Xantrex MPPT low output problems

Matthew, how old are the batteries? It should be no problem mixing different amp hr batteries, but if one or more are older and have been totally discharged, it may not be charging or holding a charge properly. This is what may be shutting the controller down, it's reading a low voltage from the bank.
  Also check for shorts or bad grounds in the wiring.
  Then try charging the batteries individually to make sure they are charging correctly. Then you may see that one or the other is not charging and causing the shut down.   
  If/When, they are both fully charged, hook them back up together and see what happens.

I've done some narrowing-down, and found out that when I leave my inverter (Prosine, 12VDC->220VAC) on overnight, I do not have the charging problem in the morning.  But when I leave my freezer or satellite internet system on overnight (through a Crydom 40A DC relay controlled by the Xantex Aux output, each with a diode installed to guard against the transient voltage from the relay), the charging problem occurs in the morning. With no loads on overnight, there is no charging problem in the morning.  So something's going on with that relay-diode setup I think.  Any ideas?

Hey Dave, thanks for the help.  I'm getting over my head, but still trying to follow you and learn in the process! My responses:

a) ok

b)could be, BUT this issue ALSO occurs when ONLY the freezer is running, and the satellite system is off.  Seems like whatever the issue is, it's common to both the freezer and satellite system (although it's possible they each have their own interference issues). Since both run through the relay and both have diodes on them, I'm thinking through whether maybe I have the wrong kind of diode?  I'm past my understanding here, so an avionics acquaintance gave me some diodes he had lying around, saying they should work.  But they may be the wrong ones. The diodes are very small black squares about the size of your finger nail, with a small metal protrusion at the top and a hole through the metal for mounting purposes.  The one on my freezer has white lettering with "LT 2420" at top, then "SBL 2040CT" or maybe it's "BBL 2040CT" (hard to see where I have it installed).  The one I've been using on my satellite system has "LT 5220" then "BBL2040CT."  Crydom's installation papers and tech tips re: DC SSRs specify certain types of diodes (e.g. "fast recovery type").  Help me out- what do I have and what do I need?  And do you agree with the Crydom instructions that "all loads should be treated as inductive even if not marked as such", and should therefore have diodes installed, e.g. DC lights?

c)(or "3") If I can figure out how to post a .bmp I'll send a diagram of how I have these diodes wired now.

d)thanks I have seen that XW bulletin and as such have not grounded the batteries elsewhere except through the XW's GFCI circuit. So that should be fine.

re: your edit: "we do need to check that the load is on the low side (negative) for those n-channel fets though?" Wow, now you've lost me.  If you can, bring it down to laymen's terms on that one. Thanks!

 readers at this point may be wondering what mosfets are and why we are talking about them...sorry about that...they are the chips inside the ssr (solid state relay)

 hey matt,

 should be OK to just use one of the outside tabs (half the diode)

how about that metal tab on top? are we touching anything with that? it's the same as the center tab electrically.

-dave

ps: good going with the sketch posting!

Mattew, you say you are leaving for a few days. I have had experience in remote villages where locals were stealing power after we left. Some even tapped in with 220 volt lines. Perhaps this is your problem too? Hopefully not. But it would explain your loss of power when everything was fine the day before and no apparant hardware problems.....anyway, just a  thought....

Matthew, just looking at your schematic. What is switching the line coming from the charge controller aux/load output to the relay? And why the relay to begin with. If you are using it for the contact rating, just to handle the current from the batteries to the freezer, why don't you just hook up the freezer directly to the batteries? It should have a themostatic switch of it's own and should be able to draw power directly from the batteries when it switches on. If there is not a secondary switch that turns on before the freezer, sending power to the relay, then the relay has continuous power coming from the controller keeping the relay contacts closed. This is where you may be losing some power. Especially if not properly grounded. You don't show any grounds in your diagram. This could also be the reason why you thought you needed to install some diodes. You have continuous power being supplied and it needs to go somewhere, so you may think that it needed to be stopped or redirected. Remember, Keep it Simple....

Hi Tom & David,
Thanks for the encouragement! 

Yes, the reason for the relay is to protect my batteries from discharging too low and damaging them or reducing their lifespan.  Yes, the idea is an automated system I can walk away from which will monitor itself to protect the batteries from over-discharge.  The Sundanzer freezer's default (not customizable) low-voltage cutout is 10.8V or 11.8V or something, which may protect the compressor, but still toasts my batteries!  So I want to keep my AGM batteries at or above the 12.2V range (~50% so I've been told), to maximize their lifespan.  Similarly, my Prosine 1000i inverter has a default (not customizable) low-voltage cutout of 10V - big help that is!  Xantrex tech support tells me if I'd sprung for the expensive XW system inverter of course I could customize its low-voltage cutout, but for my small system that extra $3600+ isn't worth it (and they agreed). 

So my plan was to use a DC relay to run all of my loads, powered by the charge controller's Aux output.  Unfortunately, finding a DC relay large enough to run my inverter through was not realistic (Amperage too high), so I opted for using 2 relays: one SSR is 12VDC up to 40A for my DC loads (some 12VDC emergency lighting, 12V freezer, 12V satellite internet system, totaling <=20A altogether); and a second SSR is 220VAC, wired between my inverter 220VAC output and my house 220VAC circuit breaker panel. I haven't tried the AC relay yet (does it need diodes too?), but first trying to work out the kinks with the DC relay.

Tom, there is nothing switching the Aux-to-relay power.  The charge controller itself switches the Aux power on or off based on battery voltage levels.  I'm currently set to "Active Lo", meaning the Aux port is normally powered (sending power to the control circuit of the DC relay, causing it to close).  The charge controller cuts Aux port power (thereby opening the DC relay) when battery bank voltage falls to 12.2V or lower for 5mins or more.  I've tested this scenario, letting my batt bank voltage drop to 12.2, and it operates correctly, cutting power to the relay control circuit until batt voltage returns to 12.4V for 5mins or more.  According to the XW charge controller's instructions, the Aux port can be used to power a relay for this very purpose. But Xantrex tech support tells me "I've never heard of anyone using a setup like this, but it should work."  Well, the XW's instructions (p. 1-8 "Auxiliary Output Functions") indicate this as a primary use of the Aux port!

Xantrex tech support suggested trying a C35 as a load controller if we can't get the DC relay setup to work, but then I still need to figure out how to manage my AC loads.  Who knows, maybe the AC relay won't have these issues! Not going there yet though!

Grounding: sorry, my simple diagram doesn't show that I have a ground running from my charge controller to a ground bus bar, then to ground.  (but not connected to the battery negative, in keeping with the XW instructions)

Relay: Is there a problem with the relay being continuously powered?

Running directly off battery: guys, we're thinking alike here! I will try running the freezer and satellite system straight from the batteries when I'm next in the village.  It's not a good long-term solution, but it will help me isolate the problem (which seems to be some kind of interference caused by the relay, but we'll see).
 
View from my power shed:


Regarding aux port output voltage: again, we're thinking alike! I just made a note to myself to try the aux port output voltage at 5VDC.  The relay instructions indicate its control circuit can be powered by 3.5VDC up to 32VDC, so I'm going to try 5V and maybe 13V (Aux port can deliver 5-13VDC) and see if there are any differences.  Hopefully something crazy like that will solve the interference issues!

Relay input Amperage:  Here's an interesting note: Crydom's instructions state that the max input current is 1.6mA @5VDC or 28mA @32VDC (I'm not sure why the current is higher for higher voltage--I would expect higher voltage to require less current...anyway...).  The XW's Aux port specifications say that it can deliver 5-13VDC and up to 200 mA!  But the amperage output is not customizable on the XW.  Could it be that the Aux port is sending way too much power (amperage) to the SSR?  Or is the current (amperage) level going to be determined by whatever the SSR "needs" or "pulls", and no more?  Next time I'm in the village I can put my Amp meter on that control circuit wire and see how much current is going through it to be sure...

Ferrite beads: could you explain? Never heard of 'em.

Charge controller status:  I need to go through the display readings in the mornings of my subsequent tests and see what the XW's says its charge state is: does it say it's in float stage when I'm getting these low readings, or does it say its in bulk, but just delivering less-than-bulk output?  I forgot to check this before.  It should only be in bulk, because my batteries have always been in the 12.2-12.4 range in the mornings after having load(s) on the night before.

That's all for now. In the mean time, trying to get back to the village soon to run more tests and finish house-building!...

Thanks for the encouragement, Dave! 

Where can I get those ferrite beads, and do you snap it on both lines?  Are you thinking I'd try putting it/them on the relay control circuit cable(s)?

Also, re: the inverter, I can run the inverter in "Powersave Mode", which will cause it to go into standby when the load is less than 10W, so I'll try it and make sure it can do that.  Sounds like it'll work well.  In standby, it draws only 1.5W.  Is your concern that with only one leg switched, I'll still have some kind of current flowing in the AC lines?  Please explain.

Looking forward to getting back to the village to test it out...

Ahh yes, keeping the fire extinguisher nearby

Yeah, man, I'm on that.  But where to get a timer for my 12V freezer?

Hey there, Tom & others. I'm back in the jungle now, messin' with my power system again.  Get this- now I've observed the problems with charging at least twice when the only load all night was my Prosine 1000i Inverter in Powersave mode.  No DC relay connection or other loads.  Since I've also observed some days with no charging problems in this same scenario in the past, and no other variables in the system that I can identify, I've contacted Xantrex to report this as probably an intermittent problem with the charger.  Interestingly, the Xantrex manual does have my symptoms in its troubleshooting section:

Problem: Low panel output. Input and output voltages read about the same.

Possible Cause: Array maximum power point is less than the nominal battery
voltage. The XW SCC is still charging, but cannot charge at Vmpp.

Solution: Check or reconfigure the array.

So one explanation may be that in the morning as the sun is rising and the
MPP is less than the nominal battery voltage (e.g. 5:30am?), the XW hangs in
this state and does not adjust as the PV array output continues to increase. Not sure since I usually start my readings at 6:30am (which is usually up to 50-60V by then), but anyway I'm waiting to hear back from Xantrex. Bummer...
 

 Gee Matthew, sounds like the darn thing is stuck in MPPT park and is not going thru sweep mode to find the ever changing max array output. there was something in the manual about setting it that way for hydro systems?

 I seem to recall a discusion somewhere about the "float" timer having weird issues on startup as well..(can't find that info again) the fix there was to switch to a two stage charging configuration.

 fingers crossed in hopes that you can get anything out of tech support...there was a bunch of info on the old xantrex forum, years worth of Q&A..all gone now after the Schneider takeover.

 isn't this the time of year when everything hits the fan out that way? the mighty monsoons!

kind regards, dave

thanks for that update matthew,

it's good to hear reports of xantrex tech support comming back to life. (been real bad for about two years now)

curious if you could tell from talking to them if this morning wakeup problem is something that they are aware of? or did it seem like they have no idea and will just replace it because it's easy?

at least you found a good "parking place" for that controller to run as a simple buck converter for now. bet it's close to harvesting the same amount as if it were doing the MPPT sweep.

cheers for now, dave