Unstable operation of AL5D

Lynxmotion Robot Arms
1

Hi there,

We got a AL5D and have it assembled.
But the problem is that the servo shakes and appears unstable for precise positioning or gripping.
HS-805BB and HS-755BB seems to be the source of shaking, what could be the problem?
Power supply for the servo is 5V instead of 6V.
We are using potentiometer as input to an Arduino101 board and digital output of Arduino101 to the servos.
Please help.

Thanks, Tony

2

The RC servos used are Hitec and the optimal voltage for lower current is 6V as opposed to 5V.
You also need to check that your power supply can provide the required current. We suggest a 6V, 3A power supply, so at 5V, you would likely need a power supply at 4A+ continuous current output.

In order to maintain their position under load, RC servos do unfortunately shake - and the AL5 series is the most precise of systems.

If you think the shaking is beyond what is normal for RC servos, can you upload a video?

3

Hi Coleman,

Thanks and I will upload a video in a moment.
If so, it is hard for RC servos to work with Arduino which operates at 5V pulse, isn’t it?
We will need to get a Bot board II then.

Another thing:
We extended the arm to about 60cm long. And we switched power driver to SSC-32U.
As I was using servo sequencer, it is okay initially. Then as the arm is extended fully straight forward. The system collapsed, and the sequencer lost control, which maybe due to the torque is out of range.
My calculation shows:
150g60cm+150g30cm=13.5kgcm, which is within the range of 805BB, which is 24.7kgcm.
This is strange, isn’t it?

Thanks, Tony

extend_60cm.jpg
extend_60cm.jpg1280×960 97.9 KB

4

BTW, while I was running SSC-32U and servo sequencer, I am using a 6V from a wall plug.
While I was running Arduino101 at 5V to drive the servo, I am using a LIPO battery at 7.4V (2S) and constant voltage regulator to set to 5V. Not sure whether it can provide 4A+ current though.
Can you please recommend a battery to use to provide 6V at 3A+ or 5V, 4A+?

Thanks, Tony

5

No, the voltage of the PWM signal is unrelated to the servo’s power.

Not necessarily.

robotshop.com/en/6v-2800mah- … ttery.html
robotshop.com/en/wh-01-wirin … ector.html

This is the reason you are having issues.

You are omitting the weight of the brackets, hardware etc. which all greatly reduce the available torque.
Note that 24Kg-cm is the maximum static torque for the 805BB - you’re really pushing its limit, which explains why it’s having trouble.

6

Hi Coleman,

Thanks! If so, it appears that I really need to reduce the arm length to 40-50cm maybe.
But I saw other kids are extending the length of the arm to a even greater length while having no issues, such as:
kennethfriedman.org/projects/robot-arm/
He is using 805BB as well.
Can you please suggest what I can do to have a longer arm AND stable control?

Another question:
“No, the voltage of the PWM signal is unrelated to the servo’s power.”
So, can I use a separate 6V power supply to drive the servos while the PWM of the servos from Arduino101 is 5V max? Shouldn’t there be a problem?

Thanks, Tony

7

The overall length fo the arm can vary, but you need to take into consideration the torque acting on each servo. When extended, the arm can reach farther over certain angles, but the worst case is stretched horizontally, where the servos are under greatest load. Vertically, the servos are under very little load.

The brackets used in the shoulder are designed to be used with the 805BB / MG (the 5805 is a digital version but provides the same torque).

We often suggest a separate power source for the servos. You can connect the servo’s signal and GND pin to the Arduino, and the servo’s voltage and GND pin to an external source.

8

Thanks Coleman!
However I do have one things to clarify:
“You can connect the servo’s signal and GND pin to the Arduino, and the servo’s voltage and GND pin to an external source.”

Then I guess that I need to split the GND from each of the servos to 2 branches, and connect 1 to external source and the other one to Arduino.
Or I can simply connect the GND of the external power supply with the Arduino to serve the same purpose with less hassle?

9

In regards to common ground, each of the servos need to be grounded with the Arduino and with the battery. Yes, you only need to connect the battery’s GND pin to the Arduino once, assuming you use the same GND for all servos.

10

Hi Coleman and all,

Please see attached video.
We reduce the arm length and the weight, yet the shake still persists.
Could you please suggest?
We are using HS 805BB as the shoulder servo, Arduino101 as the board, and separate power sources (6v for the servos).
We used about 8.4V as power supply to Arduino101
We used potentiometer as analog inputs (from 0-6V) to Arduino101 and digital output from Arduino101 to the servos.
Is so-called “digital servo” going to give a smoother movement?
Or shall I average the the analog values from potentiometer to get a smoother movement?

Thanks, Tony

11

attachment

12

No attachment. What is the max continuous current associated with the power supply?
Yes, you should average the values for smoother motion. Digital servos won’t necessarily help.

13

Hi Coleman,

Could you please check this out?
youtu.be/gIlF8GJcNoc
MP4 was not accepted and I have to use youtube.

As for the battery, we are using LIPO 2200mAh, see attached picture.

Is NiMh better than LIPO?

Thanks, Tony

IMG_20161102_095957.jpg
IMG_20161102_095957.jpg3456×4608 4.14 MB

14

We are not sure about the max continuous current of the battery.
We did use a constant voltage source (or transformer? see attached picture) to transform the 8.4V from LIPO to about 5 or 6V.
Could this be the problem?

Any suggestions?

Thanks, Tony

1478106261290_camera.jpg
1478106261290_camera.jpg3456×4608 4.61 MB

15

LiPo is generally better than NiMh these days.

What is the maximum constant output from the transformer (according to the spec sheet)?
Your best bet it to start taking average readings for the potentiometer, and also a margin of error, so if the potentiometer value varies only slightly, it won’t change the position of the servos, for a more steady signal.

16

So, in principle, the arm should NOT shake so badly, shouldn’t it?

Thanks Tony

17

Hi Coleman,

I will try 1. different power source and 2. averaging the the analog input from the potentiameters, later today because the first one needs a new battery and the second one needs some coding to do.

However, yesterday, I tried something else, and hopefully, it shreds some light on this unstable arm issue also.

  1. I used SSC-32U and servo sequencer to control the arm, and the arm appears very stable and precise and smooth.
  2. I also tried the following steps:
    a. ssc-32u to power and control the servos
    b. arduino101 to send serial output to ssc-32u through RX/TX/GND: TX of arduino to connect to RX of SSC-32U, RX of arduino to TX of SSC-32U, GND of arduino to GND of SSC-32U
    c. potentiameters as analog in to arduino101 board
    However it didn’t work out and there is no response at all for the servos

I used this simple code to test out arduino:

[code]void setup() {
// initialize serial communications at 9600 bps:
Serial.begin(9600);
}

void loop() {
Serial.println("#0 P1500 T500");
Serial.println("#1 P1500 T500");
}
[/code]

BTW, servos are connected to pin0-5.
Please help.

Thanks, Tony

18

You don’t actually need the time parameter for your project.

19

Hi Coleman,

I removed that but the servos still have not any response.
What could be the problem?
The serial communication through TX-RX should work, shouldn’t it?

Any suggestions?

Thanks, Tony

20

Same question as before - What is the maximum constant output from the transformer (according to the spec sheet)?