Building a servo question

tom of technology · Post by **tom of technology** » Sun Apr 30, 2006 3:45 pm

[quote="Andrew_Hibberd"]I belive that all 3 wires are pulsed. The faster the pulses the faster the motor goes. I am not too sure on the details but i think when 1 is high the others are low.[/quote]

i was wondering that but i am confused if all 3 wires are pulassed then is there a built in ic?????????????
please pm me i would really like to know

Post by **BeligerAnt** » Wed May 03, 2006 8:01 pm

Brushless motors have 3 windings, connected together in a "star" or "wye" format. End "A" of each winding comes out to the outside world. All the end "B"s are connected together inside. Like a letter "Y".

The controller IC (outside the motor) then pulses the 3 windings in sequence such that one is positive, one is negative and one is open-circuit.

So, at any one time current flows in 2 windings and the 3rd is open-circuit. The windings are arranged so that when they are pulsed in the correct sequence the magneic field that they produce appears to rotate.

tom of technology · Post by **tom of technology** » Thu Jun 01, 2006 5:54 pm

from what i know + and - are constant and the pulses come throughthe signal wire

EpicentrE · Post by **EpicentrE** » Mon Oct 23, 2006 7:37 pm

Yahar, I be hijacking this thread, since I don't like making new ones!

Anyway, My question to the electrical bods is this;

I know in most servos the pot is about 5k, However my question is, is the 5k value pre-programmed into the boards logic, and all it does is reads the resistance between one of the side legs and the centre leg to determine position, or does it read the total value of the pot from the 2 side legs, then work out the position by dividing that with the resistance between the centre and side leg? I guess a simple way of putting it would be, if you use a pot value other than 5k, would it still work?

Post by **BeligerAnt** » Mon Oct 23, 2006 8:53 pm

Hmm, not sure...
The pot' is used as a potential divider, and it is the voltage on the wiper which is used as the feedback signal - i.e. 0V = full left, 2.5V = centre, 5V = full right (or thereabouts).

In an ideal world with a perfect voltage sensor the total resistance of the pot' would be irrelevent and all would work just fine.
However, back in the real world, the feedback circuit does require some current flow and this has to come via the pot'. If the pot' resistance is too high, the circuit will "load" the output so that mid-position does NOT give half the voltage.

Depending on how accurate you need the system to be, and how far from 5k you want to go, you may be alright.
A 4.7k pot' will certainly be OK. I'd be slightly surprised if 2.2k didn't work. 10k might be alright but it would depend on the circuit. 100k would certainly be pushing your luck, but it might work if the feedback circuit is suitable (it's almost impossible to tell without testing!)

One thing to be aware of: Some servos have a resistor in series with one (or both) end(s) of the pot'. I know the Naro HP/BB does. This is to limit the feedback voltage swing so it doesn't go quite to the supply rails. If you change the value of the feedback pot' then you need to change the fixed resistor(s) in the same proportions, to keep the same voltages st each end of the pot'. So if you have a 5k pot' with a 1k resistor in series, if you double the pot' to 10k, then you have to double the fixed resistor to 2k.

The easiest way to check if a pot' is OK is to measure the voltage on the wiper with the pot' at mid-position (using the original pot'), then replace the pot' and any fixed resistors and measure the wiper voltage at mid-position. If the voltage is the same, the circuit is still working OK, if not, something is wrong and maybe the pot' value is critical. If you want to be really careful, measure the voltage at each end of the pot' (before and after) as well.

EpicentrE · Post by **EpicentrE** » Mon Oct 23, 2006 9:47 pm

Thanks for that Gary, seems to make sense. Reason why I asked is that I'm trying to make a linear actuator type thing with a sanyo-sized motor driving a screw, and ideally, I'd like it to be able to be controlled proportionally.

My idea was to use part of the track from a linear potentiometer, and have the contact for the track on the part of the actuator that moves, if that makes sense.

I did some experimenting with a S136L (as thats the board I'm planning to use) & its standard pot, and from fully one way to fully the other (i.e., how far I can get it go with the transmitter, not by hand) the resistance was 1.4k at one side, and 3.6k at the other (total reading across the pot was about 5k), so the total amount of travel was 2.2k, or 44%, with 28% of un-used travel at each side.

I would assume that if this will work, the way to do it would be to have the linear track in the centre as close to 2.2k resistance from one end to the other as possible, then work out values of resistors to have on each end of it from the percentages above.

I'll do some experimenting and let you know the results anyway

Andrew_Hibberd · Post by **Andrew_Hibberd** » Tue Oct 24, 2006 11:10 am

Sounds good scott, hope to see it at aws22

Post by **BeligerAnt** » Wed Oct 25, 2006 7:58 pm

Sounds like a 2.2k pot' with a 1.4k resistor at each end should do the trick. Certainly sounds interesting.

The only slight problem might be the limited range of physical sizes of pot's available. The "standard" size is about 75mm IIRC.

Try to use the pot' intact if you can, to avoid problems with reliability of the contact between the wiper and the track.

Good luck with it!

tom of technology · Post by **tom of technology** » Tue Nov 07, 2006 5:56 pm

it does read the 5 k on the trimpot and not just the resistance so 2-2.3k resistors will do nicely

tom of technology · Post by **tom of technology** » Wed Mar 28, 2007 2:37 pm

only works if resisttance is above 1K

RobotWars101

Building a servo question

Re: Building a servo question

servos