help w/ blocking diodes

[bart]

    Am putting up a mix panel array.
http://www.fieldlines.com/index.php/topic,144599.msg977441.html#msg977441
Yes its a hodge podge of odd panels. Looking back, I realize not such a good idea. looking forward, It's what I have.
   Followed dnix71 advice on about 70 v per series string (thanks, work very well) and in theory all panels, in 6 strings could be used.
For now, the plan is to use 3 of the strings, with one having the highest amps at 11.
 Due to shading at different parts of the day, it is unavoidable, need blocking diodes, as I understand.
   My problem with searching for diodes is I've got no clue what I'm looking at. When a diode is found with the correct volts, amps are to low. Correct amps, volts are to low. Is it blocking or bypass? Can one series and/or parallel diodes? Seems like I read somewhere that this was not a good idea.
  Three legs on the diode, but you only need two? which two?
   So please, someone hold my hand, point me to what will work or how one can get there from here.
   Thank you. bart

[madlabs]

Bart,

Schottkey diodes use a little less juice. You didn't specify amperage, but here is a page of some possibilities:

http://www.digikey.com/scripts/dksearch/dksus.dll

Note that it is a parametric search, so you can specify the amps and volts and package, etc. You can see I looked for 90V to have some overhead, but look at your panels and see what the max they can put out on a cold day would be and select one with a little more headroom. If you use a TO-220 package (one with three legs) the data sheet will show you which legs to use. The main reason for TP-220 is that you can attach it to a heatsink. Remember you can parallel diodes for more current and if you keep the amps low relative to the part you won't need a heat sink, like running a 5 amp at 3 amps.

HTH,

Jonathan

[bart]

Thank you, Jonathan
highest series voltage is 70 v @ 11 amps
  But zee link no workee. It just went to the main data base of all parts.
But go for it anyway. It no like Schottkey or T0-220 in the search box.
It did like diode. To many. diode T0-220. That narrowed it down.
Eeny, meeny, miny, moe, selected single, it sounded simple, and picked
http://www.digikey.com/product-detail/en/MBR1090G/MBR1090GOS-ND/918581
 Now look for diode TP-220. 0 results. Try diode TP 220, and selected single again and picked this
http://www.digikey.com/product-detail/en/STPS30SM100ST/497-12663-5-ND/2122474
   So did I totally screw this up?

[mab]

I would go for the 2nd one (30A, 100V).

remember with 3 pin to220 the metal tab is normally connected to the center pin. Looking at the datasheet, the other two pins are connected to each other, and current will flow from the the anode (two outer pins) to cathode (center pin), and will be blocked from flowing in the other direction.

[dgd]

    Am putting up a mix panel array.
http://www.fieldlines.com/index.php/topic,144599.msg977441.html#msg977441
Yes its a hodge podge of odd panels. Looking back, I realize not such a good idea. looking forward, It's what I have.
   Followed dnix71 advice on about 70 v per series string (thanks, work very well) and in theory all panels, in 6 strings could be used.
For now, the plan is to use 3 of the strings, with one having the highest amps at 11.
 Due to shading at different parts of the day, it is unavoidable, need blocking diodes, as I understand.
  ...

Shading of panels in one or more strings does not make it necessary to install blocking diodes.
The blocking diode is to block current from going in the reverse direction into a panel and just about every panel I have seen has such a diode built in - usually in the junction box. So what function an additional diode would serve escapes me.
I have had multiple different size strings connected for years with shading at certain times affecting one or two strings and have never seen a problem. My mppt controller sorts it out.
The only issue is to try and balance the string voltages but even that does not need to be perfect.
Dgd

[OperaHouse]

While all solar panels do leak, I wonder if blocking diodes are a holdout from the past when leakages were worse.  Last time I measured leakage from an array of panels far into the evening it didn't seem like enough to worry about.

[DamonHD]

There is, I think, a significant risk of damaging a cell with a large forward voltage across it (eg in shaded cells in a panel paralleled with a panel without shading): that alone seems to be a good enough reason to consider a blocking diode.  I too am less worried about leakage (or shorts) in most cases, at least with well-made crystalline panels, and use Schottky diodes for blocking in my low-voltage system.

But I have no special knowledge of the components in any particular system of course.

Rgds

Damon

[bart]

   My main concern, and it may not be warranted, was in driving the Classic 150 crazy (boB, hello). The shading will at times be severe.
   Two strings for example, will have two 55 watt panels paralleled to a series of two 180 watt panels. 67v 7a
   One string, 4 200 watt panels, 72v 11a
   one string 4 50 watt panels 69v 3a
4 different brands at very different wattages. I have no idea how a charge controller could sort this out. Maybe it can. Blocking diodes look like cheap insurance.
 Damon, your post came up while I was typing.
   So, I gather that a blocking diode should be between the two 55 w ?
Leakage? Like back feeding from batteries at night? Or from bypass diodes?
Teach me more and thanks.
bart

[dgd]

...
4 different brands at very different wattages. I have no idea how a charge controller could sort this out. Maybe it can. Blocking diodes look like cheap insurance

Different wattage strings into a controller make no difference to the controller.
However it is good to maximise wattage by getting the strings' voltages as close to each other as possible. Your strings are pretty close so that's good, they don't need to be an exact match.
The controller can cope with shading, it's designed to do this, the mppt algorithm will sort out the optimum voltage it sees the maximum wattage at. If you use the midnitesolar local app to monitor the 150 it will identify when it sees what my be shading.
The use of any additional blocking diodes is pointless, all it will do is drop the voltage by about half a volt 
Your controller will prevent any 'back feeding' of current to the panels at night.

Dgd
 

[dgd]

There is, I think, a significant risk of damaging a cell with a large forward voltage across it (eg in shaded cells in a panel paralleled with a panel without shading): that alone seems to be a good enough reason to consider a blocking diode.

This could be true with a large enough forward voltage but if the maximum string voltage is within that specified for safe use of a panel, usually 600v or 1000v then this is not a problem. All modern panels have sufficient protection built in for normal use.
The biggest concern with stringed panels that produce hundreds of volts is a sudden surge up to near max voltage with fast rising current that could  burn up one of those metal tape links  between cells.
Dgd

[bart]

    I realize now that I neglected to add a detail. Not that it appears to matter. Was only planning to add the diodes at the end of each string.
    dgd, yes I knew that the classic would take care of battery back feed, was not understanding the leakage.
Thanks you all for your help!

[DamonHD]

1) Not all panels have blocking diodes. Some of mine do, some don't.

2) The problem as I understand it is driving a large forward current through a shaded cell and causing hot spots and permanent damage.  *One* blocking diode (of adequate rating) in each string of panels is enough to avoid this issue entirely and losses are insignificant.

Rgds

Damon

[bart]

   So in the selection of a properly sized blocking diode and trying to keep it simple for this noob, one needs to look at:
1) Max. reverse voltage
2) Current rating
   Another question:
3) The voltage range that the diode "blocks" and "unblocks". In the specs. of a particular diode, is that the forward voltage?
Somehow that does not seem correct as the diode is "comparing" what the other strings are producing.

[boB]

Assume similar PV modules in the strings here, at least same Voc and Impp (or Isc)...

If you have just two series PV strings paralleled, you do not need to worry about protecting either string if
one module shorts.   With 3 or more series strings in parallel, you need to use inline fuses or circuit breakers at
with maximum trip currents at the series fuse ratings of those PV modules.

If you have strings in parallel and one or more module gets shaded, that does not cause any backfeed from
another string that is not shaded.  Backfeeding is temperature related and does not matter if the panels
have sunlight shining on them or not...   However, the ones with sun shining on them will normally be
hotter and have a lower Voc than the ones without sun shining on them just because the sun is
heating them up.

boB

[bart]

   Thanks boB.