Thinking of sending DC voltage down cable


looking at sending DC voltage down my extra 2 pair of wire in the CAT 5E cable

info on wire

cat 5e cable, stranded (max length 100m) 22

‐24 awg 0.188 ohm/meter

max current 0.577 amps

max voltage 125 vdc

my tether

length 150 ft. 45.72 meters

total resistance using above data 8.5954 ohms

since I have 2 pair not being used, I could use 1 pair for ground and 1 pair for + voltage

max current 1.154 amp from above times 2

voltage to send down at surface 22 volts

voltage drop in cable 9.919 volts = current * resistance

voltage at OpenROV 12.081

could use 2 motorcycle batteries on surface and rechargable batteries in the ROV battery pods

so I would loose half the voltage in the wire, do you think the above is right????

ROV Peak and Constant Current Draw
Hard wire for power instead of batteries on the 2.8

note that the voltage at the rov will variate with the load(amps)

with no motors running you will probably be at almost full voltage at the rov, and the "standard" esc will fry because the input circuits wont handle the voltage.

i would rather look into pumping up the voltage, and reducing it at the rov.

in commercial rovs the voltage can be in the area of 48volts and up to several thousand. at work we use 2-3KV and transfers some KW of power.

Vicor and other manufaturers deliver DC-DC converters that are small, and can deliver tons of amps.

Depending on you knowledge of electrical there are several ways of making high voltages:

rectifying mains with insulation transformer.(potential very dangerous! not recomended)

12V-120/240 inverter and rectifying this (potential dangerous aswell)

serial connecting batteries to ramp up the voltage, for the novice user i would recommend to stay under 48volts. this is the highest voltage allowed to work with for uncertified people here in Norway.

taking in consideration that the openrov can consume up to 3amps(preliminary numbers) you are forced to transfer atleast 36Watts of power(12V*3A). so with your limit in amps on the cable you need to be at atleast 31.2V to not exceed the amp-rating. and this is to high for the esc's. so you need some sort of converter to get this down to the area of 12V.

this numbers has not taken in the consideration of loss in the cable. because this needs to be calculated in relation the current consumption.

alot of these DC-DC converters accept a big variation in input voltage that helps to keep the system running if you get a high amp consumption and the voltage at the rov drops.


I am looking into using a dc-dc converter in the ROV to take 48V from a cat5 or cat6 tether.

You say that the ROV will draw about 3A at 12V.

I am just curious if this is your own empirical info, or if there is a resource I am missing somewhere, this info would be really useful in sizing the topside 48VDC power supply.

I am looking at something like this.

I am particularly interested in what kind of surges you see when all components are running at full power. I want to avoid brown outs.


the numbers of 3amps was mentioned on the forum a long time ago.

These are now the most reliable and firm numbers:

note to multiply all numbers by 3 to handle all the thrusters.


I am going to do the same kind of thing. I'm planning on 22awg CAT5, like this stuff

CAT5E insulation is rated to something like 600V, so I think a 380V converter on the topside, would let me send hundreds of watts down without much loss in the cable. I was going to use something similar to one of these military type bus converters to step down the voltage inside the ROV:

My concern is the connectors. I need a connector that seals really well at high pressure and doesn't let any water in. 380V is no joke, so even a tiny bit of salt-water in the connector could be very bad. Maybe a lower bus voltage would be better.


the cable you linked to has a rating of 125V peak.. you should not use this with the voltages youre talking about.

Dont focus on CAT5 if youre not neccesarily are transfering digital signals, no reason for a tvisted cable for power transfer. the power-line modem shouldnt need tvisted pair either(correct me if i'm wrong)

look rather for a cable with higher voltage specification, both LAPP and other companies have cables with higher voltage specs.

look at cables with PUR or rubber outer sheeting. pvc cables are hard to watertight.

PUR outer sheeting is fairly good to mould on connectors or in passtroughs.

regarding the 380V converter: is this AC or DC voltage? i guess DC since you want to run a Vicor module. are you aware of the cost of these? they are not cheap here in norway atleast...

i recently did a test at work with transfering 375VDC to a Vicor Maxi 600W brick trough 5km of 0.75mm2 cable.

with this system the voltage decreased to under 250V(minimum voltage of the brick) when the load increased over 110Watt.

on the system i tested on i also tried with 3-phase 400VAC/400Hz(from a inverter(VFD)) into a set of 400/260VAC Auto-transformers at the end of the cable, then rectified and smoothed the power before the Vicor module.

this ended up at 250Watt before the voltage dropped to low.

Note that with AC going on a long cable the power system topside will see the cable as a capacitor, this will cause a rather high current going, unless you compensate the capacitance with a drossel that makes the current go down. if you dont compensate(that is rather tricky to calculate to match) alot of power will be eaten up by the cable.

The strangest thing about a system like this, is that when the load subsea increase, the current on the cable is rather constant, and can actually decrease when the load increases!!

with a cable of 100meter or in that visinity you wont have any big issues with the cable capacitance.

what lengths of wire are you planning?

final note: DONT think about playing with voltages above 50V unless you know what youre doing and are a professional within working with voltages this high. THIS IS DANGEROUS!

"high voltage" connectors for high pressure:


You might find this on line voltage drop calculator useful:

Personally, for mini ROV's I much prefer to use on-board battery power. Keeps things much cleaner, simpler, and lower cost. But as I'm sure you already know, everything has a trade-off in ROV design. My last ROV design used DC-DC converters topside and down in the ROV (for 48 vdc). I potted the bottom side converter so it could be cooled directly by sea water. Great idea, I thought, until the converter shorted, sent 48 vdc to all my electronics and fried them. That was a costly mistake. ;)




Since I will mostly use the ROV from my boat which has 12V power outlet, I'm thinking about sending 12V down the tether. The idea is to use the external power to extend battery life. During period with low power consumption the external power will feed the ROV and charge the batteries, whilst during heavy load the internal battery packs will take care of peak loads. My only concern it with regards to charging Li-ion batteries this way since charging these batteries is a science of its own with regards of balanced charging etc. You could avoid charging of batteries by introducing a diode towards batteries so that power only goes from batteries, not into them, preferably a Germanium diode since it has less voltage drop (0.2V) compared to Silicon diodes (0.6V)

You may also need to isolate the DC component from communication boards by introducing a capacitor at each end of tether.


Hi Armand:

You have given an excellent description of the problem. The OpenROV design will eventually move to sealed battery tubes with charging over the tether, but we are only taking little incremental steps in that direction.

The next revision of the controller board (probably to be released this summer) will likely have a diode bridge that allows tether power onto the battery bus. There are already blocking diodes on the board that prevent the individual battery packs from being charged by the battery bus.

This setup will allow much longer run-times without going to the complexity of adding Lithium battery charger circuitry to the board.

For an electronics tinkerer, there are enough test pads on the existing circuit board that making these changes on your own should not be a problem.



Hi Walt, I see from previous discussions on the forum that I'm far from being the first down this road. Since I'm still new to the OpenROV community, I'll first assembly my ROV that came in the mail the other day and test it out and I'll take it from there.

I see that with the supplied 24AGW tether, you will have about 3V drop on just 200mA current on a 100m tether, so having only 12 V supply, probably won't get much useful juice.

Plenty of ways around it as discussed on forum, but all with their drawbacks in terms of increased drag due to thicker tether, more hardware or safety issues regarded high voltage. I still believe that the best is probably a combination of external power through tether, and battery backup in ROV. Otherwise the ROV will be dead in water of course if you cut tether. With batteries onboard, you could in the future program it to enter an emergency mode in the event of tether break. Once the ROV detects that tether is broken, either through lost comms or loss of power, it could be programmed to surface and blink LED's for easier detection. I take it that this is not straight forward today since I understand that today the ROV is switched on and off through power on tether. So as of today, it will be dead in the water and shut down if tether is broken. Please correct me if my understanding is wrong.

Is any power being supplied via the tether

Hi Armand:

Yes, the tether has a lot of resistance, so the voltage received at the bottom of the tether will vary wildly with the load that the ROV is providing. If you put more than ~18V on the tether you can damage the ESCs, and if you combine that voltage with the resistive drop at the tether you'll realize that you can't get much power down the tether without using a regulator of some kind in the ROV. The long-term solution is to run say 36-48V on the tether, provided by a power converter on the topside. The ROV would then have another converter, that drops the voltage back down to ~12V using a maximum-power-point tracking algorithm.

As you said, the current ROV design will turn off power if there is a cut in the tether. That doesn't mean that you have to build your own ROV that way. The schematics of the controller board are posted on GitHub, and it is relatively easy to disable this power-switching function.



You could send 50V DC (From 12s Lipo) down to the ROV. And inside of the ROV have one of these SBEC's that will output 9V/20A. It already has dual output leads for the 2 esc inside the ROV. With some more research I think you will find one that will output 12V.