I really want to understand this.
How to distinguish between a DC current that reverses frequency many times per second and AC current?

 

I know a fair amount about digital in general, but not the specific implimentation with slot digital.

But I understand scalextric uses PCM - pulse code modulation. (there are a few methods of encoding digital signals) Usually, this method involves turning the current on and off, not reversing it, ie switching between (say) +12V and 0 V, or it could be implimented to switch between +12V and +10 volts. Most of the digital signals I have come across operate in the region of 48KHz or more, 48000 on and offs per second instead of 50 on AC. rates can go up to megaHz though, and each hi voltage represents a binary 1 and each low voltage represents a zero. each 'package' or set of numbers will probably include an address or car number, so the car the rest of the instructions belong to will know to pay attention to the rest of the contents.

Since a scalextric rail with lots of joints isnt the ideal conductor for high frequency square waves, the frequency might be a lot lower, dont know. But the higher the frequency, the easier it should be to filter out for the motor.

Hope this is clear, and not further obfusciating the issue

 

Beppe thanks very much for starting this - much appreciated!

Let's see if I have any kind of handle on it so far.
I see both AC and chopped DC as 'fluctuating currents' for want of a better term.
But better check - is this a reasonable term?

Would it then be correct to say that if the current never drops below zero, it is chopped DC, but if it does, then it is AC?
eg
+18v to 0v = chopped DC
+18v to -18v = AC

I must get the hang of this!

EDITED
For the benefit of anyone trying to follow this thread - THAT IS WRONG!
I think I can handle being wrong as long as I learn something in the process!

 

Er... , electronics is not my area of competence at all, so I do recommend you look up that Australian site (I don't know why the link did not come out active....)

As a former electrical engineer, I can say that AC (single phase or 3 phase) is classically considered to be a current that varies following a sine wave, at a fixed frequency - so I wouldn't consider the DCC shape as AC

The one thing I've grasped is that DCC was adopted by model trains because the environment is electrically very "noisy" - so that if you classically superimpose the control signal to power it will be drowned out - with DCC, signal and power coincide

Hope this helps a bit.... at the moment, I'm just too excited from astro's revelation of Scalex having used "my" method for LC activation !

Beppe

 

chopped is slang, the term is 'pulse code modulated'. Modulation is usually the adding of one signal to another, and what is added to the dc is 'pulses'. In theory, the digital modulated signal will be a square wave with distinct changes of state from hi to low voltage, at irregular coded times.

What is usually ment by ac is a continuous regular sign wave, or oscillation, always changing voltage between +12V and -12V (for example). Main reason we have AC mains is that for technical reasons, less electricity is lost down the extremely long distances between power stations and ppls houses.

 

According to the NMRA, Digital Command Control uses pulse-width-modulated bipolar DC for both the command signals and power. This means that the voltage ranges from a positive (of some level) to a negative (of the same level). Think of it as square wave AC if you like. However, the duration of the square waves (pulses) is not constant - it is the width of these pulses that are interpreted as binary signals by the decoders. To actually power motors or other devices, the input voltage is rectified and the decoder determines the amount of that rectified DC that should be fed to the device.

Probably clear as mud, right?

 

The mud is beginning to clear a BIT!
I know it's too much to ask for a degree course in electronics on a Slot Forum, but you guys are helping, and I, for one, appreciate it.

 

thanks fergy, thats clearer

 

Just to clarify a bit further (at the risk of boring most)...

If the range of the bipolar DC is +12V to -12V (24V peak-to-peak), the rectified voltage will be 12V. Note that it doesn't matter what the pulse widths are, the rectified voltage will always be constant since the voltage that the rectifier sees will always be either +12V or -12V. A rectifier doesn't care about pulse widths or pulse duration - it just changes the + or - V into a positive voltage level. There will always be either a +12 or -12 voltage present, so the rectifier will generate a constant 12V level, subject to further notes below.

The power fed to motors, etc., will use 0V as the common or "ground" voltage and will use 12V as the maximum positive voltage.

Due to ineffencies in rectification, though, the motor will see less than the max voltage. To compensate, the DCC system will usually operate at a voltage slightly higher than the target motor voltage. For example, if 12V is the target voltage for motors, the system may actually run at something like 25.2V peak-to-peak (12.6V positive and 12.6V negative). By DCC standards, the voltage can actually be as high as 27V peak-to-peak.

This means, though, that if any slotcar system exceeds 27V peak-to-peak, or approximately 12.9V DC for motors, then it is non-standard by DCC conventions.

Not sure how this translates to Scaley's version. If they are following true DCC conventions though, they can't exceed this voltage level. To feed motors at 15V would require a peak-to-peak DCC voltage of approximately 31.2V, outside of the current DCC standards limit.

As an aside, the DCC convention is one of the reasons why digital control is not currently considered appropriate for large numbers of devices (cars) where rapid changes in instructions are relevant. Each car receives its instructions via "packets" of data sent from the DCC source. Each decoder recognizes it's own set of instructions. Thus, there are a finite number of "data packets" that can be sent each second and this information cannot overlap. Each car sees instructions at intervals, not constantly. It's entirely possible that if a large number of cars are involved, that a given car may not see every input that its driver initiates. Consider how fast a driver can change trigger position - it would be desirable to send a packet to the decoder for ANY change in trigger position. If too many cars are on the system though, some of those inputs may not get sent. Believe it or not, the bandwidth of DCC has practical limitations. It works well for trains, but slot cars depend on much faster, and more frequent, changes in inputs.

Ooops... getting carried away, I fear....

 

If anyone is bored, they don't have to read it - this is good stuff!
No Fergy - you are not getting carried away at all - this is a very valuable resource and I do follow most of it. I was just getting thrown by the AC/DC business, mainly because everything I have seen on DCC until recently clearly said it was AC current. Maybe it was thought easier for our less electronically educated minds to grasp. Actually it WAS until you guys started to straighten me out. Well, I THINK you are straightening me out!

 

sorry for saying it wasn't AC, tropi, looks like i was essentially wrong!

 

Errr . . . I THINK you were right, Astro!

In any case, if you hadn't said anything, we wouldn't have this topic now and it's a good one. So ->

 

Astro, is ISN'T AC!! You were right - it's DC, but it is DC that rapidly switches from positive to negative. Many don't realize that DC can have a negative value. Unlike AC's sine wave form though, bipolar DC switches almost instantly from max positive value to max negative value, and vice versa - it's a "square" wave that AC does not have. Also, AC has a constant repetitive wave form, usually 50Hz (UK and Europe, for example) or 60Hz (North America, for example), but the wave form of DCC signals can vary drastically - the width of the DC pulse can vary radically (theoretically from millionths of a second to more than a second) while AC forms have a constant width (50 Hz or 60 Hz).

 

I'm glad you stepped in there Fergy - I thought I had understood properly but wasn't confident enough to say any more than I did! I've got it, by George I've got it!
But am not surprised that it was confusing.

Is it possible that other systems HAVE used genuine sinusoidal AC?
(That sounds as though it wants to kill itself!)
My thinking here is in connection with full blown AC power supplies, 220-240v AC in UK, that have simultaneously used the same power cables to network computers.

 

digital signals can be modulated onto ac, prob not as DCC or PCM (there r a lot of modulation schema), and i believe on the mains systems they use very high frequencies approaching RF. For this to be legal, u have to filter ur mains very thoroughly, cos its illegal tho very easy to contaminate the grid with ur modulated signals. My understanding is that the modulated frequency has to be a good order of magnitude greater than the carrier frequency (50hz in the case of mains), and my suspicion is that due to scaly rails, joints and braid contact, the frequency for this must be relatively low

 

Tropi, I'm not aware of any serious modern system that has used AC to communicate data. The problem is that AC cycles from max to min voltage with a sinusoidal wave form. Theoretically, it is possible to alter the frequency of the AC wave form, but it is more efficient to do that with the square-wave of DC. Particularly with today's digital electronic systems - by definition, digital is on or off, not slowly changing from one to the other and therefore yielding states of 'maybe' .

Digital components react to changes in voltage levels or peak voltage levels, and changes must happen very quickly to avoid erroneous interpretations of the signals - sine waves are very imprecise and lead to digital interpretations like "yes, no, yes, no, maybe, no, yes, no, yes" during the transitions. DC signals, however, change state rapidly, from one extreme to the other, and thus are compatible and friendly with digital circuitry. It's "yes" or "no" and there is no other possibility.

Writing this as Astro replied, so: yes, other signals can be added to either an AC or DC power form, but not on the AC line voltage, because, as you said, it will be fed to other mains users. The main voltage must be stepped down for our use anyway, and the motors are DC, so any attempt to use AC is pretty pointless anyway. The actual average DC frequency (DCC's PWM is imprecise as to actual frequency) can be quite a bit higher than the line frequency, but is still limited by many factors - as you say, track connections, braid contact, etc. And don't forget that a "packet" of info includes an "address" for each car, along with a "function request" (such as speed control), and a "value" (what speed?), etc. It's complex stuff! But DC is the only efficent way to address it.

 

Here is one interesting link
Digital Power Line Communications

 

Tropi, any particular aspect of that site that we should look to?

Seems like lots of people patting themselves on the back, but I can't really find out what they're talking about.... maybe I'm lazy...

 

Yep you're lazy!

Internet via power line
The connection is that the system transmits digi info via an AC supply.
It's 1 am here and I am crawling off to bed - will probably have digital dreams.
Do androids dream of electric sheep?

 

Okay, sorry, didn't have the stamina initially to look.... yeah, the AC line has been used before - in my last house, it was used for intercom units - we could talk or call from several points in the house - cool but not particularly secure. This was done by overlaying a signal on the AC that could only be detected by decoding units - it was out of phase from the standard AC and very low-level so it didn't affect standard AC devices. Still, for a system as complex as digital slot cars (worse actually than an internet connection because of the real-time urgency), and problems from external and internal noise sources, it just isn't as ideal as a complete DC system that is totally isolated from the mains voltage.

...and I don't want to know about UK electric sheep...

 

I'm glad the thread is now in much more capable hands
Just two notes :

- I gleaned that the legal frequency limit to avoid EMI is 10 kHz - this comes from a Finn, Julius Wilkko a.k.a. lordjw, who as a matter of fact would be a very nice addition to SF's talent pool

- as for the maximum number of cars compatible with little response delay under DCC, I'm a great believer in the bumblebee theory (I suppose you know it) : since Davic runs well with 15 cars, I'm not worrying

Beppe

 

Yes, Julius would be very welcome here - the more talent the better (especially better for the untalented like me)
He just seemed to disappear from SCI in mid-Nov-03.
I have messaged him via the EZBoard system and hope that he eventually takes a visit here.

 

Hi guys,
Fergy and the others do a very good job explaining digital slot tracks in a few lines.
But it seems you are not aware of some very fine sources on the internet about the subject:
At the university of Queensland some students wrote a thesis about digital slot racing end of 2003. It is all there for download as pdf files.
For example:
http://innovexpo.itee.uq.edu.au/2003/exhib...4106/thesis.pdf

There are more of them if you go to the following page :

http://innovexpo.itee.uq.edu.au/2003/exhib...byresearch.html

in order to find the documents on this long page do a search on the page (CTRL-F in IE) and look for 'slot car' or 'digital'. it will take you to the links that point to the pdf's.

The documents are +/-60 pages each but you do not have to read everything.

Happy reading.

 

Hi Xlot, others

I finally registered myself to this forum. Nice to see people still discussing about DCC. If Scalextric system is cheap enough it can be that nobody bothers to build DCC though. It will be very interesting to try n' reverse engineer Scalex digital, but my first impression is that people who have designed the system really mean business. It may contain ASIC chips with heavy signal processing.

Julius

PS.
Xlot, our first RF link works in lab. It will hopefully be installed to a slotcar shortly.

 

GREAT to hear from you again Julius !!

Look, we need to talk about a number of issues, which I beleve will lead to the much desired open source/ NMRA environment (you listening, Thomas ?)

The starting point is that Scalextric have adopted the LED emissor principle for lane changing. It's totally immaterial whether they saw my proposal or if they thought of it concurrently, what matters is that I believe the concept is not patenteable, and they have proved that it works - so this can be adopted universally
(OK, my only doubt is that there may be an even simpler method - Tropi was hinting darkly at ball-less, wireless optical mouse technology)

The next point is signal coding - it's not as essential, personally I'd be happy to live with a Scalextric monopoly - still, it would be nice if people could make aftermarket decoder chips
Here, I'm absolutely out of my depth - since you already mentioned it, could you elaborate? Oh, and are you saying that Scalextric's might not be DCC as we know it ?

Another point is lap counting : until now, nobody has been able to tell me how SCX and Scalextric make it work - any clue ?

Finally, how do you see your RF link alternative in this context ?

Ciao
Beppe