1
\$\begingroup\$

enter image description here

Hello! I was trying to control an MG996R Tower Pro servo motor using Matlab and I need to get some feedback from the internal potentiometer of the servo. However, I don't know where to connect the wire so that I can get some feedback from the potentiometer. I have attached a picture of the board of my servo. Thanks in advance!

\$\endgroup\$
1
  • 1
    \$\begingroup\$ I'm voting to close this old abandoned question as off-topic because it has never been sufficiently documented to be answerable. The feedback potentiometer is not even visible in the only photo offered. \$\endgroup\$ Jun 3, 2019 at 18:18

3 Answers 3

0
\$\begingroup\$

Generally speaking questions about modification of undocumented devices are not allowed here.

That said, it is likely that the servo amplifier wires the variable resistor as a potentiometer, which is to say applies some relatively constant voltage to the outer leads and measures the voltage on the inner lead.

If you want to "snoop" on the position you would need to connect a high-impedance ADC input to the middle lead, and perhaps also the the higher voltage outer lead to measure the scale.

Do take note that some MCU ADC inputs are not high impedance at all - especially in faster conversion modes, they can draw a fair amount of burst current. If that is the case you may need a buffer amplifier to avoid disturbing the servo amplifier's own reading.

There are also various retrofit designs/kits out there which replace the electronics of common servos with an MCU based solution, often offering a more digital communication scheme. You could potentially use one of these and modify it to report position back over a bidirectional interface scheme.

\$\endgroup\$
2
  • \$\begingroup\$ "Generally speaking questions about modification of undocumented devices are not allowed here." -- something as ubiquitous as a model aircraft servo would hardly be considered undocumented. \$\endgroup\$ Jun 3, 2019 at 13:41
  • \$\begingroup\$ @ScottSeidman - normal operation is relatively common, but that is now what the question is about - the internal implementation of a given example is exactly the kind of undocumented mystery that is not allowed here \$\endgroup\$ Jun 3, 2019 at 18:16
0
\$\begingroup\$

Every single servo that I've taken apart or built* has the V+ lead connected to one side of the pot, the GND lead connected to the other side of the pot, and takes output from the middle. There's only a few (two that I know of, or perhaps just one) analog servo amplifier IC designs out there. The new "digital" servos may do it differently**, but probably don't. (So @Chris Stratton -- now they're documented!!).

If you've replacing the servo electronics entirely, then just run three leads to the pot (it's under the electronics board in your picture). If not, then your second-best bet is to grab the center lead and ADC convert it -- this will be noisy. If you really want a good reading, then run all three pot wires back to your data acquisition system, treating them a signals independent of ground. Then measure all three (or apply ground and center tap to a differential amp, and +V and ground to a different differential amp and measure those two). The ratio of (center tap - 'gnd')/(+V - 'gnd') is the number you're looking for.

* I'm old enough to have built half a dozen Heath and Ace R/C servo kits.

** For best performance you'd want to at least RC filter the incoming power for the purposes of driving the pot, or regulate it down -- both of these would take up physical room and voltage headroom.

\$\endgroup\$
0
\$\begingroup\$

As you mentioned DC servo motors usually use a potentiometer to return position feedback for their own driver. The only thing you need is to follow potentiometer track on driver PCB. Next solder a wire to it and bring it out. Then you need to do an identification of your motor position feedback. To do this only connect the potentiometer wire having been already brought out to a voltmeter. Finally command your servo motors to some specific positions and write their related voltages. Now you will be able to obtain a linear equation which relates position and voltage. I have already done a similar project. This page helped me a lot. I hope it would be helpful for you too.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for your worthy comments. I edited my answer and have posted it as a new one. \$\endgroup\$
    – Pana
    Jun 3, 2019 at 13:07

Not the answer you're looking for? Browse other questions tagged or ask your own question.