Aditional 4200w of PV

[dgd]

All good info I will go with two series 20 panels total.
Kurt

Hmmm, ok, since you already use PVs in serial pairs to charge a48v bank an  you see no issues then this must be good.
I could never see this working for me, I suppose it would be the equivalent of my 24v bank being charged by a single 36v panel (or lots in parallel)
To keep the classic charging that's 24v plus a third of 24v or 32v input needed  or even higher when the bank is up to 27v. With variable weather conditions that PV would not always be near 36v.
So what I'm saying is there is vary little voltage headroom for the mppt to work with.
As for those posting that input voltage should be closer to battery voltage to increase the Classics efficiency (and generate less heat someone said) I would say 'show me the evidence'.
boB posted some time ago when asked about the efficiency of the Classic with different battery V versus different input V that it had not been tested and anyway such measurements would be very variable. I think it was in reference to the efficiency graph that was supplied with the MX 60 and why there was not a similar graph for the Classic.
Heat is not an issue AFAIK since I generally only see my C150 and C250 get warm when near max current output. Certainly negligible heat when in BULK MPPT  70A output at 28v with 75v PV input.
Maybe someone from MN can comment factually on this.

And I still think to get decent charging with enough voltage headroom for mppt to work efficiently that a 3 series string of 36v (nominal 24v) panels is the way to go 

Dgd

[Westbranch]

To clarify what I meant by 'closer to the required bank charge voltage' is:
it is more efficient to down convert from 36V (nom.) to 24V than it is to convert 96 or 72 V to 24V.

As DGD has pointed out, when you are using Gt-type panels with a Vmp that appears to be, for example, 12 v,  {vs the ones truly classed as 12 or 24 V which yield the proper voltage needed to charge one of those 2 categories of battery} one needs to make sure that in inclement weather that you will  get an adequate voltage for charging, particularly  the Absorb phase where voltage is held constant and Amperage drops.

I believe it was boB that stated a good value to target when determining how high a PV generated voltage you will use, is to target 80% of the max the CC can handle, so for the Classic 150  it's =~ 120V.   This is not an absolute but a safe level that does not stress out your equipment.
Mind you Chris Olsen has posted his experience with maxing out his classics hooked up to his Wind turbine(s). Eventful ! to say the least.

Operating voltage is not a 'use this number' as there are all the site specifics that change location to location, the one you pick is specific to you, and the one that will charge your batteries, properly and safely.

I concur with dgd as to 3 panels appears to be a good fit for a 48V battery.

[mtdoc]

They will be installed in Subtropical QLD Australia. Winter min temps 2 deg C (average min more like 12 deg C over winter) with average day time winter temps of 18 - 25 deg C. Summer temps min 20 deg C max 38 deg C and average day temp 30C. Cold weather isn't a issue in my area.

Looking at your panel data - based on the temp coefficient and VOC - with a string of 3 you'd be at 149.4 Volts at 2 degrees C.  So you'd be ok voltage wise with a string of 3 - just barely.

Apples and oranges IMO.  A string of 2 and the controller will run cooler and be more efficient but you'll need larger wire for your 30m run or lose some efficiency there and have less headroom in case of partial shading etc.

So bottom line - either 2 or 3 in a string will work - tradeoffs either way.

[vtmaps]

As has been mentioned, having three panels in series might push you into hyperVoc.  Having two in series will be easier on the classic.  Either way you will need some pretty thick cable to have an acceptable line loss. 

I forgot to mention, you might consider aluminum wire for the cable.   Thick aluminum is less expensive than thick copper.  However, there is some cost in making safe connections from the aluminum to the copper on each end of the cable.   --vtMaps

[zoneblue]

I just looked at controller losses here:
http://midniteforum.com/index.php?topic=1316.msg10518#msg10518

and found that they are actually quite a bit higher than expected for higher pv voltages, at least with the CL150.

Given the size of your arrays, and the distances, working at much higher dc array voltages seems where you are headed though...

I would say that your application sounds to me like as good an example of where ac coupling would work as any.  We live in changing times, and ac coupling is gaining ground.

The downside is starting again on all your gear though.

[jimbo]

So what are they like under 90V?  I played around with my wiring yesterday and had just under 70V feed the classic 150charging my 24v bank

[offgridQLD]

I wont be getting any shading on the pv in my location. When I need to make 60v is during the (absorb stage) usually mid morning and I don't see a situation where I couldn't make 60v from two 36.9v pv's that I listed before. Though I could be wrong.

This is the specs of the 24 panels wired in two series that have now on my roof. they have been working fine in my local conditions on my classic 150.




From what I can see they are actually slightly lower voltage than the new PV I want to install.

The new pv has a Power temperature coefficient    -0.40%/K so dose this mean for every 10deg C rise in temp over 25C. I will get 4% drop in voltage?

Heat is way more of A issue in my location (subtropical QLD Australia) than cool weather. My thinking is typical summer weather is 30 deg C - 38 deg C. Now how hot would the pv be on the roof 55 deg C or more? that could be a 15% voltage drop. so 62.7V for the new PV.

As for the voltage of 3 series being on the cuff for the classic 150 at 2 deg C. It would never be 2 deg C in the day time in my area. Perhaps a once in 100 year record would be 2C over night .

I have ordered 21 panels pick them up next Tue. If I have to go 3 series I can with 3x7 . If not I have one spare PV (not such a bad thing I guess if one got damaged)

Kurt

[vtmaps]

I have ordered 21 panels pick them up next Tue. If I have to go 3 series I can with 3x7 . If not I have one spare PV (not such a bad thing I guess if one got damaged)

If you go three in series I predict that your Classic will be running VERY hot. 
This is because:
1) the Classic is less efficient at the higher voltage
2) you are operating at its upper power limit and more power = more heat (for any given efficiency)
3)  the Classic is less efficient at its highest power (for any given input voltage)
4) you are operating in a relatively high ambient temp

--vtMaps

[Halfcrazy]

Boy that is a tough one. The comments made by VTmaps is spot on "But" being in Australia I also worry the 70ish volts VMP may sag enough to be an issue in the heat. That said I would agree with VTmaps and say definately wire 2 in series and try it.

Ryan

[offgridQLD]

Well I am counting on the fact that I haven't had any low voltage issues with my 3960W system that's wired in two series and Last summer was a hot one.

That said perhaps the old Polycrystalline BP panels I have up there now are better in the heat than the new additional Mono ones I am getting

The old BP Polycrystalline Cell specs (that did fine in 2 series over summer  hit 38deg C a few days over summer)

Temperature Coefficiency of Isc    0.065 %/ºC
Temperature Coefficiency of Voc    -0.36 %/ºC
Temperature Coefficiency of Pmp    -0.50 %/ºC

vs new Mono cell specs

Power temperature coefficient    -0.40%/K
Open-Circuit Voltage Tempreture Coefficiency  -0.31%K
Short-Circuit Voltage Tempreture Coefficiency  0.06%K

The new Mono's are 36.7V x2 = (73.4V)     VS    35.2V  x2 = (70.4V) for the old Polycrystalline's

Kurt

[dgd]

If you go three in series I predict that your Classic will be running VERY hot. 
This is because:
1) the Classic is less efficient at the higher voltage
2) you are operating at its upper power limit and more power = more heat (for any given efficiency)
3)  the Classic is less efficient at its highest power (for any given input voltage)
4) you are operating in a relatively high ambient temp

--vtMaps

1). Midnite have never produced efficiency figures/graphs  showing batt volts vs input volts  so where is the evidence of this?
2). operating the classic at the upper power limit should have no bearing on whether the PVs are in strings of 2 or 3. All the voltages of a 3 string are well within the operational parameters of a classic 150
3). Eh? Again no evidence for this within its working limits. Where has MN published this info?
4). Ok, but this should not be factor in the decision on 2 or 3 nominal 24v panels in series.

I would say your predictions are unlikely and the advantages of giving the classic some voltage headroom for mppt to work efficiently far outweighs any perceived unproven classic temperature
Issues.

Dgd

[dgd]

Boy that is a tough one. The comments made by VTmaps is spot on "But" being in Australia I also worry the 70ish volts VMP may sag enough to be an issue in the heat. That said I would agree with VTmaps and say definately wire 2 in series and try it.

Ryan

Spot on eh?
So does this mean that classic 200 and 250 designed to allow much higher input voltages are significantly less efficient than a 150?
Does Midnite recommend keeping PV string voltages just above battery voltage to get maximum efficiency from a classic?
You seem to be recommending two nominal 24volt panels in series to charge a 48volt battery bank

Dgd

[Halfcrazy]

Boy that is a tough one. The comments made by VTmaps is spot on "But" being in Australia I also worry the 70ish volts VMP may sag enough to be an issue in the heat. That said I would agree with VTmaps and say definately wire 2 in series and try it.

Ryan

Spot on eh?
So does this mean that classic 200 and 250 designed to allow much higher input voltages are significantly less efficient than a 150?
Does Midnite recommend keeping PV string voltages just above battery voltage to get maximum efficiency from a classic?
You seem to be recommending two nominal 24volt panels in series to charge a 48volt battery bank

Dgd

Significantly less no not at all but slightly less efficient yes. Think of it this way take a 10 watt light bulb and close it inside an aluminum case. after 2-3 hours the case will get hot. So 0.5% in efficiency on a 3000 watt array would be like sticking a 10 watt light bulb in the case so that would be a fair bit of heat and this is exactly why the 200 and 250 drop of in current ratings. So as you can see a 10 watt hit in efficiency with a classic running at max load of 3-4KW to run a smaller wire for a long distance would be trivial but it is also heat so that is why we had to make the 200 and 250 handle less Amperage. So for those systems with long wire runs a slight hit in efficiency is well worth the major savings in copper. I worked on a system design the other day that was like 750 ft of wire and needed 4 Classics. We where talking thousands of dollars difference in Copper when looking at using Classic150's or 200's or 250's so it really simply comes down to what is correct for that particular install. I have used statements like "If the array is 0-100ft away use a 150 more than that use a 200 or 250 depending on wire size etc"

In a perfect world I would go 3 panels in series on a 48 volt battery when using the standard 37VOC panels but in this case the panels are 44VOC so that puts it very close to 150 and if it gets much below 77F there the Classic would be in hyper VOC. What I tell people is to design the PV array based on several factors Distance from the controller, battery voltage and ambient conditions being a few of them. I always start my designs around a Classic 150 when possible do to it being more powerful, less expensive and slightly more efficient (Efficient do to the lower VOC not hardware) but at the end of the day DGD is correct it is a good idea to have some headroom to allow the Classic (or any MPPT controller for that matter) to work.

In my own 48 volt system I have 3kw with a VOC of 89volts into a Classic 150 and a second 3kw running a VOC of 185 volts into a Classic 200 and the 200 definitely runs hotter. I will try to look mid day when they are both cranking and get temperatures. I am not concerned about the heat though, one thing about the Classic is the case is the heatsink so unlike the Outback where we buried the FETs inside and shielded the heatsink inside the metal box the Classic intentionally drives the heat to the casting. So the user will Feel like it is running hotter when in reality the electronics are probably cooler?

Again to back up what I would ideally say for this particular system would be 3 in series but I would suggest a Classic 200 just to eliminate unwanted time in Hyper VOC. My concern on 3 in series is not heat related at all but HyperVOC related. I have no real hard fast rules on VOC vs battery voltage but I can say that to low a VOC is common and then you suffer losses do to the controller not being able to MPPT the array. This is especially pronounced when someone takes a C40 of the wall and drops an MPPT controller in its place. Typically if the array was designed correctly for the C40 the performance may actually drop with MPPT because of the low voltage

Ryan

[Vic]

I believe it was boB that stated a good value to target when determining how high a PV generated voltage you will use, is to target 80% of the max the CC can handle, so for the Classic 150  it's =~ 120V.   This is not an absolute but a safe level that does not stress out your equipment.
Mind you Chris Olsen has posted his experience with maxing out his classics hooked up to his Wind turbine(s). Eventful ! to say the least.

Hi Wb,

It is not exactly clear what voltage you are referring to  with the 80% you noted.
Certainly,  this  80%  is not for the string Vmp,  into the CC,  which would be  TOO HIGH,   IMHO.

Generally,  believe that running three of the new PVs in series would be just a bit too high for a Classic 150.

One of the systems here runs three 35.4 Vmp PVs per string which yields 106-ish V strings.   In this moderate climate,  the highest recorded Voc on the MX-60 (which we still use on this system) is 141 V.

For this system,   loss of efficiency on the MX CC is striking when in Float with light loads.  With only about 4 - 5 Amps load on the output of the CC,  the Vin is about 122 V,  even in Summer afternoons with nominal ambient temps of about 35 degrees C.  The MX,  even with this light load,  runs its internal fan at times,  even with the temps inside the power room at about 21 C.  This is with an efficient CC,  but with a relatively high Vin delivering few watts.

One way to mitigate this situation,  obviously,  is to place a relatively large load on the inverters,  causing the almost Voc condition to  be reduced to a value much closer to string Vmp.   Here,  the noted large load(s) are usually A/Cs.  The A/C that has priority is the one in the power room,  trying to keep the batts and electronics as cool as possible.

And Kurt,  most of the time,  the highest Voc is experienced early in the morning,  around Sun rise,  where the illumination from the sun is very low -- insufficient to heat the PV's cells very much at all,  and when the lowest temps are usually experienced.  With a great CC,  like the Classic,  with its HyperVoc circuitry,  you might just miss some charging in the early AM with strings of three of these PVs,  but likely,  this reduction should be small.  To me,  the most dramatic effect of high string V,  is efficiency reduction.

I have no measured data to provide on the observed efficiency reduction,  just anecdotal observations,   which to me,   are fairly striking,  and not at all a Linear function -- appears some form of exponential function.

As HC Ryan mentioned,   perhaps considering a Classic 200,  given your PV wire run could be in order.

Just opinions,   from the   Center Of Nowhere,   Vic

[zoneblue]

1). Midnite have never produced efficiency figures/graphs  showing batt volts vs input volts  so where is the evidence of this?

Yeah they have, kind of, check it out:
http://midniteforum.com/index.php?topic=1316.msg10518#msg10518