Problem solved - using external power + servos

I build static life size robots and usually use external power supplies instead of the vex batteries. My current project drove me about crazy trying to get the external power supply to work.

I’ve used external power supplies and vex on a full scale Robby the robot and on several other robots in the past with no problems. But on this project when I turned on the controller, the controller goes bezerk and then shuts down. I finally figured out the reason is because I’m using servos. My other projects only used vex motors but this project uses several servos.

I don’t understand why there’s a difference between regular motors and servos but that’s definitely the problem. And … if I hook up only one servo it works. And … if I turn the controller on with one servo and then plug in the second servo it also works. But if you turn the controller on with two servos and powered with an external 7.2v power supply it goes bezerk.

After hours and hours of trying different servos, different controllers, different wires, different voltage regulators and power supplies, etc. I finally decided to experiment dropping the voltage. I have an adjustable voltage regulator with an LED readout of the voltage. When I dialed it down from 7.2 volts to 5.9 volts all of a sudden the problem went away. Oh and something I neglected to say is it always works with vex batteries. So maybe the batteries aren’t 7.2 volts after all. I also tried 6.1 and 6.2 and that seemed to work but what I know for sure is an external power source of 7.2v is not liked by vex controllers using more than one servo.

I just thought I’d pass this along in case anyone else had to deal with this very strange problem

Well, the servos jump when you first power them up, which means they will briefly draw a lot of current when powered up. The motors don’t do this, which could explain the difference.

Powering it up with one servo attached, but then attaching the 2nd one probably works because you are preventing them from spiking at the same time.

Note that all this is happening in the first fraction of a second after you power on the Vex controller. This is also the time that the capacitors in the Vex will need to charge up, and the voltage stabilize. I suppose power spikes during this time could cause problems running the startup code.

Batteries tend to have better surge current capacity than most bench supplies, so I’m not surprised you are seeing a difference.

As you lower the voltage, you are also reducing the amount of current the servos draw.

I wonder if you are running into a problem with the regulated supply not reacting quickly enough to changing load: The regulator will compensate for the initial current demand, but then when the servos all shut off, it may put out too much power briefly as it again reacts to the changing load. If so, then reducing the voltage would limit these excursions.

Do you have a large filter capacitor (1,000uf or greater, up to 10,000uf) on the output of your regulated supply? If not, then I suspect you will get better results by adding one, preferably close to the Vex controller. This will help smooth out any spikes, and will make the bench supply act more like a battery acts under load.

Let us know how it goes,

  • Dean

Thanks for the explanation Dean.

I’m not an expert when it comes to electronics and am still learning so can you recommend a simple large filter capacitor that I can place between the controller and the power source? I googled large filter capacitor and searched some electronics sites but it all looked greek to me.

I could just leave it at 5.9v but it would be better to eliminate the dc/dc converter and just use the external power source like I do with my other robots.

Thanks

Sure thing. This 4700uf cap from Radio Shack should do the trick. Here are a few links to capacitors from other suppliers: Jameco, Digikey, All Electronics.

You will want to wire it across the power leads, in parallel with the Vex controller. Make sure that the wires between the capacitor and the Vex are short (a foot or less) and thick (at least 18 gauge), or it will reduce its usefulness.

This capacitor is polarized, so you’ll have to make sure to connect the lead marked (-) to the negative supply and the one marked (+) to the positive supply. Other than that, just make sure it is all insulated properly and you should be good to go.

Oh, and when the capacitor is charged up, it holds a charge like a battery, and it can discharge it very quickly! It won’t be enough voltage to shock you, but if a wire or screwdriver shorts out the leads, the spark can be very surprising, and might even spot-weld or take a chunk out of the unfortunate metal object.

Right - there is a lot of information to sift through. For your purposes, an “Aluminum Electrolytic” capacitor is the right basic type. The rated voltage should be at least double the voltage you plan to run it at; 25V-50V is the right range for your needs. The capacity, measured in micro Farads (uF) just indicates how much energy it can store, and anything in the 1000uf to 10,000uf range should help smooth out any power spikes.

If there are any electronics surplus shops in your area, you might visit them. You can often pick up surplus “Computer Grade” power supply capacitors that have nice screw terminals that are easy to attach to. They will be fairly large (anywhere from the size of a D battery up through a soup can), but if space isn’t an issue, they will work very nicely for your needs. Just stick near the ranges I mention above and you should be fine.

I should also say that I’m not 100% certain this will help with the problem you are seeing, but at least it can’t hurt, so it is worth a shot.

Cheers,

  • Dean

Wow, good stuff Dean. You are a wealth of information about electronics and I appreciate you taking the time to post all the information.

I won’t try the capacitors right away as I’m about to permanently install these robots on our front porch this weekend. But it’s on the todo list when they get upgraded with several planned improvements in the near future.

Thanks again,

Try running the servos off your seperate power supply instead of going through the vex controller. Vex controller has an internal limit of 4A for all motor ports - not very useful when running high power servos.

  1. external 7.2v supply + -----> servo +
  2. external supply GND - -----> servo - AND vex controller motor port -
  3. vex controller motor port signal ----> servo signal

For safety reasons you may want to add a fuse to your external supply to limit the current or something. Depends what servos / current it normally draws, then factor 125% of that.