Classic 150 ignoring ABSORB voltage setting?

[k7zs]

Hola from Sunny Baja.

A friend has recently installed his 2nd system using the Classic 150 as it's CC per my recommendation.   

The batteries are 8 Rolls Surrette S460 in series for 48v, there are 8 250w panels, wired as 2 pairs of 4 in series an input of around 110vdc.

Obviously, being in Baja, we don't worry about cloudy days, so we get into float every day.

We are having a problem where we cannot get the batteries to charge up to proper specific gravity, and in adjusting the ABSORB voltage to raise it, the charge controller ignores the new settings, and continues to top out right around 58 volts, in spite of my continued upward adjustments, currently set @ 59.8v.

Ambient temperature is an almost ideal 75 degrees F, the charge controller reports 24.6 degrees celcius, so there is no temperature compensation going on.  (I do wonder what the REMOTE temps means on the TEMPS screen??) 

We have not had to alter the voltmeter, it's a pretty straightforward installation, but no go on setting up a higher absorb voltage???

Firmware is a little dated, for the Classic it's 1849 dated 4/21/2014 and I will be updating that soon. 

ANY CLUES as to why this would happen, and what to do to fix it?  A complete 'power cycle' or restarting the programming sequence from a 'cold boot'?

Thanks for any hints.

Kevin XE2/K7ZS

[Vic]

Hi Kevin,

WOW!!   Strings of four 250 W PVs.   This would give a String Vmp of about 122 - 124 Volts.  Many of these 250 W PVs have a Voc of about 38 V.   This could give an unloaded Vin of about 152 volts,  which would cause the Classic to Rest,  perhaps,  all day !

What is the exact model number of the PV modules on the Classic?

AND,  you really do not want to run strings of four 60-cell PVs on a Classic 150.  That is just TOO HIGH a voltage!   It is unlikely that the 250 W PVs are 72 cells,  which would be even worse than 60-cells.   And strings of two on a 48 V battery is really a bit too low ...  So,  you might need to find one more 250 W PV,  and run three strings of three.

One other guess is to check to check the voltage settings in the Limits menu,  under Charge.

With Flooded batteries,  you really do not need these,  and at times,  they get in the way for FLAs.

Of course there are always questions,  as some things are not quite clear ...

So,  both CCs are charging the same battery,  but each of them has its own separate PV array ?

What is displayed on the Classic MNGP  --  your post implies,  that perhaps the other CC is not a Classic.  What is the Vin,  Battery V,  charge Stage,  etc?

Is EA being used,  either on the Classic itself,  or as set for the WBjr?

What brand/model is the first-installed CC?

What charge Stage,  Vin and Vbatt is shown (if it has a display)  on/from the first CC?

TIA for answers,   73,  Vic

[Westbranch]

In addition to what Vic listed, are you using the Local App to make the changes? 
If so there are some settings that need a Vulcan Mind Meld (Factiory Reset) that needs all power to be shut off , hold both   ''< ''  and   ''>''   keys down and then power on battery and then PVs.  After checking Vic's points, I would try the VMM next and see if that straightens things out.

Let us know.

Oh, it helps to write each change and result down as there are so many you can change on the fly...

[Vic]

Hi Kevin,

It looks like my last EDIT occurred after your last visit ...   I tried to add the following,  and you might not have seen  this ...

WOW!!   Strings of four 250 W PVs.   This would give a String Vmp of about 122 - 124 Volts.  Many of these 250 W PVs have a Voc of about 38 V.   This could give an unloaded Vin of about 152 volts,  which would cause the Classic to Rest,  perhaps,  all day !

What is the exact model number of the PV modules on the Classic?

Thanks,   Vic

[Westbranch]

Hmm, maybe the hot Baja sun is keeping the Vmp low?

The batteries are 8 Rolls Surrette S460 in series for 48v, there are 8 250w panels,

wired as 2 pairs of 4 in series an input of around 110vdc.


K7zs what are the PV specs?

[Vic]

Hmm, maybe the hot Baja sun is keeping the Vmp low?

The batteries are 8 Rolls Surrette S460 in series for 48v, there are 8 250w panels,

wired as 2 pairs of 4 in series an input of around 110vdc.
   ...

Yes,  WB,  that IS where I started,  with the reported around 110 V input  ...   but when the PVs become unloaded,  the Vin rises to the String Voc for the temperature of the PVs ...   this IS  NOT "around 110 V".

It does revolve around what is the PV's Voc for the cell temperature.

And,  of course,  too much guessing on just what IS the String Voc,  and Vmp is just noise,  now.

Running these Input voltages to a Classic 150 in a warm environment is not the best when thinking about efficiency and Classic operating temps,  IMO ...  We do run a String Vmp of 106 V,  and  MPPT CC's efficiency does take a hit,  especially when the PVs are lightly-loaded.

FWIW,   Vic

[k7zs]

Hi Vic and Westbranch.  Thanks for the quick replies.
(several of which arrived while I was writing this)

To clarify, only ONE Classic is managing the array.

I don't have the exact panel model info, they are made in Mexico, but I did have the specs prior to install, and used the Sizing Tool.  Came up with a .9 Classic 150's result.  Close to the edge, but within range and we've never seen it in resting mode. 

** I will endeavor to get all the panel specs.  It's a climb up a hill with lots of cactus and thorny things to navigate! **

Westbranch, I wasn't involved in the panel install (done by local professionals), but the cable run to panels is almost 200'.  At first, these 'professionals' had them wired as 4 pairs of 2, using a combiner AT THE PANELS and running a single pair of wires down to the CC.  At that point, we only had 57 volts or so showing, which of course would result in MPPT not being possible, nor reaching my voltage target.  I pointed this out and they rewired it to 2 strings of 4 (only option with 1 pair of wires coming from combiner, plus panels mounted with concrete footings etc., not practical to add 1 panel)  But, your observation of 'hot baja sun' dipping voltage is a good observation.  I will check controller for max Voc and see what that is.

I am not using the LOCAL APP to manage the settings, just the menu on the controller's MNGP.

I am sure I checked the voltage limits, but will check again and then force a reboot using the ARST ON then NEW DAY command to see if the VMM has anything to do with it!  I plan on leaving the ARST command on, - never met a microprocessor controlled machine that didn't benefit from a routine reset.

I will be back up there later to try to glean the panel specs, but that shouldn't be the root cause of the problem.

Thanks, more later!

Kevin

[Vic]

Hi Kevin,

Thanks for the added info.

OH,  and the Temps page shows the battery temp,  which is what you noted as 24.6  I would guess.   The other two temps are the FET temp (these are the switches that run the Buck Converter,  as you know),  and the other temp is that in one area of the control electronics.

It IS good that you ran the Classic String Sizer,  that implies that the Voc for those 250 W PVs is not excessive.   There is a range of Vocs,  for a given PV power level  --  almost all PVs in this power range are 60-cells,  but there is always a chance of an outlier or two.

And,  most of us see "Baja",  and think only of sweltering heat,  but believe that this is not always the case,  so we almost always ask for more information,  and so on.

73,  Vic

[Halfcrazy]

So if you adjust the Absorb voltage up, You should have to press Enter and it will say "Data sent and saved" this should be all that's needed. if it hangs still then we would want to reboot it

[k7zs]

Hi HalfCrazy!

Thanks, that is my next step...  I will be setting ARST to ON ... I do that with my Linux kernel based Routers on the LAN here and it has made for years of 'hang free' service!

-Kevin

[Vic]

So if you adjust the Absorb voltage up, You should have to press Enter and it will say "Data sent and saved" this should be all that's needed. if it hangs still then we would want to reboot it

Well,  we still run Cl FE 1849 (as does Kevin) here,  and changes to Vabs take effect immediately if the Classic is in Absorb,   BUT,  it will not be in effect for the next Solar day  --  after Arst.   Veq,  if Target EQ has been reached, also takes immediate effect,  unless there is insufficient PV power available,  or  the upper Vlimit is reached ...   just my experience.

Did mention this before,  but,   would probably be  a good idea to check the Limits menu,  to check to see if  the upper Limit is still at Default,  which could be getting in the way.  Forget the Default value ...   was it the Vabs setting when the Classic was first Commissioned ....   ??  Forget.

FWIW,   73,     GL,   Vic

[k7zs]

Hola

Update on panel values, accurate voltage readings and wiring data.

Need to have my NVRAM checked, my recollection of 110 Vin was way off - this afternoon it was 132v, against a Voc of 135v..

Panel Data:  Voc 37.6, Vmpp 29.8, Impp 8.39, Isc 8.92.  SO, Voc for 4 in series is right at MAX, 150.4v.  Therefore, losing 15v on the long wire run.  Discovered the 'pro' installer used #8 wire.  6 guage needed for only 3% loss, we are giving up 10 percent.  Might be a blessing in this case, as this is keeping the controller out of HyperVOC all the time.  Still, the provider should have recommended the 200, and we may ask for that. 

I also re-did string tool, came up with a MARGINAL on the voltage, and indicates HYPERVOC all the time.  only a .5 150 required, but again this points to swapping out the Classic 150 for a 200.

To solve the original problem ... I performed a full power down of the controller, then applying battery voltage first, then the PV.  I did put ARST to ON, and have adjusted the ABSORB to a level higher than it has ever reached before, by .5v, so tomorrow we'll see if it obeys!

The battery voltage limit was set to 61.0v, the factory default (which is also the default EQ voltage on 48v system), so not a factor.

Tomorrow I'll see if we reach a higher absorb level and report back.

Thanks  Kevin K7ZS

[Vic]

Hola

Update on panel values, accurate voltage readings and wiring data.

Need to have my NVRAM checked, my recollection of 110 Vin was way off - this afternoon it was 132v, against a Voc of 135v..

Panel Data:  Voc 37.6, Vmpp 29.8, Impp 8.39, Isc 8.92.  SO, Voc for 4 in series is right at MAX, 150.4v.  Therefore, losing 15v on the long wire run.  Discovered the 'pro' installer used #8 wire.  6 guage needed for only 3% loss, we are giving up 10 percent.  Might be a blessing in this case, as this is keeping the controller out of HyperVOC all the time.  Still, the provider should have recommended the 200, and we may ask for that. 

I also re-did string tool, came up with a MARGINAL on the voltage, and indicates HYPERVOC all the time.  only a .5 150 required, but again this points to swapping out the Classic 150 for a 200.

To solve the original problem ... I performed a full power down of the controller, then applying battery voltage first, then the PV.  I did put ARST to ON, and have adjusted the ABSORB to a level higher than it has ever reached before, by .5v, so tomorrow we'll see if it obeys!

The battery voltage limit was set to 61.0v, the factory default (which is also the default EQ voltage on 48v system), so not a factor.

Tomorrow I'll see if we reach a higher absorb level and report back.

Thanks  Kevin K7ZS

Hi Kevin,

Thanks for the update.

Believe that you are saying that,  this afternoon the indicated Voc was 135,  and the Vin was 132 V.   This could be mainly because the PVs were very lightly-loaded  (low current from the array).   Although,  if it was late in the day,  this could be part of it.

As you probably know,  the voltage drop in the cable run from the PVs is maximum when the PVs are most heavily-loaded  --  in Bulk,  and in other charge stages with some larger loads on the inverter.   Voc is not really affected by resistance in the cables feeding the PVs into the CC,  as,  by definition,   there is no load on the PVs when they are at Voc.

AND,   both  Vmp and Voc are affected by the cell temperature of the PVs.   In bright sun the cell temps rise,  and reduce these voltages.   Bright sun and high ambient temperatures further decrease these voltages.

Thanks for the specs on the PVs,  and good that you re-ran the String Calc for these data.   Your idea about a Classic 200 should be a good solution.

And glad that you checked the Max voltage Limit,  and this is not a problem.

Previously I mentioned that your CC might never come out of HyperVOC,  but in warm weather this really should not be a problem for a Classic 150.    But if the site is in the mountains there could be days where this might be the case ...

Thanks.   73    Vic

[k7zs]

OK, this is interesting (well, kind of sad really) and I am either missing something by a mile, or there is a fairly significant problem in the unit and/or firmware.

I performed a full VMM on the 150, reset the desired absorb, float and EQ voltages, made sure the limit would allow them.  That's all I did.  Come to find out it's not totally IGNORING the absorb, but it triggers absorb 1.6 VOLTS LOWER than the entry!  I set the ABSORB level for 59v, and it went  into absorb @ 57.4.  So, hmmm .... lets set the LIMIT up to 61, and set ABSORB to 61 - what happens???  ZOOM, off the controller went until the batteries reached 59.4 - 1.6v below the setting.

The voltages shown ARE the correct voltages, all agree using a DMM, a Trimetric 2025A and the 150.

SO, while I can set the ABSORB setting higher by doing the math, sure seems odd that a sophisticated piece of gear like this can stumble on such an easy task....  UNLESS I am missing something?

Anybody offer a suggestion as to WHAT might that be???

-Kevin

[Westbranch]

do you have any of the 'adjustments' turned  on or ?. .. you can check using the local app, hmmm, you are running Linux IIRC....  access to a microslop machine? or an android?  the LApp makes it real easy to look at all the factors quickly...