Well since I've been wanting to get into electronics I ended up buying a scope. It's definitely ugly output off the shunt:
Guessing the inverter or some outside factor is causing that. Though, The last one (testing the voltage between + and - on the bank) is kinda odd, for a big battery plant is it normal to see that on the voltage side? I would expect a smooth line that very slowly goes down as batteries discharge. The voltage test was done in discharge mode, so were most of the others, but TBH I kinda lost track, I should have wrote down the tests in order as I took them. The data looked fairly similar either way though. If anyone wants to see a specific reading I can always go back and take it. To test I had the probe on the far end of the shunt and the ground on the battery side (shunt is on negative side as per meter's instructions, so I could do that without causing a short through the scope).
Also, I'm wondering if this shunt meter is not unidirectional, could it be an issue or would it just go to 0 when power is going "backwards". Since it's an inverter charger the source changes direction. On AC fail the inverter-charger is the load, when AC power is on, the inverter-charger is the supply and the batteries are the load till fully charged.
It seems to me this meter is just garbage though, instead of doing some kind of averaging it just does nn readings per second and displays the literal value of that particular reading so that's why the numbers are all over the place.
I was looking at current transducers on Digikey and found a few interesting ones, this one has the most useful data sheet where I get to know the output voltage range:
http://www.digikey.ca/product-detail/en/L01Z200S05/MT7176-ND/529408These seem nice, as I don't need a shunt and it is non invasive. Removes a potential point of failure in the system as well.
Though I'm a little concerned about the actual working range. According to the graph I basically have .5v of play between 0 and 200a, and my expected load is well under 50a and currently it's a 100a system. Connecting that to the arduino not sure what the accuracy is going to be like. I want to be able to read up to 200a though, I may potentially upgrade the system in the future, and it's a good idea to be able to read past the fuse rating so I can set a proper alarm point for a unlikely situation that the fuse does not blow when it should.
There's a lot of others too but the data sheet is not clear on range, not sure how to interpret them. Open to ideas with better explanation. More can be found here:
http://www.digikey.ca/product-search/en?FV=fff4001e%2Cfff801ed%2C564006f%2C5640070%2C56400c0%2C56400c3%2C10d0001a%2C10d0001b%2C12040010%2C12040013%2C1204005b%2C12040091%2C12040096%2C120400b3&k=current+transducer&mnonly=0&newproducts=0&ColumnSort=1000011&page=1&stock=1&pbfree=0&rohs=0&quantity=&ptm=0&fid=0&pageSize=250They also have stand alone ones that are invasive but much cheaper, just not sure how you'd go about introducing that physically to #2 wiring and I don't even see how those two prongs are rated for 200A. I'm thinking these may be meant to be used with a shunt and not actually directly. Not closed to the idea of a shunt though, I think my issue is that the meter is crappy. I can either passively smooth the output or read it with an arduino or similar device and smooth it out with averaging.
I also read about op amps, could I use one in a case like this with the existing shunt? Perhaps with a potentiometer so I can tweak the range with a known load then introduce it to an arduino.
As for the safety issues, perhaps a few clarifications I need to make:
- Those plastic clips are designed for PEX pipe, it was the most appropriate thing I could find. In a proper battery bank bus bars would be used to interconnect supply/load/devices but not sure where to find those and don't have equipment to make them. PEX pipe can handle much more heat than they will ever see there. (think: hot water line)
- The system is rated for 100A but the actual load is about 30 so everything is actually quite oversized. I want to upgrade to -48v eventually which will quadruple my capacity without increasing amperage.
- RE. Fuses being closer to the battery: You do not want anything that could potentially create sparks inside the battery cabinet. I wanted to ensure that each run from the batteries are a single cable with absolutely no junctions. The purpose of the switches is to turn cut one bank out so they can be disconnected for semi annually maintenance. Checking connections for any corrosion, checking electrolyte etc... Though I imagine I could figure out a better way of routing the cabling so that the fuses are near the exit of the hole. I think I may get rid of that 2x6 and put plywood, it will give me a nice flat surface to mount wires and things on. I could even put drywall over the plywood for fire rating, but really not needed but guess it would not hurt. If I keep the shunt method, I need to keep the shunt from touching anything though, or it could alter the reading. That's why it's in mid air like that, but I guess there could be a better way of doing it.
- For those who really think this gets hot enough to cause a fire, while I don't have a thermal imaging camera to prove it I can use a rather non scientific method to prove that it's not: Coconut oil. It melts at about 25C. I can almost guarantee if I put a chunk on those cables it will not melt while the system is under load. (when AC power fails, I can throw the breaker for test for a couple hours)
I do eventually want to redo the wiring though, just so that it looks nicer. Telco flex is just really tough to work with and I should have taken more time and effort to make nicer routing. In a telco application the turn radius needed tends to be much larger than a small system like mine where everything is closer together. It typically travels on cable ladders and held in place with string or tie wraps. In my case I don't really have anything to tie wrap to so went with the pex clips as the cable fits nicely in there.