Question:what Would Cause Excessive Heat To Build Up On Hd Servos?

 
#1

Had a weird thing happen while I was testing out my HD-7 Arms the other night. The servo located at the elbow where my spring shock is located was getting real hot to the point it softened the PLA bracket. the servo was stinking of burning metal. The voltage regulator was at 7.2v. there was no stress on the servo, the spring assisted perfectly as seen in the videos on the thread(filmed after I changed the burning servo) also I noticed the voltage jumping above the regulated 7.2 I set for it. it was jumping from mid 6v to almost 8v. I was under the impression that the regulator keeps a steady flow of voltage that is pre-set. any help would be appreciated.

#2

Stress. If it's being held and it's struggling (near to or maybe over it's max torque rating) then it will use more current constantly to try and lift/move, this will cause heat. You would hear the servo buzzing/humming in this case.

Short circuit. If something inside the servo is shorted it can cause excessive heat.

Incorrect voltage. If the servo is rated for 6v and you apply 7.2v it can cause excessive heat.


Those are all pretty generic (done that way for anyone else who searches for similar and find this topic). Your situation sounds like it should be fine... well, except for the regulator. Disconnect the regulator's load (the servo side), connect your volt meter to the load and supply the regulator with whatever voltage you are using. Watch the volt meter, does it move or is it steady? The regulator should provide a steady voltage and should not go above the set voltage (it may go below when under heavy load) - at least all of my 6v regulators give a constant 6v.

What voltage regulators are you using? If they were cheap China made ones from ebay then I would look there first personally (not saying all cheap Chinese made stuff is bad but some is)

#3

Anthony, that is bad news.

I sure hope that it was simply a bad servo and not too much load on the arm. My regulators keep the same voltage until you unscrew or screw the little brass screw to adjust the voltage. I had mine set to 7.2v because that was supposed to be the most efficiant voltage.

Maybe you had a bad regulator. Out of the three that I ordered, two worked fine and the other stayed on 12v all the time. Maybe you had a bad regulator. Maybe you should adjust the voltage under a load so it doesn't change too much. Or, Adjust it without a load and see if it changes. I don't know, it is worth a try.

Maybe the arm was left on for too long. There is a Release function that you can turn on right after you get the arm where you want it to go. But if there is too much pressure on the arm, it will move.

#4
#5

@ Rich I think it could be the regulator. The servo can take upwards of 8.4v, I stay at 7.15v. So it could be the regulator. Another weird thing happened last night as I was assembling the other claw. one servo was going back and fourth on its own. no regulator. but when I applied a regulator and gave it 7.2v it was ok. what caused that? using the m995 servos.

@ mel I use the release always in the scripts to relieve stress. with the spring setup it rests in the upwards position. it only fully extends when moving to a set pattern or command, doesn't stay stretched out that long and besides as you saw in the video it worked nicely

#6

@ technopro the arm isn't that heavy. a lot of the parts are hollow to reduce weight but I use just enough infill and shells to compensate the stress.

#7

The only thing that's caused any of my servos to move on their own (other than the power supply being inadequate) was when I used a serial LCD display - when sending serial commands the servos would twitch. This was MG995s and also the smaller blue SG90 micro servos. I never did get to the bottom of that, I changed LCD to I2C in the end.

The MG995 has reports that it can take up to 8.4v but also reports that anything above 7.2v will damage them, I've not been brave enough to test which are correct.

To be honest, to save guessing and a lot of messing around the first thing I would do is remove the EZ-B (or use a spare if you have one), remove the servo that was getting hot, and remove the regulator. Connect the servo to the EZ-B so it is the only thing on the EZ-B powered by the regulator, write a simple script that loops movement and see if it gets hot. If it does then the servo or regulator are likely to be the issue. Then try without the regulator and see what happens, if it doesn't get hot then it's the regulator, if it does it's the servo.
If it doesn't get hot with the regulator and servo attached it's neither and it's time to move on to the other components.

Simple script;

Code:


$count = 0
:start
Servo(D0, 100)
Sleep(1000)
Servo(D0, 50)
Sleep(5000)
Servo(D0,0)
Sleep(1000)
Servo(D0,50)
Sleep(5000)
If($count > 25)
HALT()
EndIf
$count = $count + 1
Goto(start)



That will move the servo around from one extreme to the other, wait at 0 and 100 for a second and wait at centre for 5 seconds. It will repeat 25 times which is 5 minutes (25x12 seconds). Increase the $count>25 to increase the time it runs for if you want to test it for longer (repeating 5 times is 1 minute so 25 is 5 minutes, 50 is 10 minutes, 400 is an hour, 9600 is a day)


As for the arm's weight, have you done proper calculations for the "weight" of the arm taking in to account the length of the arm from pivot point and the weight? If it's long it doesn't necessarily need to weigh a lot for servos to struggle. Melvin's arms just about manage on MG995s powered at 6v and they are only standard Omnibot 5401 arms so just under a foot long weighing very little. I've not had the chance to read through your topic on the arm properly yet but I wouldn't rule out the weight if the calculations haven't been done.

#8

Also, Anthony, if the Battery voltage drops to where it needs to be charged. That will also cause weird symptoms on the EZB card. A fresh Battery and then good to go. That is what I experienced.
Winky

#9

@rich the servo that got hot wasn't the mg995, it was the robot HJ S3315D Torque Metal gear Digital Servo 15kg.
specs:
.Brand: HJ
.Color: Black
.Material: Alloy
.Compatible models: R/C helicopter
.Application: Run
.Size: 40mm x 20mm x 50mm
.Weight: 60g
.Wire length: About 315mm
.Speed: 0.18sec / 60 degrees at (6V)
.0.16sec / 60 degrees at (7.2V)
.Torque: 14kg.cm. at (6V)
.15kg.cm. at (7.2V)
.Voltage: 4.8V~7.2V
.Dead zoon setting :4 microseconds
.Rotation angle: Maximum 180 degrees (plug / output wheel / radio control system compatible with Futaba JR Hitec)

#10

@aameralis On the first 5DOF arm that I designed for the EZ:1 robot (now abandoned to the new Bosch motor design), I needed a minimum of a 40Kg.cm Turnigy servo at the elbow. It worked well, but when under load and made a lot of noise this was at 7.2V. This arm must have about the same weight (as your arm) further down line from the elbow ie 2 x claw servos and a wrist rotate servo, and with the 40Kg servo it just about worked, so it could well be that your elbow servo is underpowered even with your spring arrangement, and picking up any target weight (objects) is going to make it even worse. I think the servos you are referring to are 7.2V max, so they will not like > 8V across them for long, my guess is the regulator is shot, but I also think the the elbow servo could be underpowered for the job required here. Try a new regulator to see if that helps, If it does and your arm can pick up a decent weight without overheating and making noise, then your chosen servo is obviously ok.

Tony