by

|   Jul  9, 2009 at  2:15am

From the schematics it looks like your version of these boards is based on the OpenServo V2 branch. Out of curiosity, what made you decide to go for that version, instead of V3?

by

|   Jul  9, 2009 at  2:17am

Note: The firmware link currently points to the example control code.

by

|   Jul  9, 2009 at  6:33am

Great to see these. I've been following this project for a while. Pity they don't seem to have the back EMF sensors like v3, would have been sweet. Also a proper datasheet would be nice.

by

|   Jul  9, 2009 at  9:49am

Why doesn't anybody offer an OpenServo board pre-attached to a servo? Maybe at few different servo models of various strengths?

This would directly compete with CrustCrawler's AX and RX offerings, at a lower price point and offering hardware compatibility with existing servo hardware (brackets etc). I would buy it.

by

|   Jul  9, 2009 at 11:36am

Can I stick this on a DC motor instead of a servo?

by

|   Jul 14, 2009 at 12:25am

If you put two equal resistors on the potmeter pads, it should be possible to use a DC motor. I would not use large DC motors though, because I'm not sure how much current the H-bridges can supply.

The good news though is that you must power this board with more than 6V. There's an internal 5V regulator for the logic, but the motor is powered by the raw input.

by

|   Sep 26, 2009 at  5:08pm

Does this come flashed with the bootloader and application?

by

|   Oct  4, 2009 at  8:26pm

Has anyone else had problems selecting suitable gains for the openservo, The servo either overshoots or has lots of steady state error. Also Integral control does not appear to work, the motor does not move with just integral control. I am using the pre loaded firmware

by

|   Oct 14, 2009 at  6:54pm

Yes!

Please if anyone has gains that work with the "TowerPro" servo that sparkfun sells please post them. The only 'kind of' suitable parameters I've found is P-200 D~240, but as AK47 said, that leaves huge steady state error. Also MAKE SURE TO SET "SWAP_PWM_DIRECTION_ENABLED" or your servo will try to destroy its self.

Sparkfun, please a little support on this one, and please "Release" your CVS code module before you zip it up for the public.

by

|   Oct 19, 2009 at  7:05pm

I have tried with 3 different servos (high torque metal gear, $2 micro servo and some standard plastic gear servo) and had the same outcome with all 3. Huge offset from desired location with gains of P300 D600, or massive out of control oscillations with P600 D2000. Integral control is what is really needed but it still does not appear to work. Derivative control only makes the control worse above about 800 (suspect its due to increased noise when taking derivative). Any help would be much appreciated.

by

|   Jan  7, 2010 at  3:28pm

You should look up PID tuning techniques to properly tune the servo control loop in your application. A good linear controls textbook is your friend. The Ziegler–Nichols method is the most commonly used for PID systems, but PD systems are easily tuned by hand.

Here is a simple method of manually tuning a PD controller: Set the derivative gain to zero, slowly increase the proportional gain until you get oscillations, then start increasing the derivative gain until the oscillations stop. If you want, you can increase the proportional some more and compensate with more derivative, but you will get diminishing returns and more jerky motion at some point. Also as d3smo said make very sure your signs/directions are right before increasing the gains too much.

Note that unless there is a constant load (torque) on the servo, any integral term will only mess it up (with delay and overshoot). That is because integral terms are only needed when a nonzero input (motor voltage) is required to have a stationary output, like in an oven, but that is clearly not the case for an unloaded servo. Increasing the proportional gain is the only way to remove steady-state error in this kind of system. I learned this the hard way with my own servo mechanism long before my first control theory class.

by

|   Jun  7, 2010 at 11:23am

Ok, just got some of those devices. To start a list of servos suitable for this product: this version of the OS v2 can _NOT_ be installed into a Hitec 645 MG, simply because the motor is too tall. I may be wrong, but it appears that the area of the PCB occupied by electronic components is larger, compared to the v2.1b PCB posted on the openservo web-pages (here: http://www.openservo.com/PrintedCircuit2). In particular, the Mosfets and the 33uF capacitor (C8) are closer to the motor contacts.

Looking at the photographs on your web-pages it seems that your PCB can be installed into a TowerPRO SG-5010. Any other servo, maybe something comparable (considering torque) to the 645 MG?

No offend meant, and best regards.