Do I really need these diodes?

[doug7x]

OK, folks...I built a couple 196W panels from cells I bought on ebay. Both are 6x6 cells, 7 strings of 7 cells in series.(1 panel is poly, 1 mono) They are both parelelled to a 300W GTI that I also got on ebay. Since I'm grid tied & not charging a battery bank, do I need the diodes between the strings? ( I'm using 5 diodes per panel) I was thinking this would help the panel live longer when operating in partial sun/shade conditions, but if the diodes are costing me power (theyr'e NOY schottkys...just some 10A silicone diodes also from ebay), I'd like to eliminate them. waddayaall think???   thanks in advance...Doug

[Flux]

Not sure where you have the diodes.

It is a good idea to have one  blocking diode per series string especially if you have a mixture of mono & poly. This only costs you one diode drop and in a high voltage string is not worth worrying about.

Parallel diodes across parts of the series string are a good idea to prevent total loss of production of one string if one cell gets shaded but these are not contributing any volt drop or loss in normal operation and I see no need to be wanting to dispose of them.

If you have more than one series diode per string then it is pointless. If you have built it bases on individual low voltage panel ideas you may have included these redundant series diodes.

Flux

[doug7x]

Flux, when you sat "series string", are you talking about each string of 7 cells, or the entire 49 cells in series?...either way, my layout is 7 X 7 cells in series for 24.5 OCV & @ 4W per cell, 196 theoretical watts. That is one panel. For this post, I'll just talk about the poly panel...panel has 7 strings of 7 cells...diodes between 1&2, 2&3, 3&4, 4&5, 5&6. I don't have diodes on 1 or 7 because they are accessable externally. (the red & black wires). Would I be better off, say, if I just left 1 in towards the middle of the panel...I could also add externally to the leads if I have to. How much wattage do you think the extra diodes are costing me?  When I was building it, I DID see a slight Vdrop across the diodes, but I didn't think it would add up to much once the sun hit it....now I'm thinking I may have shot myself in the foot (LOL)  

[Flux]

I think you only need one series diode per string of 49 cells.

Shunt diodes across each string of 7 cells would be advisable but these add no volt drop.

I would put the series blocking diodes at the positive end with the common going to your positive feed to the inverter and for such a high voltage I would use something with a decent reverse voltage rating, not schottky in this case.Regular diodes are ok but some of the fast ones seem to have a lower volt drop but with only one diode drop it really won't matter.

If you have diodes as you suggest with one per block of 7 cells you are loosing upwards of 10v.

Flux

[doug7x]

Thanks, Flux....pretty much whai I was thinking.(leaving 1 on the + to the GTI)...If I really am losing 10V, I'm gonna REALLY like the output after I make the changes this weekend. I've beev using just a kill-a-watt for my output monitoring. This weekend I'll break out the DMM & get some before & after values. I'll post the results...thanks again...Doug

[ghurd]

If this is the 12 "and 24V" GTI, you do want the bypass diodes too.

Probably want them anyway.

If it is mostly a 12V GTI, check the max input open voltage allowed.

The panels Vo could get past 35V under some conditions.

May need a snubber on the input, if the grid goes down on a cold sunny day at noon it could smoke the inverter.

G-

[doug7x]

the input voltage recomended is 14-28 VDC...not sure what you mean by "snubber"  thanks again...Doug

[doug7x]

I still haven't had the time to pull the poly panel apart, & I'm a little leery of doing the testing outdoors. Any suggestions on an indoor light source for R&D purposes...right now, I have a 500W halogen work light & 4 incandescent floodlights (appx 300W total) pointed at an under construction mono panel (196W theoretical rated output) I'm only getting about 20W AC output.(kil-a-watt & GTI) (light source is appx. 3 ft. from the panel). I'd REALLY like to be able to get the panel up to operating temp. & output, too "cook-off" any remaining moisture before I seal it up. With the lights that I have, I thought I'd be pretty close to the "1000 W per sq. meter" that I would need to simulate sunlight, but appearantly I've missed something. (BTW...panel shows 26.2 VDC open circuit, so I'm pretty sure all the cells are working)THANKS....

[ghurd]

Light bulbs will not make any power to speak of.

It needs direct sun.

G-

[doug7x]

Ghurd, it looks to me like you might be right. I wonder what they use at the factory foy their "light tower"...arc lights, maybe?...either way, I'm gonna have to come up with something else...kinda hate to resort to a hair dryer...Doug

[doug7x]

OK, folks..I FINALLY got around to pulling the poly panel apart...indoors, under the lights I talked about earlier, I was getting 20.88 OCV with the diodes installed...after I pulled them, it went up to 24.36 OCV. I'll see if there's a difference in the AC output wattage once I get it sealed back up & out in the sun...Doug7x

[doug7x]

Just got it out in the "sun" (it's been solid overcast since I brought it outside)...Seeing appx. 15 AC watts . Under  these conditions, should I be seeing more power, or do I have problems? (It's been in the  50's - 60's with no real sun)(May in Chicago)  Thanks...