Comment by James Ross on June 6, 2008 at 9:45pm

This is labeled as Beta2 but there is a Beta3 floating around. Could you upload the latest schematic and board layout?

Anyone building this will also need the AVR programmer when they order their stuff.

Comment by Chris Anderson on June 6, 2008 at 10:47pm

Erk. Beta3? That's a mistake--do you have a link? I'll correct it.

Here's the latest board and schematic (Eagle format):

ardupilotBETA.brd

ardupilotBETA.sch

Comment by James Ross on June 7, 2008 at 7:15am

On the BatchPCB website, the old version is titled ArdupilotBeta3 currently linked on the original ArduPilot post

Also, I'm not a EE, but it appears the diode short also plagues Jordi's Open Source MUX (AntiFail) V2

Comment by bGatti on June 7, 2008 at 9:49am

Chris,
"mem" and I have been working on a scheme to avoid the spaghetti mess of servos in and servos out on such a board.

While only half-tested, The scheme replaces the 27 servo pins and 75LS157D in the current circuit with two arduino pins and four wires.

Specifically, we interrupt the composite radio signal - interpret it via interrupts, then re-inject a new composite signal back into the receiver so that the servos remain connected directly to the receiever. Fewer pins, parts, and wires.

Progress: we are looking for a receiver with clear and clean access to the composite signal trace.

Forum Thread

Ben

Comment by Chris Anderson on June 7, 2008 at 12:13pm

bGatti,

I think that's exactly what Paparazzi does. Check it out here.

Comment by Chris Anderson on June 7, 2008 at 12:15pm

James,

Good catch. I'll fix that earlier post and delete the BatchPCB board with the incorrect version number.

And yes, that diode error came when we integrated Jordi's original MUX. He's updated that one with a much improved board, which we'll be posting soon.

Comment by bGatti on June 7, 2008 at 2:31pm

Chris,
Yes, in principle at least.
Although the schematic shows the receiver decoder unused. and they include the 4017 on board - the alternative is mentioned at your link - but they list an all-or-nothing limitation.

mem has worked out doing both the encode and decode using hardware timers and interupts for low-cpu loading.

We have also overcome the limitations - one can use receiver based servos, arduino based servos, and "virtual servos" ie channels for software control which are never exposed on any pin.

One might consider - where space is an issue - to place the servo connections on a part of the board which can be clipped off...

Ben

Comment by bGatti on June 7, 2008 at 6:55pm

related to this, I might also propose a form of fail-safe which has been used in CNC machines.

A watchdog signal can be pumped into a capacitor, if the voltage on the capacitor falls, then the receiver is reconnected to its output stage. This would provide rc control in cases where the microcontroller requires a reset.

Comment by Geetesh Dubey on June 8, 2008 at 12:16am

do we have some sort of autopilot code for avr...if so...im looking forwaRD TO it very eagerly... in my project im using an AVCS gyro to stabilize only the roll of F-22 Raptor Hangar 9 airframe...hope ardupilot is gonna be a lot useful with some tradeoffs(parts availability constraints)...i dont have an FMA co-pilot...so gotta look 4 sm alternative...

Comment by Brian on July 10, 2008 at 11:41pm

Cool Beans!
Regarding the update "Three female-to-female RC cables (you'll cut them in half to make the pigtails to the RC receiver) " doesn't the link show Male-to-Male RC cables? What are you doing for strain relief? Will future boards have a spot for the cables to be zipped to the board?

Comment

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page