Continuous servo zero movement point of 2 servos!

[roineust]

Hello!

I have converted 2 servos of the same kind, into continuous servos.

Then i connected them to the same signal source, but was unable to make both of them NOT move at all, at the same time, using software configuration. Always one of them will move a little bit, while the other stops. Is there an electronic component, that could help me solve this problem, on the hardware side of things?

Thanks,
Roi.

 

[chuckey]

As you are keeping the servo type secret I cannot help you. It could be noise being picked up somewhere.
Frank

 

[roineust]

Oh! Sorry!

here it is:
https://todbot.com/blog/2009/04/11/tiny-servos-as-continuous-rotation-gearmotors/

And another model, very similar, a bit smaller.

BTW, from the place that the signal wire splits towards both the servos, everything is quite symmetrical, does it eliminate the possibility of noise pick up?

Thanks.

- - - Updated - - -


Well,

Apparently, you might be right - if i change a bit the position of the wires, where there is soldering, at the place that they split to both servos, sometimes both the wheels stop rolling, sometimes one, sometimes another - so the problem yet, is that it is not consistently non-movement.

Is there some way to isolate the wires better? They are already warped in shrink, each and every edge. Could one of these small ring magnets, that are found on the end of the wires of RC regulators, help somehow?

Thanks.

 

[albert22]

Once you eliminate the feedback there are not servos anymore they become just gearmotors. In normal operation the servos will stop turning when the commanded position is the same as shaft position. When you fix the feedback potentiometer to say 90 degrees the motor will stop only when you command exactly 90 degrees. This can be fine tuned in software for one motor. But if the two devices are getting the same command then you have to match the potentiometer settings.
I suggest the following>
1-Convert one of the servos as the link shows by a fixed voltage divider.
2-Adjust in software for stop.
3-Convert the other servo using a potentiometer or trimpot instead of fixed resistors.
4-Connect the 2nd motor and adjust the trimpot to stop with the same software settings.
Now both motors should stop at the same commanded position.
Take into account that keeping them moving in sync with each other is not possible with this configuration.
Hope that I was clear enough to help.
Regards

PD. you may use trimpots for both of them and and adjust both for a stop

 

[roineust]

Hey albert22!

Yeha! that's a great idea!

Wouldn't a trimpot on the signal wire itself do the job, just as well, though? (i mean on the wire, outside the servo)?

 

[albert22]

>Wouldn't a trimpot on the signal wire itself do the job, just as well, though? (i mean on the wire, outside the servo)?

No. The signal on the inputs to the RC servo is digital. The only thing you can adjust with a trimpot on it is the amplitude. The servos only respond to the pulse width.

If you find a way to disengage the feedback potentiometer from the gears you dont need anything else, just use the original potentiometer instead of the trimpot.

Anyway. I dont what they gain with that modification. I think that it may be more reliable and perhaps easier to just strip all the electronics from the servo and use and external H bridge and PWM if necessary to control the motor.

 

[roineust]

Please forgive my lack of knowledge in electronics, you might have already answered the question just above, but i don't fully understand the use of all the terms -

Yes i have stripped off the small potentiometers from the servos - you are saying that it is possible to solder it back again from the outside wires of one of the servos (the one that is still moving a bit) and achieve the stableness i am looking for?

Should i solder back the 3 wires - black red and white or do i need only the white or did i understand completely wrong the whole thing?

 

[chuckey]

I believe that these servos use pulse position feed back, but what ever. They work by your remote control signal moving a pulse inside a window. The amplifier then powers the motor, so its pot moves a different pulse to overlap the first. If they do not overlap then there is an error voltage which powers up the amplifier and moves the motor until overlap is achieved. If you connect TWO pots to the feedback circuit, you will muck up the timing completely. The easy way is not to use the feedback on one motor, hopefully it will just follow the connected one. The sophisticated way would be to build a second pulse forming/comparison/amplifier to slave the second motor of the first.
Frank

 

[albert22]

Please forgive my lack of knowledge in electronics, you might have already answered the question just above, but i don't fully understand the use of all the terms -

Yes i have stripped off the small potentiometers from the servos - you are saying that it is possible to solder it back again from the outside wires of one of the servos (the one that is still moving a bit) and achieve the stableness i am looking for?

Should i solder back the 3 wires - black red and white or do i need only the white or did i understand completely wrong the whole thing?

There is nothing to forgive.
When you modified the servo according to the instructions posted on the bloqspot. You replaced the internal potentiometer with fixed resistors. This is the same as disengaging the potentiometer from the gears. Which the guy at the blogspot didnt do because he couldn’t remove or freeze the pot without messing up the gearing.
If you were able to stop and turn the motor in your software then everything is ok.
Then you converted a second servo. I suppose that you tested the 2nd servo and worked the ok too. Probably the software settings were not the same.
Then you paralleled the two by sending the same control signal to both. And now you can stop either servo but not both.
The problem is that you have a mismatch in the fixed resistors of one servo with respect to the other. So I suggested to replace the fixed resistors with a trimpot in one of the servos or in both of them. And follow the procedure posted above.
The trimpot should be connected in the same way that the original potentiometer was connected.
The cables that connect the servos with the controller remain as they were (in parallel).
Make sure that each servo behaves as you expect when tested alone.







.

 

[roineust]

Hey Albert, thanks,
Now i think i understand completely what i should do inside the second servo!

Since weight and size are important with my application, would one of the following tiny 0.4 gram trimpots be good? :

https://www.dipmicro.com/store/RM063-204

Would the component called "variable resistor", do the same thing, as a trimpot? for example, this one:

https://www.ebay.com/itm/3-x-50K-OH...195?pt=LH_DefaultDomain_0&hash=item3a6778e45b

What kind of tech spec parameters, do i need to check out on the trimpot, for these kind of small servos we are talking about?

Hey chuckey!

The servos seem to work almost perfect and i am trying to get to perfect.
Maybe the reason that the servos are generally doing what i need, is that i took out both their trimpots? i don't know...i am experimenting with the stuff, but i don't understand a lot....

Thanks.

 

[albert22]

According to the blogspot, the servos use a 5k potentiometer so that is the value that you should use. If there is any doubt measure the original pot. Then install whatever size and shape that fits. Both of your links or similar are fine, except for their values that are very high.
Trimpot is a trade mark for a TRIMmer POTentiometer or preset and it is is just a variable resistor.
Usually the shaft of of the trimpot is locked in place with a drop of nail polish once you find the final adjust to avoid vibrations.

 

[roineust]

Great!
Just ordered a bunch of them 5K trimpots!

Thanks!