TOU (Time of Use) using 1.5.33 FW (firmware) if anyone knows (and More)

[sollap]

I hope the minimum SOC% change helps you. That would be a relatively easy fix. You mentioned adding batteries, are you doing an SOC reset each time you add? I recently added fully charged batteries to my existing fully charged batteries. I updated the amp hours, but I did not click the SOC reset. When I started discharging the batts, the SOC meter stayed stuck at 100%. It never changed the whole time I ran them. Once I clicked that reset a charge was forced and the SOC counter worked normally.

If you do find out that the SMA feeding the grid is a problem, I noticed something in the programming manual. See page 39. It seems you can hook up the SMA to the load terminals of the skybox and it will feed up to 7.6Kw to the grid. That's something I would not have guessed!

Thank you for the info on the CT's. That is a great price on ebay. Unfortunately, I don't think the CT's are going to prevent the selling. After watching the skybox yesterday I feel that it's using the grid as a ballast to quickly adapt to power changes that the battery is too slow in responding to.

The web interface that Mike was referring to is probably the same one I was referring to. The skybox itself hosts the pages. The display is the same as on the skybox, but it responds quicker. You can connect to whatever IP the skybox is assigned with port 3000. For instance - http://10.0.0.100:3000

[Mike Curran]

Is there another computer interface besides Optics RE? Something that could be directly tied into my local network straight to the skybox? (You refered to "web interface")

Yes. First find the local web IP address for your Skybox. Should be something like 192.168.1.XXX (fill in the X's). Then add the number 3000 with a colon, like this: 192.168.1.XXX:3000. Type that into your browser and you'll get to the Skybox web access. You'll have to enter your password.

[Chuck E.]

@Sollap and anybody else interested. I am having some moderately good luck using "demand cap enable" unfortunately the last two days have been poor days for PV, yesterday only 1/3 of typical production and today is looking to be about the same. I do want to see how this reacts under sun (not clouds) before I report any solid findings. ATM though, yesterday, charging started BECAUSE of changing my SOC to a higher value, this appears to be what triggers the charge to start (the battery actually hitting your lower value setting of SOC) without manually initiating a charge. Also appears to be a different parameter charge when Skybox initiates the charge vs manually initiating the charge, this may have been some of my issue with the skybox ignoring my loads and not keeping charge at a level that does not draw grid power (or at least minimal grid power vs a LOT).

Work has me bogged down so when I get a chance I will share a longer post with all I have discovered in the last 72 hours or so.

Again thanks to Sollap and Mike Curren for all their help and thoughts that keep driving this forward.

Chuck

[Chuck E.]

Some resolutions I have discovered that may help others: (maybe this was in the skybox manual(s) but maybe I missed it)
( I have disregarded TOU for the moment, simply because I have found something that is finally working with Demand Cap Enable)

Charge will not start on its own unless the battery hist the lower value SOC (which is adjustable and should be used with high regard of you actual battery capacity if you have any way to know.)
I will list some Empirical data here on what I know on voltage vs capacity (SOC). This data was calculated when I was cycling my battery cells to obtain capacity (and match the batteries in 14S packs). Cell discharger is an RC battery charger called an iCharger. Discharge rate is 15Amps (so know this will be a lower voltage than static voltage because of the load)
72AH cells Lithium Ion NMC cells nominal voltage rated at 3.7v:
3.3v per cell is almost empty this voltage occurs with 70AH (of 72AH capacity) of battery capacity depleted (under a 15 amp load remember)

3.62v per cell is right near the 36AH discharge point so is about 50% capacity (15A loaded)

3.77v per cell is at 21.6 AH discharged out (50.4AH remaining) so this equates to about 70% of capacity remaining.

These are my numbers for my Li-Ion NMC cells, Note they will not apply to Li-Fe AT ALL! Also its important to remember that a lower amp draw will equate to a higher read voltage and the inverse.

This is just going to be a ramble so please forgive me

A sincere Thanks to Sollap and Mike Curren for the info on logging directly into the Skybox vs Optics RE, wow, what a difference in response time. Direct connect is WAY faster.

My issue with manually initiating charge is behind me for the moment, I am running on "self Consumption" with Grid sell set to zero. I am also using Demand cap enable with the minimum grid demand set at 1.5KW and Charge limit set at 1.5KW. These seem counterintuitive to trying to have zero grid CONSUMPTION but (at sun up in particular) if I set them closer to zero there is a bouncing of using the battery to provide power, the pv curtailing and remaining on battery much longer than I want it to (wasting PV) For my application, I have found Biasing the system toward a small grid charge allows it to try to charge before entering this mobius loop, until PV fully takes over the charge.


Here are the differneces in charging I have discovered: If you initiate a charge, either by the charge button or Reset SOC button the system will charge at your maximum PV (in self consumption mode anyway) irregardless of loads or grid useage. It will continue to charge and NOT discharge again until it reaches 100% SOC or your absorb voltage (this is a mistake I was making on a regular basis!!!) HOWEVER: there can be another way I have tripped into.

With the following conditions met I have observed my battery switching between charging and discharging depending on the loads:

1.) I have CT's at my service entrance
2.)I HAD to take everything (Everything but my String inverter) off of critical loads panel (apparently Skybox cannot server two masters, it appears that when there was loads on critical loads that the skybox would DISREGARD the CT's loads reading at my service entrance)
3.) I have demand cap enabled at settings I listed above
4.) I have my battery discharge amps set to 20 amps (this will determine how much power is allowed to discharge when your demand cap is exceeded)
5.) My battery discharge is set such that my battery goes through the entire night WITHOUT reaching my low SOC (this is important!). Apparently this leaves the system in a Bi-Directional mode on the battery (which will not occur once the low battery SOC is reached which initiates a straight charge cycle Mono-directional only). What I am seeing in the AM is the battery fade from discharging, to PV starting to charge slowly, BUT if a load comes on, like a microwave oven, the skybox will immediately cover that load by reducing the charge and EVEN GOING BACK TO DISCHARGING THE BATTERY if needed. This will continue into the morning with the charge getting higher and higher and even if at 12 noon the house Air conditioning comes on, skybox will do exactly the same thing, it will reduce charge (if PV isnt enough available) and even go as far as discharging the battery again if its needed (say its cloudy an not a lot of PV). This behavior will only occur if the battery has not had a charge cycle initiated manually or by the battery reaching the lower SOC (or LBCO but hopefully not)..

Sorry for the rant, but its been a wild, journey to this point for me, maybe I'm a moron and all of this was in the manual. I didn't comprehend this from reading it. So forgive me if everybody already knows this. I didnt.

I do think one discovery is not in the manual and that is the critical loads and CT loads conflicting with each other.

Cheers
Chuck

PS I will rant again soon unless I have offended by being WAY off topic at this point.

[Mike Curran]

FWIW, here's the programming manual's take on using CTs: