Hi,

I'm building a boost converter which incorporates the circuit below for a feedback reference. As expected, the voltage at OUT declines with the capacitor voltage which results in a predictable reference "range" for the output of the converter (the set voltage decays). How can I optimize this circuit with passive components only to minimize this range? Alternatively, I am interested in basic circuits that can provide a non-temperature dependent voltage reference. The goal of this project is to build from scratch and make use of minimal power replenished by a solar panel. The present solution is to design within the declining reference voltage which is acceptable as the final outputs are LEDs.



Regards,
Mark

 

The diode current will be between 100 μA and 270 μa, which is a pretty small diode current. What voltage rand would you expect on the non-inverting input of the comparator? The range will probably be 450-550 mV. Is that what you were expecting?

If I was doing this, I would be looking at something else.

 

The diode current will be between 100 μA and 270 μa, which is a pretty small diode current. What voltage rand would you expect on the non-inverting input of the comparator? The range will probably be 450-550 mV. Is that what you were expecting?

If I was doing this, I would be looking at something else.

Yes, I have observed a range of 400-600mV. The inverting input is connected to 10:1 voltage divider of the boost voltage (~300mV) so I suspect there is an optimal combination of the divider ratio at the inverting input vs diode current at the non-inverting input. If I could reduce this range to about 50mV or less would be alright. What else would you be looking at?

 

Yes, I have observed a range of 400-600mV. The inverting input is connected to 10:1 voltage divider of the boost voltage (~300mV) so I suspect there is an optimal combination of the divider ratio at the inverting input vs diode current at the non-inverting input. If I could reduce this range to about 50mV or less would be alright. What else would you be looking at?

I should add that I failed to illustrate a 10k trimpot between OUT and the comparator in the voltage divider configuration which sets the boost voltage. (OUT to left pin, comparator to wiper pin, GND to right pin). I can redo the schematic if necessary.

 

Yes, I have observed a range of 400-600mV. The inverting input is connected to 10:1 voltage divider of the boost voltage (~300mV) so I suspect there is an optimal combination of the divider ratio at the inverting input vs diode current at the non-inverting input. If I could reduce this range to about 50mV or less would be alright. What else would you be looking at?

How about the whole thing? It is hard to know what to suggest when all you've show is a fragment. I'm not even sure what voltages are available for the input and output of the converter. I'd recommend the incorporation of a TL431A as part of the feedback and output monitoring of a boost converter.

 

How about the whole thing? It is hard to know what to suggest when all you've show is a fragment. I'm not even sure what voltages are available for the input and output of the converter. I'd recommend the incorporation of a TL431A as part of the feedback and output monitoring of a boost converter.

Ok I will prepare the whole schematic. I only did a portion for this topic as there are other bugs I'm yet to work out in the design that are unrelated to this topic. In the meantime, thanks.

 

Ok I will prepare the whole schematic. I only did a portion for this topic as there are other bugs I'm yet to work out in the design that are unrelated to this topic. In the meantime, thanks.

It is hard to answer questions about a subcircuit without the context in which that subcircuit exists AND without actual performance requirements. More information is better than less information.

 

The circuit could be improved by using a constant current source supplying the diode.

 

An LM4040 is only 20p. I think a poor man could afford that - it's better than a poor reference!
The LM4040 can work down to 60uA, and it is perfectly stable with temperature.