Here is something interesting **broken link removed** I haven't had a chance to read the whole thing, but they are talking about quashing "load dumps" which is very similar ( and more severe ) to what you're dealing with. I'll need to read it all later.
Thank you. I will read it later I promise. Right now I am immersed in reading up on Existentialism (via Pascal) which although he doesn't realize it was a train of thought from MrAl's mention of Von Neuman.
Here is something interesting **broken link removed** I haven't had a chance to read the whole thing, but they are talking about quashing "load dumps" which is very similar ( and more severe ) to what you're dealing with. I'll need to read it all later.
I am beginning to get a suspicion! for 2 days now I have not had a single trip occurrence. On contemplating any realistic differences this past two days one thing stands out; NOT A CLOUD IN THE SKY ALL DAY BOTH DAYS. I am wondering if a sudden cloud passing across the PV's and dropping output from say 50 amps to 5 amps is the culprit that is causing by battery bank to "bounce" ? I am clutching at straws I guess.
Deal with it like an automotive electrical system. Reverse polarity protection which would prevent negative transients. If any relays are being used on the 12 V side, be sure they have reverse biased diodes on the coil.
Use Transorbs or (TVS) for transient suppression, close to the suspected device (inverter?).
You could also try adding a small inductance in series with the offending devices.
No that's pretty much what you were talking about earlier. When the PV drops out the batteries kick in, that switching causes the transient. It's more like a source dump than a load dump but the same generic rules apply, a sudden change in source or sink current causes a rebound effect in the passive/active filtering of the power output. Filters are all you need but the hard part is finding out where they need to be, a native diode isolation might prevent a slapped on cap from sourcing current.
SirSparks, only if it's truly a transient.. And here we go again with these wild speculations because you can't quantify what's actually going on! There have been at least a half dozen good suggestions so far and none of them are valid unless you can determine what the actual problem is in the first place not just the effect you're perceiving.
SirSparks, only if it's truly a transient.. And here we go again with these wild speculations because you can't quantify what's actually going on! There have been at least a half dozen good suggestions so far and none of them are valid unless you can determine what the actual problem is in the first place not just the effect you're perceiving.
Without a multi channel recording oscilloscope I don't think I can do anything BUT speculate and use trial and error fixes.
If it weren't so darn annoying it would be interesting. LOL.
Water heater? Doesnt that require water running through it? I assumed you would not be home when the trip normally occurs sometimes so you would not be able to run the water, and that you would not want to waste any water, but again i dont know your entire setup so im guessing a lot of this.
If it is true that you can get away with turning the water heater on without worry no matter what time day or night or if you are home or not, then sure, it's a definite possibility It would make a really good load that's for sure, as long as it doesnt burn up for some reason like a lack of water at some time.
BTW i corrected my last post that read "three resistor banks" there would really be ten resistors banks made up of 3 ohm 50 watt resistors each bank, with ten low cost transistors to turn on each bank. That's the last resort i guess but it's something that would work at least even with the higher power you're talking about.
Good morning MrAl.
The way my heater is designed it would work fine, a pressure relief valve would blow off but ideally I would want to include a (few second) timer as a safety should the heater load jam on. I have decided to attack this problem in the following order;
#1 caps
#2 caps + choke
#3 water heater dummy load for 1 second. (remember by the time I look at battery voltage ( in less than 1 second after alarm) V is already back down)
#4 New batteries I should have a payment coming in soon and my present ones were just a cheap start up option.
Honestly i really dont think you have to buy new batteries unless the old ones wont hold a charge anymore, or for too short a time. There are other ways that are much cheaper. I hate to see you go through that expense for something this elementary.
BTW i recommended a shunt regulator not a series regulator because a series regulator will always drop some voltage, while a shunt regulator will be effectively out of the circuit when it's not needed.
I (almost) always new that day would (will) come. A few facts;
Auto batteries are good for about 200 charge-discharge cycles(calculated to 80% remaining capacity) for me that means 200 days which I am already 50% through.
RE (Renewable Energy) batteries last for 1500 cycles. (to 80%)
Auto batteries should not be discharged below 50%
RE batteries may be discharged by 80%.
A little math will show us that RE batteries are the less expensive long term deal however intial cost is very high.
I'll do a quick calc here;
My batteries are 585AH when new, will last 200 cycles to 50% discharge and cost $400.
Hence 585* 0.5 * 200 = 58500 total amps for $400 = 0.068 cents per amp.
RE batteries ; 675 AH 1500 cycles to 80% discharge and cost will be $1100.00
Hence 675 * 0.8 *1500 = 810000 total amps = 0.0135 cents per amp.
So comparing the two batteries over their respective life times the RE is 5 times less expensive.
Now that said we should remember that just a very few accidental 100% discharges will kill either battery type so that in the short term until all bugs get ironed out with the system it may be that Auto batteries would be the way to go. Anyway I must expect to be in need of replacing mine in another 100 days.
I'm glad I did that math, the results surprised me !!! I now feel better spending the money on RE
Your query about the water heater needing flow lead me to an experiment just now with surprising results; I energized the heaters w/0 water flow for 2 periods of 5 seconds each total The (tiny) copper tank for the elements increased about 15F on each 5 second burst and with no discernible pressure increase (from expansion) so it will be no problem firing this for the one to two seconds needed.
HOWEVER the first 5 seconds caused the inverter to alarm and the second one tripped it!! The log once again shows 16V for the first time in 3 days. I think we are getting closer to the root cause here and probably must exclude a dummy load as the way to fix this problem?
I did the experiment a second time and was more careful taking readings. Intially the controller was in ABSORB mode and outputting 14.8V. after 3 5 second bursts with each burst reducing V to 12.5V (only with the heater on) the controller switched into BULK mode for 30 seconds and charged at 13.2V (after heater load removed) After the 30 seconds it switched into ABSORB again and the voltage climbed to 16V tripping the inverters.
It returned to 14.8V in about 2 seconds. All this was with a PV input of about 40 amps.