How to prevent power supply wire from over heating

Hi, I use the 5 Volt out from a PC PSU. the 5V output is rated 5V 21A. Now I connect it to 6 996R servo, it works fine. Then I added another 6 996R servos, the 5V wire starts the get warm. then I added another 6 996R, so now it has 18 996R servos, the 5V wire gets really really hot to the point it starts to melt the plastics insulation. To solve this problem, is it simply use a large gauge of wire or do I need to use a bigger PSU Thanks in advance.

check power supply rate ( Watt)
check total Watt of your servo.
use more then one 5V cable ,like connect 4 servo per cable..

what is amp on single servo ???

If the wire is getting hot, but the current is within power supply specs, you need thicker wire.

I am very surprised that you're having issues, though - what kind of wire are you using?

Yes, "fatter" wire. Or multiple wires.

And, make sure you have good terminations because there is often more resistance at the terminations than than in the wire.

You can find current ratings for various gauges of wire on the Net.

The length of the wire can make a difference too. Not because of the heating, but because of the voltage-drop across the wire. A longer wire will have to be thicker (lower AWG) even you are below the current rating for the particular gauge.

18 servos at once means about 18 amperes. You should probably be using gauge 10 stranded wire from the power supply. American Wire Gauge Chart and AWG Electrical Current Load Limits table with ampacities, wire sizes, skin depth frequencies and wire breaking strength

Thank you all for the replies, I will have to try out the ideas. :slight_smile: the wires I use are AWG22 or 24 solid wire from Radio Shack. Would the stranded wire be better?

Also it remind me of another question, I used a multimeter, set it to 10A to try to test the total current drawn. As soon as I hook it to the multimeter, the servos starts to act irradically then stop, I could never get a precise reading. When I check the 996R specs, the running current is 500-900mA@6V, So it seems
5V21A from the PSU is enough for the 18 servos, am I right?

Thanks again. :slight_smile:

So it seems 5V21A from the PSU is enough for the 18 servos, am I right?

Perhaps. It depends on the servos, the loads and how many are running at the same time. Fix the wiring and let us know.

rich1812:
5V21A from the PSU is enough for the 18 servos, am I right?

In theory yes, but only just, you're too close to the limit for my conservative taste..... I'd be inclined to use 2x PSUs and split them 50:50, thus giving yourself a nice margin.

Also depends if that 900mA is stall current or not. I wonder if the stall current could be more?

PS.... are you saying that you used an ammeter which you know only reads 10A on a setup that you expect to draw nearly double that? Hope you have spare fuses....

Hi thank you all for the excellent inputs, I separated the servos into two groups, each group has its own rail on the breadboard. Then I took two separated out put from the PSU, connected them. This time it seems mitigate the issue a bit, I had it running for ten minutes, both set of wires get warm or on the hot side. but it hasn't reach to the point it will melt the wires nd the breadboard. :slight_smile: I suppose to eliminate the heat is impossible, it 's a matter of reduing the heat, am I correct?

@JimboZA, according to the specs, the 996R stall current is 2.5A!!! So when the 18 servos are under load, I would need a PSU of at least 5V/45A!! For my school project, eventually it would have to be powered by battery, I am not sure where I can find a battery with that kind of power, a motorcycle battery perhaps? It seems like a catch 22 situation, the bigger the battery, the more weigh, the more weigh, the more current it takes...

I wish they make atomic battery like the one in the movie Autómata. :slight_smile:

Its easy to calculate the power dissipation in copper wire, since the resistivity of copper
is known and the dimensions (length, cross-section area) are known.

power = I^2 R - I is current in amps
R = r L / A - r is resistivity ohm-m, L length in m, A cross-section area in m^2

The resistivity of copper is 1.68 x 10^-8 or thereabouts at room temperature.

With area in square-mm its easier to use

R = 0.017 L / A

You will also find charts giving safe upper current limits for wire of each gauge
(in free air and enclosed)

rich1812:
the bigger the battery, the more weigh, the more weigh, the more current it takes…

?

JimboZA:
?

Hi, I should have stated more clearly, the project is for a walking robot. One of the requirements is that the robot has to be self-efficient, so it would have to carry its own power source i.e. the battery, it cannot be wired to an external PSU. That's why I thought with the weigh of the battery loaded, the servos would have to consume more power, hence more current. is this assumption correct?

Right now I am in the research state, so I use the PSU for testing. Eventually I would have to ditch the PSU and use a battery when I find it. :slight_smile:

Pull the battery along in a trailer behind the robot?

Ah right, I see what you mean: the heavier it is, the more power it takes to carry it self, so needs a bigger battery, which is heavier, so it needs a bigger battery etc.

CrossRoads:
Pull the battery along in a trailer behind the robot?

LOL :slight_smile: It did come across my mind. :slight_smile:

JimboZA:
Ah right, I see what you mean: the heavier it is, the more power it takes to carry it self, so needs a bigger battery, which is heavier, so it needs a bigger battery etc.

Precisely! What to do what to do... And I haven't even found a battery of that capacity yet. :frowning:

rich1812:
Precisely! What to do what to do... And I haven't even found a battery of that capacity yet. :frowning:

rich1812:
Precisely! What to do what to do... And I haven't even found a battery of that capacity yet. :frowning:

After you mentioned using Rat Shack's 24 gauge wire I would suggest to go back to design stage and not just hooking things up and watching for smoke.
You got lucky this time! But I would get a different teacher ! ( I assume it is a school project).

Also using multimeter to measure current of servo pulses is futile.
Figure out the REAL power ( current / time / load etc.) you need to get correct battery capacity.

And is NOT just Ah capacity you need to know.
For example - lead acid car battery can supply few Amps , but for SHORT time.

SAFETY FIRST!

Could look into LiPo battery backs. You can get some pretty high capacity units, and they are about the highest energy density you can get.
Example
http://www.tenergy.com/31185
Use a couple or 3 of these and to spread out the motor load. Then you can keep the wiring smaller too.

Vaclav:
After you mentioned using Rat Shack's 24 gauge wire I would suggest to go back to design stage and not just hooking things up and watching for smoke.
You got lucky this time! But I would get a different teacher ! ( I assume it is a school project).

Also using multimeter to measure current of servo pulses is futile.
Figure out the REAL power ( current / time / load etc.) you need to get correct battery capacity.

And is NOT just Ah capacity you need to know.
For example - lead acid car battery can supply few Amps , but for SHORT time.

SAFETY FIRST!

LOL, I thought that's what breadboards are for so that I can bake bread. :} Actually it was my dumb and cheap idea to use the 24 gauge wires. Now I changed to AWG 20, it works better. I still need to go buy some gauge 10 as one of the replies suggested.

CrossRoads:
Could look into LiPo battery backs. You can get some pretty high capacity units, and they are about the highest energy density you can get.
Example
AT: Tenergy 7.4V 6000mAh Heavy Duty LIPO Battery Pack w/ PCB
Use a couple or 3 of these and to spread out the motor load. Then you can keep the wiring smaller too.

CrossRoads:
Could look into LiPo battery backs. You can get some pretty high capacity units, and they are about the highest energy density you can get.
Example
AT: Tenergy 7.4V 6000mAh Heavy Duty LIPO Battery Pack w/ PCB
Use a couple or 3 of these and to spread out the motor load. Then you can keep the wiring smaller too.

Thanks for the link.

I will get one of these to try it out.