TPS61200 Solar LiPo charger questions

The IC will work with voltages as low a 0.3V.  You may be running into power limitations.  The datasheet equations will help you calculate how much current you get get out as a function of Vin and Vout. You need to monitor the input voltage as well as the output voltage and current to make sure the IC is working properly.  It could also be turning on and off due to fluctuations in input voltage above and below UVLO.  If charging a battery, you need a series resistor to limit the current.  Without the resistor, the TPS61200 sees something close to a short circuit condition.  In general, I do NOT suggest using this IC to charge a battery.  It does not contain the necessary protection needed to physically keep the battery from either electrical, chemical, or physical destruction.

Thank you very much for responding.  I suspect your are correct about the low current and will be testing much higher mA solar cells shortly.  I certainly understand that this circuit is not ideal for charging a LiPo, but the mention of low ouput solar cells caught my attention.  I have been looking at the bq24071 as a potential choice as I found an article and schematic where this was used with a solar cell.   I have a 4.5v cell arriving soon and will work on laying this out.  I would be happy to hear if you have any other chip recommendations.  Thanks. 

After thinking about it a little bit, I think your best solution is going to be to run the solar cells into the TPS61200,  Then run the TPS61200 output into a standard Li-Ion battery charger.  TI has many of these and you can look the selection guide to find one that meets your circuit requirements.  Then charge the battery from the charger.  You can set the charger current low enough that it does not drag the TPS61200 output voltage down.

That should work to use the TPS61200 output as an input power source for the battery charger such as bq2408x, which is a standalone linear battery charger for low charge current applications, or bq2415x/bq2410x switch mode battery charger for efficiently use the solar energy or high current the applications. You can program the charge current, battery charging temperature ranges, and safety timers...

That sounds like a good approach.  I have the bq24086 chip to test  with the basic diagram in the supporting docs.  I also found an article that uses the bq24071 with a solar cell and will be getting that chip tomorrow to test.   Thanks for all the feedback.

I read your article on using th bq24071 with a solar cell and that also intrigued me.  I have the bq24086, but as I will be charging a 10mA or 20mA LiPo, I was wondering if I could use the bq24085 and remove the need for the thermistor?  the solar cell I plan to use is 4.5V at 22mA - not the most efficient, but fits my size constraint.  Any thoughts are appreciated.

Well, I wanted to follow up on your suggestion and started by creating my own PCB.   I have attempted at least 6 times to design a board that mimics the layout of the evaluation kit.  I cannot get any results from the voltage out.  I have tested all my connections for shorts and am using the parts listed in the BOM for the TPS61200EVM-179 yet I have no results.  I have connected a .5V up to 4V source and get about. .020 on the voltage out.  Oviously something is wrong.  I am not an engineer, but have designed other boards with parts this small and have had great success.  I beleive my layout to be correct, but perhaps the fact that I am using only one layer is the issue. 

Any thougths on my layout or designing this circuit for a single-sided board are greatly appreciated.

Thanks,

Mike

You are going to need to do some basic troubleshooting.  Look at the voltage on each pin of the IC to make sure the circuit behaves as expected.  I am very concerned about a single layer layout.  The TPS61200 is a switching power supply.  Layout is critical.  The external componets must be in the correct location AND have proper routing.  If you change the routing, you must be able to understand parasitic inductance, capacitance, etc.  Does the IC provide the same results with and without load?  My suggestion is to order a TPS61200 EVM for comparison.  You might also try changing the IC incase you damaged it due to ESD during your assembly or testing.

Thanks for the quick response. I did some further testing and while I followed the best possible layout with a single layer - I do beleive that might be the issue.  I just tested a solar cell that puts out .5v/ 175mA and got .389 at the voltage input and .036v at the output, but when I measured pin 1 right at the chip before the .1uf cap I got 2.534V but on the other side of the cap I get the .036v.  While the voltages vary on different test boards the results are the same - a reading at pin 1 but not past the caps - not sure if that is a sign of something specific.   This is a test without a load attached.  I agree about the eval board and I have one coming on Monday which should help in the trouble shooting.  I also think there might be an issue with the way I have connected GND and PGND.  Based on the eval PDF - it looks like both are connected to the PowerPad. 

I have received an evaluation module and that has proved used to test out solar charging.  When I hook up a 0.5v /175ma cell, the voltage at the input pin is about .345 which is enough to run it, but since the TPS61200 requires .5v to start the process, if I increase the light a bit, it is usually is enough to kick it in and right now it generates 3.3volts.   if I hook up any of my other solar cells that are anywhere from 3.2v to 4.2v, but only range from 10-18mA, I barely get .105v at the input without a load.  

Are you aware of a minimum current requirements for this chip?   I also want to make sure I match the current out to the cell specs I am charging.

Look forward to suggestions.  Thanks.

You must monitor the input power to the TPS61200 to make sure you are not pulling too much power out of the solar cell.  At startup, the TPS61200 pulls more current than it does during steady state.  This extra current is needed to charge up the TPS61200 output capacitors.  This additional current may be too much for your solar cells to deliver.  If the solar cells can not maintain a stable input voltage, they do not have enough current capability.  Note the power difference between your two different solar cells.  0.5V*175mA = 87.5mW.  3.2*0.018=57.6mW.  You must also consider at what point on the solar cell power curve you are operating.  Proper testing and analysis will require an oscilloscope and current meter that can be monitored by the oscilloscope.  You will need to monitor the currents, voltages, and power versus time during startup to see if the solar cell is capable of operating under those conditions.  One option that may work is to start the TPS61200 into a no load condition until the output voltage is in regulation.  After all voltages are up and stable, then enable the load.  You will need to be able to monitor the input voltage when you enable your system to make sure the solar cell maintains a stable input voltage.

Hi,

I have just run a couple of experiments with a TPS61200 eval board (3.3V) and a nominal 3ma load.  When slowly increasing the input voltage from a bench supply I did not see the device start until the input voltage was around 0.8/0.9V.  If the input was switched on at 0.5V then the TPS61200 did start but it does not seem to at 0.5V if the voltage is brought up slowly.

1) Has anybody else observed this or can repeat it.

2) If this is correct then it seems at odds with the intended use with single solar cells (~0.5V) where due to clouds etc the voltage is likely to vary slowly??

Thanks

Guy

That was very helpful and certainly makes sense.  I was finally able to design a two sided board that did work as expected.  Thanks for all the clarification!

I have not measured it the way you have described - but with all my testing I am not surprised that was what you found.   The solar cell I am using now is rated at .5v, but when connected puts in about ..423V.   As long as it is close to a light source it does jump so my adjusatble output voltage of 4.2V is reached  and does charge the LiPo cell nicely.  Keep in mind I am only charging small 30-50mah lithium polymer cells.   I agree with your second statement and limited sun.  I was hoping this device to serve as a "solar energy harvester" of sorts to make us of available light.  Perhaps different input caps might help? 

I was able to design and fabricate my own board that is about a quarter the size of the eval board and it works great. 

Gomez Adams said:

Hi,

I have just run a couple of experiments with a TPS61200 eval board (3.3V) and a nominal 3ma load.  When slowly increasing the input voltage from a bench supply I did not see the device start until the input voltage was around 0.8/0.9V.  If the input was switched on at 0.5V then the TPS61200 did start but it does not seem to at 0.5V if the voltage is brought up slowly.

1) Has anybody else observed this or can repeat it.

2) If this is correct then it seems at odds with the intended use with single solar cells (~0.5V) where due to clouds etc the voltage is likely to vary slowly??

Thanks

Guy

 

 

I understand the need to monitor the input current levels but what I cannot seem to find is the minimum current required to start-up and then maintain operation.  The power difference you explained certainly makes sense, but in choosing solar cells that fit in the physical space, I am trying to determine what current constraints I have. 

Since  I know the circuit will operate once engaged are their any thoughts on somehow using a quick burst from the LiPo battery to charge the input capacitors at ramp up and then disconnect that burst  - not sure how that can be done automatically or if it is at all prcatical - just a thought.

As far as monitoring the input voltage when the system is enabled.   Would it be possible to place a low current draw SMT LED in the circuit to indicate when it has turned on.  The LED would not need to stay on or perhaps blink periodically to indicate it is operating.  I am looking for some visual way of knowing when it has engaged. 

Look forward to all thoughts!

Mike

Michael Day said:

You must monitor the input power to the TPS61200 to make sure you are not pulling too much power out of the solar cell.  At startup, the TPS61200 pulls more current than it does during steady state.  This extra current is needed to charge up the TPS61200 output capacitors.  This additional current may be too much for your solar cells to deliver.  If the solar cells can not maintain a stable input voltage, they do not have enough current capability.  Note the power difference between your two different solar cells.  0.5V*175mA = 87.5mW.  3.2*0.018=57.6mW.  You must also consider at what point on the solar cell power curve you are operating.  Proper testing and analysis will require an oscilloscope and current meter that can be monitored by the oscilloscope.  You will need to monitor the currents, voltages, and power versus time during startup to see if the solar cell is capable of operating under those conditions.  One option that may work is to start the TPS61200 into a no load condition until the output voltage is in regulation.  After all voltages are up and stable, then enable the load.  You will need to be able to monitor the input voltage when you enable your system to make sure the solar cell maintains a stable input voltage.

I just wanted to follow-up again about incorporating a low current draw LED in the output to idicate when the converter has engaged.  This LED could be a mechanically activated momentary indicator or perhaps a periodic flash to save energy.  It just seems that in practical field use without a meter, there is no way to determine if the light source was strong enough to engage the converter since greater that .5V is required. 

Any simple circuit ideas are greatly appreciated!

Thanks,

Mike

Try the attached circuit. You'll need to use this as a starting point.  The relative values between the TPS799 input and output caps should set  the pulse duration and duty cycle.  The LED should flash any time the input voltage is high enough to generate a VAUX output voltage.

Here is the attached file.

Michael Day said:

Try the attached circuit. You'll need to use this as a starting point.  The relative values between the TPS799 input and output caps should set  the pulse duration and duty cycle.  The LED should flash any time the input voltage is high enough to generate a VAUX output voltage.

Great - thank you very much.  I was out of the country and was unable to reply sooner.  This sounds like exactly what I am looking for and I will give it a shot.

Thanks,

Mike

nancy! said:

Here is the attached file.

 

I have built the circuit and have the following results:

I have not adjusted the flash frequency yet as it is much to fast but what I have noticed is that it begins a very fast and faint flicker when the output voltage is around .4 V and then when it reaches the target 4.12 V the LED appears to stay on and stop flashing or it appears that way to the eye. I am guessing that the flashing and then solid on is correct behavior?

I am most concerned in knowing when the 4.12V is reached. Any thoughts are appreciated as well as any suggestions to lengthen the flash as programming is my background and I am not sure how to best "adjust the relative values between the input and output caps."

 

Thanks

Mike 

Try the attached circuit. You'll need to use this as a starting point.  The relative values between the TPS799 input and output caps should set  the pulse duration and duty cycle.  The LED should flash any time the input voltage is high enough to generate a VAUX output voltage.

nanobot9000: Mike said:

I understand the need to monitor the input current levels but what I cannot seem to find is the minimum current required to start-up and then maintain operation.  

Below is an example how to ensure that the input current does not exceed the cell capacity.

The idea is to monitor the input voltage to ensure it does not fall to low with regards to the knee of the solar cell. This is achieved by controlling the the feedback pin of the TPS61200 (via OPA349 output).

Further details, though in german, available at (including a higher def version of the schematics, just click on the picture in the article):

 http://www.elektroniknet.de/home/stromversorgung/fachwissen/uebersicht/uebersicht/gleichspannungswandler/sonnenenergie-direkt-in-die-batterie/4/

Bigger picture available in the article or by copy/paste below link in browser (script blocked on the e2e site)

http://www.elektroniknet.de/index.php?eID=tx_cms_showpic&file=uploads/pics/7120204_tm_03.jpg&width=1000m&height=750m&bodyTag=<body%20bgColor%3D"%23ffffff"><!--%20SZM%20VERSION%3D"1.3"%20--><script%20language%3D"JavaScript"><!--%20var%20IVW%3D"http://elektnet.ivwbox.de/cgi-bin/ivw/CP/98100";%20%20document.write("<IMG%20SRC%3D\""%2BIVW%2B"%3Fr%3D"%2Bescape(document.referrer)%2B"\"%20WIDTH%3D\"1\"%20HEIGHT%3D\"1\">");%20%20//%20--></script><noscript><img%20src%3D"http://elektnet.ivwbox.de/cgi-bin/ivw/CP/98100"%20width%3D"1"%20height%3D"1"></noscript><!--%20/SZM%20-->&wrap=<a%20href%3D"javascript:close();">%20|%20</a>&md5=e40842e52ed1b89e724d01a921bf2405

Hi,

Hi,

I want to power my CC2430 using the following structure:

Solar cell -> Li-Po batterie(or the best CC2430 power possible) -> End device.

ie, power the CC2430 with the batterie and using the solar cell to recharge the batterie.

I want to do that with TPS61202.

How have to be wire conection?

thank you 

As described earlier in this post, you would power the TPS61202 from the solar cell and use that to power the battery charger.  Possible battery chargers include the bq24080 and the bq25010.  The bq25010 has an integrated DC/DC converter that can be used to power the CC2430.

Excuse my ignorance, but I'm not familiar with these devices.

you could give me an example of wiring?

You can click on the part numbers in the previous post to look at the datasheet for each device. 

http://focus.ti.com/docs/prod/folders/print/tps61202.html

The output of the solar panel becomes the input to the TPS61202.  Vout on the TPS61202 would then go to the IN pin on the BQ24080.

http://focus.ti.com/docs/prod/folders/print/bq24080.html

OUT on the BQ24080 would go to the battery in parallel with the CC2430.

thank you very much and do you know what kind of solar panel I would have to use?

ie, the electric parameters to power the cc2430.

thank you again

The power consumption of the CC2430 will depend on lots of things specific to your application and use case.  Experimental testing using the CC2430 in your end application is the best way to see what its power requirements are for your design.

Hi,

One more question. If I have 5V in the IN bq24080, I will need 4.1 o 4.2V(this depends on the type of battery to be fully charge) in the Out pin and this have to go to the possitive batterry pin and the battery in parallel with the cc2430.

If this is correct, how I have to configure the pinout for achieving that value if the cc2430 supply voltage range is 2.0V – 3.6V.

Thank you very much 

Berto

If the CC2430 needs 2-3.6V, then a step-down circuit will be needed.  I recommend an LDO after the battery to power the CC2430.  A dual level LDO, such as the TPS780330220, could be beneficial to your system run time.

You may want to look at the latest product release for solar harvesting charging, the bq25504.