As the sun get's lower in the sky the atmosphere blocks more solar radiation.
Yes, that accounts for the drop from 25W to 20W, but not the drop to 6W which happens pretty quickly over the space of 30 minutes.
I've seen this thread only now - I'm keeping track of my solar power by some homebrew data logging stuff.
The public interface is here:
http://wc.wunderkis.de/webcam1.php(sorry, it's German labeled), the last chart is the kWh / installed kW of my solar panels.
So got a few really nice and sunny days after very dark December and January.
For my purposes (not available to the public), I've got more charts, like these telling me the amount of produced solar power and the ratio to power used by the household (daily for the last 30 days and over one year):
Yah! Another RRD Graph user
As the sun get's lower in the sky the atmosphere blocks more solar radiation.
Yes, that accounts for the drop from 25W to 20W, but not the drop to 6W which happens pretty quickly over the space of 30 minutes.
I am wondering if this has something to do with diffraction through the glass or special glasses often used in double glazing (high iron content) is that possible ?
I've seen this thread only now - I'm keeping track of my solar power by some homebrew data logging stuff.
The public interface is here:
http://wc.wunderkis.de/webcam1.php
(sorry, it's German labeled), the last chart is the kWh / installed kW of my solar panels.
So got a few really nice and sunny days after very dark December and January.
For my purposes (not available to the public), I've got more charts, like these telling me the amount of produced solar power and the ratio to power used by the household (daily for the last 30 days and over one year):
OK I can see my primative graphs with GNU look entirely out of date wrt you guy's so when we hit the end of the month and I suck some fresh data from the logger I will have to try RRD too
Things are definetly improving as the days get longer again, here is my weekly average output to date in Kwh/day, horizontal axis is elapsed hours (500hrs = 21Days) still in GNUplot
Not sure the window reflection accounts for what you see. Low angle sun going through thick air and more water vapor?
It could have been. What surprised me was how sharp it fell. I had the little query monitor open in a terminal on my desktop. I could see the bright sunshine coming through the window (well a gap in the curtains), but the Panel Amps was really low. So I went to investigate as it had been sitting happily around 1.8-2.0 amps from 10am through to 2pm. Yet at 3pm it had fallen to like 0.5A. Pushed the panel back up to the window and it went back up to 1.5A.
I've left it against the window, I might lose 10% or so in the morning when the sun is bang on, but it should keep the watts higher further into the afternoon.
And today we had - nothing at all! Not an electronic sausage
Clearly I need to add some snowlar panels
And today we had - nothing at all! Not an electronic sausage
Clearly I need to add some snowlar panels
To right but suprisingly mine have been hauling in up to 250W through 4 inches of white stuff!! (25% peak power).
Actually I seem to remember reading long ago that the poly panels I have are better at diffuse light than amorphous even though the latter have a bit more efficiency in full sun.
I lost energy today. Another day costing more electrons running the charge controller than the panel produced.
Maybe pay some attention to the thirst of this charge controller, I spent a lot of time when I first started reducing the power needed to run converters to the barest minimium so my pickup light level was as low as possible. Not sure how much I really gained from it apart from the design excercise and being able to see my converters hang on in the evening right down to a measly 1 watt
I measured it. 50mA with pulses of 75mA when the ESP8266 fires the data out on the Wifi every 5 seconds. <-- Should drop that back to once a minute.
Doesn't sound like much but when you multiply it (avg. 60mA) out to 24 hours, it's 1.4Ah.
I switch the whole system off at night, so at the moment it's probably only about a third of that.
One idea, are you using a switching or linear regulator to get from your battery voltage (13.8V?) to your mpu (3V3?) a switcher could save you a lot of power if it's reasonably efficient and I expect Ebay is loaded with the things ready built
Switching MPPT DC/DC. 50mA/75mA isn't an awful lot to run a charge controller, it's LCD (no backlight) a RS485 to Serial adapter and an ESP8266 and a wifi client.
According to the controller it's efficiency is suspiciously near perfect. If its showing panel power of 6.1W the battery charge watts show as 6.1W.
Still have to shorten the panel cable, or replace with some mains flex.
You just told me that was the consumption you measured sorry if I was wrong but the mpu supply doesnt need to have anything to do with MPPT surely it can just run off the battery I thought.
Maybe the reselution of measurement is such that it's internal consumption just doesnt show.
So your cables are still acting as room heaters ?
Yes
I was in homebase and lifted a 5 meter bundle of 15A mains wire, but set it down again when I read the price. £12.99. Ouch.
I'll order some on ebay, I have run out of "kettle leads" to salvage. When you are looking for other cables there are fecking dozens of them, but when you are looking for power leads, you can't find any for USB leads and wall warts
The way it goes.
I mentioned the MPPT, DC/DC bit in case you thought it was a PWM duffer or linear regulator.
Yes the controller and monitor runs off the battery, but it doesn't seem to show internal load. It registers watts and amps to 2 decimal places though, so it should. However it does other weird things too.
Over RS485 if you have 1Amp of solar, and 0.5Amp of load it will show:
Panel current: 1A
Battery current: 1A
Load current: 0.5A
There is a separate register for "Charge Current" which would read 0.5A. It seems the battery sense is before the load gets split off. On the panel LCD it does the sums though and subtracts the load from the battery current to be fairer. Over RS485 you have to remember to do it yourself.
I believe solar panels are more efficient, at lower temperatures, because the voltage drop of the diode junction increases.
It's been cold in the UK recently, so has anyone noticed an increase in output, above what would be expected for the increase in day length? I suppose it's been cloudy recently so may be not. The MPPT also changes with temperature, so the inverter would need to monitor the temperature, in order to increase the efficiency of the panels, when they get cold.
Yes I was in homebase and lifted a 5 meter bundle of 15A mains wire, but set it down again when I read the price. £12.99. Ouch.
I'll order some on ebay, I have run out of "kettle leads" to salvage. When you are looking for other cables there are fecking dozens of them, but when you are looking for power leads, you can't find any for USB leads and wall warts The way it goes.
I mentioned the MPPT, DC/DC bit in case you thought it was a PWM duffer or linear regulator.
Yes the controller and monitor runs off the battery, but it doesn't seem to show internal load. It registers watts and amps to 2 decimal places though, so it should. However it does other weird things too.
Over RS485 if you have 1Amp of solar, and 0.5Amp of load it will show:
Panel current: 1A
Battery current: 1A
Load current: 0.5A
There is a separate register for "Charge Current" which would read 0.5A. It seems the battery sense is before the load gets split off. On the panel LCD it does the sums though and subtracts the load from the battery current to be fairer. Over RS485 you have to remember to do it yourself.
Ahh sorry of course it's all one lumped in unit with the charge controller, I had an idea you had a seperate MPU that you were powering. Seems like you are slowly learning this things quirks
I agree on the cables I was looking for a straight Ethernet the other day and all I could find were crossovers that are normally like rocking horse poo!!
I believe solar panels are more efficient, at lower temperatures, because the voltage drop of the diode junction increases.
Other way around a silicon junction voltage reduces at -2mV/degC.
It's been cold in the UK recently, so has anyone noticed an increase in output, above what would be expected for the increase in day length? I suppose it's been cloudy recently so may be not. The MPPT also changes with temperature, so the inverter would need to monitor the temperature, in order to increase the efficiency of the panels, when they get cold.
The whole point of MPPT is Tracking, so it adapts to changes in the maximium power point all by itself, this is the advantage of MPPT compared to some who claim constant voltage is adiquite!
Interestingly the plot thickens, when I checked the spec of the panels I use (enclosed) although Voc indeed has a negative coeficient Isc has a small posotive one but the overall Pmpp has a negative one!
How quickly things turn around.
I had resorted to mains power for charging my devices for a week. Ran the solar battery down to 12.0V and watched it drop running just the charge controller to 11.8V over a week of snow and heavy cloud.
Wednesday this week the sun peeked out once of twice and it rose back to 11.9V
Thursday the sun graced us for a few hours through broken cloud and it rose to 12.3V
Today it was sunny blue sky from 7am till 4pm and I came home to 12.8V, damn near full charged.
How quickly things turn around.
I had resorted to mains power for charging my devices for a week. Ran the solar battery down to 12.0V and watched it drop running just the charge controller to 11.8V over a week of snow and heavy cloud.
Wednesday this week the sun peeked out once of twice and it rose back to 11.9V
Thursday the sun graced us for a few hours through broken cloud and it rose to 12.3V
Today it was sunny blue sky from 7am till 4pm and I came home to 12.8V, damn near full charged.
Maybe we can go into business selling the met office historical data to improve there forecasts
Not to bad over here in the east but cloud cover at least every other day so averageing 1Kwh/day (from 1Kwe) ATM
Here is my system's data from 1/1/2016 to date. I supply the house with some of it's energy needs from a 3kW solar array. I have a 18kWhr battery, and only cycle it about 6 kWh a day. Sometimes I can run large loads on solar AND charge the battery which explains the 15kWhr peaks in inverter energy.
Early Sept '17 the inverter died, hence zero energy during that period. Location is Melbourne, AUS.
You can see the seasonal dips and peaks.
Hi Poida thats a great system
Sure wish I was in Melbourne except maybe the hurricane! Do you have a make & model for the inverter or is it a DIY ?
I woke to ZERO POWER today, seems the panels are part snowbound and although the inverter happily reports zero watts it doesnt want to start and guess what, in minus4C I dont want to go out there with the laptop and pic debugger to ask it why hahaha
Most likely there is just enough voltage for the mpu to run but the 12V isnt solid enough (it checks that to ensure there is enough to turn the mosfets hard on).