OpenServo.com

[andylippitt]

I was chatting with Mike last night and it seems we both thought of the same unusual idea. Every servo we've opened has limited 2d board space, but has plenty of 3d space. Motor on one side, nothing on the other. We kicked around the idea of mother/daughter board pairs. Perhaps a unified logic board with swappable hbridge boards.

I think it would be best if the connector between them did not have to handle high current. So perhaps the daughter board is the logic board and the bridge has the external connectors on it. Any connector would have to handle vibration, yet can't have too large of a footprint or else we've undone any gains made by splitting the boards.

Thoughts?

[prolinuxfan]

The layout of the control section is not real critical, and can be offloaded to a daughterboard without too much problem, but the MOSFETs can be affected by stray inductances and capacitances often introduced by such an arrangement, so they should not be.

That is also the reason why breadboarding mosfets isn't a hot idea, since circuits that look really good on paper often end up failing when breadboarded, but may actually work very well on a PCB.

You will have to watch the gate capacitance for the MOSFETs; it can affect your turn on times, and too much will keep you from being able to drive them to fully on fast enough to prevent thermal problems and switching losses, but this is doable
_________________
Allen
(My R2 *WILL* climb stairs when I'm done)

[prolinuxfan]

You probably won't really want sockets, though -- for reliability and control of the circuit parameters, it may be better to solder an edge connector or header.
_________________
Allen
(My R2 *WILL* climb stairs when I'm done)

[ginge]

I just though I would leave my 2 pence on this one, as I have actually constructed a L shaped board using the Openservo MPU+software.

Personally, at this stage, I don't think the idea is worth exploring to any degree. Without the obvious complexity it involves to the user, it still leaves large technical issues to resolve for the designer.

Here is a small list of issues that I have either come across, or could happen;
*Mechanical vibration can cause dry joins and bad connections, leading to intermittent connections and general bad karma.
*None soldered connections can cause Impedance and capacitance problems if not designed carefully.
*It's not as aesthically pleasing as the current designs
*Not all servos are the same size. For example the resistor board on my servos sits at a slightly lower level than the driver PCB. This leaves no space for any extra boards.

I had a few small problems that are not impossible to overcome using a daughterboard design. It's just that over time I grew to dislike the daughterboard, eventually eliminating the need with a different H-Bridge driver chip.

Feel free to disagree

Barry