Solar panels for less than ideal conditions - bypass diodes?

Nice to hear. I've been waiting to hear more about the tracers.

Salvo,

Unfortunately, the mppt charge became a wash, as cloud cover only increased as the day went on. I was hoping to try out the new Tracer 3215bn on a 65% soc bank. I ended up switching to paralleling my new 40a vec1093dbd with the pm4b-45 for 85a, which went very well, and as expected.

However, I did manage to solar charge for two hours, from 11am to 1pm, under mostly cloudy (whitish-grey)skies. I checked panel Isc occasionally, and compared it to what the mppt was getting to the bank, amp wise, and the mppt was definitely doing better than what the pwm would have passed thru. Obviously a much more controlled comparison needs to be done, but my initial observations apparently speak for themselves.

One example was at 12:45, when there was a break in the clouds. Isc was right around 10a (x2), so say 19a - 21a pwm, but the Tracer showed 9.7a @ 30.9v at panel array, and 22.5a @ 13.4v to the bank. Now that does work out to >100% eff., but that's only because the Tracer rounds up voltage quite early. But I was doing a 12v on 12v, so eff. should be quite high.

About 10 minutes later, heavy clouds rolled in, and the Tracer showed 3.1 panel amps @ 31.6v, and 7.6a to the bank. Not sure what Isc was at that moment, but likely less than 3.8a (x2). One thing is for sure though... amps took a much bigger hit than voltage.

Tomorrow I'm doing another Tracer run, weather permitting, of course. But in order to do a proper side by side, I'll need to reconfigure some wiring and split the bank... this will be forthcoming, I assure you. So far, I'm liking the Tracer mppt.

Can you show the data for the two scenarios?

jrnymn7 wrote:
I just did a 280w mppt run yesterday, using two 12v panels in series, in cloudy conditions, and it definitely did better than any pwm/parallel would have.

*You probably won't be happy with a PWM and nominal 12 volt panels in difficult low light conditions. You might do better with a 24 volt panel and pwm or with MPPT.*

"This is TOTALLY backwards. MPPT will never out perform a PWM in low light conditions.

Also, MPPT is NOT a better performer than PWM for <600-800 watts of panels. "

In another thread, it was clearly shown in a side by side test that a single 12v panel thru mppt did way better than thru pwm, in rainy, cloudy conditions. This is because panel amps take a nose dive in such conditions. PWM relies on panel current more so than panel voltage, while mppt relies more on panel voltage to achieve its gains. I just did a 280w mppt run yesterday, using two 12v panels in series, in cloudy conditions, and it definitely did better than any pwm/parallel would have.

Contrary wise, pwm has the advantage in hot sunny conditions. This is because high panel temps result in higher Isc but lower voltage.

grizzzman wrote:
I agree you dont want to use the old panels with the new ones. Perhaps a test of the old panels are in order? A short curcuit amps test on a sunny day around 12 and 1 mid day. I would expect to see 4ish to 6ish amps between the two panels using the ten amps setting on your multi meter. This is on the roof close to the panels. If it passes move the controller and redo the wiring and if it fails chassis battery charger it is. What are your thoughts?

I'll take a few measurements and post what I find.

Wayne

On edit: Well, that will need to wait a while. By the time I got things together, the sun disappeared and black clouds moved in.

smkettner wrote:
In reality the diodes are to protect the panel from damage more than provide shade tolerance.

But I love it when a plan comes together ie mine and and the diodes yield additional output.

However my installation was one of elevated panels to eliminate rig shadows. But controller logging of panel voltage with available shade is revealing in terms of bypass diodes and other "factors".

Sal,

I agree but there is shade and then there is shade. If "partial" shade reduces the amps but the voltage is OK then the parallel amps will add. If the "partial" shade causes a bypass diode to conduct then that panel in parallel won't contribute.

So we have full sun, overcast in 10 different shades, partial shade, leafy shade, hard shade, night, etc. Anyone want to attempt to define "partial" shade?

Empty Nest, Soon wrote:

grizzzman wrote:
. . .
It seems odd to me to at least in my eyes to waste 100 watts as a float charger for the chassis battery. . .

Good point. But while it is nominally 100 watts, with the unreasonably long cable run to the front of the coach and back, I seem to be getting only a fraction of that output to the batteries.

I suppose I could move the existing controller back near the batteries to improve these panels’ effectiveness, but that means pulling a second new set of wires. I hate thinking of maybe mixing these old panels on the same controller as new panels. I would probably just buy another new 110 watt panel rather than sort out compatibility questions for a 20 year old panel.

Wayne

I agree you dont want to use the old panels with the new ones. Perhaps a test of the old panels are in order? A short curcuit amps test on a sunny day around 12 and 1 mid day. I would expect to see 4ish to 6ish amps between the two panels using the ten amps setting on your multi meter. This is on the roof close to the panels. If it passes move the controller and redo the wiring and if it fails chassis battery charger it is. What are your thoughts?

You have a point. Two things are happening at the shaded cell. You got the reduced irradiance of the shaded cell forward biasing the cell and you got a much higher current from the other cells trying to get through this cell, causing a negative voltage drop. If the negative voltage drop wins out then the bypass diode will conduct.

http://www.pveducation.org/pvcdrom/modules/bypass-diodes


I based my initial thoughts on a series test conducted by NBB (??). I believe he has 3 series panels. If one of the 3 panels is angled 45 degrees away from the sum (simulating shade), current production is drastically reduced.

LScamper wrote:


Salvo wrote:
"Taking this logic further, in a series configuration, the shaded panel t shat has 400W/m^2 irradiation will not conduct its bypass diodes. This string will have a total current about 2A, severely limiting production."

Not sure that is correct. The other cells are still producing their max current. Seems that the currents will add at the junction of the shaded cells and the not shaded cells. The current from the not shaded cells will forward bias the bypass diode so the short circuit current will be that of the higher current cells.?

smkettner wrote:
"Example:

Two 24v panels with 60 cells each (120 total cells)for a total of about 60 volts. There are 6 groups with 20 cells each. Each group produces 10 volts. (each cell produces about 1/2 volt and maybe 8 amps)

If all groups are producing power you get 60 volts 8 amps 480 watts.

One group is shaded you get 50 volts 8 amps 400 watts.
Two groups shaded 40 volts 8 amps 320 watts.

Actually it is a little less as each diode will lose 1/2 volt.
So with two groups shaded it is really 39 volts 8 amps 312 watts."

I agree. But, I note, as smkettner states that the total current is the current of the higher current, not the shaded cells! Also the voltage across the bypass diode of the shaded cell is about -.5V not +.5V. This could still charge if there is enough head room for the controller to work.

Salvo wrote:
"Taking this logic further, in a series configuration, the shaded panel t shat has 400W/m^2 irradiation will not conduct its bypass diodes. This string will have a total current about 2A, severely limiting production."

Not sure that is correct. The other cells are still producing their max current. Seems that the currents will add at the junction of the shaded cells and the not shaded cells. The current from the not shaded cells will forward bias the bypass diode so the short circuit current will be that of the higher current cells.?