1

I use Adruino Uno micro-controller powered with USB 5V from PC and Adafruit 16-Channel 12-bit PWM/Servo Shield on it with 1000uF capacitor, powered with 6V 2A power supply.

Seems like it is not correct power source. Need your help to choose single proper power adapter. I use 8 servo motors:

6 units of Power HD High-Speed Digital Micro Servo DSM44:

Stall Torque: 1.2Kg.cm (16.66 oz-in)@4.8v | 1.6Kg.cm (22.22 oz-in)@6v
Speed: 0.09 sec (4.8V) │ 0.07 sec (6.0V)
Operating Voltage: 4.8 ~ 6.0 DC Volts
Digital: YES

and 2 units of FEETECH Mini Servo FT1117M

Stall torque @ 4.8V: 3 kg·cm | @ 6V: 3.5 kg·cm
Speed: @ 4.8V: 0.13 sec/60°  | @ 6V: 0.11 sec/60°
Digital: NO
Ai_ve
  • 13
  • 5
  • There is no specific information in your question from which this could be determined. The problem is that the stall current of servos is far higher than their normal current. If the MCU browning out is unacceptable it may need its own supply, in any case power for the servos should not run through the Arduino, and if the Arduino regulator is used you must respect its large dropout voltage. – Chris Stratton Mar 03 '19 at 15:50
  • @Chris Stratton Hello, I can't find Stall current or any other information from this manufacturer – Ai_ve Mar 03 '19 at 15:56
  • Yes, that is the point, you are working with parts that do not have exactly known behavior. – Chris Stratton Mar 03 '19 at 16:09
  • How many servos are active at once for a large step? – Tony Stewart EE75 Mar 03 '19 at 16:20

2 Answers2

0

The vendor pages that you link to do not provide any indication as to what the running current is for the servos. The manufacturer web page is no better. There is probably a good reason that they do not provide this information considering the the current draw of a servo is so variable depending upon applied voltage and the mechanical load linked to the spindle of the servo.

The manufacturers also make consideration of the typical audience that uses these type products. If they specified a current requirement it may be used as a hard fast number by some users and the then turn into a customer support nightmare when users blame them for the numbers being either way too high or way too low.

Without actual current draw information it will be difficult for you to calculate the size of power supply that you will actually need. So you have two basic options to use.

  1. You can use a brute force technique to provide a power supply that can deliver way more current than you would ever need and wire it up with hefty sized wires to prevent voltage drops.
  2. You can characterize the current draw for your servo in your application by making measurements. Set your voltage and then operate the servo in a manner that mimics its actual usage. In some cases you may be able to make these measurements using a digital multimeter in current measuring mode which will work for steady state conditions. To get a grip on the surge current requirements you typically place a small valued resistor in series with the GND line of the servo (0.1 ohms for example) and then use an oscilloscope to monitor the voltage drop across this resistor in dynamic conditions. Once you know the voltage drop you divide that by the resistor value to determine current flow.

Lastly you will need to look at the operation scenario of your whole setup with all the servos. Not all servos are likely to be drawing worst case current at the same time. Some are likely to be idle drawing nominal current while others are levering against some load and drawing more current. There may also be some that have reached a mechanical limit and stalled drawing maximal current.

Michael Karas
  • 56,889
  • 3
  • 70
  • 138
  • Hello, here is missed by me and found by Sunnyskyguy EE75 datasheet for [FT1117M](https://www.pololu.com/file/download/FT1117M.pdf?file_id=0J1279) and [DSM44](http://www.chd.hk/UploadFiles/Att/2012111512034844.pdf?fbclid=IwAR3TntqzZWFsQvkTcseU7ifctLw-ZZFkqKnASYnVgAAVvK2ksr_ev_c0Wq4) – Ai_ve Mar 03 '19 at 18:46
0

A servo of this size, stall torque and max speed are all variables, but the half-bridge driver Ron and Motor Coil DCRs.

Ultimately determine peak current \$I_{bat}= \dfrac{(V_{bat} - V_{emf})}{(R_{on} + DCR_{coil})}\$ for the duration of time required to reach max speed then reverse current at speed might be greater with added BEMF voltage at speed velocity. This "start current" is the same as "Stall current" except it reduces as motor ramps up in speed generating -Vemf. That is it is not stalled and thus current ramps down with rising speed towards the load current.

Stall/Start current is typically 10x rated current for 100% duty of a motor, so a battery RMS current and energy Ah depends on the work being done by the servo.

When a digital command to change positions, Newton's Law for acceleration mass and slew time are limited by the torque and speed curve. The gear motor with servo feedback responds to angular position error from initial position to new command position.

I expect the or random seek or average seek over the entire range is equal to 1/3 max stroke. This is the same as in disk drive servos, however I the maximum rotational velocity depends on the load and it only reaches it with no load.

The 2nd Servo does have stall current specs ( error in your links)

No. 工作电压 Operating Voltage Range         4.8V    6V
4-1* 静态电流 Idle current(at stopped)         5mA   6mA
4-2* 空载速度 No load speed          0.13sec/60deg   0.11sec/60deg
4-3* 空载电流 Runnig current(at no load)     130mA  150mA
4-4 堵转扭矩 Peak stall torque               3kg.cm 3.5kg.cm    
4-5 堵转电流 **Stall current                 700 mA 800 mA**

Conclusions

Estimate: DSM44 servo Isc = 1A @ 6V +/-50% ... x6 motors= 6 A Specs: FT1117M servo Isc= 800 mA @6V ...... x2 motors = 1.6A for a worst case 7.6A

If 6V supply has an ESR of 50 mΩ then 7.6A x 50 mΩ= 0.38V drop from 6V.
If a supercap was added to a DCDC supply, for a 100 ms seek using an average seek.

hypothetical problem....
The equivalent capacitance for a 0.6V drop from 6V with 6A $$_C=\frac{I_c*dt}{dV}=6A/0.6V * 0.1s= 1 Farad$$

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182
  • Hello, all servos working at the same time for different range. So, 7.6 or to be sure 9Amp with 6V, but finally I found datasheet for [DSM44](http://www.chd.hk/UploadFiles/Att/2012111512034844.pdf?fbclid=IwAR3TntqzZWFsQvkTcseU7ifctLw-ZZFkqKnASYnVgAAVvK2ksr_ev_c0Wq4) which is 700 mA, should be 6A – Ai_ve Mar 03 '19 at 18:33
  • It is also possible to operate at 20% higher voltage and current if the duty cycle of motion is reduced.. – Tony Stewart EE75 Mar 03 '19 at 18:37
  • A better solution might be to determine average power and use 6V battery with float charger e.g.7.2V Li-Ion but I don't know your application – Tony Stewart EE75 Mar 03 '19 at 18:38
  • Or you could use PC 12V to 6V with a DC-DC Converter 20A 300W Step Up Step Down Buck Boost Power Adjustable Charger ($9) low cost , advantage adjustable Vout – Tony Stewart EE75 Mar 03 '19 at 18:43
  • Well, I think I will get 6V 8A DC power supply. What information would you like to know on application? I want use full speed 0.07 sec/60° of DSM44 with complete or mostly shorter degree range then 60° and acceleration of mechanical part mass between 23-223g max for each, all 6 units at the same time, and 445gr for each with 2 units of FT1117M 0.09 sec/60° by same scenario – Ai_ve Mar 03 '19 at 19:09
  • 70ms/60deg is the no-load speed . The frequency of seeks will affect motor temp rise as it cannot handle continuous oscillating seeks with that mass. Seek time may be 5x to 10x. Did you expect this? – Tony Stewart EE75 Mar 03 '19 at 19:14
  • Yes, I hope, I took into account the margin for torque and speed with choosing of this devices by size and ability process 2-3 messages per second from control application mostly for 0-30° range, and I've checked simultaneous work of all motors with load and exist 6V 2A power supply for 2 hours test, and as a result, it is true, a slight heating was felt on several most active motors, but it was not hot. I have notice that most mechanical parts for DSM44 about 23-50g with one maximal 223g, and FT1117M has a huge stop intervals in process. Anyway what would be your advice? – Ai_ve Mar 03 '19 at 19:48
  • It depends on speed vs power vs temp rise specs. I might choose DC-DC converter to vary the voltage and thus torque and sense motor temp. to improve seek time 25 to 40% e.g. 7.2V until temp rises then regulate Vout with temp. PC PSU's have the power already, you just need a converter with variable Vout and 30A vbl. Current limit for $10 – Tony Stewart EE75 Mar 03 '19 at 19:51