Author Topic: Live truck data feed  (Read 4867 times)

0 Members and 1 Guest are viewing this topic.

Offline MadScientist267

  • Impossible Condition Curator
  • Moderator
  • Hero Member
  • *****
  • Posts: 1514
  • Karma: +44/-4
  • Rules? What rules?
Re: Live truck data feed
« Reply #15 on: May 29, 2016, 04:10:48 am »
SN -

Apologies for the delayed response... I saw this, meant to come right back to it... and well... :-\

About all I can offer up at the moment about the monitoring is that I now view it as "not optional"... There's just too much that such a simple concept can do for a system... Most notably are balance info and safety.

For example, my 3/4 set is back and forth on behavior, and tonight, the monitoring confirmed that an internal event occurred in #4 on the way back from my early memorial day outing. Opening the door, I smelled strong sulfur, as if they'd been equalizing. Not spotting anything physically visible in the battery box, I would have been left to wonder if my nose was just playing tricks on me or had something else happened. The monitor confirmed one of the scarier scenarios, one of the cells had shorted at least long enough for the monitor to see as it swept by, indicating there's the potential for a dangerous condition inside that block. Had they been charging (thankfully they had not been for a few hours by then), that could have easily resulted in an explosion. It isn't the first time I have seen unexplained spikes on this pair, but was the first time one had been associated with the acrid scent of sulfuric acid.

Moral of the story, when you see a spike, don't ignore it. It's important to determine it's exact cause (hopefully as benign) or have reached all dead ends in the search before going into diligent mode and keep an eye out. If connections are all good, there's a hidden cause, and if it's in a cell.....

5743-0

The spike at ~23:30 represents a potentially very dangerous situation within block 4. By comparing this plot with the main rail voltage and current flow plots (not shown), an external cause has been eliminated. The "bubble" around noon was simply part of continued balancing effort on this troubled string, and is not indicative of an unknown (or dangerous) cause.

That said, my layout for monitoring isn't ideal, but I'm getting away with using a 4 channel differential input ADC board and carefully chosen divider resistors as seen above in the pics in previous posts. It wouldn't work if I were to say, rearrange the blocks as a 24V string, because of the limitations in the common components and design of the board.

String monitoring needs special considerations, and so usually there's a specially designed board or set of boards that have isolation built in to deal with these numerous voltage taps at the interconnects. I can't recommend a specific one straight out by personal use, and wouldn't recommend the board I'm using because of its cost vs versatility. I had it on hand left over from another project, so it cost me nothing to implement.

Don't let that stop you from looking for one and asking around for advice from others that have experience with a specific design - as I already mentioned, there's potentially more at stake than just satisfying curiosity :o

Sorry I don't have more to provide on it directly... this is one of those things where the quality and suitability are equally important. Bad info is worse than no info. :-\

If you do go after a roll your own approach such as what I've done, at the very least you need differential input ADC. The ADC in say, an Arduino isn't directly useful for this purpose because it is common ground (however you can make converters to go from differential to single, but IMO, it's probably easier and cheaper overall to just find a monitoring board that's suitable for your bank and be done with it.

As far as the crunch box, I'm using a Raspberry Pi2B that also performs the other truck related functions.

Steve
Wanted: Schrödinger's cat, dead and alive.

Offline solarnewbee

  • Full Member
  • ***
  • Posts: 231
  • Karma: +1/-0
    • Cayenne Dashboard
Re: Live truck data feed
« Reply #16 on: May 30, 2016, 11:45:11 am »
thanks Mad!

Interesting read indeed. I just covered an arduino lesson involving resistor voltage dividing that I immediately recognized in your pic. There are converters available on eBay that convert for various needs to rs and ttl functions that communicate serial input to arduino and raspberry. I have 2 Raspberry Pi 2 v1.1 (ordered a cap touch 10.1" screen for local use)that I'm considering using solely for battery and power monitoring and think i'll use the arduino specifically to monitor inverter functions and take actions accordingly, most likely need more than one and a data logger of some sort. I have a gsm module for big failures, maybe set it up so I can remotely reset the system. I will still be in the states after the system goes online in January '17.

I've got a long row to hoe but I'm sure I'll get there.

SN

SN

Any day above ground is a day for potential mishaps

Offline solarnewbee

  • Full Member
  • ***
  • Posts: 231
  • Karma: +1/-0
    • Cayenne Dashboard
Re: Live truck data feed
« Reply #17 on: May 30, 2016, 11:57:36 am »
PS.

My battery and high voltage panels for simple monitoring, just some cheap desktop computer boxes.

Shunts for monitoring current and voltage and donuts for inverter and grid electric(i am leary of grid tying my inverter if i'm not there)output on a simple method. May set up auto switch in case of inverter faileure. Something my daughter can use, ie., switch to grid if the inverter just refuses to play right.

Waiting with baited breath for more Mad!

SN
SN

Any day above ground is a day for potential mishaps

Offline MadScientist267

  • Impossible Condition Curator
  • Moderator
  • Hero Member
  • *****
  • Posts: 1514
  • Karma: +44/-4
  • Rules? What rules?
Re: Live truck data feed
« Reply #18 on: June 11, 2016, 06:07:47 am »
Looks like you've got a good start there...I'll be curious to see how it progresses ;)

Steve
Wanted: Schrödinger's cat, dead and alive.