Suggested mod to design to enable string of 4

[djsharpe]

I am in Australia which where I am is not as cold as east coast US. I have bought at a low price 16 off Canadian Solar 230W poly modules 60 cells for a customer. Want to run a 150 Classic (already bought) and obtain max output  possible at 48V. These are similar to the hyper example given. I want to run them in 4S4P. Plan is to bypass one module of each string using 12V coil relay in normal close position and open when panels warm up to OCV<150V. Contacts will be bridged by cap/resistor network to avoid contact arcing up to 40V 32A
Will apply a temp probe to underside of module likely to warm up last  and use output of dig temp sensor ($25 on Ebay)which has a hysteris setting to do this sensing. Suggestion is that design could enable this. Perhaps the built-in 12V supply could feed the relay.
At least this arrangement will provide 3/4 of array in cool weather and is better than waiting until panels heat which will be no output until OCV<150V. NB no rules about voltage limits here.

[Vic]

dj,

Seems to me that your proposed configuration -- 118.4 Vmp will result in a significant reduction in efficency of the CC,  without regarding the need to workaround the HIGH Voc issue.

I, personally,  have looked at this same configuration (different modules),  and decided on strings of three for about a 93 Vmp.

Have been running two separate arrays for just six years now at 106 Vmp using MX-60s.  This results in the CC getting surprisingly warm at very low power delivered,  as under light loads,  my Vin would run at about 120-123 volts.  You are proposing to do the similar thing,  but your Vin will almost always be higher than this.  Agree that the Classic is a different CC,   but,  why not make it easier on yourself and customer,  and improve your efficiencies,  too.

Study these power Charts:
http://www.midnitesolar.com/images/classicFrontPage/graphs.php

Just MHO,  Good Luck,  Vic

[djsharpe]

I have looked at the charts but efficiency is not shown . However I have seen the Outback graphs that indicate I will loose a few % operating in this way (assuming the classics have similar data) so I ll go to 2S8P as the PV modules are not far from the batteries. Thanks for the info Vic

[phonetic]

I almost did the same, had Outback MX 60 charge regulator fed by 3x4 190watt mono arrays, the location had a low temp of -5`c, to overcame the high Voc I had via a thermostat and relay, switched in 500watt dummy load across the string, to bring the Voc below 150 volts, as soon as the temp rose to +5`c the dump load was disconnected.
Im located in Ballarat in Victoria gets cold in winter.
Just updated to a Classis 150, no need to worry about the dump load now.
Regards
Frank

[Vic]

dj,

I had assumed that your battery voltage is 48.  If this is the case,  believe that you will not have enough input voltage to run the cc.  Recall that your modules have a Vmp of about 29.5 volts.

Perhaps I am missing something here.  If you were able to run strings of three you would have about 90 volt string voltage,  altho,  of course,  this would only use 15 of the modules.

I do not want to pick at your design too much.  Think that your original setup could be made to work,  but it might stress the Midnite Classic a bit much.  Here,  have been running A/C in the power room to keep the electronics happier.

More later,  Good Luck,   Vic

[djsharpe]

You are quite right it is 48V and Vmp is 29.6V so two in series would be 59.2V. The bats are flooded so any equalising would need to use the Diesel battery charger (actually an interactive inverter called an SPPro) as I assume that the Classic cant boost. However there is still more volts available from the array (Voc 36.8V) so at reduced output it might be able to equalise.
I would prefer to to run 3S & will have to find 2 more PV with 230W or equivalents to do this as there were only 16 units left from my supplier. Ill keep you posted on the outcome. When I do the job Ill allow extra wiring so I can reconfigure the PV modules down the track. Of course when looking at efficiency the losses in the wiring is part of the exercise & we dont want to run enormous CSA wire. We are attracted to MPPT devices to reduce wire size! David

[Vic]

Hi David,

OK,  Well,  the CC needs a few volts drop in its operation,  and with cool/cold batts,  you might run out of Vmp at good power levels.  The curves from the PV modules are fairly flat.  Three per string would be ideal IMHO.

Just now,  60 cell modules are quite common,  so you should have some choices,  altho,  the dimensions might not be close enough.

Good luck with your new project.  Vic

[djsharpe]

Ill do 4s4p intially & if unhappy with efficiency will go 3s6p later when I can get 2 more 230s. The high volts issue will be dealth with by a home made clipper that will start to operate at around 145V. It will be like an active zener diode but will in fact be a number of Mosfets operating in the active range & behaving like heating elements. I dont envisage much power being lost though. Further info when I try it. D

[laszlo]

David,

You are right, if the temp derated VOC is higher on the pv than your equalize voltage, the cc will elevate the voltage via pwm but you will have less than peak output.

Your project sounds fascinating -- I'd love to hear how you made out.

I am in Texas so climate is similar to yours, and have not had any issues so far. What I did is put a small  box fan on a timer and run the air on the CC via a timer between 11am and 3pm in winter and 9am and 7pm in summer.

.

[djsharpe]

I have built and installed a series regulator for a Classic 150 48v system that has 4P4S 230W  60 cell Canadian Solar Monos. The reg is based on  a 200A  IGBT that tries to maintain 150V on its output via 150V zener reference. There is a 4-5V drop across the IGBT when in its active op. A temp probe attached to the underside of one of the lowest module drives a temp controller such that when a predetermined temp is reached around 20degC 12V DC is applied to the IGBT gate which then goes to about 0.5 V VCE Sat and a relay closes same time such that V drop is then zero and power flows straight thru. Power losses in this mode is around 2-3 Watts. The relay is equipped with steering diodes and a 470uF 200V cap such that ii will never interrupt high DC V in case of failure. 12V power is sourced via a DC -DC converter direct from the incoming PV array Volts. This is needed for the temp controller, DC injection to the gate, 12V relay and a 12V fan which only runs at 60 deg C. The whole lot is built into an H section heat sink so the fan is prob not need. The controller sits on top of the H. The benefit is of course that you will get power when otherwise hypervolts would prevent op. It works well but we havent seen very cold weather yet being in southern Aus. The probe indicates that the panels heat up to 20-25 deg above ambient as predicted.
I built the reg from scrap parts (my cust encouraged me to recycle equipment)
I can send pics and cct diagram if anyone is interested. David
ReplyQuoteNotify

[laszlo]

Well done! So you have for of these total on the entire array, one for each string? If it can be cheaply made, this is something other folks may want to build and try out as well.

[djsharpe]

There is one only series regulator. It gets its solar power from the Combiner box after the string switches where the arrays are commoned. My unit was designed for 30A input limited only by the 12V relay. This relay is only there to reduce losses. There is only about 0.5V drop across the IGBT when in sat mode. The rating for this one could be up to 200A so a large array could be treated if multiple Midnites were in parallel. If I were making more of these I would do the electronics differently. This one was a working prototype. David

[boB]

How much of a voltage drop does your IGBT device have to drop at most ?

And was a Classic 200 not enough current for your array ?

boB

[djsharpe]

The IGBT has about 4-5 V drop when it is active ie when panel temp less than set point and for lower voltages than 150V but after that it is bypassed. It might see 14V at -5 degC at sunrise.  If Id used an op amp or comparator  the drop would have been only 0.5V until approaching 150V at which point it would have inc as Varray exceeded Vset (150V). It is really just a simple linear regulator with a few tricks. I dont think the 200V version would have been OK with 16 off 230W modules 3680W. David