Can I branch the CAN to go to two separate places?

[jee7s]

Originally Posted by wireties

All the normal rules still apply but there are additional rules (terms in the equations).

This is taking the thread a tad off topic, but, I respectfully disagree. It's not that there are additional rules. The rules are the same but you can't use as simple a model of the rules.

The so-called "normal" rules are really a number of simplifications and assumptions that apply at DC that don't apply when there is change in the signal. Even the complex impedance is a simplification that applies to lumped element models. Fundamentally, both of these are case specific subsets/simplifications of Maxwell's Equations, which are the set of laws that govern electromagnetism at non-relativistic conditions.

I say this because I think it further illustrates the bad thinking I highlighted earlier: one must acknowledge the assumptions that are used (like assuming conductors are ideal) before applying a concept. It's one thing to use a suitable simplification because a set of conditions are met. I don't think any engineering would happen if we needed to do vector calculus when determining a load instead of using V=IR. But, as a good engineer, you need to know when you can simplify the model and when you cannot.

As an aside, @ratdude: even Kirchhoff's Laws don't always apply, particularly if a changing magnetic flux is present.

[wireties]

Originally Posted by jee7s

This is taking the thread a tad off topic, but, I respectfully disagree. It's not that there are additional rules. The rules are the same but you can't use as simple a model of the rules.

Thus the "terms in the equations" comment ...

[philso]

Originally Posted by wireties

QFT - I should have mentioned this. But on the roboRio the CANbus is running near full-speed. You would have to slow the bus down and all kinds of things might stop working (PCM control loops not closing, not enough bandwidth to query/set Talons etc).

What is "full speed"? A quick search did not turn up any specs on the data rate of the RoboRio's CAN Port. In fact, NI's spec sheet totally neglects to mention the CAN Port in any way.

What do you mean by "slow the bus down"? Do you mean a slower data-rate or do you mean a slower edge-rate?

[ratdude747]

Originally Posted by philso

My coworkers and I are puzzled by some your statements.

That's how I was taught in Electromagnetics class. Of course resistance is still present, but it's not 100% of the picture and looking at it alone will lead to results more puzzling that the statements I made.

Originally Posted by wireties

All the normal rules still apply but there are additional rules (terms in the equations). You are looking at the complex impedance of the load (which did not factor in at low frequencies) rather than simple resistance. The size and shape of the wire and connector pins all matter because they help determine the reactance portion of the impedance (resistance plus reactance). Balancing the impedance of the source and the load become important to reduce mismatches that can cause ringing (reflected waves bouncing around willy-nilly).

That's more or less what I was trying to say. But that means the normal rules do NOT apply, as V doesn't nessarily equal IR any more. As mentioned earlier in this post resistance doesn't go away, but it's not the only factor and in many situations is negligible.

Originally Posted by jee7s

This is taking the thread a tad off topic, but, I respectfully disagree. It's not that there are additional rules. The rules are the same but you can't use as simple a model of the rules.

The so-called "normal" rules are really a number of simplifications and assumptions that apply at DC that don't apply when there is change in the signal. Even the complex impedance is a simplification that applies to lumped element models. Fundamentally, both of these are case specific subsets/simplifications of Maxwell's Equations, which are the set of laws that govern electromagnetism at non-relativistic conditions.

I say this because I think it further illustrates the bad thinking I highlighted earlier: one must acknowledge the assumptions that are used (like assuming conductors are ideal) before applying a concept. It's one thing to use a suitable simplification because a set of conditions are met. I don't think any engineering would happen if we needed to do vector calculus when determining a load instead of using V=IR. But, as a good engineer, you need to know when you can simplify the model and when you cannot.

As an aside, @ratdude: even Kirchhoff's Laws don't always apply, particularly if a changing magnetic flux is present.

I was taught in electromagnetics class that Kirchoff's laws do apply in RF, and specifically those alone are what do translate from circuit theory to RF unchanged. While a changing flux can cause all sorts of fun voltages, in a pure snapshot of an RF system all voltages will be accounted for someplace. It may be someplace you don't expect, but it will be accounted for.

Otherwise, obviously I agree.

[wireties]

Originally Posted by philso

What is "full speed"? A quick search did not turn up any specs on the data rate of the RoboRio's CAN Port. In fact, NI's spec sheet totally neglects to mention the CAN Port in any way.

1Mb/s but the actual throughput may vary. CANbus uses NRZ encoding and bit stuffing etc so realized throughput is usually in the 850kb/s range.

Originally Posted by philso

What do you mean by "slow the bus down"? Do you mean a slower data-rate or do you mean a slower edge-rate?

one implies the other (but not really sure what you mean by edge rate)

[techhelpbb]

Originally Posted by ratdude747

I was taught in electromagnetics class that Kirchoff's laws do apply in RF, and specifically those alone are what do translate from circuit theory to RF unchanged. While a changing flux can cause all sorts of fun voltages, in a pure snapshot of an RF system all voltages will be accounted for someplace. It may be someplace you don't expect, but it will be accounted for.

Otherwise, obviously I agree.

I agree. In any active AC circuit at a given instant in time Kirchoff's Law ought to work. Now over time, that's a whole different matter. Once one starts decoupling the voltage and the current over time (phase) with capacitance and inductance that transition makes interesting things happen like resonance. Then we gotta start pulling in math like Fourier transforms and Laplace transforms.

Since CAN signals are NRZ digital (with the 1&0 voltages dependent on the PHY layer design) they represent a typical digital square wave of odd-integer harmonics with differential signal. The issue, and the TDR neatly demonstrates this concept, is that an analysis over time (not at any given instant) can produce reflections when the termination or legs (even the legs of the circuit inside the devices) get too long. So it's not merely the real component of the terminator resistance at work. It's the capacitance, inductance and impedance of the circuit as a whole. (Active AC components tend to introduce mathematical imaginary or polar numbers into circuit analysis so when I say real resistance I mean DC resistance.)

Long story short - I don't think anyone in the last few posts disagrees that students shouldn't follow the guidelines. No one wants to have intermittent phantom problems that only happen when a certain pattern of traffic is sent over that CAN bus. I don't think I've ever seen a FRC team test their CAN bus with a BERD/BERT even though the speed is pretty high on that circuit (a T1 digital telephone circuit is basically 1.544Mbps and these are usually tested with something like a T-BERD). There's limited rational reason in the last few years of robots that the CAN bus configuration as recommended can't be achieved. Even if you somehow put a Talon 14 feet up an end-effector you could avoid doubling back with the CAN buss wire (which given you'd be running the power up there seems sort of strange) by simply ending the buss up there.

There are some good examples of why a star configuration has advantages if you suspect your devices might fail to be connected. However, like in computer networking, how far does one want to go to get a star network without the circuit implications? Active CAN hubs? How about a CAN switch? (NOTE: I am making no suggesting the 2 links I provided are FRC legal or even recommended, just pointing out that such devices exist with evidence.)

[adciv]

Originally Posted by techhelpbb

Even if you somehow put a Talon 14 feet up an end-effector you could avoid doubling back with the CAN buss wire (which given you'd be running the power up there seems sort of strange) by simply ending the buss up there.

30 feet, in 2015. And we did end the canbus there (had to find the right resistor for it too).

Related notes, from discussions with CTRE and beta testing the CAN BUS is fixed at 1Mbps (raw rate, not payload) and will not auto negotiate down. I don't think this has changed since then.

Related note if I'm understanding the CAN spec correctly, if a star configuration is used, the termination resistor values need to change so the effective resistance is still ~100 ohms. This would be interesting in conjunction with the fixed values inside the roboRIO & PDP.

[philso]

Originally Posted by wireties

1Mb/s but the actual throughput may vary. CANbus uses NRZ encoding and bit stuffing etc so realized throughput is usually in the 850kb/s range.

one implies the other (but not really sure what you mean by edge rate)

Where does this 1 Mbps figure come from?

Most of the CAN transceiver chips, like the ones I am using at work, are rated by their manufacturers for operation up to a maximum data-rate of 1 Mbps but that does not mean the system they are installed in are running at that data-rate.

Data-rate is a measure (bits per second) of the instantaneous rate at which the data is transmitted over the bus, when there is data. One can also measure the average data-rate taking into account the time when the bus is inactive (unused) and any error correction made necessary due to noise or other problems (resending of packets).

Edge-rate is a measure (volts per microsecond) of how fast the high-to-low and low-to-high state transitions take place. It is the high frequency energy contained in fast transitions that causes reflections. It is possible to have a CAN bus system with a low data-rate (say 1 bit per second) that has significant reflections because there are impedance mismatches AND the edge-rate is very high (the parts I am using have a spec of 35 nsec. minimum).

Once a CAN Bus system is assembled, the line lengths, actual line characteristic impedances, actual termination resistance values, impedance mismatches and actual edge-rates are what determine what the peak voltage and the duration of the reflections will be. The characteristics of the reflections will generally remain constant unless the system is changed in some way.

The CAN Bus receivers are connected to a circuit (usually incorporated in the microprocessor) that roughly synchronizes with the bit edges. The state of each bit is sampled in the middle of the bit period, often multiple times. As long as the reflections after a bit transition has died down by the middle of the bit period, the detected state of the bit will be accurate. This technique was developed a long time ago to make it easier for communication systems like CAN Bus to minimize the effects of reflections in the system.

Many of the CAN transceiver chips can be set to produce lower edge-rates, often using an external resistor. This allows the system/circuit designer to reduce the energy in the reflections (and hence their peak voltage and duration) in the system based on his/her knowledge of the length of the transmission lines (bus length), the amount of impedance mismatch expected and the maximum data-rate required.

The long and the short of it is that the CAN Bus standards were set up with the ability to tolerate some amount of reflections since the people developing the standards understood that the world is not perfect and that the systems will not be manufactured and installed perfectly. Therefore, while it is best to avoid using the star configuration since it is non-ideal, it is not "certain death" to use a star configuration or to add a branch as long as one is taking some simple precautions. Because of the inherent robustness, CAN Bus systems have found many applications outside of the original intended use in cars.

[wireties]

Originally Posted by philso

Where does this 1 Mbps figure come from?

From looking at the CANbus driver code

Originally Posted by philso

Most of the CAN transceiver chips, like the ones I am using at work, are rated by their manufacturers for operation up to a maximum data-rate of 1 Mbps but that does not mean the system they are installed in are running at that data-rate.

That is because 1Mbps is the recommended maximum rate.

Originally Posted by philso

Edge-rate is a measure (volts per microsecond) of how fast the high-to-low and low-to-high state transitions take place.

This is more commonly called rise time. And you are correct, it does cause reflection issues.

Originally Posted by philso

Once a CAN Bus system is assembled, the line lengths, actual line characteristic impedances, actual termination resistance values, impedance mismatches and actual edge-rates are what determine what the peak voltage and the duration of the reflections will be. The characteristics of the reflections will generally remain constant unless the system is changed in some way.

Thus the generic recommendation to wire it in series with a 120 ohm termination.

Originally Posted by philso

The CAN Bus receivers are connected to a circuit (usually incorporated in the microprocessor) that roughly synchronizes with the bit edges. The state of each bit is sampled in the middle of the bit period, often multiple times.

CANbus does not work exactly like asynchronous serial methods. CANbus receivers attempt to synch synthesized clocks with the transmitter. That is the entire purpose of the NRZ encoding and the RLL coding - to have enough transitions to phase lock a synthesized clock.

Originally Posted by philso

Therefore, while it is best to avoid using the star configuration since it is non-ideal, it is not "certain death" to use a star configuration or to add a branch as long as one is taking some simple precautions. Because of the inherent robustness, CAN Bus systems have found many applications outside of the original intended use in cars.

With respect using a star configuration is an unwise and strictly unnecessary technical risk. It is a good thing to teach students to follow the recommended use of communications links, sensors and other subsystems on the robot. Otherwise, a FIRST robot can quickly become unreliable. And nearly impossible to diagnose.

[philso]

Originally Posted by wireties

This is more commonly called rise time. And you are correct, it does cause reflection issues.

I had referred to rise-time and fall-time in my original post in this thread. They both contribute to reflection issues.

Originally Posted by wireties

Thus the generic recommendation to wire it in series with a 120 ohm termination.

I referred to "actual line characteristic impedances, actual termination resistance values, impedance mismatches and actual edge-rates" because the 120 Ohms is only an ideal value for an ideal 120 Ohm transmission line. It is the non-ideal characteristics of a real world system that will cause some minor reflections even though one is using the recommended termination resistance.

Originally Posted by wireties

CANbus does not work exactly like asynchronous serial methods. CANbus receivers attempt to synch synthesized clocks with the transmitter. That is the entire purpose of the NRZ encoding and the RLL coding - to have enough transitions to phase lock a synthesized clock.

And the phase locked synchronous clock is used by the UARTs to sample the state of the bit in the middle of the bit time to avoid the ringing that can occur with each bit transition. Yes, the coding increases the number of transitions making it easier to implement the PLL. The low-level circuitry used to detect the state of each bit is not really affected by the coding. A bit is a bit.

Originally Posted by wireties

With respect using a star configuration is an unwise and strictly unnecessary technical risk. It is a good thing to teach students to follow the recommended use of communications links, sensors and other subsystems on the robot. Otherwise, a FIRST robot can quickly become unreliable. And nearly impossible to diagnose.

One of the lessons that the team members learn is management of risk. You are correct that there is risk in using a star configuration. Since there is not really anything with zero risk, the challenge is to quantify the risk and determine if it is acceptable.

For an example, consider a branch on a CAN Bus that is 10 ft long to reach a Talon SRX that is mounted on some arm on the robot using the same wire as the rest of the system. With a 1 Mbps data-rate, the bit time is 1000 nsec. With a rule of thumb value for the propagation time for a pulse down the transmission line of 2 nsec/ft, a 10 ft long branch would cause a reflection that returns in 40 nsec. after the bit transition that caused it. Assuming that the proper termination resistances at the RoboRio and the PDP are in place, it is reasonable to expect that the amplitude of the reflections would have diminished to become insignificant by the middle of the bit time, 500 nsec. after the bit transition.

Based on the calculation shown above, I would determine that the risk of using the 10 ft long branch is minimal. Since each person/team is willing to accept different levels of risk, I can understand if you choose not to use such a branch.

[wireties]

Originally Posted by philso

For an example, consider a branch on a CAN Bus that is 10 ft long to reach a Talon SRX that is mounted on some arm on the robot using the same wire as the rest of the system. With a 1 Mbps data-rate, the bit time is 1000 nsec. With a rule of thumb value for the propagation time for a pulse down the transmission line of 2 nsec/ft, a 10 ft long branch would cause a reflection that returns in 40 nsec. after the bit transition that caused it. Assuming that the proper termination resistances at the RoboRio and the PDP are in place, it is reasonable to expect that the amplitude of the reflections would have diminished to become insignificant by the middle of the bit time, 500 nsec. after the bit transition.

Based on the calculation shown above, I would determine that the risk of using the 10 ft long branch is minimal. Since each person/team is willing to accept different levels of risk, I can understand if you choose not to use such a branch.

A better choice is to put the CANbus termination at the top of the arm. If you had two such arms live with the increased wire length. CANbus can be 40m long, plenty long enough to wire anything on a FIRST robot properly. Lesson #1 in risk management is to avoid risk whenever possible.

[techhelpbb]

As I am a CSA: I usually make my recommendations and explain that the default recommended configuration includes ample consideration of system traits, that a star configuration adds risk.

If a team proceeds, has issues, and refuses to change course there is nothing I can do than to keep suggesting that using the branches/star puts additional potential for issues into that system. Sure if all the connections are great the reflection shouldn't be an issue, but then again if those CAN bus wires are moving around...

With the telephone style Jaguar CAN/serial connectors and a Jaguar say 15 feet away on an end-effector are you really sure the connector will not move at all so you can predict the characteristic of that network? Maybe on a bench test this would be okay but what about when that extended end-effector gets smacked by another robot?

[philso]

Originally Posted by wireties

A better choice is to put the CANbus termination at the top of the arm. If you had two such arms live with the increased wire length. CANbus can be 40m long, plenty long enough to wire anything on a FIRST robot properly. Lesson #1 in risk management is to avoid risk whenever possible.

Yes, it would be better to put the termination at the end of the longest branch. I am not sure if the termination resistors in the RoboRio and PDP can be disconnected easily so this might not be possible.

Originally Posted by techhelpbb

As I am a CSA: I usually make my recommendations and explain that the default recommended configuration includes ample consideration of system traits, that a star configuration adds risk.

If a team proceeds, has issues, and refuses to change course there is nothing I can do than to keep suggesting that using the branches/star puts additional potential for issues into that system. Sure if all the connections are great the reflection shouldn't be an issue, but then again if those CAN bus wires are moving around...

With the telephone style Jaguar CAN/serial connectors and a Jaguar say 15 feet away on an end-effector are you really sure the connector will not move at all so you can predict the characteristic of that network? Maybe on a bench test this would be okay but what about when that extended end-effector gets smacked by another robot?

In your example of the wiring on a moving arm, I would be more concerned about an outright interruption in the connection (connector or wire breaking) than the changes in the characteristic impedance of the cables on the moving arm. My personal experience with moving wires around on such networks (while monitoring with a scope and using proper high-frequency probing techniques) is that the "lumps and bumps" move around in time a little but the overall peak amplitude and the decay time of the reflections are not affected in a significant way.

I may have overlooked them but I don't recall seeing strict limits on the wire size and the twists per foot for the CAN Bus wires to make the characteristic impedance of the wiring more uniform in any documentation from NI, CTRE or any other FRC related sources. I also don't recall seeing requirements for shielding of the CAN Bus wires. There is no requirement that the termination resistors are high accuracy types. Those are the sorts of things one would have to do to make the CAN Bus characteristics more stable but doing those things probably takes one past the point of diminishing returns and greater/unnecessary cost/effort.

With regard to risk management, please keep in mind that the size of the CAN Bus network in an FRC robot is a small fraction of the maximum size intended in the standard. Digital communication systems like the CAN Bus are developed with significant noise margins to guarantee successful implementations by relatively unskilled installers. That is where the restrictions such as the maximum length of the bus come from. As long as one is only using a fraction of the noise margin that has been built into the technology, one should be pretty certain of success.

Looking at the big picture, by following some fairly simple recommendations, the risk of the CAN Bus being the cause of a failure on the field is significantly less than other sources of risk such as choosing the wrong sort of drive train or scoring mechanism.

[techhelpbb]

Originally Posted by philso

Looking at the big picture, by following some fairly simple recommendations, the risk of the CAN Bus being the cause of a failure on the field is significantly less than other sources of risk such as choosing the wrong sort of drive train or scoring mechanism.

Normally, with test data in hand I would agree. However the CAN bus as implemented by the Jagaurs sort of demonstrated how what you get that uses a CAN bus can greatly impact the risk of CAN bus. There were a lot of people that had Jaguar issues and when those were the only ESC with CAN one gets one problem leading to another.

I am not being overly critical of the Jaguar, but FRC11 finally decided to stop fielding them and they now are in a box in my one shop as a result. They were too complex to debug in the 6 week season, changed manufacturer too many times to get fixes and we had some design issues that the Jaguars took heat for. There were some teams that stood out as Jaguar success stories and credit goes where it was well deserved, but it was not the best investment of FRC11 time and resources.

I look at branches and star in a similar way. It would seem to be an interesting rabbit hole. The recommended design does work so why invest in the alternative?

[Andrew Schreiber]

Originally Posted by techhelpbb

I look at branches and star in a similar way. It would seem to be an interesting rabbit hole. The recommended design does work so why invest in the alternative?

It's good to know the limits of the system so you know what you can get away with should you have the need to. Or there's the argument of knowing points of failure... Or there's just the general concept of liking to know things.