Rogers postbag

Tell us about the cool things you've recently got
User avatar
RogerClark
Posts: 7459
Joined: Mon Apr 27, 2015 10:36 am
Location: Melbourne, Australia
Contact:

Re: Rogers postbag

Post by RogerClark » Tue Nov 14, 2017 9:44 am

@Pito

I'm not sure how a step up converter would help.

I found the charge controller actually only needs 0.1V above the battery voltage, for it to function correctly.
The charge controller is Linear so just dissipates any voltage difference as heat.

I was hoping that the LM2596 would provide a more efficient way to convert from 5V to VBat + 0.2V e.g. from around 3.2 V to 4.4V

However I have been doing some more tests, and the efficiency of the LM2596 is not very good at low currents and low input and output voltages.

From my tests its around 80% (or possibly less) efficient, which makes it only 20% more efficient than the panel its self, which seems to be around %60 efficient at 3V output (compared with its power output at 5V)

The efficiency of my panel seems similar to the graph posted by @steveStrong.
e.g at 3.7V his efficiency was 70% (0.39W instead of max of 0.55W)

As this is only a maximum of 10% worse than using the LM2596, it seems pointless using it

MarkB
Posts: 7
Joined: Sun Oct 02, 2016 11:40 pm

Re: Rogers postbag

Post by MarkB » Tue Nov 14, 2017 12:50 pm

The voltage at the maximum power point of a solar cell is proportional with light intensity, but not so strongly that the panel is going to be delivering useful power at very low voltages. For load working voltages in the LiPo range (<=4.2 V) from a nominal 6 V solar panel, I don't think a step up converter ever makes sense. Note that the figure below is a single cell while a nominal 6 V panel is 12 such cells in series.*

Image

I have a little setup that's been running off a solar panel similar using a supercap (4.7 F?) as the power storage element with a zener diode to limit maximum charge voltage. If the application doesn't require a great deal of energy storage, that's probably a simpler approach than most anything involving a battery. The efficiency is sub-optimal of course, especially if the capacitor is significantly discharged between charge cycles.

Coming up with an efficient battery based solution is on my "someday list", but that list grows far faster than items getting checked off.

* - I waffled between 10 and 12 on the number of cells before coming to the realization that both are available on eBay etc. Maybe I'm the last to realize this. I'm pretty sure mine are nominal 6 V panels and 12 cell. The photo Roger posted up thread appears to have 5 V and 10 cells. In any case, that provides a potentially useful free parameter for project design.

User avatar
RogerClark
Posts: 7459
Joined: Mon Apr 27, 2015 10:36 am
Location: Melbourne, Australia
Contact:

Re: Rogers postbag

Post by RogerClark » Tue Nov 14, 2017 7:56 pm

MarkB

Thanks.

I have done a rambling long post on my blog, and my conclusion is that your solution is the best, and even if using a LiPo rather than a supercap it would probably be Ok.

For a LiPo I seem to get good results just be changing the charge current resistor on one of those TP4056 based modules.
As they seem to only need 0.1V above the battery voltage to operate.

I used a 8.6 k resistor in the end, as it gave a reasonable charge current under full sunlight

User avatar
RogerClark
Posts: 7459
Joined: Mon Apr 27, 2015 10:36 am
Location: Melbourne, Australia
Contact:

Re: Rogers postbag

Post by RogerClark » Tue Nov 14, 2017 8:34 pm

BTW.

This thread on Stack overflow has some interesting posts on autoregulating the TP4056

https://electronics.stackexchange.com/q ... extraction

But no one pointed out that its a linear device and its better to go for max current at the voltage required by the battery, rather than maximum power.

Also, no one actually tried a buck converter. But I found they are pointless as the effeciency is 80% at best, which is only marginally better in theory than the panel outputting at a lower voltage.
And in real world usage, the buck converter was far worse.

User avatar
RogerClark
Posts: 7459
Joined: Mon Apr 27, 2015 10:36 am
Location: Melbourne, Australia
Contact:

Re: Rogers postbag

Post by RogerClark » Tue Nov 14, 2017 11:44 pm

BTW.

Someone contacted me via my blog about this module

https://www.ebay.com.au/itm/CN3791-Lith ... 3094202363

So I've ordered some for testing

However, as they cost $3 USD each. which is twice the price of the solar panel, its more cost effective to buy a bigger or better panel than use this module, when the panel only costs around $1

User avatar
Slammer
Posts: 255
Joined: Tue Mar 01, 2016 10:35 pm
Location: Athens, Greece

Re: Rogers postbag

Post by Slammer » Wed Nov 15, 2017 12:02 am

Roger, watch the video of Andreas Spiess about small solar chargers, you will find measurements for most common modules, is a very good review.

PS. Dont forget to watch and the second video of review : https://youtu.be/ttyKZnVzic4

User avatar
RogerClark
Posts: 7459
Joined: Mon Apr 27, 2015 10:36 am
Location: Melbourne, Australia
Contact:

Re: Rogers postbag

Post by RogerClark » Wed Nov 15, 2017 12:55 am

OK. Thanks

ag123
Posts: 798
Joined: Thu Jul 21, 2016 4:24 pm

Re: Rogers postbag

Post by ag123 » Fri Nov 24, 2017 1:25 pm

Google ads:
This is off topic, but I'm thinking given the traffic you may like to host google ads, they could in some ways help to lower the cost of running this, just 2c

User avatar
ahull
Posts: 1650
Joined: Mon Apr 27, 2015 11:04 pm
Location: Sunny Scotland
Contact:

Re: Rogers postbag

Post by ahull » Fri Nov 24, 2017 2:36 pm

ag123 wrote:
Fri Nov 24, 2017 1:25 pm
Google ads:
This is off topic, but I'm thinking given the traffic you may like to host google ads, they could in some ways help to lower the cost of running this, just 2c
You wont get much revenue from me I'm afraid, my adblocker will bin the ads. :lol:
- Andy Hull -

ag123
Posts: 798
Joined: Thu Jul 21, 2016 4:24 pm

Re: Rogers postbag

Post by ag123 » Fri Nov 24, 2017 2:40 pm

ahull wrote:
Fri Nov 24, 2017 2:36 pm
You wont get much revenue from me I'm afraid, my adblocker will bin the ads. :lol:
agree that google ads only pay tiny pennies, a lot of traffic generates only very little revenue, But if there is already quite significant traffic, it may make some difference, not too sure if that would be worth it though.

Post Reply