High VOC and charge controller, please help

[Paulcheung]

Thank you guys,

However my system has never gone to float yet, according to Rolls battery Manual, the absorb charge time would be 4.5 hours. My battery bank is fairly small compare to my load. I consume 20 to 24KWH daily and my battery bank is 750AH @48 volt. My battery supply around 10 to 12KWH daily and where I live is kind hot and always has clouds passing by.

Thanks again.

[Vic]

IMO,  there is a sweet spot at around 90-ish Volts for Vin for the Classic 150 on a 48 volt battery.

That is certainly correct for typical 'grid-tie' panels with a Vmp of 30 volts. 

I think that if you were constructing your array from true 12 or 24 volt panels, you could aim for a Vmp that is about 76 volts as a ever-so-slightly sweeter spot.   

Of course, every system is different... if you live in a very hot climate and refrigerate your batteries, a Vmp of 76 volts might not be high enough to give the controller the headroom it needs.
--vtMaps

Well,  first,  the OP indicated that 60 cell PVs were being used with a Classic 150.  I did not qualify my statement about My Opinion regarding the sweet spot.

In my limited experience,  believe that (generally)  the Classic seems to like a bit higher string Vmp than some other MPPT CCs,  like the OB MX-60.

AND,  since we are discussing Classic CCs,  which implies medium/large systems (again IMO),  is seems less likely that true 12/24 V PVs would be used,  at least for new installations.

Furthermore,  True 12 V PVs run about 17.8 - 18.25-ish STC Vmps -- 35.4 to 36.5 V for true 24 V PVs.   This is a bit short of "76 V"  when making strings of 144 cells (four 12 V or two 24 PVs).

One system here,  with Classic 150s  does run True 24 V PVs (35.4 V STC Vmp),  in strings of three for a nominal 106 V string Vmp  ...  yes,  this is a bit high,  but this system has been in service for 10 years,  and the Classics replaced a venerable MX-60.

To me,  about the best string V for a Classic 150 might  be two 80 cell Pvs (like the Kyo 315 - 325s),  but these are rare,  in the scheme of things.

Again,  just My Opinions,  YMMV,  etc.  Thanks,   Vic

[zoneblue]

I think that if you were constructing your array from true 12 or 24 volt panels, you could aim for a Vmp that is about 76 volts as a ever-so-slightly sweeter spot.

Nice vt!. Theres lots to like about 72 cells if you can cope with their phyical size. Shippings a bit of a b*tch.

seems less likely that true 12/24 V PVs would be used,  at least for new installations.

When i was shopping here 72 cell panels were cheaper than 60 cell, in tier one panels at least.

Vmps -- 35.4 to 36.5 V for true 24 V PVs.   This is a bit short of "76 V"

I woud have thought (hunch) that 40V ought to be enough headroom? Any DC converter ive ever come across was happiest when the output wasnt too far from the input. My laptops run of a 24-19V converter, and that thing is barely tepid with 3 laptops.

Here the headroom is about 30V. But i have no way to monitor this more accurately, because the classic PV amps is too squishy. It would need 2 shunts and a four channel 1% ADC logger to get a decent throughput effiency measure. I suppose we could design something and mail it to each other to test each configuration.

[offgridQLD]

I have two classic 150's both with 4000w arrays on them.  One has 2s (80v) the other 3s (120v). The 80v classic runs so much cooler for all the reasons already mentioned. I run Lifep04 lithium batterys so Im dealing with a charge voltage scope of 53V-56V. So less than the typical FLA that needs 60v+ for EQ . I will be reconfiguring the 3s arrays down to 2s this winter in preparation for a cooler running controller over summer.

My vote is keep the voltage as close to the battery voltage as you can get away with. keep the heat out of the controller.

Kurt

[mike90045]

I was limited by the wire gauge I could add to my existing conduit, and went with a 170VMP array for nominal 65V charging (NiFe 48V bank) and just with a 2Kw array, controller is toasty hot. The fans are managing it, and I'm glad I don't have to listen to them.

[boB]

I was limited by the wire gauge I could add to my existing conduit, and went with a 170VMP array for nominal 65V charging (NiFe 48V bank) and just with a 2Kw array, controller is toasty hot. The fans are managing it, and I'm glad I don't have to listen to them.

Mike, how hot is "toasty" hot ?

boB

[mike90045]

Mike, how hot is "toasty" hot ?
boB

FET's run about 48°C, almost painful to touch the case/heatsink.   

[boB]

Mike, how hot is "toasty" hot ?
boB

FET's run about 48°C, almost painful to touch the case/heatsink.

48C for the FETs ?  That's very cool actually.   UNLESS the temperature is
not reading correctly.  What is the PCB temp reading ?

Some people have more of a problem with heat than others do of course.


boB

[zoneblue]

You seem to be pretty much ideally set up to test this. Comparing current and voltage readings in and out of both controllers should give us a quick and rough guide to the actual difference in efficiency. Even just a clampmeter job. The tricky thing when i tried it was that you really need 2 clampmeters plus two voltmeters. To get readings at the same time. However if you did this around noon, on a clear day, and at a time when both controllers were mid bulk, thats about as stable a condition for the controllers as youll get. Then as quick as you can, measure all 4 currents with the clampmeter, then all 4 voltages with the same meter one at a time. Mutliply them out, should be interesting.

I have two classic 150's both with 4000w arrays on them.  One has 2s (80v) the other 3s (120v). The 80v classic runs so much cooler for all the reasons already mentioned.

[boB]

Clamp meters aren't very good for efficiency testing but would give you a ball-park figure.
They must be very close in readings to each other and of course input current is going to
be less than output current.

I like to use a camera to take pictures of the readings so then the calculations can be done
easier.

But I would use some kind of temperature measurement first because 48C just isn't
very hot for a charge controller.

boB

[CDN-VT]

Mike, how hot is "toasty" hot ?
boB

FET's run about 48°C, almost painful to touch the case/heatsink.

I know that feeling of hot & 48C on the FETs .. Im thinking my next build is to have a heat sink that's cooled into the ground rather than a wall of wood . Big steel sunk into the soil & then cemented  & drill tap the controllers to that plate & extract the heat into that steel mass. If need be have a steel heat fins above for the heat exchanger ..

VT