LTC3105 not regulating properly

Hi,

I am attempting to use the  LTC3105 to regulate 5.25V. I am using 4 solar cells in parallel as the power source. At the max power point of the solar cells, they are rated at 1.67V and approximately 18.4 mA (so for four in parallel, that would be approximately 73.6mA). I set the mppc on the LTC3105 to 1.67V using a 167k resistor, however the output voltage does not reach 5.25V. Instead, it reaches approximately 2.5V. I have double checked my resistor divider that sets the voltage, and that looks ok, however I measured the FB pin and that was at 0.5V instead of the nominal 1.004V.

I tried lowering the mppc value to see what would happen, and I discovered that when I lowered the mppc to 1V, the output remained at 2.5V and the input remained at 1V (as one would expect). However when I flooded the solar cells with light the input voltage jumped from the programmed 1V input up to approximately 1.67V, which is the mpp of the solar cells. Once this jump occurred, I was able to achieve the regulated 5.25V output. 

I don't want to have to flood the solar cells with light and break the mppc limit, as that is not how this part is supposed to work, so does anyone have a recommendation for how I can get this system to operate properly? Could the issue potentially be with the startup of the part? It seems like once I flood the system with energy it begins to work, but it does not work under "normal light" conditions. My input capacitance is 10uF - is this to low for an application such as this?

For reference, I have attached a pdf of the solar cells I am using.

Thank you very much,

PaulPDF

  • 0
    •  Analog Employees 
    on Oct 29, 2019 12:37 PM

    Hi,

    I looked into the waveforms you provided. Since Vaux is more than 1.4V and LDO voltage is 2.2V the part is not stuck in the startup mode. Instead it is stuck in the region where AUX has enough energy to charge the LDO but it cannot charge Vout till its regulation.

    Can you find out the exact load on the circuit? The 2.5V buck regulator might be drawing current more current and leading to overloading.  The part seems to switch normally but is not reaching regulation because of overloading or the input itself is not able to provide enough energy for the output to reach regulation. You can refer efficiency vs output current graphs on page 5 of the datasheet for approx. maximum output current possible.   

    How does the circuit operate without the attached load?

    Have you also tried testing the circuit with a normal power supply instead of the solar cell? If it works fine then it means that the input source is a weak source and paralleling several solar cells depending upon the required load for the application should help.

    Let me know the results after testing the above two conditions.

    Usually higher input cap is required for lower input voltage source during startup. In this case since input voltage is high enough and is able to pass the startup mode.

    Also can you try the tests with Pin 4 (~SHDN pin) unconnected. This will negate the possibility of the excess overloading of the pin which can stop the part from reaching regulation.  

     

    Thanks

  • The load on Vout is approximately 2mA. I've not tested it without the load, but since the load is so small I don't think this could be the issue. I tried testing the system with a power supply and I found the following:

    I have the mppc set to 1.3V (130k resistor) and I set the power supply to 1.6V to mimic the solar cells. In this configuration, Vout still only gets to ~2.6V and the current draw is ~110mA. When I bump the power supply high enough (up to around 1.8V) Vout gets to the target 5.25V and the current draw drops down to 3mA. 

  • 0
    •  Analog Employees 
    on Oct 29, 2019 6:38 PM in reply to pl21engineer

    This seems strange as I tried the similar setup on the demo board DC1587A and was able to reach regulation. Also I assume 110mA current you mentioned is the measured output current at LTC3105 output pin. 110mA of load current is too large for the step up from 1.6V to 5.25V and might not be possible (LTC3105 has max peak current limit of 0.4V).

    I was able to regulate the output voltage of 5V with Vin=0.9V,Vmppc=0.8V, Vldo=2.2V and resistive load of 10mA. Similarly at 1.6Vin , I could draw load current of 50mA.

    I think the buck regulator connected at the output is drawing more current during its startup and leading LTC3105 to hit its current limit.

    You can try your setup with a resistive load to be sure if LTC3105 works as expected.

  • Hi Anisha,

    The 110mA was what I measured as being drawn from the power supply (input, not output on LTC3105). Regardless, I just pulled the buck regulator off and that is exactly what was happening. After I took the buck regulator off, the cells were able to boost up to 5.25V no problem. Since I still need to use a regulator of some kind, is there anything that you might recommend to help it push past its startup routine? I suppose I could change regulators if necessary.

    Thanks,

    Paul

  • 0
    •  Analog Employees 
    on Oct 29, 2019 9:09 PM in reply to pl21engineer

    Hi Paul,

    Does the buck regulator has an enable pin ? I think if you enable the buck converter only after vout from LTC3105 reaches regulation, the circuit might work as expected. Additionally if it has soft start feature, then using that option will limit the initial inrush current of the converter by increasing the startup time.  

    Thanks