Continuous load drops when panels come on and off line

[mrstargate]

Hi all,

I'm new to the forum and am a first-time user of the KID. A little background; I am with a small Wireless Internet Service Provider (WISP) in southwestern Colorado. We have a few remote towers that are powered with solar. We recently decided to move up to some better quality components and a local solar outfit recommended the KID to us. We have since purchased and put one into service and overall we really like the unit.

We spent a few weeks on the bench ensuring we had all the information needed to properly set it up for our batteries, panel, etc. and everything seemed to work fine but since it wasn't actually in service there was one phenomenon happening that we didn't see until installed at the tower. The load is a set of radios that run continuously, around the clock, with about a 1A load that doesn't vary much. Everyday at sunrise and sunset, the entire load (all radios and a couple of network switches) drops for about ten minutes. We don't believe the outage is that long as the radios have to go through a power up sequence and then our monitoring system has to do it's thing to determine it is back online. We know power is being lost because of the timers in the radios. We presume this is happening when the panels switch on and offline, as they gain and loose sun. I noticed elsewhere in the forums some mention of the load connection being determined by the location of the fuse, but I could find no more information on this...it made me wonder however if there was some additional information that I wasn't able to extract from the manual or find here. I believe what we want is a 'battery connected' load, but whether or not we are properly setup for that is the question.

We considered just connecting directly to the batteries, but my concern there is that, especially with the WBJ which we are using, the controller will not properly gauge the current between the load and batteries and therefore charge functions will not happen properly.

That's pretty much the overview of our issue. If any of you have some info or input it would be greatly appreciated. Thanks much in advance!

~Orion Lukasik
AlignTec Inc.

[xsnrg]

Hi Orion,  welcome to the forum. 

If you have a WBjr, as long as all load negatives are connected such that any current flow through the shunt, you are in good shape for monitoring.  Said another way, only 1 wire should connect the shut to the battery negative, and nothing else should touch that lug on the shunt, the battery negative cable, or the negative post on the battery.  This ensures all charging and discharging is measured by the shunt.  Set up this way, you can be directly battery connected.

If you are set up on the load terminals in the KID, what are your load settings?  In particular, what is your low voltage cut off set at,  and cut in voltage?  Knowing the settings under the load menu will help diagnose what might be happening.  Also, what battery voltages are you seeing before the panels start charging in the morning?

[vtmaps]

If you are set up on the load terminals in the KID, what are your load settings? 

I think you may be onto something... If Orion is powering his radios from the load terminals, that is likely his problem.

I am much less interested in knowing his load settings than in knowing what he expects to accomplish by using the load terminals.

--vtMaps

[mrstargate]

xsnrg's question raised my own questions, that being of a pair of batteries whose health I doubt. The low voltage cut off is the only real feature that I thought I would benefit from by using the load terminals. The cutoff was set for 24V (24V system) and the battery voltage in the morning before the panels kicked in was typically sitting around 24.6 to 24.8. But since I don't trust these batteries, I don't know what actually may be happening with the voltage the instant the panels connect and disconnect.

We mean to get a replacement set up there this week anyway, and before we change anything I want the new ones up there to see what, if anything, changes. The load output is set for manual, always on except for the cutoff. I strongly suspect we will wind up just connecting the load to the batteries, and we have a separate shunt for load current that we use with a site monitor. Then our snmp software gives everything at a glance from any web browser, and sends alerts to our phones. So all in all, the low voltage cutoff is probably just unnecessary.

I'll keep you posted on how it all turns out. Thanks gents for responding.

~Orion