I have added a new page to the gallery section of this website.

The servo gallery page is a servo wall of shame, if you have servos which have detonated under normal use conditions due to poor design or faulty operation, photograph them, send me the details and I will name and shame them!

Check the link for a couple of examples.

Matt.

Hi Matt - I would be interested to hear make/models of those that have failed. Those look like some fairly massive failures, I must say!

Right now I'm using Hitec HS-81MGs on my hexapod. They work OK I guess, though they are far from perfect (as all servos are).

Out of curiosity - What frequency are you driving your servos at? For a while I was driving mine at 50Hz. I then did some experimentation with changing the drive frequency on the fly and found that the faster the frequency, the better the reference tracking is. However, if you change the pulsewidth signal drastically (say by a quarter of a ms) it can induce oscillations. Thus I always power them up at 50Hz, then when the robot is in stance, I up the frequency. I found 80Hz is about as fast as I want to go typically. If I'm moving slowly, 100Hz works well, but it can destabilize some if the robot walks too quickly. 125Hz is great for holding a stance, but can't seem to handle much movement.

I've also noticed that current increases as you increase the drive frequency, with a shape looking something like:
current = C1 - C2*e^(1/frequency). (where C1 and C2 are constants)

This worries me a bit as at the higher frequencies current consumption is fairly high, which could lead to any of the failures that you have experienced.

Of course, this is just for specific servo that I'm using. I've heard that some servos can handle up to about 400Hz!

-Michael

Hi Michael,

At the higher frequencies you will probably find that the current consumption is more constant, and if you are using a multimeter to measure it it will appear to be higher. At lower frequencies the current would usually have a higher peak drain, and on a multimeter appear to be lower. You really need to measure the current with an current probe for a DSO.

I run my servos between 50 and 70hz.

Matt.