Quote from: Jim on 03/20/2015 06:11 pmQuote from: meekGee on 03/20/2015 06:01 pmWhich is where a more automated design comes in.And how are they going accommodate all the connectors with scores of these fat fingered pins?This kind of work is what I do.There are ways, and it's a combination of interface design and automation.You certainly don't want to try to design robots that try to mate standard MIL connectors which were designed for human hands.But if designed for automated processing, it's not a hard challenge.
Quote from: meekGee on 03/20/2015 06:01 pmWhich is where a more automated design comes in.And how are they going accommodate all the connectors with scores of these fat fingered pins?
Which is where a more automated design comes in.
Chris, a question regarding this from the article:QuoteWhile efforts to achieve this goal are in the pipeline for the Upper StageMy understanding was SpaceX has abandoned the idea of recovering S2 on F9. Is that not the case?
While efforts to achieve this goal are in the pipeline for the Upper Stage
Unlike other manufacturers SpaceX gives each stage its own avionics. That greatly reduces need for interconnnection. One serial data link can be enough. Make that three for triple redundancy.
This kind of work is what I do.There are ways, and it's a combination of interface design and automation.You certainly don't want to try to design robots that try to mate standard MIL connectors which were designed for human hands.But if designed for automated processing, it's not a hard challenge.
Quote from: meekGee on 03/20/2015 07:24 pmThis kind of work is what I do.There are ways, and it's a combination of interface design and automation.You certainly don't want to try to design robots that try to mate standard MIL connectors which were designed for human hands.But if designed for automated processing, it's not a hard challenge.Just takes massively large connectors.
Or you use a different connector type.
Why hundred? Other ways to do it.
Here ya go:
Quote from: Space Ghost 1962 on 03/20/2015 08:20 pmWhy hundred? Other ways to do it.True. It's always seems odd to me that rocket stages seem to like a lot of separate wires (the "parallel" data approach) while transmitting data on serial buses. I can only presume this design decision (and it is a design decision, not a law of the universe) dates from the time when any electronics were complex, heavy and power hungry. Direct wiring from sensors and actuators made sense.Today sensors and actuators with built in networking through things like the CANbus standard are common, lightweight and support up to around 2000 devices.
Quote from: guckyfan on 03/20/2015 06:37 pmUnlike other manufacturers SpaceX gives each stage its own avionics. That greatly reduces need for interconnnection. One serial data link can be enough. Make that three for triple redundancy.Still doesn't change anything. The upper stage still controls the whole stack. And there are range safety connections and telemetry too.
Quote from: Robotbeat on 03/20/2015 07:45 pmHere ya go: Just for your information Fulematics has been at this since before 2003 and has gone through more than one company reconstruction. That they have not been able to sell this "revolutionary" product after all this time should be a warning that it is not easy.
1. No need for hundreds of pins. That's got to be historical baggage. Serialize the data.2. There's no reason to have the second stage drive the first. The first is already fully autonomous, has identical avionics, and knows all there is to know. It can make the same decisions the second stage does, can drop it off within an envelope that the second stage can continue from. Given that the first stage has to continue flight anyway, I don't see why it has to go through the shock of "switching commanders" on stage separation. The first stage should be working like a carrier airplane.This is another case of historical baggage. An EELV is a single vehicle that's dropping parts until only the upper stage is left. That's why it's built the way it is. An F9R is a different type of beast. A highly reusable first stage that continues flight and RTLS right away, and a an upper stage that even if it comes back, does so much later. They have different operations cycles, and are their own self-contained entities. Fewer failure modes this way, too - rockets have been lost because of failure of inter-stage connectors. (can't remember which right now)3. Look at spacecraft dockings. Connections are made, after the two heavy bodies mate. And those are free-flying vehicles. You can do much better if the bodies are guided. You can do precision guidance by the jig, or you can have pilot pins and such on the flight hardware.4. The connectors can be rigidly connected to the master bodies and everything connect at once, or you can have them execute a secondary motion after the mechanical mate. Either way, if you pre-plan for automation, it's alot easier than if you try to make automation work in a less structured environment.
Honestly, the electrical connector is the easy bit of this whole thing.Something like a self-aligning Magsafe connector with more (and stiffer) pins and a solenoid or servo locking mechanism to ensure a tight fit.