Servo Recommendation

[DKNguyen]

http://www.tribotix.com/Products/Robotis/Dynamixel/RX/RXCatalogue.htm

I need 4 servos (one per wheel) for steering my rig (the 25lbs one) and would get the RX-28 if I had enough money, but I don't. Would the RX-10 be enough you think? THe torque works out to be 139-136oz-in of torque depending on the voltage. THe RX-28's torque works out to be 393-523oz-in (after the converison from metric).

The torque of two RX-10s adds up to be the same as a HS-5955TG, except with the 5955 if one wheel has no resistance then all the extra steering torque goes to the other wheel. With two servos that are half as weak per wheel all the torque is stuck on the wheel it's connected to. THoughts?

 

[JohnRobHolmes]

The rx-10 would work well, but it wouldn't get you out of a bind. Do you need independent wheel control or could you drag link the tires together? Can you use servos with PWM control?

 

[DKNguyen]

Yeah, the problem is it's independent wheel control (or else I would just get a single RX-28 and gang it to both wheels for the same price).

I can use servos with PWM control, but it's highly preferable to use servos that accept a set-and-forget serial protocal (and feedback!) It's more important for me to know that I'm in a bind than to get out of one...it' just down to cost really. $400 vs $800 is really tempting.

 

[JohnRobHolmes]

could you modify a hobby servo for serial control?

 

[DKNguyen]

I would normally modify a 5955. The problem is the case is so small and I am uncertain if I could fit my PCB inside (actually I know if I stuck everything I wanted in it would not fit.) THere is differential serial communications (1.5pt), flyback diodes (1pt) temperature feedback (1pt), voltage feedback(0.5pt), current feedback (1pt), gate drivers (5pt)...I have enough room for a total of 2 points worth of stuff. If I'm going to go out of my way and take the effort to modify something, I'm going to stick everything I want into it and I don't think I can with a standard servo case.

If they had the 5955 in a 1/4 scale case, then yeah I would give a run at it. But the RX-10 is the price of a 5955 anyways and all the work is done for me, but it's half as strong with no titanium gears (the only reason other than price I would choose the 5955 over the RX28).

PLus...it's a ridiculous amount of work to build that! (well not as much as a BLDC driver, but the BLDC driver doesn't have such strict size consraints and is a cooler project) This brushless drive is keeping me super busy already. It's also nice that the RX-10 operates on the same voltage levels as a brushless motor. I got a lot of stuff on my plate and I've found that if I have too many things waiting in line for a project, I get frustrated and lose interest very quickly.

 

[DKNguyen]

Bah. Maybe I'll just have to suck it up and pay the extra $400. It would suck to buy the half torque, half price ones only to realize they're not strong enough and then have to buy the another set of the stronger ones.

Well...there goes the money I had budgeted for batteries, sensors and the primary controller. Now I Need to scrounge up antoher $400 from somewhere.

 

[JohnRobHolmes]

Could you put the pcb outside the servo case and just run however many wires you need to the motor and feedback?

 

[DKNguyen]

Could you put the pcb outside the servo case and just run however many wires you need to the motor and feedback?

I've thought about that. Obviously it's not as neat or as rugged, and the potentiometer signal will get distorted somewhat. It still comes down to me being lazy though. I'll gauge the price it will take to build those boards. If it comes out to $75 or more a board, then I'll just buy the darn things.