Classic 150 Vin Sticks Just Above Vbat

[Vic]

On one system here,  with the latest Production FW -- 1779 -- although have noticed this on prior FW as well,  have noticed that in the mid afternoon on a full sun day,  that the Classic 150 Vin will get hung at about 5 - 6 V above the battery voltage,  and appears to be missing out on about  50% of available power from the array.

Have tried,  many times,  pressing "Enter",  which caused the Vin to rise near the nominal string Vmp of about 90-ish V,  and show significant additional power produced momentarily,  and then seems to immediately sags back to 5-ish V above the battery voltage, -- about 56 V in this particular case.

This Classic is running with defaults,  except for battery charge parameters.

It appears that the two arrays on this single Classic could produce about 3200 Watts (at this time of day) a higher Vin,  but it appears that the Classic really wants to produce 1800 - 2200 W at this low Vin,  instead.

This Classic does not appear to be limiting,  and is not particularly warm.

This particular event is Bulking the bank after cycling to about 60% SOC,  so the Bulk was being done in the midafternoon.

So,  am I misreading what the array really wants to produce?  Is there a better configuration option --  am running Solar mode.  Each PV array is "identical",  on the same roof,  with no shading,  but the average cable length for each array is probably about an 8 foot difference.

Suggestions?   Thanks,   Vic

[dgd]

Vic,

I think I may have a very similar situation with one of my Classics, a Classic 250.

Several times now I have noticed the input voltage shows 33v which is 5 to 6v above the BattV and power production is well under half of what I would expect on bright sunny day with 1400watt PV array and mpv 90v.

When I start looking at MNGP and pressing keys the input rises up to over 80v and input power rises.

As I have surplus of PV power I have tended to ignore this but on one occasion went checking all connections were good and none were showing signs of heating.

dgd

[Vic]

Hi dgd,

Thanks,  this system usually has lots of excess power available.  So far,  when cycling the bank is the only time that real PV power is required.

NOTE:  after the original post in this thread,   thought about the two added Common-Mode chokes that have been on this system (although,  these chokes date to the original installation in 2011).

There is a small choke on the PV input  three or four bifilar turns on a stack of three Type 31 cores,  and  a LARGE one on the Classic's output cable to the battery -- about ten bifilar turns on two hockey puck sized cores (unknown type of probable Ferrite material).  This choke is probably considerably more turns than is required -- have not tested to see turns vs choke performance.

Had thought,  subsequent to the original post,    that these might disturb the Classic's  view of the correct Vin for Max Power ...   dunno.  All FWIW,  Thanks,    Vic

[boB]

A try this, go to that secret screen from the main status you know the one that you look for the reason for resting number...  check the numbers in the lower left corner what does it read?.

Also, does legacy mode track good?

[Vic]

Hi boB,  Thanks.

This system is remote and unattended.   Will return there within a week.

Have never needed to look at that secret status screen,  as have not ever seen any Classics Resting on the job,  but will search for where it is,   and note what it says.

Will also see how Legacy mode behaves,  and report.

Not a huge deal,   Thanks,   Vic

[boB]

Thanks Vic !

To enter the secret screen from main status...

Hold down left arrow key and then press the enter key.

boB

[Vic]

Thanks boB!!     _   .._

[dgd]

A try this, go to that secret screen from the main status you know the one that you look for the reason for resting number...  check the numbers in the lower left corner what does it read?.

Also, does legacy mode track good?

When the Classic shows input voltage just above battery voltage the secret screen shows in left side top the correct input voltage, usually over 100v, the open circuit for my 90v MPV array and the lower left number quickly cycling 4 to 3 to 2 to 1 to 0 and repeating this. After I press ENTER the main screen shows input back over 85v but this soon creeps down to about 31v again.

I just looked at the logs for the last week and the Classic doing this is reporting really low energy, usually about 0.5Kwh from a 1400watt array compared to 4.8Kwh from a 2100watt array.
I would have expected a more 40% and 60% split between them during BulkMPPT when most of the power is used.

Not sure if this is the same as Vic has 

dgd

[Vic]

Update:

Went to the site mentioned above.

Much to my chagrin,   was NOT running 1779 FW.

The Classic mentioned is still running 1608/09 from 10/4/13 (could swear that this was updated not so long ago ...)  sorry for LYING about this.

Legacy P&O  does NOT appear to get stuck,  FWIW.

In Solar Mode,  various sets of number on the screen Left --

176
0
30-52

or:
216
0
0

This may all be moot,  if the latest Production FW  has changed Tracking ...

Thanks,  SRI for misstatement on the FW Version.     Vic

[dgd]

I have been monitoring my 250 this last week, latest FW1779, and in solar mode it will often drop the input voltage to 4 to 5 volts above battery voltage and continue charging at much lower wattage. It does this when brightness reduces slightly and stays at low input voltage pv=90v, in=31v batt=27,  power 123w from 700w array. when brightness returns pv steps back up to about 90v with power over 550w.
In testing with a 150lite, fw1779 and same 700w pvs (I can reroute pvs between Classics with breakers) when brightness reduces the pv in drops to about 80v, batt=27,  power 350w.
With 250 in legacy p&o it does not do this down o 31v in thing and power is same as Lite 150.
On very similar part cloudy days the 250 energy in solar mode was 0.8 and in legay p&o 1.9,  on real bright days solar mode made 1.1 and legacy p&o made 2.1

Another 1779 150 with 2.1Kw pv never does this in volt drop to just about batt v and it runs solar mode with 5kw in per day.

I just can't understand this as it seems not a FW issue.

dgd

[Vic]

Ten days ago,  went back to the site with the system with Sticky Vin.

Legacy P&O Mode does behave better -- Vin goes to Vbatt +5-ish V less often,  and when it does get there,  spends less time there.

Updated FW to 1779 on this visit.  Spent a short time watching the performance with this FW.   The day was intermittently clouds/bright sun.   In Solar Mode,   it seemed that while Vin would still get dragged down to Vbatt + 5-ish V,  it spent less time there.   The observation period was only about 45 minutes.  There is NO monitoring of this site at this time.

Will need to spend more time watching this system.

On the visit two days ago,  installed the WBjr,  and spent no time watching the Tracking.

This site has 5.88 KW STC array,  and currently runs a single Classic,  current Limited to 73 Amps out. The string Vmp is 93 V STC.   YES,  this array size with a single Classic is not ideal,  and there is a sister Classic on the wall,  waiting to be placed back in service,  to share the work,   but just have not yet done the Follow Me cables,  etc.   Have not had the time to do that and monitor the performance.   This site is close to being on the internet,  but not quite there yet.

Will try to continue to watch the performance of this system.

As was mentioned before,  this system runs a fairly high L Common-Mode choke on the output of each Classic,  and a relatively small CM Choke on the PV in leads to the input of each Classic.  Had wondered if this L hanging on a six-foot chunk of #6 AWG cable could have some effect on either the way the Classic "sees" the PV input,  or the battery ...   dunno.   These Ls are hard-wired into the cabling to/from the Classics,  and is not easy  to just bypass or eliminate them.  Would not be the end of the world to do so,  just not easy.

dgd,  it was my guess that this was a Tracking FW  issue,   but whatta I know?

Thanks,     Vic

[dgd]

Vic,
I think that you are correct and it is some sort of tracking error.
Once the solar mode drops the input voltage to a few volts above the BattV then it gets stuck there and does not seem to attempt an mppt scan. It then appears to be locked in some sort of simple controller mode with mppt processing broken.
In effect the controller probably is working at well under 50% efficiency.
The confusing issue for me is that some Classics work well in solar mode and some do not. I do not think Legacy P&O is a reasonable alternative as Solar should be significantly more efficient.
But I do wonder how many Classics are only working better in Legacy P&O mode because of this problem with solar mode.

screenshot shows two classics in solar mode, not a bright day in bulk mppt. Note ClassicW has input V just above BattV whereas other Classic is normal inV. PV arrays on each are beside each other, no shading, equally illuminated.

dgd

[Vic]

Hi dgd,

Interesting.  For the system that sticks at low Vin,  it really makes no functional difference at this time,  as the system is lightly used,   and almost every day that is scheduled to Bulk (and fully charge the bank),  it is in Float in fairly short order.

But this system is over-PVed,  and perhaps this could make the Classic track differently (?).

The other main system has only 3.15 KW STC PV,  and basically all the rest is identical to the other system,  EXCEPT that the string Vmp is 106.5V,  vs 93-ish.  Both are 48 V.   I have never seen this system seem to stick at low Vin.

Just FWIW,   Thanks,   Vic

[boB]

Yeah, I think I know why this is happening.  (sticking close to Vbatt).

But, if the Voc is MUCH higher than battery, this sticking will have a MUCH worse effect
on power output when it is down there.

So, in most  all of your cases, I believe it sticks mainly at say, 150 watts or less ?

And above ~150 watts is usually better behaved ?

This may be fixable by firmware.  Legacy P&O will also work well if your system is having a problem
with SOLAR mode.  Either mode, when working properly, has their disadvantages because of time
spent away from the MPP voltage.  I don't think your kW-Hours will be all that much different
one way or the other, typically.

More after I have a look at something.


BTW, I have been working on a special application the last couple of days for a customer that needs
to charge an 86V Lithium Ion battery (with internal BMS) from a single 300 watt PV module at
around 40 volts Voc connected to the Classics' battery terminal.  No low voltage battery !

Remember a long time ago we talked about a boost mode ?  Well, I guess you could call this
a boost mode. I don't have an 86V Li Ion battery but am using an adjustable load on the
PV terminals to act like one.  Seems to work great.

[Vic]

Hi boB,

Thanks for the info.

On this one system with sticky Vin,  the arrays would "like" to produce about 3 KW in Bulk.   When Vin gets drawn down  just above the battery V,  the production is almost halved -- about 1700 W delivered.    The E/I curves are fairly flat,  but you are correct,  this is quite a distance from the probable ideal  Vmp  of about 83 - 85 Vmp for the solar conditions at that time  (STC String Vmp is 93 V on a 48 V battery).

Again,  for this system,  it is not a big deal.  It could produce much more power than it is asked to produce.   And,  as you stated,   Legacy P&O  reduces the frequency and duration of these low Vin events.  Just thought that I should toss this out for your consideration.

Nice to see that you have a form of Boost conversion working with a Classic!  Back to work.   Thanks again,  boB,  for all your great work.        Vic