I just received a new HS-645MG servo that I am using with an SSC-32 servo controller to tilt the head of my robot. I purchased it to replace an HS422 that was clearly under-torqued to handle the load of the head, especially when tilted far from vertical. (I solved that problem temporarily with a pair of springs but those tended to induce oscillations.)
Unfortunately, now that I have installed the HS-645MG, it produces a horrible jittering noise when supporting the head at almost any tilt angle except vertical. This occurs with or without the auxiliary springs attached so it is not a simple load issue. And besides, the lesser HS422 never had this problem. I tried several different voltages from 4.8 through 7.2 but got the same result. I am using the LynxTerm application to control the servo.
Are there any other tests I can try before concluding it’s a bum servo?
I don’t have a picture handy at the moment, but the setup uses the Lynxmotion pan-and-tilt kit with the new servo being used for the tilt. Even with the head removed from the pan and tilt brackets, so there is very little weight being lifted by the servo, I get the noise. I’m beginning to suspect that this is expected from a metal-gear servo. Which makes me wonder–are those of you out there using metal-gear servos on arms like the Lynxmotion SES arm just putting up with the noise?
I have 6 - HS-5645MG servos and they do make a low volume high pitched hum sound, but usually when there is a slight load, and the servo is trying to hold position. If the servo is working as commanded, I don’t know what to say about the noise you hear. All servos can jitter to some degree, it depends how they are instructed to move. My servos run smooth when speeds are nice and swift or slow. If I instruct them to move at a medium speed, then they tend to jitter a small amount which is normal. You can experiment with the servo rates to try to get smoother performance.
Thanks for your feedback. The noise occurs when trying to hold a given tilt position. I command the servo to move to, say, 1800 (not too far off vertical), and then it sits there fine buts makes a racket. Listening carefully to my HS 422 servo (which I believe has non-metal gears), I can indeed hear a similar sound when I try to rotate it gently away from its hold position. But it is a much, much quieter sound than the 645. If I assist the 645 a little with my hand to hold its position, I can make the noise go away. But it really isn’t being asked to heft much more than the weight of the tilt bracket at this point.
I have a fairly long bracket (2-3"?) on the tilt portion of the pan-and-tilt so that when the head is angled way down (it’s looking at the ground), the HS 422 couldn’t quite muster up enough torque to move the head back up. Hence the springs initially, and now the trial with the 645. If I get a chance this weekend, I’ll make a video so you can hear the sound and see the setup.
Fantastic, this will help others “see” and “hear” what’s going on. I believe what you are saying, but I think it will benefit the community by posting an example.
After some more tinkering, I believe I have resolved this issue.
First, I confirmed that even the HS-422 can make a significant buzzing noise if you put enough torque against a given hold position. The metal gears of the 645 simply make this buzzing sound louder and sharper.
Second, I was able to reduce the buzzing in both cases by making sure the servo voltage on the SSC-32 was close to the low end of their range–i.e. 4.8 volts. I had been powering the entire board with an 8.4 V battery with the logic and servo power jumpered–for some reason I was assuming that the servo side of the circuit was down-regulated by the SSC-32 to something closer to 6 V. So now I have my own regulator in place to give the servos just 5V with the logic-servo-power jumper off.
Finally, with the voltage now in the right range, I found a spring tension on the tilt platform that creates very smooth motion without oscillation.
So in the end, a combination of over-volting the servos and a high lifting load (about 14 oz at 5" from the rotation point = 70 oz-in torque) was the source of my trouble. With a lower voltage and proper spring tension, I can use the original 422’s without trouble.
Wow! Most all servos freak out when powered from higher than 7.2vdc. Some freak out at 7.2vdc as well. But a freshly charged 8.4vdc battery can be over 9vdc, so you’re lucky it didn’t melt down. Glad you got it working.
Yeah, it’s interesting because I did measure the 8.4 V battery at just over 9 V after a charge and was thinking in terms of the 9 V logic circuit on the SSC-32 and, as I mentioned earlier, assumed the servo circuit had its own regulator. So it’s pretty amazing that the only glitch I noticed was this buzzing noise and some oscillations. And I’ve been running 5 or 6 servos off the SSC-32 like this for several months!
)