SlotForum banner

CarDCC hardware requirement specification

6K views 36 replies 11 participants last post by  Newbie1/32 
#1 · (Edited by Moderator)
Greetings fellow slot car racing fanatics


I've put up a draft CarDCC hardware requirement specification on a website kindly established by Randy Brown ... thank you for your kind help.


I'm hoping to get some feedback from you about what is god and what is bad, or alternative solutions that you feel would provide a better solution.

Please feel free to comment on any aspect of the proposed digital system.
Your help in making this system the ultimate slot car experience, is highly appreciated.
 
See less See more
3
#3 ·
QUOTE (drummer @ 9 Jan 2005, 06:50)This is a good start Soren, well done. Particularly exciting is a power base that could easily switch betwen analogue and digital racing.
<{POST_SNAPBACK}>Thank's a lot drummer.


If there are areas in the specification that needs a better (perhaps non technical) explanation, don't hesitate to ask.

I'm wondering if I have been a bit too technical in some of the descriptions?.


We need to get the specification done in a way that everybody can read it and form their own opinion. Not everyone is into electronics, but their slot racing experience is just as valuable to this project.


So please let me know if I need to explain parts of the specification in a different way.
 
#4 · (Edited by Moderator)
Hi there


I was thinking about an alternative way to communicate between car and lane change driver that could overcome some of the potential problems of the IR opto link used by Scalextric.

The first idea was to use the guide blade for car id coding. Here is a draft design of a CarDCC guide blade, that would allow up to 16 cars on the track. I've used the normal 18,5x5,5mm Scalextric guide blade as basis for the design, but the overall size is not important ... just the relative dimensions.

IMO there are the following advantages to this solution:

1) Analog cars doesn't need to be modified to race in analog mode, where you race one car per lane. The lap and sector timing will work just as in digital mode. The guide blade address could be ignored, and the system could use the lane address instead.

2) You need no sensor electronics in the car, you just make some slots in the guide blade.

3) The sensor output can be used for an extra LED light control instead?, like flashing headlights or the flashing lights on a safety car.

4) There's no need to drill a hole in the car body.

5) It's easy to fit the car controller in the car, because it can be placed anywhere.

6) Cars without free space behind the guide blade can be converted just as easy, like the narrow F1 cars or cars with a front motor.

7) It doesn't matter if the car is sliding, so you don't have to have a straight section before the lane change.

8) The solution has low sensitivity to stray light with a fork sensor arrangement. Outdoor racing should not be a problem.

9) The fork sensor arrangement can cover different lane spacings with the same sensors. It should be easy to cover several centimeters distance between LED and photo sensor in the fork arrangement.

10) In digital mode the system could program the address automatically by reading the guide blade address. This reduces the number of programming keystrokes on the track controller. You just switch on programming mode and drive the car across a lane change, and the system will do the rest (saves you the inconvenience of entering the car address on the track controller).

I think that this method of detections has been tried in real life ... I vaguely recall reading about a system with something along these lines somewhere?.

What do you all think about this?.


Edited:
Oops ... I forgot the best argument: It's a keep-it-simple solution, and I really like that.
 
#5 · (Edited by Moderator)
I found the review.

QUOTE (SWoRd @ 9 Jan 2005, 13:45)... I vaguely recall reading about a system with something along these lines somewhere?.It's the Carrera Pro-X system.


Looking at the pictures they have a huge guide blade with four square holes. But they also have an IR opto link ... why is that?.

Perhaps they don't use the square holes to identify the car?.
That should be easy to answer for someone with a Pro-X system ... looking at the holes in the guide blade, are they similar, or is there a distinctive hole pattern on each cars guide blade?.

The guide blades look huge, but I guess that the turn radius is larger on a Carrera track.
 
#6 ·
A bit of news (although nothing spectacular
)

I received the long awaited motor controller today.


Althoug my telephone camera is really lousy at taking close up pictures, I thought it could at least give you an idea what the prototype CarDCC controller looks like.


I'm still waiting for the four rectifying diodes, but I expect to have the first car controller assembled any day now. The embedded software is still far away though.

I've checked the overall size, and the controller will fit into the sidepod of a formula one car. I don't know if there are cars on the market with less space inside, but the dimensions of the controller are 15x23mm.
 
#7 ·
Fabulous work as always. At first the DOC document did not come up, but it did today.

Suggestions and comments detailed. Obviously this is a list of things that I feel deserve comment or could imho be done differently, it is not intended to come over negatively.

Firstly regarding the guide method of ID -

This would work elegently for me on the old scalextric blue guide system - simply snap out the guide number 7 and click in the guide for car ID 12 when needed. But changing the ID of cars with other guides can be fiddly. The new scalex type is screwed in from inside the car, others have the wires soldered to the braids or use those bullet things - I hate changing those guides cos the braids never go back the same!

also - the variety of guide types means that the system presumably would not come with a set of guides, but people would have to saw their own. I am just trying to envisage how this would work in clubs where the cars need to be recoded etc.

Sector Timing -

It is my understanding that you are planning to use the LCs to divide the track into sectors. Since at the moment people are generally thinking of evenly spacing LCs, this makes a lot of sense. However, maybe a sector track should be a seperate thing to an LC? It would make track/sector design more flexible (eg you might want to set up a couple of sector dividers so you can time the velocity along the long straight of your track)

High Voltage Low current -

I understand what you are saying and the advantages of this. Thers a few things I don't quite follow:
Say a given car in analogue goes top speed at 12 volts DC, drawing 2 amps. power used = 24 watts, and it goes at a certain speed.
Now with the high voltage system, we are sending the car 24 volts, but only allowing it 1 amp (again 24 watts) Is the chip inside the car supplying the motor with 12 volts 2amps, or 24 volts 1 amp?
If the chip is supplying 24 volts but limiting the current, then the car motor will be running in a current starved situation. Even though it is recieving the same power, will it behave the same? and what about at 22 volts? Has this method of motor control with current starvation been tried in the analogue (or digital) environment before it is implimented?

I just know that home racers get frustrated with power when there is not enough current, and in analogue the best driving is unlimited amps, adjust voltage for top speed..

Hand controller connector

jacks or mini jacks - although cheap and small, are not the most reliable of connectors, and the cheaper sockets sustain wear and corrosion quite quickly. Something with seperate pins (XLR/cannon connectors are excellent but pricy and bulky - telephone connectors as used by SCX and Carrera are probably the best except they are fiddly to wire) would in my opinion provide better longterm reliability

ok - only 25% through - need a break! hope this is useful

cheers

Dave
 
#9 ·
My understanding of the track drivers is that each lane is kept electrically seperate, but in digital mode, they are fed (from different chips) the same signal. Part of the advantage of this solution is to make it easy to swap between digital and analogue?

Whilst this adds flexibility in some ways, the track pieces from all manufacturers will allow unusual topographies of digital track. The circuit can narrow to one lane in places, and widen to 3, or 8 if you so wish, in places. Pit lanes are a vital part of it.

Assuming most people will want/accept 2 lanes across the start/finish line, how does the proposed system cope with such circuit designs?

(I hope these are the sort of testing questions you are after!)
 
#10 ·
Hi Astro

Great feedback.

QUOTE (astro @ 11 Jan 2005, 17:11)Fabulous work as always. At first the DOC document did not come up, but it did today.Thank you.


QUOTE Suggestions and comments detailed. Obviously this is a list of things that I feel deserve comment or could imho be done differently, it is not intended to come over negatively.Don't worry ... I won't take it personally.


The reason I put the specification up here, is because I want the weaknesses in the system design to be identified before I get too far in the design process.

QUOTE Firstly regarding the guide method of ID -

This would work elegently for me on the old scalextric blue guide system - simply snap out the guide number 7 and click in the guide for car ID 12 when needed. But changing the ID of cars with other guides can be fiddly. The new scalex type is screwed in from inside the car, others have the wires soldered to the braids or use those bullet things - I hate changing those guides cos the braids never go back the same!You certainly have a point here ... perhaps making slots in the guide blade is more cumbersome than mounting the LED sensor in the bottom of the chassis?.

I'll try if I can think of a better idea ... perhaps if we could design a system where you can mount small photo blockers in the guide holes or slots, and set the address without the use of tools.

That way you always have to make 4 holes or slots, and block the ones you don't use. I think the major problem would be to make a design that won't fall off in the slot, but who knows ... it might be possible.


QUOTE also - the variety of guide types means that the system presumably would not come with a set of guides, but people would have to saw their own. I am just trying to envisage how this would work in clubs where the cars need to be recoded etc.Yes, I agree ... we need someting that's simple to configure.

IMO the ability to place a lane change anywhere, regardless if the cars are sliding, is rather nice though.

QUOTE Sector Timing -

It is my understanding that you are planning to use the LCs to divide the track into sectors. Since at the moment people are generally thinking of evenly spacing LCs, this makes a lot of sense. However, maybe a sector track should be a seperate thing to an LC? It would make track/sector design more flexible (eg you might want to set up a couple of sector dividers so you can time the velocity along the long straight of your track)Yes, I thought we could use the sensors that have to be there for lane change, for lap and sector timing as well.

There should not be any problems in placing lane change drivers anywhere on the track. If you only need the timing functions, you can leave the solenoid outputs unconnected. Also there should be no practical limit to the number of lane changes you can put on the track


We might even be able to do something clever with a lane change sensor at each pit garage in the pit lane?.

QUOTE High Voltage Low current -

I understand what you are saying and the advantages of this. Thers a few things I don't quite follow:
Say a given car in analogue goes top speed at 12 volts DC, drawing 2 amps. power used = 24 watts, and it goes at a certain speed.Exactly.


QUOTE Now with the high voltage system, we are sending the car 24 volts, but only allowing it 1 amp (again 24 watts) Is the chip inside the car supplying the motor with 12 volts 2amps, or 24 volts 1 amp?This is a bit tricky, but I'll try to explain it. Please bear with me if this doesn't come across in a logical way.


The car will deliver 24V 2A 50% of the time. If the signal change quick enough, the motor will see this as 12V 2A 100% of the time.

This is because a motor forms an inductive load. If you apply a voltage to an inductive load, the voltage will appear without delay, but the current will rise slowly according to the time constant of the inductance. Likewise the current will disappear slowly when you remove the voltage (the inductor will continue to deliver current while the magnetic field in the inductor collapses).

This is opposite a capacitive load, where the current appears immediately and the voltage rises slowly.

The power supply will deliver 24V all the time, but it will deliver 2A 50% of the time, and 0A the other 50% of the time. In average this will draw 1A from the 24V power supply.

QUOTE If the chip is supplying 24 volts but limiting the current, then the car motor will be running in a current starved situation. Even though it is recieving the same power, will it behave the same? and what about at 22 volts? Has this method of motor control with current starvation been tried in the analogue (or digital) environment before it is implimented?If the voltage drops to 22V, the car regulator must increase the duty cycle from 50% to 55%.

QUOTE I just know that home racers get frustrated with power when there is not enough current, and in analogue the best driving is unlimited amps, adjust voltage for top speed..You are right. I think that adressing the power supply and power sharing issues, wil be one of the most important features with the new CarDCC system.

QUOTE Hand controller connector

jacks or mini jacks - although cheap and small, are not the most reliable of connectors, and the cheaper sockets sustain wear and corrosion quite quickly. Something with seperate pins (XLR/cannon connectors are excellent but pricy and bulky - telephone connectors as used by SCX and Carrera are probably the best except they are fiddly to wire) would in my opinion provide better longterm reliabilityIt's certainly possible to use other connectors ... I'll check if it's possible to make a dual mounting option, so that you can chose if you want minijacks or RJ telephone connectors.

Those vere very usefull inputs sofar.
 
#13 ·
QUOTE (drummer @ 15 Jan 2005, 17:07)Hi Guys, just thought Id say, I think the chassis mounted sensor is really a simple, cheap, effective and readily available solution, I'd stick with it.I'll check out if it's possible to implement both options in the prototype lane change controller?. That would allow us to evaluate both methods.
 
#14 ·
Hi GB
QUOTE (gboulton @ 11 Jan 2005, 17:17)Downloaded the spec doc today, will give it a good reading some time this week, per our discussion in the other digital thread.That's great.
... I'm looking forward to hearing your opinion.
 
#15 ·
Hi Astro
QUOTE (astro @ 11 Jan 2005, 17:42)My understanding of the track drivers is that each lane is kept electrically seperate, but in digital mode, they are fed (from different chips) the same signal. Part of the advantage of this solution is to make it easy to swap between digital and analogue?Yes, making the switch from digital to analog by software control is one of the advantages.
Another advantage is that you increase the current capacity of the system each time you add another output driver.

QUOTE Whilst this adds flexibility in some ways, the track pieces from all manufacturers will allow unusual topographies of digital track. The circuit can narrow to one lane in places, and widen to 3, or 8 if you so wish, in places. Pit lanes are a vital part of it.It would be possible to connect all tracks together, if you don't want to use the extra power available from multiple drivers.

One of the reasons for choosing a multiple driver approach, is that it's more difficult to design a 30A short circuit and overload proof driver, than three 10A drivers.

There are also problems involved with handling 20A or 30A on a single PCB.

The modular approach will allow you to add an extra track controller to power the lane changes, the pit lane and the track signal lighting. That would free the other track controllers of the burden of supplying anything but the cars.

QUOTE Assuming most people will want/accept 2 lanes across the start/finish line, how does the proposed system cope with such circuit designs?I would prefer a flexible system concept, that would allow the price conscious buyer to stick with one track controller. At the same time this shouldn't prevent those that want to improve the performance of their system, to add more units that would improve the capacity and power handling capability of their system.
 
#17 ·
QUOTE A bit of news (although nothing spectacular smile.gif)

I received the long awaited motor controller today. thumbsup2.gif

Althoug my telephone camera is really lousy at taking close up pictures, I thought it could at least give you an idea what the prototype CarDCC controller looks like. smile.gif

I'm still waiting for the four rectifying diodes, but I expect to have the first car controller assembled any day now. The embedded software is still far away though.

I've checked the overall size, and the controller will fit into the sidepod of a formula one car. I don't know if there are cars on the market with less space inside, but the dimensions of the controller are 15x23mm.

Gee Sword looks great, can you tell me what controller chips you have decided to utilise? thumbsup2.gif
 
#18 ·
Hi Drummer
QUOTE (drummer @ 17 Jan 2005, 06:41)Gee Sword looks greatThank you for the kind words.


QUOTE can you tell me what controller chips you have decided to utilise?In the prototype car controller I'm using the Atmel Tiny13 microcontroller and the Motorola MC33887 motorcontroller.

Here are the latest descriptions of the prototype CarDCC hardware (see under "Digital slot car controller").
 
#20 ·
Hi Beppe
QUOTE (Xlot @ 17 Jan 2005, 20:49)It does look impressive, SørenThanks a lot.


QUOTE Could you please explain in lay words where output current measurement would be available in the car controller ?I'll give it a shot, but don't hessitate to ask again if my explanation sounds like nonsense.


The motor controller U3 has a current feedback output FB on pin 20. This output will source the motor current divided by 375. If the motor draws 5A, the FB output will source 5A / 375 = 13,33mA.

By connecting a 100Ω resistor from FB to ground, we convert this current to a voltage. Using Ohm's law we get U = I * R = 13,33mA * 100Ω = 1,33V. In other words: when the motor current goes from 0A to 5A, the voltage on the FB pin goes from 0V to 1,33V.

This voltage is fed to an analog to digital converter input PB5 on the microcontroller through a 10KΩ resistor. The 10KΩ resistor will not influence the current measurement voltage, because the ADC input as a very high input impedance (typically 100MΩ).

The microcontroller will be able to measure this voltage by using the internal ADC. It can then calculate the motor current by using Ohm's law I=U/R:
Motor current = 375 * FB current
... knowing that I = U / R and the voltage U is measured by the ADC, we get:
FB current = ADC / 100Ω
... hence:
Motor current = 375 * ADC / 100Ω = 3,75 * ADC

Example: the ADC measures 1,33V ...
Motor current = 3,75 * 1,33 = 5A

The 10KΩ resistor will allow you to use the PB5 signal for debugWire in-circuit emulation when you're programming or debugging the microcontroller software. The only draw back is that you can't measure the motor current when debugging.

The C2B capacitor is only there to filter out some high frequency noise on the current measurement signal.
 
#21 ·
Thanks a lot, Søren - I think I got it (apart from debugWire in-circuit emulation !!)

So, will your motor controller have current limitation (PWM override) as a way of defining car classes ? Or is there another output overcurrent protection ?

I realize on a normal set-up the potential for a downstream short is limited - but there may be other people in addition to myself whose brain has gone into hibernation - on the other hand, no protection would generate additional sales


Ciao

Beppe
 
#22 · (Edited by Moderator)
Hi Beppe
QUOTE (Xlot @ 18 Jan 2005, 08:57)Thanks a lot, Søren - I think I got it (apart from debugWire in-circuit emulation !!)Great.


... the in-circuit emulator (ICE) is a debugging equipment you attach to the microcontroller, that allow you to download software, view the internal registers, RAM and ROM, singlestep your program one instruction at a time and set breakpoints when the code reaches a certain point.

In short, the ICE allows you to look inside the microcontroller and change registers and program execution ... a very nice tool for debugging.

... and an Olimex AVR ICE cost only $39,95.


QUOTE So, will your motor controller have current limitation (PWM override) as a way of defining car classes ?Yes, the current limitation is software configurable ... that's the big advantage by using this method compared to the hardware controlled current limitations.


QUOTE Or is there another output overcurrent protection ?Apart from the software current limitation, the MC33887 motor controller also have a built in hardware current limitation. The hardware current limit will cut in at 6,5A, and is there to protect the motorcontroller from overload (such as short circuits), if the software limiter isn't fast enough.

QUOTE I realize on a normal set-up the potential for a downstream short is limited - but there may be other people in addition to myself whose brain has gone into hibernation - on the other hand, no protection would generate additional sales
Combined with a built in thermal overload protection, the MC33887 should be well protected from any kind of abuse such as short circuits and other overloads. In all practical terms, the MC33887 should be "destruction proof".
 
#23 ·
Hello Soren,

as mentioned here in the Forum (CarDCC Status Thread), I'm absolutely impressed by your CarDCC concept. I read the hardware requirement specification about four
times and now I'm understanding most of your ideas.

But to some issues I've questions, comments or suggestions
. At first I thought it would be the best to post that directly in the SlotForum, but it would be difficult to describe the references to this specification document. So I decided to put in the comments directly in the specification with red colored text.
By no means, I want to be impolite but this is the easiest and fastest way to tell you my opinion about the topics. I hope you comprehend my feedback not negative, the only thing I want to do is to talk about possible problems that I saw in the concept, alternative solutions or sometimes "nice to have" features that might be integrated in the future.

How can we exchange this document? If you give me your email address via Forum Privat Mail i can send you the specification including my feedback.

Best regards
Guenter
 
#24 · (Edited by Moderator)
Hi Soren

Have just revisited the CarDCC thread. Having spent some time reverse engineering the SSD Powerbase and in car electronics, there are not suprisingly a lot of similarities between SSD, CarDCC and of course the NMRA DCC system where it all comes from.

Seeing what is and is not there provokes numerous ideas for additions and improvements. At that point one has two choices. Take all the ideas and produce a better system. Or produce some add ons and improvements to an existing system.

So an idea. Have you, or anyone else, enthusiasm for a SSD compatible mode for CarDCC. Hardware would probably need no changes. All the change would be in the software. Existing cars could be used with slightly reduced functionality, but with the benefit of an enhanced control system.

Alternatively the CarDCC car controllers could be used with the existing SSD Powerbase with the benefit of better performance and control.

It would be a stepping stone approach with parts being interworkable, but still leaving open the prospect of a full CarDCC system in the future.

What do you think, could be a best of both worlds way forward?

Finally a detail on your in car electronics. On the cct is the PWM signal for the rear lights generated in the microcontroller, and then brake from the motor driver? Also a detail comment. I realise you can add a lighting controller to the car if more inputs and outputs are needed but I feel that more inputs / outputs in the base controller would be of benefit.

Richard
 
#25 ·
Hello Richard,

I am just in the progress designing a replacement of the Carrera Pro-X car controller. It will be compatible with Sören's CarDCC and Pro-X. We discuss the related items in various threads in the German "freeslotter forum".

Sorry I do not know anything about how SSD works. Do you have a link to a description ? Especially interesting would be the protocol of the data sent from the track controller via the track to the car controller and also how this data is coded on the track supply (i.e. Pro-X uses +17V/GND, DCC uses +xV/-xV).

Concerning the questions about Sören's car controller design my understanding of the schematic is:
The PWM of the dimmed rear lights is generated by the micro controller. The transistor(s) for the brake is/are included in the MC33887 (H-Bridge made by Freescale).

Regards

Juergen
 
#26 ·
Hi Juergen

Thanks for the reply. Unfortunately I cannot read German so cannot contribute to Freeslotter.

I am having a "Dim" moment on the rear lights.

I can see that PB0 of the microcontroller has a 10K pull up and is connected to -FS of the H-Bridge, which from the data sheet I see is the Fault Status output from the H-Bridge.

However PB0 also seems to be connected to the rear lights? Is PB0 switched between being an Input for Fault Status and Output for brake lights?

I have understood the SSD protocol but it is not documented, just in my head at the moment!

I am not sure what you would call it but the SSD Powerbase has an H-Bridge in the output stage driving both rails with antiphase signals. The 0 and 1 timings are very similar to DCC. (Is this what you call +xV/-xV ?)

Is your Pro-X care controller electrically the same as Soren's but with modified firmware for the Pro-X protocol? Are you going to put both CarDCC & Pro-X protocols in the car at the same time?

Richard
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top