RC Model Reviews

As far as I know the S.bus was a Sun Microsystem design and it is very fast. It is a huge step forward to get rid of the 1960s servo analog interface and it is great for RC. I suspect you don't like iPods either as they are different than the Walkman?

Jim

PS Couldn't see how to start a new topic one level up

Futaba's S-Bus is their own proprietary system, I don't think it owes anything to Sun, except perhaps sharing a name.

In essence, it's pretty simple and you could convert any RC system (especially the FrSky 7-channel with a PPM output stream) to a serial-bus operation. The real question for most people is "why would you bother?"

In the computer world, the most reliable networks are star networks -- where each computer has its own connection to the server. This means that if any individual cable fails, only the connected workstation is affected -- the rest continue to operate without any change.

The kind of network the S-Bus represents is a multi-drop one where all the computers effectively get hooked in parallel to a single cable. If that cable fails anywhere along its length, all the computers which are "downstream" of the break are affected. And what's just as worrying is that if *any* of the connected computers develops a fault that places noise or a heavy load on the bus -- *all* the computers in the network will be affected.

Replace "computers" with "servos" in the above and you'll see that the Futaba S-Bus does have the potential to *reduce* the reliability of an RC installation by creating extra points and modes of failure that don't exist in our current setups.

There *are* instances where S-Bus can be useful but, for the vast majority of users, it's a backwards step that seems designed as more of a way of introducing a proprietary standard that Futaba can use to lock its competitors out of its market.

I suspect you are right. The only ones that would appreciate it are those who want better resolution and faster updates and more channels. The old PPM concept is somewhat limited but most folks only care if the servos move from one end to the other.

Anyway, I appreciate what they are doing and believe it is the right direction.

Jim

ROFLMAO

Hi Jim

How have you been?

Hey, what are you flying these days?

michael

310 878 7018

human21364@hotmail.com
malibu, ca

RCModelReviews wrote:In the computer world, the most reliable networks are star networks -- where each computer has its own connection to the server. This means that if any individual cable fails, only the connected workstation is affected -- the rest continue to operate without any change

I'd have to disagree with you there and say the most reliable topology is a ring, if one cable fails the signal comes around from the other side.

I've seen many security and card access systems use this design and RS485 serial communications. Each slave was hardcoded to an address and there would be one master sending out addresses and instructions or addresses and status queries, then the slave at that address would have a fixed time to respond. (also used for very long distances, 1200m between devices with the proper cable)

Another thing to keep in mind is that consumer radio control is a small part of Futaba Corp, some of the other technologies are going to filter in, like automotive displays and mining radios (two that I know of).

ergocentric wrote:I'd have to disagree with you there and say the most reliable topology is a ring, if one cable fails the signal comes around from the other side.

Maybe for one break, but with a star you lose 1 client per break, with a ring you could lose all the clients with just 2 breaks. C'mon guys, what are we communicating on? A web is easily the most resilient.

rogueqd wrote:

ergocentric wrote:I'd have to disagree with you there and say the most reliable topology is a ring, if one cable fails the signal comes around from the other side.

Maybe for one break, but with a star you lose 1 client per break, with a ring you could lose all the clients with just 2 breaks. C'mon guys, what are we communicating on? A web is easily the most resilient.

unlikely to ever attach card access systems to the internets, ever

a ring will report the break immediately (as will a star) but the ring will keep everything connected in the downtime, very useful for card access, no client likes it when they cannot update the access codes (the card access panel itself is battery backed up and independent)

think of it like this: your phone goes to your two neighbours instead of the central office direct, this actually saves the phone company wire and money, and if one wire gets cut by digging they will know immediately and know which wire and the phones will all work until the van comes out on Thursday, if two wires get cut they will know which phones are disconnected and know which wires are cut and know to send the emergency repair crew (would they send an emergency repair crew if it was only your phone?)

web/mesh: more wires, more nodes, more resilient, more money, more maintenance

besides all that does anyone else find the name S Bus funny?

we used to have jokes about the kids in the short bus

ergocentric wrote:web/mesh: more wires, more nodes, more resilient, more money, more maintenance

yes, but you originally said most reliable, not most practical. Having every servo connected to every other servo as well as the receiver would increase reliability. I know it wouldn't be practical. On a small plane a ring isn't that practical either, so we're back to the star.