I assembled an AL5A kit to evaluate before ordering 12 units for electronics class here at Irvine Valley College. I laid out a simple interface PC board to connect the servos to power and our existing BasicATOM Pro 24 microcontrollers (running Basic Micro Studio). Unfortunately, I’m having major problems with the shoulder and elbow servos being unable to support the arm’s own weight, much less lift any external load as in the demonstration videos. When either servo is commanded to the center position with 1.5-ms pulses (hservo [0\0]), it can’t make it without manual help. Then, after the pulse width is incremented to move the arm up, it won’t return to the original position when decremented by the same amount. I.e, the servo is failing to null out the position error, making arm positioning completely unrepeatable. The servo just stalls and hums without any apparent holding torque either, as you can move it with your hand.
We are wondering if our power supply is adequate. We have tried a 1-amp rated switching regulated 5-V wall wart supply and found that, even with only a single servo connected, it puts big glitches on the supply at the 50-Hz rate put out by the controller. The glitches pull the voltage down below 4 V for a ms or more each cycle. We then tried a 2-A rated lab bench supply adjusted to 5 V–although the glitches are less, the servos are still not positioning correctly. (The Basic Atom is running off of its own separate wall wart sharing only a common ground with the arm power). Can’t understand why these servos are acting so wimpy. The shoulder servo, an HS-755HB, looks pretty hefty and is supposed to have a lot of torque. Any suggestions? Thanks!
That’s exactly what to look at. The 755 servo on its own can consume up to 1.5A at stall (4.8V) and 230mA no load at all. The wall adapter we suggest is WP-04 which is a 6V, 2A supply. For the larger arms, we include a 3A supply. Also, be sure to use the springs to balance the load. Last, check that the PWM from your controller is being repeated at least every 20-30ms.
Unfortunately the HS-755HB shoulder servo just doesn’t seem to be cutting the mustard. I got a 2.6-A 5-V wall adapter and it seems to be holding up fine under load, according to the oscilloscope. None of the servo’s internal gears were stripped when I looked, and it moves normally on the test bench with no load. I installed the arm helper springs, and checked that the control pulses were coming from the Basic Atom every 20 ms. Nevertheless, when installed on the AL5A, the arm literally droops under its own weight, even when under power. If I command the shoulder servo to lower the arm, it does so, but then won’t raise it again when commanded without some manual help. The other 4 servos all seem to be moving repeatably. I got the HS-755HB new, but is it possible that I got a dud? Any other suggestions? Thanks.
Do you have any sort of alternative electronics which can send an R?C signal such as a hobby remote control? It’s incredibly rare that a servo would fail in this way. Are the other servos experiencing any issues or is it just the 755?
Last, can you send a clearn image (or images) of the setup so we can see if there are any issues with the connections?
Yes, the other 4 servos are working OK, and the oscilloscope shows that the Basic Atom is sending out the correct PWM control signals. Most recently, I changed from the wall wart to 4 new alkaline D cells in series, to get the supply up to 6 V under load. The attached photos illustrate the complete setup. The circuit board simply provides connections to the Basic Atom I/O lines and signal ground. Only the shoulder servo is connected to the Basic Atom in the photo for clarity. I haven’t tie-wrapped the servo cables yet.
The shoulder servo is still not able to raise the weight of the arm back up to the home position. Maybe the bearing is shot, because I notice a little eccentric motion (wobble) in the shaft when it’s chugging along trying to raise the arm back up. I got this servo from a hobby supply, but it was brand new.
I am thinking of getting another shoulder servo and trying it. Any other thoughts/suggestions? Thanks.
The resistor on the signal line (100 Ohms) is for student proofing. I did check the voltage waveform on the servo side and it’s the same amplitude as on the I/O pin side of the resistor (no appreciable voltage drop). I will check the battery pack voltage under load to make sure it’s not dropping.
