UCC28070:ucc28070

Hello Vitalij, 

Two points concerning the UCC28070 PFC controller: 

1.  This device does not easily work with DC inputs.  To do so requires that artificial zero-crossings are periodically injected into the VINAC input, so that VINAC can follow changes of input voltage.   Because this input was designed for rectified AC signals, it is a little more complicated to mimic this signal than simply grounding the VINAC input with a transistor for a few microseconds. 

2.  The UCC28070 current control loop is designed to wave-shape the PFC input current based on the IMO pin voltage.  I am not an expert on MPPT, however, I believe that your goal is to vary the input current to maximize output power.  So I suggest to use an analog circuit to modify the IMO signal.  I don't recommend to chop the CAOx output signals with grounding comparator outputs.   The IMO signal will influence both A and B channels at the same time to the same magnitude, anyway.  

Since the IMO output is a current source into a resistor (Rimo), another resistor on IMO can connect to an external Op-amp which can raise or lower the IMO voltage around the nominal operating point to achieve MPPT.   And it can do it smoothly and continuously (except for during the artificial zero-crossings).

Unfortunately, I am not able to help you with details on designing the MMPT controller.  I can only help with using the UCC28070 within its capabilities.

Regards,
Ulrich 

Hello Vitalij, 

This is what I mean:
 

Rimo1 is the normal Rimo that you calculate for your normal maximum power at lowest input voltage. 
Rimo2 is a adjustment resistance that couples a modification current into the IMO node to raise or lower the IMO voltage to maintain MPPT. 
LM358 is connected to some circuit (that I don't know how to design) that tracks MPPT and causes the LM358 output to raise or lower its output to affect Vimo. 

My thinking is that the UCC28070 is always boosting and regulates output voltage and changes input current as output power (load) rises or falls. 
In this solar arrangement, I assume that Pout is held constant (or changes very slowly) and Vout is regulated by the control loop as usual.   
Meanwhile the solar panel voltage (Vin) goes up or down depending on clouds or whatever, so if Vin goes down Iin must go up, so Vimo must go up. 
Actually, this is the way the boost PFC normally works, so while "explaining" this, I think I don't understand MPPT correctly.   

I'm beginning to think that the LM358 should control the output loading, not the IMO voltage.  
Please ignore Rimo2, above, and set Rimo1 to program the maximum power that can be processed from the panel(s) regardless of input voltage. 
So now I think that you have to be able to raise or lower the output load current (because Vout is regulated to a constant) to achieve MPPT. 
I'm sorry for the confusion earlier. 

Regards,
Ulrich

Hi Vitalij, 

Yes, if the Pout jumps and changes IMO, there will be a transient change to Vout until the voltage-loop control can react and restore regulation to Vout. 
The amount of transient ΔVout depends on the amount of ΔPout, the rate of change of Pout, the V-loop bandwidth, and how much Cout you have.
In most PFC designs, this ΔVout is typically < 10% of nominal Vout, but your results will depend on your design.

As I mentioned earlier, I am not an expert on MPPT, so I am not able to help you further on how to adapt the UCC28070 control to accomplish this.    
Maybe some other reader of this forum may have some experience with that and can contribute. 
Otherwise, experiment with your ideas and see how they work. 

Regards,
Ulrich

Hello Vitalij, 

Yes, I once worked on a "paper-design" using the 555 timer to generate the low-going pulses, but never completed it. The complicated part of the design is not generating the pulses, but shaping the edges of the pulses. 
It has been found that straight up and down edges do not work on the UCC28070. Something about the VINAC sensing does not "like" the straight edges.
  
The complication is to add some angle to the falling and rising edges, to be similar to the edges of a rectified AC signal at VINAC. 
Besides the "soft" edges, the device also needs to see VINAC < 0.7V for at least 50us.  

Recreate that signal shape and the artificial zero-crossings should allow the UCC28070 to detect falling changes of DC input voltage. 

Regards,
Ulrich