Tuesday, September 3, 2013

Interesting data point on "TrustFires"

http://britnerf.co.uk/sutra7671.php#7671
According to BritNerf user Hammy:
   
I am now modifying a RS (commisson) and this one has 4 trustfires with solarbotic motors

I used my fluke multi metre to make the measurement.

When the 4 trust fires are not loaded, then they read 15.86V, however once the flywheels are spinning, this voltage drops to 6.5V. and the current flowing is 1.7A...


Whoa! That is a stupendous amount of voltage sag under a very minor load!

With the variability of Fire brands of cells and the fact that the rest of the system (including the cell holder, which is a VERY significant amount of resistance) was never described, this cannot be a useful numerical data point.

It's still a stark reminder of why these batteries and this practice are not a good idea, though. I will cover and explain the situation with 14500 cells in more detail in one of the upcoming battery guides, but suffice to say, they are a very poor choice, if you didn't already know.

The gist of it is that this type of Li-ion cell on its own, even in a best case (such as cells from the likes of Sanyo) is totally unsuitable for the uses people keep putting them to in nerf! This type of cell is designed for a 1C or 2C discharge rate - 1.6 amps at the very most - and even at that rate, the cells are sagging a good bit and heating up. Its internal resistance is far too high to perform anywhere REMOTELY near how a proper battery choice will on a mod - and it is also being abused quite severely. The quality problems of Fire cells and resistance of cell holders only compound that, perhaps significantly.

Carry on, and seriously, lay off the TrashFires!

8 comments:

  1. Thanks for the heads up. I'd seen many, many bad reviews on Amazon for them and was somewhat concerned about the "doesn't work, caught fire" posts. o_O How much was stupid user and how much was terrible QC? Little bit of both, it sounds like. Either way, definitely avoiding them for my Rapidstrike mod.

    Speaking of which, might I ask for some battery advice? I'm modding my Rapidstrike with Shark 40K's, so I'm thinking I need to hit around 12V for the flywheel circuit (splitting it off from the pusher circuit to simplify things). I'm not good enough to solder together my own cells, so I want to find a pack I can wire up off-the-shelf. Reviewing the ones available, I'm thinking LiFePO4 seems like the safest option while keeping power density high enough to be usable.

    There's quite a few pre-assembled 12V packs around: some are drop-in replacements for lead acid batteries, others for power tools, and a few of the USB battery extenders that also have a 12V output (though these are probably LiPo, they're a bit safer since it's a contained unit).

    I'm thinking a pack with 4S2P 4S2P LiFePO4 18650 cells, but with a built-in PCB for over/under voltage protection. Thoughts?


    ReplyDelete
    Replies
    1. Many of the reviews of TrustFire and similar brands of cells are not from nerfers or people trying to do what we are doing in general (which is power motors and supply rather large currents) - but from users of flashlights and similar. I would say operator headspace factors a lot into that but the quality of the cells is the ultimate cause of many of those problems. It's an overall recipe for disaster; cheap unknown cells, serious mismatch of capacity and IR, and then that is combined with cheap chargers which charge improperly, and users who overdischarge and otherwise abuse the cells.

      In nerf, I would say the reason we haven't had any serious fires or explosions from these is luck. People use them in a very risky way, and often without knowing much about Li-ion.

      Now as to your battery advice, to keep things short, you need to remember the considerations of the application. One is ability to supply the required current safely and without an excessive and performance-inhibiting voltage sag (this is why it is incorrect to use Trustfires on a nerf gun). I am not sure what the stall current of the Shark motor is at those voltages, but I would shoot for something rated for a 15A continuous discharge or higher, which indicates a low enough internal resistance to pull a high-current application and perform well. The second criterion is capacity - again, not knowing the particular motors too well, I cannot give you anything in terms of shots fired per mAh - but in general you want to be 1000mAh+ and you will probably be limited more by current capability than capacity when choosing batteries.

      Said another way, you need an RC/airsoft/power tool type of pack with a high current capability - not a low C-rate type that you would use for a light or electronics.

      I recommend you look at good old NiMH. 2/3A cells in particular can pack a big punch in a small space, sometimes rivaling lithium in energy density when the pack is all assembled. You can order custom NiMH/NiCd packs from cheapbatterypacks.com and the like using a simple online tool, and a 12.0 or 14.4V pack made to fit your RS would be a top level solution, if you cannot find an off the shelf NiMH/NiCd pack (or set of packs, used in series) to fit.

      You also have RC LiPo, which is worth considering if you need a compact, light battery. Check out HobbyKing (and if you get into any kind of Li-ion, I recommend the Turnigy Accucell 6 charger from there). LiPo takes care and has a number of requirements for safe use to avoid hazards, however. Do your research and take proper care.

      LiFePO4 is a great chemistry and there are great cells, but there are also a lot of low-discharge cells which are crap. That pack you linked, unfortunately, is designed for a 3 amp discharge. It is also a very large pack and if you are going to use a 4s2p brick of 18650, you may as well just use NiMH.

      In general, you are not going to be able to use a protection PCB on an application like this, so just forget that.

      Power tool packs as they come off the tool may or may not be suitable. I have seen people put a pack mount from a drill on a gun so that a pack can be clipped right on. That is a very slick solution, but tool packs are also large and heavy and many of the Li-ion packs have complex protection/control circuitry that may get in the way of use. So, unless you can determine that and/or mod these yourself, I would stick with RC type packs.

      Delete
    2. Yeah, I'm starting to see the current thing being my main limiter here, especially with the stall current of the motor being a huge unknown. Based on skimming some of the slot car forums, a ball-park outer limit of 15A seems good. Based on that and wanting to stuff it in the existing battery compartment (sleeper!), I'm thinking a basic LiPo RC pack should do. And since this is all in the range of auto 12V systems, I figure some kind of simple breaker should do nicely. If I'm spiking more than 20 or 30 amps, pop! But a smaller one with much less burning plastic shrapnel. >_>

      Man, that custom battery site is amazing. It's too bad they only do NiMH/NiCd (who even uses NiCd anymore?), since if they had any Li chemistry cells, I'd totally buy one from them. Getting a custom 12V NiMH with enough current looks to be around $50, while I can get an a basic Venom LiPo pack for $40.

      If someone hasn't posted a circuit diagram for a split power source setup, I may once I get it working. Pain in the butt if someone wanted to manufacture this, since people would hate changing two sets of batteries. Easier for me to debug, though, especially if I'm also throwing in a PWM controller for each circuit. Want to find the optimal settings for practical use without too much jamming/fishtailing etc.

      Delete
    3. I am guessing, based on the specs of the Shark motors, that they aren't a terribly high current setup. Most people using them in the UK are running 4S lipos. Definitely not a really low turn, current hungry motor (an example of such motor would be the Tamiya series). It's still not worth using a borderline battery, though. RC lipo and NiMH are both worth looking into; don't get too drawn by how new and cool lithium is to overlook the actual variables here.

      You can get lipos with those specs cheaper than 40 bucks. Not that I recommend it - IMO there is just too much risk with this volatile type of cell to be gambling with quality - but HobbyKing, just saying for completeness.

      On the subject of PWM controllers and separate batteries for the pusher, I recommend you take a look at using diodes placed in the positive lead feeding the 2 pusher control switches to drop some voltage and control ROF. I used 1N5551 diodes in my triple 180 motor RS build (6 of them in series to get 500RPM on 3S) but the 1N5400 series is also a good choice for feeding pusher motors. The problem with many of the PWM controllers is current handling. An undersized one would work for the pusher most likely, but I would keep absolutely everything out of the current path to the flywheel motors. You want to let those take all the current they want on startup to get crisp response. That is, after all, what the big battery is for in the first place. So, decide your battery voltage based on what the flywheels require to spin at a desired speed (sidenote, fishtailing is going to happen, unless you slow these WAY down, due to high-velocity stock darts' nature) and then use speed control methods like diodes or PWM on the pusher to make it run on that same battery.

      Delete
  2. Excuse me Toruk, I'm a big fan of your work and since I don't have any accounts on the HVZ site or anywhere I thought I'd ask you a question here. I'm looking to modify my Rayven, based off of Sg Nerf's results I decided on 12V, but I'm having trouble finding any suitable batteries. I don't think trustfires are a good idea especially after reading this. I have a place nearby that can assemble battery packs but will not do lithium, now, I have almost no experience with batteries or the like so could you recommend to me the type of battery pack I should use? any other suggestions for my Rayven would be greatly appreciated. Btw the Rayven is being used in an outdoor war without many people at all but some pretty good weaponry.

    ReplyDelete
    Replies
    1. Before I get to battery recommendations, let's talk about the Rayven.

      You will find an old mod guide (of sorts) that is on Nerf Mods and Reviews called "What's in a Rayven build" and covers the typical Rayven overhaul. Not much is involved - a simple but almost mandatory reliability mod, an optional trigger improvement, an optional motor swap, and a rewire.

      On the subject of motors, I assume you plan to use stock motors supplied with either the Elite or green model. I suggest swapping in Barricade motors; this will gain you more durability and responsiveness without being, or requiring the use of, anything expensive or exotic. Higher end motors can only gain marginally in any regard on the venerable "cades".

      One thing to remember about SGnerf's testing is that he reports battery voltage as the resting voltage of a fully charged pack - so 3 lithium-ions would be called 12.6V. This is not accurate nor comparable to battery industry standards. The voltage equivalent of that test you saw is a 3S LiPo or 9 cell NiMH (nominally 11.1 or 10.8V respectively and very close in voltage). Furthermore, trustfires and spring terminals, of course, will not even deliver their nominal voltage under the IDLING load of a flywheel gun - that idle current for Barricade motors on 3S is approximately 1.4 amps measured with one of my SAX packs and my old cade-motor Rayven of my own, and at that current, a ***Fire pack has sagged severely!

      End result here, if you use stock motors you may hit SG's results on the range and then some with one NiMH cell less voltage - 9.6V. Furthermore, cade motors rev higher than green model Rayven, and will further add velocity. Your call, but I would recommend 10.8 + cade motors for the war situation. People running any faster flywheels speed than this (think 4 trustfire) are NOT gaining any velocity over you, only noise and motor wear! Don't push cade motors (or stock motors) higher than 3S (11.1).

      Chemistry-wise, I would recommend you use NiMH if you have little experience with batteries. LiPo requires some vigilance, as well as a hard protective case for a Rayven application. NiMH can be safely velcroed on the side of the gun without fear of a fireball from a punctured lipo cell.

      If you have a pack assembler nearby, I would ask them about NiMH AA and 2/3A cells. If they don't have anything rated for at least 15A, go to cheapbatterypacks.com and get a 9 cell flat pack of Elite 1500 with deans, XT60, etc. Alternatively, you can try checking around on the internet and hobby shops for 8 or 9 cell NiMH packs.

      Don't forget the connector and charger!

      Delete
  3. One more thing that I forgot to add is that I need if possible for the pack to be rechargeable since I probably won't be able to buy a multitude of these battery packs.

    ReplyDelete
    Replies
    1. Everything we are talking about (NiMH, NiCd, LiPo, Li-ion...) is rechargeable. No one uses primary (Disposable) batteries in this kind of application. None of those can supply the current anyway.

      Delete