Lithium Polymer Battery Packs?


So the LiFePo4's would not need additional circuitry? I'm much more comfortable modding the design than I am with electronics.



For PoE would need 2-pr. Still some debate ove the effects of the additional drag with 2 pr cat-3 for various depths (should try some tests).

There could be custom approaches with Ethernet over powerline (,1.html ) that use 2 wire with modifications for lower voltage. Probably would need to have access to the chip sets for this (Qualcomm is one supplier) or work with one of the suppliers on a custom low-voltage approach.

Once I am operational in the water, I am interested in these topics.



After seeing a post about gauging capacity by mAh, I realized that I was being foolish in my comparisons of batteries. an alkaline battery with 8000mAh is different than a LiPo battery with 8000mAh. Specifically, since the LiPo produces 3.7v with 8000mAh it has more than double the power of the equivalent 1.5v Alkaline.

What really needs to be looked at is the total Watt-hours, and after some napkin math, I found that the C-style batteries picked out by the ROV team right now are actually really good.

1.5v*8000mAh*8 batteries to get 12v gives 64Wh for the usual 8 C battery setup

3.7v*4000mAh*6 batteries to get 2x11.1v gives 88.8Wh for the LiMnNi C-style batteries; better than the Alkalines and with better current supply capabilities. This means they've got ~40% more capacity than alkalines alone. Eric, have you noticed this difference? What charging setup do you currently have?

For the setup that I had planned, 4x 3.2v*10000mAh gives a whopping 128Wh, albeit for double the weight of the C batteries. Taking all of this into consideration, the LiMnNi batteries look to be the best weight/capacity ratio of them all at the moment.



What kind of power protection circuitry would we need to incorporate? Best I can tell, slipping one of these:
into each battery pack (you would have to use a slightly larger diameter tube) would provide the necessary protection with minimal alteration to the overall design.



The specifications show that the size is 10x50x3mm, so it might be possible to incorporate it within the current battery tubes (do some fun tangent math with conflicting diameters and such). The protection circuitry wouldn't be completely necessary as long as the ESCs are programmed to cut off power at low voltage levels, and the charger they supply cuts off voltage at high levels during charging.

If at all possible though, you should incorporate that circuit into the design, as it will increase the reliability of the batteries as a whole, considering they're Lithium based. What I'm worried about is over discharge protection, which shuts off the power if the load gets above 15A peak (5-15ms response time) so although 30A (2 packs) will probably never be drawn, it's always possible that it will and you'll get a whiteout on the ROV.

Since the guys at OpenROV HQ already have these batteries on-hand, it would be excellent if they would try out putting a 1cm wide by 3mm thick piece of plastic into the current tubes with these batteries to test for fit. If it doesn't work out, then yes, a larger tube must be incorporated.

Since that PCB essentially ties each battery together, it would probably be a good idea to get some large-diameter heat-shrink tube and just make 3-cell battery packs with two wires coming out and connectors on those ends, so it's easier to remove and replace the packs for charging and such. Good catch with the circuitry.



Hey guys-

I'm still spending most days out on the ice here, but the next time I get the chance, I'll try to fit a similar sized object in the battery tube with the batteries and let you know how it works out.

Since the OpenROV Cape has voltage sensing capability, it might be possible to just make the low voltage protection system work in software. Since this particular type of lithium battery doesn't fail as catastrophically as others when it is undervolted, it might even be possible to have the under-voltage shutdown be overrideable (for instance if you were in a pinch and decided it would be worth it to ruin the batteries for an extra few minuets of run time).

As for over-voltage protection, that has to be regulated during charging. The charger I'm using only charges one of the six batteries at a time, so it's not very practical at the moment. On top of that, the current battery end cap design is just barely good enough for holding out water, and whats more, there is no easy way to turn the ROV off and on, so I've resorted to unplugging the batteries each time. It seems to me that making the battery packs unpluggable using the tether connector would solve all of these problems at once. Battery packs could be oil filled since they wouldn't have to be opened (oil filling would make the current end cap design work) and it would be much easier to charge the battery packs before a dive. You could even have spare battery packs that you could just swap with. Plugging in the tether (or charger) would also be how you turn the ROV on.

I envision a tether connector on the ROV with extra conductor pins- two that relay data and two (or four depending on if the packs of three are in parallel or series) that close a circuit to turn the ROV on or charge the battery. To illustrate what I'm talking about, I've drawn a quick doodle in MSPaint. Yes I drew this with my right hand, and yes, I'm right handed.. and no I wasn't drunk when I drew it. Anyway this is just a budding thought so it may have some issues- for one, I acknowledge that this design would turn the ROV on while charging in both cases, and (this is probably more of an issue) the design for batteries in series would also put the ROV circuitry in series with the charging batteries. Just some thoughts to get the ball rolling. Let me know what you think!




I've looked a bit closely on the bulgin buccaneer connectors that just got released ( or I think these connectors can work good if we mould them on the inside. and possible play around with some thicker o-rings to get a bit more squeeze on them, usually IP68 connectors dont squeeze the seals enough to get a pressure thight seal, with increasing the string-diameter a bit we could get better squeeze and more pressure tolerance.

These are just starting to be available on RS(allied electronics in US), and when everything is available, i'll order some and do some trials.

The plastic connectors are VERY affordable, the metal cost a bit more.




I did a quick sketch in Inventor, and it looks like we'll need to increase the pack size after all, but not by much. 1 1/4 " ID should work just fine.



I think id go for a slightly bigger tube, because you need to solder some cables to the board aswell, and the solderpoints is on the edges..



What I've found is that as soon as you get larger than 1 1/4" nylon tubes, it becomes hard to find ones with thin enough wall thickness to make them economically viable; their price rapidly increases, so I searched around on mcmastercarr and found that impact resistant polycarbonate has essentially the same specs as nylon while being considerably cheaper. The endcaps would have to be changed to something different though, as this has a different ID.



reply to Springhalo;

What I've found is that as soon as you get larger than 1 1/4" nylon tubes, it becomes hard to find ones with thin enough wall thickness to make them economically viable; their price rapidly increases, so I searched around on mcmastercarr and found that impact resistant polycarbonate has essentially the same specs as nylon while being considerably cheaper. The endcaps would have to be changed to something different though, as this has a different ID.

With 8585K17 ( 1/2" ID and 1 3/4 OD ) the caps 9753K78 should fit.



For the sake of simplicity, a single six-pin connector might be the best option: two pins for data, two for battery pack access, and two for e-chassis power. Then you could have specific tethers for

  1. Battery powered operation by jumping the battery connection to the chassis power supply at the connector
  2. Surface power using a four-wire tether to send power directly to the chassis
  3. Supplemental power via a surface-side battery of equal voltage connected in parallel to the battery packs

An additional two-pin connector would allow swapping of battery packs.



Moving to a 1 1/2" ID would allow room for this PCB,

Which allows higher amperage.



The thing is that the batteries we've chosen are only rated to work at 4A, so having a 10A board will go against that and possibly harm them. The 4A cutoff is the best choice right now, plus it's got 15A burst current cutoff, so it'll allow for motor startup surges within 30A, which is plenty reasonable for 3 small motors.



Those look good. 1/8" walls should be good and the endcaps are extremely affordable (so much so that I wouldn't want to get that 25 pack!). We might have to add a tad more ballast in the front to compensate for more buoyancy, but aside from that it should be splendid.



Have you guys looked into high voltage lines? If we had a base station that output ~80v (the RMS of rectified 120v, which could be produced by a 12VDC-120V car inverter) you could get a lot more current to the ROV without having heavy power lines. All you would need on the ROV is a switching power supply that brings that 80v down to 12v or so, 'boosting' the amperage in the process.

Note that in order to trickle charge Lithium batteries you need a special charger, so a design using such an approach would probably work best with NiMh batteries linked earlier in this thread.



With the current shell design, the diameter of the battery tubes is a tight fit between the main electronics tube and the cross bars. I wonder if it might be better to look into moving power regulation electronics into the main electronics tube instead.




not to be picky, but a recitfied 120vac line gives ~170vdc (VDC=VAC*Square of 2)

Ive seen this be done on a small borehole camera system we got at work made by ECA Hytec. They use a 1:1 transformer and then rectify the mains. on the camera/crawler end they use vicor dc-dc converters that accept this voltage and transform to 12,24 or other voltages.(

downside is the cost of these modules, they cost alot in retail. but can be acquired a lot cheaper on ebay.



That's good to know. Perhaps the cross bars could be re-positioned lower in the chassis design (by making the chassis taller in the plans). I don't think that would change much aside from re-arrangement of the acrylic cutting layout, which may or may not be possible depending on how tightly packed you have the components on there (I don't have the reference on me). All We'd be looking at is an extra 1/2" of vertical room.

The power regulation PCB has to be separately connected to each battery, meaning if we were to mount it externally, there would have to be four wires coming out of each tube, which we definitely don't want.



I wasn't aware that when they call it 120VAC they are referring to the RMS of the supply. That is good to know, and even better for the tether. A switching power supply wouldn't be too hard to conjure up for inside the ROV, heck, any laptop charger that outputs 12v is a candidate; just bypass the transformer and diodes and wire it in directly to the switching regulator.

Considering I already plan to use 4-wire ethernet instead of 2 wire, an extra two for voltage transmission shouldn't be much more of a burden.