OFF TOPIC - Op amps, power... that sort of thing.

[Gabriel]

hi all, sorry for posting an off topic question, but after some googling and wikipedia searches I am still faced with doubts.

I am building a solar power supply.
(to run PICs programmed with CCS hahahaha )

The solar panels (3.5 volts total 500mA) feed NiMh batteries and a MAX 756 ( 5V boost chip).

I want to strap a 5.5V 1F (yes, one Farad) capacitor to the MAX756 output.

The MAX756/MAX757 combine a switch-mode regulator
with an N-channel MOSFET.
According to the datasheet it can source 200mA.

I was thinking of using a dual, single supply op amp, with both amps configured as voltage followers and placed in parallel, to buffer the boost chip output to the load and super cap.
(Maybe a resistor in series to prevent oscillation caused by the high capacitance?)

Seemed to me like a good idea.... that whole impedance matching thing..... or was it unity gain?....

Any suggested ideas?.... reading material?

My intended loads are intermittent and might pull high currents bursts, and the boost chip is intended to only be on to refresh the supercap if its voltage falls below a certain threshold. (The load circuit will be "sleeping" most of time).


..screw that Thevenin dude :(

Thank you in advance.... sorry for the OT question.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[SherpaDoug]

A unity gain single supply buffer has two inputs. One is the signal input, whose voltage is copied to the output. The other input is the power supply which supplies the current for the output. Where do you intend to connect these two inputs. If you connect them to the same point the buffer will do nothing for you.

Do you need any regulator between the supercap and the load? Calculate how much the supercap voltage will drop and see if it is within spec. If you charge the supercap to 5.25V and the load burst draws it down to 4.75V you may be OK with no regulator. You have to look at the voltage tollerances of your chips and circuits.
_________________
The search for better is endless. Instead simply find very good and get the job done.

[Gabriel]

hello and thanks for your reply,

Let me see if I can clarify some things:

The boost chip feeds of a rechargagle battery.

It outputs 5V@200mA regulated and current limited to the formentioned values.

I was thinking of using a op amp configured as a voltage follower as
a buffer between the boost chip and my load.

The op amp is powered by the boost chip, and the "signal" its buffering is the output from the boost chip..... ( is this what you meant ? ) (hope not)

I was thinking/hoping this might lower my power output´s impedance....
and keep my boost chip stable....

Is this even nessesary? ? ?

As for the voltages/ cap charge issues you mentioned...

All my ICs ... well the entire circuit is 5-3v tolerant....so I can discharge my super cap to ~3V....(but i dont intend to)

But... I forgot to mention that the boost chip is only OFF while the circuit is sleeping. Load circuit turns on the boost chip upon wake up.
Also the boost chip is turned on automatically when the super cap voltage falls to 4.75V ( funny you used that value in your example).

The super cap purpose is to provide current for any bursts needed and to keep VCC at apropiate levels when the circuit is sleeping. ( like a PIC and a RTC)

Hope that helps you help me ! hhahahahaha.... no, really, I do apreciate the help.

Thank you
_________________
CCS PCM 5.078 & CCS PCH 5.093

[FvM]

Usage of an OP buffer in the power supply path is completely absurd in my opinion. By the way, where do you get the OP's power supply?

By connecting a boost converter to a solar panel, you'll face the problem of setting an optimal duty cycle for the actual light intensity. This is basically a MPT (maximum power tracking) problem, and I fear, the converter isn't designed for it. If the supercap is charged to 5V, the converter will simply lower the power delivery, that's the simple point.

[Gabriel]

The booster is conected to a set of batterys that are charged by a solar panel.

The booster only "boosts" when 5V are actually needed. when the booster is off, the voltage on the cap can only drop to (Vbat - diode drop)... which ~3V... if it gets there.... ( booster recharges capacitor when the cap voltage drops to 4.75 automatically..... and the target circuit activates the booster upon wake up).

The op amp is to buffer the Fet on the boosters output....
because of the high capacitance on the output , the op amp will provide a lower output impedance (thats my idea).

The op amp is powered by the booster... and the op amp powers the super cap and the target circuit.


So... solar panel -> battery-> booster-> op amp buffer-> super cap-> circuit..

In that order.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[FvM]

O.K., the NiMH batteries should allow stable operation of the boost converter, but I don't see a purpose
of the OP buffer. At best, it causes a voltage drop. I think, you can simply omit it. You should provide a suitable
current limiting.

[Ttelmah]

The OP-Amp, is just a way of wasting power....
The output elements, will be less efficient, than a simple FET selected for the job, without all the extra circuitry, designed to give high voltage gain.
I'd suggest you 'step back', and if the output elements in the MAX booster, are not man enough for the job, go to the National semiconductor site, and use their design tool, to select a better IC.
However, I can't really see the 'point' of a lot of what you are doing. A regulated boost converter, will automatically switch itself off, when the output reaches it's target voltage, and will switch on when the voltage drops - no need to get the PIC involved here at all. Small, low ESR capacitors close to the switcher, will give greater efficiency, than a super-cap. Just have the solar panel charge the battery, and a boost inverter generate the voltage you require. When you don't need the supply, just use the shutdown pin on the regulator to turn it off.

Best Wishes

[Gabriel]

thanks for your replies....

FvM, Ttelmah
point taken. drop the op amp.... seemed like a good idea... but i guess i was overkilling somthing that didnt need to fixed.


Ttelmah

[NLXH6T]

If your output circuits only need uA, then use a uA capable boost circuit (there are a lot around).
Your approach is a bit like the run/coast approach to power saving, for economy on motor vehicles. It can work, but the vehicle needs to have a larger motor than is needed for constant power approach. If the course has hills and things, where the higher power motor is needed, it makes sense, but if the course is pretty flat, the greatest economy is produced by having the smallest possible motor, and designing this to run at it's best efficiency all the time. So, unless you need to have a bigger boost circuit, because your output circuitry has intermittent higher demands, then instead, use a smaller boost circuit, and design it for efficiency at the power needed. I have boost circuits producing 5v, from 3.3v lithium batteries. With care, efficiencies over 90% can be produced, and with output circuits drawing uA, only uA*1.5 are needed from the source supply. If however you need the larger boost capability, because the output circuit has intermittent higher demands, then it is worth going with the run/coast approach.

Best Wishes

[Gabriel]

exactly... running and coasting....

i couldnt have said it better!.... thats exactly what i am doing.

i need uA while sleeping... which is most of the time. but i have peak currents of 150 mA.... for about a minute... every 30 minutes. ...and in ocations... even at a larger interval.... 1 minute per hour...

fully charged 1F@5V Supercap, and 5V@200mA should cut it no?....
im running my booster off of NiMh 3.6 battery pack... 4500 mAh.


running the booster permanently.... seems like a waste of power when my ckt is sleeping.... what am i boosting power for?

my solar panel has now been updated to 5V@500. (i just bought an aditional cell)..... the whole panel is tiny...half a sheet of paper at best.
why so small? .... they steal them around here... im hoping the tiny size and power will deter thieves since its useless for any real power aplications.

so... thats kinda the reason im runing and coasting, as you so elocuently put it!

.... well, i think we are digressing from the main issue. the op amp in question.

i do apreciate further comments on design suggestions and so on....


thank you all.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[FvM]

I think everything has been said regarding the OP. I revert to posting on topic.

[Gabriel]

fair enough...


thank you all.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[mkuang]

How much power is the rest of your circuit (the part that you want to drive with the outpus of the unity gain buffers) drawing? Opamps usually can only sink or sources several tens of milliamps. Exceeding that is a very bad idea.

And exactly how are you powering the opamp again?

The output of a switchmode regulator, when combined with a capacitor value recommended by the manufacturer, should be able to provide a low impedance over all frequencies of interest. I have not seen one power design that uses an opamp to buffer the output of a power supply.