at $30.
I think your estimate is a bit on the low-ball side. As a general rule,
the cost to produce an item can be assummed to be about twice the cost
of the components. One problem here is that the "run sizes" are likely
to be too small for automated assembly, unless you can afford to tie up
a whole bunch of money for a long time. Which adds to the cost.
I see Marklin 2-rail digital decoders, which I assume are what you are
proposing to duplicate, advertised for about $70. Generally, the
distributor gets a mark-up of 20% of the retail cost; the retailer gets
another 40%. The retail price is therefore 2.5 times the wholesale
price, which must include the manufacturing cost, including the manufacturers
profit. It may be possible to beat this, but it's not gonna be easy.
Unless you consider your time and effort, and the use of capital, to
be worthless.
One can "cut out the middleman" by attempting to sell by direct mail order,
but there are significant costs invovlved in that route as well.
>Now, if you can find me a microprocessor with built-in EAROM that fits
>in an 18-pin DIP or smaller, I can get this to fit in a Z-gauge tender.
>I can do all of the software work and most of the hardware. Anyone want
>to help?
You bet! It looks like you're located in Richmond. We should probably
move this to e-mail or telephone.
Disclaimer: These opinions are my own and have |
nothing to do with the official policy or the | -bill
management of L.B.L, who probably couldn't | wlb...@lbl.gov
care less about employees who play with trains. |
Stan R. Ames
>In article <1992Aug12.1...@pixar.com> Br...@Pixar.com (Bruce Perens) writes:
>>In article <46...@blue.cis.pitt.edu.UUCP> eb...@pitt.edu (Elmer Beachley)
>>writes:
>>
>>> with a open system, multiple sources and competition, will the demand
>>> for this type of stuff every drive the prices low enough to meet the
>>> goal of a $50 decoder?
Seeing as the thread of this discussion relates to cost of receivers I thought
that I would pass on some numbers for those who are interested. Lenz
Electronics sells the C82 decoder to Marklin for about $28. This includes
his profit and all cost of materials.
Now before someone jumps the gun consider that Marklin contracts this in lots
of over 10,000. Before the end user sees this it goes through many hands and
each hand adds cost. In the US where the market is still small, the cost of
marketing and user support adds considerably to the cost. These same units
cost much less in Europe where the support costs are shared amonst many more
users.
Now consider the costs of the small Arnold receiver. This unit fits nicely
into "Z" scale but because of its 1 amp capacity it is also very useful in N
and HO scales. It is basically the same receiver as its HO cousin but it
uses a different production method. While the larger receiver uses surface
mount technology the smaller receiver uses chip on board technology where the
actual chips are mounted and then the entire package is encased in epoxy.
Now it turns out that Lenz does not manufacture this receiver. Instead he
sends it to someone else for assembly, which adds more cost.
How do we get high quality receivers at low costs?
The solution is more volume and standard components that anyone can use.
However more volume does not come until they cost less.......
One solution is to manufacture the receivers in the US, where most of the
chips come from anyway. If someone wher to manufacture these receivers
here they could sell the smallest receiver for around $40 and still make a good
profit. Now before someone writes that that cannot happen, Lenz while he was
here offerred exectly that. To date no one has excepted his offer. So
if anyone out there has the interest to manufacture the chip on board receivers
contact Lenz Electronics.
Now this is completely outside the scope or interest of the NMRA committee that
I am on but I thought I would pass it on so as to put more reality into
the thread of this discussion.
Stan Ames
Robert Turner
>>in an 18-pin DIP or smaller, I can get this to fit in a Z-gauge tender.
>>I can do all of the software work and most of the hardware. Anyone want
>>to help?
>
>You bet! It looks like you're located in Richmond. We should probably
>move this to e-mail or telephone.
>
Did you say software on the WONDERFULL Motorola microprocessors? Count me
in as well.
Word of warning: Author has much to gain by increasing the use of Motorola
microprocessors (microcontrollers).
Robert
--
Robert Turner (602) 897-5441 Semiconductor Systems Design Technology, Motorola
tur...@ssdt-tempe.sps.mot.com OR ...!uunet!dover!turner
GOD is real, unless declared integer. --J. Allan Toogood
Elmer Beachley
Bruce:
I think your assessment of the parts cost for the decoder is correct
with the exception of one point: they don't use an off-the-shelf (i.e.,
high volume = inexpensive) chip for control. Instead, the chip (which]
admittedly I know very little about) is a custom, specialized one
the incorporates the eeprom, control logic, etc. So far, there are
maybe one or two sources for this chip, though there is an effort to
independently develop a comparable chip here in the U.S.
From what I know about the economics of the semiconductor business, you
need volumes in the 6 figure range before the costs become reasonable.
I question whether this volume will ever be available in model railroading.
In Europe, where these things have been in use since the mid-eighties,
and where it would appear there was significant market volume, the
decoders still list for around $100 US.
Most of the competing command-control systems on the market appear to
be designed around off-the-shelf components, though I'm sure Stan,
who has studied this much more than I, will correct me if this is not
true.
Don't misunderstand; I'd like to see a Marklin-equivalent system
succeed. I'm just not sure the economies of scale are there to support
it.
Elmer Beachley
Elmer Beachley
Bruce:
It is more like $500-$600 to get started with Marklin, somewhat dependent
on the type of decoder. If you study the system, you really only need
two of their boxes, the 6027 and the 6050. Of course, you have to
control your trains from a serial line (or buy software which I'm not
sure about the costs). This is assuming purchase from U.S. Marklin
dealers at better than 30% discount, which, if you spend some time on the
phone, is not hard to do.
Elmer Beachley
Chris Webster
>
>The CTC-16000 is Keith Gutierrez's (father of CTC-16) new system. The new
>system retains compatibility with old receiver's a plus in my book but allows
>more channels and does some error correcting to give better performance on
>dirty track.
>
>stewart
Is it a commercial system, or is it going to be something you can build
yourself, as the original CTC-16 system was?
stewart l dejournett
>Is it a commercial system, or is it going to be something you can build
>yourself, as the original CTC-16 system was?
>
as he did with the sound system.
From a custom chip point of view for the Lenz system Keith would be the best
point of contact within TI since he is in ASIC products but I believe there
would be a conflict there between his system and Lenz
stewart
Stan R. Ames
>
Please note that while the Lenz receivers use a custom VLSI chip, none is
required. It is possible to build receivers without this chip but the
receivers will be larger. At least one US manufactures believes that they
can adapt a US chip and make a HO version of the receiver that while larger
than the Lenz design will be fully functionsl. Alas to fit into N or Z
guages a custom design is needed.
>From what I know about the economics of the semiconductor business, you
>need volumes in the 6 figure range before the costs become reasonable.
Actually the numbers are in the low 4 digit range to make the cost tradeoff
reasonable. By the time you make it to 6 digits (as they have in Eurpope)
the volumes allow you to make significant design improvements (ie different
versions of the chips) without effecting the manufacturer cost significantly.
>I question whether this volume will ever be available in model railroading.
>
>In Europe, where these things have been in use since the mid-eighties,
>and where it would appear there was significant market volume, the
>decoders still list for around $100 US.
>
I do not have the numbers with me but they were posted on the net by someone
it Eurpoe earlier this year. Generally the Europe prices are about 50% of
the US cost for Marklin equipment. I seem to remember a C-82 decoder somewhere
in the $60 range. We have been told that by October the Lenz C-100 decoder
(same size, better performance) will sell in the US for around $48.
>Most of the competing command-control systems on the market appear to
>be designed around off-the-shelf components, though I'm sure Stan,
>who has studied this much more than I, will correct me if this is not
>true.
>
actually there is much custom VLSI in use today and 1993 will see a great
deal more. The new CTC 16000 chip is custom, the onboard sound chip is
custom and the new onboard digital system uses a custom chip. Once you
design it it costs 15K to produce the first versions.
>Don't misunderstand; I'd like to see a Marklin-equivalent system
>succeed. I'm just not sure the economies of scale are there to support
>it.
>
Just to clear up one misconception. The proposed NMRA standard is not a
Marklin-equivalent system. The Marklin system implements a good subset of
the standard. The command stations offered by Marklin while of very good
quality may or may not be useful to many US modelers due to the operation
philosophy. The standard only covers what goes over the rails. Calling it a
Marklin like system scares off both users and more importantly new
manufactures. To be technically correct the standard was based upon a system
designed by Bernd Lenz. Numerous changes have been made and more are comming.
To date Bernd Lenz has incorporated our changes. Lenz Electronics makes
equipment for Marklin, Arnold, SBmodelbau, Roco, and Lenz.
We live in a world market. If our requirements are different we should have
different systems. Where we can take advantage of existing markets we can
greatly reduce cost. I have seen the numbers while visiting Lenz and although
optomistic a end user cost of $25 for a HO scale receiver is not out of line
by 1994 or 1995. I think that you will see receivers that implement subsets
of the standard in the lowwer $40 range by spring.
I also believe that we will see at least two competing but COMPATABLE systems
by mid 1993. If we can both reduce cost and achieve multiple compatable
manufactures than I will consider the efforts worthwhile.
Stan
Herb Poppe
gre...@edisto.mro4.dec.com (Michael Greene) writes:
> I too was one of the attendees at the presentations in Columbus. Here's
> a few of my comments/observations:
>
> 1) I attended the:
> CTC 16000 clinic on Tues evening
Can you or anyone else comment on what the on-track signal looks like for
CTC16000?
Did Keith say whether technical info is available from his company?
Did he mention when Model Railroader is going to publish an article on his
system?
Is he shipping receivers, command stations, etc?
Herb Poppe National Center for Atmospheric Research (NCAR)
hpo...@ncar.ucar.edu 1850 Table Mesa Dr.
(303) 497-1296 Boulder, CO 80303
Herb Poppe
Perens) writes:
> In article <46...@blue.cis.pitt.edu.UUCP> eb...@pitt.edu (Elmer Beachley)
> writes:
>
> > with a open system, multiple sources and competition, will the demand
> > for this type of stuff every drive the prices low enough to meet the
> > goal of a $50 decoder?
>
> Gosh, I don't see why not. Let's look at the hardware we are talking
about
> here:
>
> First, a simple embedded-controller microprocessor...
>...
> Now, if you can find me a microprocessor with built-in EAROM that fits
> in an 18-pin DIP or smaller, I can get this to fit in a Z-gauge tender.
It was suggested elsewhere that one of the Motorola 68HC05 might fit the
bill. Unfortunately, the low pin count devices (the K1 and L2(?)) don't
have the same timer system that the high pin count devices have. The
latter timer system has very nice capabilities for measuring the timing of
an input waveform (the CC on-track signal) and simultaneously generating a
PWM output signal (motor speed). I found the timer system in the K1 and
L2(?) disappointing. As I recall, none of the 68HC05 devices have on-board
EEPROM (unlike many members of the 68HC11 family). Also, I don't think any
68HC05 devices have Schmidt trigger inputs (which might be all that is
necessary to clean up the on-track signal - after it is level-shifted).
I haven't yet found the "perfect" micro for this application (I've had
this in mind for several years now), but it might be worth taking a look
at the National Semi COP8722C (a member of the COP800C family). This is a
20-pin, 8-bit device. It has 64 byte RAM, 64 byte data EEPROM and 1K byte
program EEPROM. (Its sibling, the 8622, has 1K byte masked ROM.) It has 4
Schmidt trigger inputs. It has a 1us instruction cycle time. The timer is
a little weak (it can measure an input waveform or generate a PWM output
waveform but not both at the same time). However, the timer could be used
to generate a 50us (say) software counter which could be used to measure
the input waveform (using edge-triggered interrupts from the on-track
input signal) and generate the output PWM (every 50 us the software would
check if it was time to toggle the PWM output - higher overhead that using
the timer directly to interrupt the processor when the next output
transition is desired). The 1K program EEPROM makes program development
easier (for us low budget types) than other approaches. I don't know if
the 8722 is available in a SMT package. Another feature of this chip is the
serial MICROWIRE PLUS interface which would allow it to interface to
additional micros - a digital sound system, for example.
Another useful IC would be a motor driver chip. I have used a 16-pin
Sprague/Allegro device (forget the part number at the moment) for this
purpose. It is an H-bridge with current-limiting and over-temperature
shutdown. Not available in SMT to my knowledge. A smaller pin-count
package would be desirable as well.
Stan R. Ames
>CTC16000?
>
same as in CTC 16 but with shorter time slots.
>Did Keith say whether technical info is available from his company?
>
Product information is available from CTC Products (Keith's company)
>Did he mention when Model Railroader is going to publish an article on his
>system?
>
I some how doubt that Model Railroader will print a new article but
I may be mistaken
>Is he shipping receivers, command stations, etc?
>
He is shipping new command stations, and upgrades to existing command stations
Shipments of receivers are scheduled for late fall. He has not actually
invested the money to manufacture the new chip.
If you are looking for CTC compatability or all you want to do is
operate trains (speed and direction) or you must have a good walkaround
system today than I think that Keith's system is now at the top of ones list.
If you do not fall into one of those catagories I might suggest waiting as
Keller will come out with a digital version of OnBoard, Keeler will come
out with a digital version of CTC 80 Lenz will come out with a new command
station and in the background is the proposed standards effort. This next
year promishes to be interesting. Some fallout will likely occur in the
not to distant future.
Stan Ames
Michael Greene
In article <1992Aug13....@newstand.syr.edu>, cweb...@rodan.acs.syr.edu (Chris Webster) writes:
|>Newsgroups: rec.models.railroad
|>Path: ryn.mro4.dec.com!nntpd.lkg.dec.com!pa.dec.com!decwrl!uunet!psinntp!newstand.syr.edu!rodan.acs.syr.edu!cwebster
|>From: cweb...@rodan.acs.syr.edu (Chris Webster)
|>Subject: Re: Command Control Standard Status Update
|>Message-ID: <1992Aug13....@newstand.syr.edu>
|>Organization: Syracuse University, Syracuse, NY
|>References: <1992Aug11.1...@linus.mitre.org> <46...@blue.cis.pitt.edu.UUCP> <1992Aug12.1...@mksol.dseg.ti.com>
|>Date: Thu, 13 Aug 92 12:08:40 EDT
|>Lines: 12
From the presentation I saw in Columbus, the components will
be available in both kit form and "RTR" form from Keith & company.
I have a copy of the 1 page product sheet with estimated prices, but
unfortunately it is not here in the office.
Michael
Michael Stimac
>
>If you do not fall into one of those catagories I might suggest waiting as
>Keller will come out with a digital version of OnBoard, Keeler will come
But don't bother waiting if you're not in H0. Keller has no plans,
currently, for bringing out 0-scale capable receivers. I've no
idea if the digital receivers will be small enough for N-scale.
Stan?
>out with a digital version of CTC 80 Lenz will come out with a new command
>station and in the background is the proposed standards effort. This next
>year promishes to be interesting. Some fallout will likely occur in the
>not to distant future.
The so-far unique sound capability of Keller stuff makes it
a product that has no competition, in some ways. Thus, the competition
is not "pure". In any case, having a multi-thousand dollar investment
in Keller products, I'm not about to change in a casual fashion.
I suspect that many of the modelers who will have command control
already have an installation of some kind, and conversion is a
major issue, not just how good the newer systems are.
Michael Stimac
--
Michael Stimac
(415) 355-8889
ro...@tymnet.com
These opinions are not necessarily anyone's but my own.
stewart l dejournett
>Can you or anyone else comment on what the on-track signal looks like for
>CTC16000?
>
>Did Keith say whether technical info is available from his company?
>
>Did he mention when Model Railroader is going to publish an article on his
>system?
>
>Is he shipping receivers, command stations, etc?
Keith believes in having technical info available so I'm sure he will
provide it. An MR article usually takes about 9 months from beginning of
writing to publishing. I know this since I proffed his sound article material.
I'm sure Keith will be shipping soon i.e. before the article. In a previous
post I stated you can use existing CTC reciever's but I also know he has got
his new receivers ready. Before Colombus Keith told me he would feed me
some info to post on the net since he does not have access. I will try and
get all I can, technical info as well.
stewart de journett
Texas Instruments
stewart l dejournett
>I haven't yet found the "perfect" micro for this application (I've had
>this in mind for several years now), but it might be worth taking a look
Herb
You might not find the perfect micro but might need to go to an ASIC. Most
of the silicon compilers these days have libraries of processor cores that
would be usable. A masked ROM also costs $$ to get set up so maybe the $$
for an ASIC are comparable. You can do some ASIC work on PC's these days as
well so a high end workstation is not always needed. Just a thought
stewart
Chris Webster
>
>The so-far unique sound capability of Keller stuff makes it
>a product that has no competition, in some ways. Thus, the competition
>is not "pure". In any case, having a multi-thousand dollar investment
>in Keller products, I'm not about to change in a casual fashion.
>I suspect that many of the modelers who will have command control
>already have an installation of some kind, and conversion is a
>major issue, not just how good the newer systems are.
>
>Michael Stimac
>
scale and prototype only; no tinplate, large scale layout) a few days
ago. The other customer (it was a slow day) asked about the Onboard
system's sound quality. The dealer --one whom sells a lot of these
systems -- commented that he doesn't use it and doesn't plan to. When
he was pushed farther about the sound, he said "if anything, it sounds
like a GMC bus!" His point was that it didn't sound like the diesels
he models nor does it have any capabilites that allow him to change the
sound on it."
I guess the point is that the "advantage" of it having sound isn't if the
customer isn't satisfied by the sound.
Personally, I've only got a small HO shelf layout, and am using conventional
cab control. Being young (21) and planning to be around a long time, I'm not
going to invest in a system that's going to be dead when I'm still young
(under 40!). I suspect most of my equipment will still be usable and
satisfying me then because of the effort I'm putting into it now, such as
superdetailing and tuning my motive power.
Jeff Mizener
my favourite train store owner about the Lenz system and it's pricing.
Given that the store is in Berlin (Vogelbusch Eisenbahnen), I'll give the
prices in dm and then we can convert.
To get started you need a DigitalPlus central unit, a power transformer
(not critical, any 16vac 57VA transformer will do), a hand controller and a
programmer track (notmal piece of track connected to the programmer outpur
on the central unit -- used for programming loco addresses.)
Cost: Central unit dm300 (~$200)
hand controller dm180 (~$120)
receivers dm 95 (~$63) each for N, 79 for HO
-----
dm575 (~$383) to run one engine.
This assumes a 1.5 exchange rate, something that is, right now, highly
optimistic.
Add to this:
RS232 interface dm 69
rcvr for 4 turnouts dm 80
'feedback' unit dm 69 (TO position, block occupation, etc).
I don't think that $63 for one of these N-babies is particularly cheap,
considering that I'm going to need quite a few of them. But they are
beautiful and tiny. And in stock.
Oh yes, the system has 64 speed steps, and has five extra function outputs
per locomotive and is compatible with Maerklin digital and Arnold.
Now all I need to do is to convince all my N-Trak friends to plop
down big bucks for this system and we can remain compatible. But
until that happens...
Cheers,
Jeff
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Jeff Mizener / Siemens Energy and Automation / Intelligent SwitchGear Systems
/ Raleigh, NC 919-365-2551 / j...@sead.siemens.com