I need help on designing a schematics for solar study light project for Africa using BQ24040 and TPS61070

John,

I can't design it all for you, but I can definitely help you along the way!

From your list above, it seems like the bq24040 will be an okay choice. One immediate problem I see will be thermal performace - the efficiency of your charger will be very poor. Assuming a sunny day and your solar panel is at it's full 6V, and you are starting from a "dead" or almost dead battery (call it 3V), you're efficiency will be 50 percent! Assuming that the ambient temperature outside will be "hot", and the solar panel will be "hotter", it is likely that you will hit the thermal regulation of the charger.

One thing you might look into is using super capacitors as a buffer between the solar panel and battery charger, although this will add cost.

Hi David,

Thanks a lot for your reply. The charging efficiency (as well as power consumption efficiency) is extremely important, we should not waste even 0.1W of power. That is why the issue you have pointed out is very important. However I can not understand why the charging efficiency will be very poor (about %50). Is it because of the input voltage is low (6V / 3W) ? If so, I can use 9V -3W panel instead. If this is still not enough, then may be I can try to use 9V - 4W or 5W. But I am not sure if this will improve the efficiency, because this action will only increase the amount of power supplied as input. So could you please help to explain the main reason? so that I can be able to figure out the issue.

Another issue is the temperature which I don`t really have any idea of its logic. Yes, I know that during the sunny day it will be hot, and the panel would be much more hotter. But will this affect the temperature of the IC or charge control board? How can the IC understand that the panel is very hot? Besides as I have observed personally, the temperature of the panel will not exceed 100 C degree(I haven`t measured by thermometer, but it will not burn my hand, I can stand more then a minute), so in this case will it still hit the thermal regulation of the charger?

As the last issue that you have mentioned, could you please advice a specific super capacitor that I can use? Also, shall I use it directly in between solar input and the charger IC?

Besides, I have decided to use MC34063 instead of  TPS61070 since 34063 is a buck & boost inverter (but not buck converter only like TPS61070), so it can manage to provide a fixed voltage/current to LED lights regardless the battery voltage. It also can provide up to 1.5A which will be more suitable for USB charging and LED bulb which are supposed to supply 1A or more.

Finally, we have designed below preliminary schematics. Any comments or suggestions would be most welcomed.

Thanks,

John

John,

The reason the efficiency will be low for the charging IC is that it acts like a linear regulator. The way a linear regulator works is by using a resistive element (usually a mosfet), to dissipate the energy needed to go from one voltage to a lower voltage. In general, efficiency can be thought of as power out divided by power in, or (Vout*Iout)/(Vin*Iin). In the case of a linear regulator, there is only one path for current to go, so Iout and Iin are the same value. This means that the efficiency will be the output voltage divided by the input voltage, or eff = Vout/Vin. In the case of a 1 cell battery being charged and a 6V input, this means that your efficiency will probably range from 2.9/6 at its lowest (<50%) up to 4.2/6 (~66%). A switching regulator is much more efficient (usually between 85-95%), but is much more complicated and usually more expensive.

Usually, a solar panel will be hotter than the ambient temperature outside (it is a black surface sitting in the sun, so it gets hot). If it is an unusually hot day (or a normally hot day in Africa, depending on where you are), then the panel will be very hot. If you have the charging circuit glued onto the back of the panel (I have seen this before), then the IC will generate more heat (the temp of the panel + the heat generated by the charger). So on a very hot day, or during the hottest part of the day, it is possible that you reach the charger regulation temperature very quickly. It may not happen, just something to think about.

As for how the charger "knows" the temperature, there are ways to create integrated circuits so that they have a temperature coefficient. If you know the temperature coefficient, you can compare that voltage to a steady voltage (with no temperature coefficient), and turn circuits on/off with that signal. This is extremely oversimplified, but that's the idea behind it.

As far as a supercap goes, I can't give you a good suggestion at the moment. The biggest concern will be to make sure you have a high enough voltage rating (blowing up a supercap would be very bad), plus some margin of safety. Digikey, Mouser, or your local electronics supplier will most likely have some supercaps. You can also read more on the topic
here:

www.sensorsmag.com/.../using-a-small-solar-cell-and-a-supercapacitor-a-wireless-sen-7310

I think the MC34063 will be a good choice. It is a fairly simple application circuit as far as buck-boost converters go.

As far as your schematic goes, I don't see any glaring issues. I think the charger circuit looks good. I can't give you 100% certainty on the other items, but I think they look alright. My biggest concern is with the usb charger, but don't see anything immediate. My suggestion would be to make the separate parts first, and then put them together.

Also, if you move onto layout, make sure you read about best layout practices for thermal performance. You will want lots of copper on both layers, especially around the linear charger. Since you aren't relaying much information, switching noise from the MC34063 probably won't be much of an issue.

Hi David,

Thank you very much for amazing clear explanation. I am planning to use the charger board inside of the study lamp, so it will not place close to solar panel, in this case I guess temperature will not be an issue so I will not need to consider a super capacitor, if I am not wrong.

However the efficiency is a serious issue for me that I didn`t know the difference between linear and switching chargers. After your reply, I had a quick look on the switching chargers, and the one with the lowest cost is BQ24157, but this seems extremely complex, and I don`t really know how can I implement this in my system. Even though I would only need its charging function, it seems I will still need to deal with a lot of unnecessary items. So I have decided to continue with BQ24040. Here I just would like to know if I use a MC34063 as buck converter to reduce the voltage but increase the current to maximize the total efficieny, will it work? As I know MC34063 has a good efficiency, but I am not sure if BQ24040 will be happy to get the input from MC34063 instead of solar panel or not.

Besides, I just heard that MC34063 is not automatically switching buck&boost, so this will be a challenge for me.

Thank you also for your suggestions on the schematics and layout design.

Regards,
John

John,

No problem! Hopefully it helps.

The supercap suggestion was not for thermal, but to improve the charging functionality. One of the problems with solar is that the capacity will change with the conditions outside - this in turn effects the chargers ability to charge the battery. Using a super-capacitor in between allows you to slowly charge the caps with whatever solar capacity is available, then cleanly charge the battery from the capacitor bank. This may be better left for a "2.0" version of the project.

Unfortunately, your suggestion for changing the current to maximize efficiency will not work. With a linear charger, the charge current is irrelevant for efficiency, but will impact the temperature.  If you wanted to maximize efficiency, you would want to keep your input voltage (Vin) as close to the battery voltage (vout) as possible. For a linear charger, this means you would have to keep your input voltage about 0.3V above the battery voltage, and you would have to have a way to track the battery voltage as well. This definitely would not be easy to implement cheaply and would require some thought.

I once had a professor tell me that the best way to tackle these tougher projects, the best way was to get something to work, usually the core of your project. This will be the "1.0" project if you will. This is usually the toughest milestone. Once this is complete, you can focus on improvements. In any projects, there will almost always be iterations.

As far as the MC device, I apologize for not noticing that it was a buck or boost only.

If it's possible to get away with only 5V for usb charging (which is the standard, I believe. Either 5V&1A or 5V&0.5A), then you could do something like this:

For the boost, you could use something like the TPS61253. For the buck converters, you could use the LMR10520 from the simpleswitcher family. These devices are easy to use, and you can still implement your resistor switching idea to change output voltages. The IC's are a little more expensive than the MC device, but you will save overall because of the smaller component count. Your efficiency will also be higher, and the design will be more simple.

Hi David,

I agree with you that the first step would be the most difficult one, and the next versions would be much more better.

I will try our current design first to see performance, if it can satisfy our minimum requirements, then we will start with that and keep improving with better parts to enhance the efficiency.

Just one question: do you think it will be a good idea to choose a solar panel with a voltage of 4.5V instead of 6V, while keeping the same max. power of 3W? Will this action help to improve the efficiency? If you think this will help, then what is going to happen if the sky is not very clear or during the early morning & late afternoon? I don`t have intense knowledge about it but I think 6V max should still be a better choice. Here I would be happy to know your suggestions.

I will try to share the results when we got the boards sample.

Thanks again for your kind helps.

Regards,

John

John,

No problem! We are here to help.

4.5V would definitely increase the efficiency of the charger. When you look at solar panels, you will want to look at their I-V curve.

For solar panels, they will have an open circuit voltage (Vsc) specified in the datasheet. Then, as you draw current from the panel, the voltage will eventually begin to drop. The manufacturer will supply this curve, and from that you will be able to see your panel characteristics. Their is one point on the curve where the I*V product is at its greatest - this is called the Maximum Power Point (MPP). That is the "optimum" operating point for the panel, but it is okay to draw less current than the MPP. Drawing more current will result in the panel "crashing", and no charging will occur. For that reason, I would actually suggest using the bq24090 for your charger. The bq24090 has what is called VinDPM, which will reduce charge current to keep your panel from "crashing". It is functionally the same, but also two cents/kilounit cheaper. I apologize for not catching that earlier.

One problem with dropping to a 4.5V panel, as you mentioned, is when conditions are not perfect. The panel voltage/current capacity will drop. So then you must make a choice between higher efficiency in good weather, or a more robust design.

Hi David,

I have learned a lot so far, by your great helps, and I think I am going to be a professor soon :))

After your suggestion, I will try with 4.5V panels to compare for sure when I got the boards.

You have pointed a very important item -DPM-, which I have also wanted to ask you at the beginning, but somehow I forgot.

The charger I have used in my design is bq24040, which is almost same like the one you proposed-bq24090-, both of them have the DPM function.  Actually one of the main reasons I have chosen this charger was the DPM function. In the datasheet of bq24040, it says that the charger reduces the current when it senses the power is not sufficient, so that it can maximize the charging power by assuring the voltage is enough for charging the battery.

But what I learned so far tells me that, this will still not be the best charger because it uses linear topology. But if the DPM function has a good contribution in terms of charging improvement, especially during the cloudy sessions which the input voltage would be lower then average, then I guess bq24040 will satisfy our needs for the beginning. But if you think there are some other advantages of bq24090, then I may also consider to use it.

Thanks,

John