Hi

I am trying to build something equivalent to the MAX6433 low battery monitor system, but without the 140ms high batteyr, delay that the MAX component has.
I think I have the latch working- but now I need to set up a reference voltage of .615V. I do not want to use a V regulator. Can someone please tell me how to do this voltage reference?
(The circuit must draw as little current as possible).

Thank you in advance.
Regards
arrow

 

I didn't go and look at that part. How stable do you need that reference? Is 0.615V that critical? It couldn't be for example 0.65V? I find it hard to believe it need to be that accurate. COnsider these ways to create a reference:

1) Buffer a Vbe drop with discrete transistors. This way is very low cost and you can easily make the circuit temperature compensated. It is not the lowest current draw approach however.

2) Crudest of all would be to buffer a voltage divider with an opamp or discrete transistor follower.

3) Find a uA reference (national, texas instruments et al.) and a sub uA powered opamp(again check TI) a couple of high value resistors and you can make an accurate 0.615V reference with low output impedance. I think you can certainly do this complete circuit running on 5uA or LESS depending on how much current the reference has to source. This is a typical way arbitrary references are created. It is not the lowest cost way but it offers the potential to be the most accurate and stable.

 

Hi

Thank you for your post.
My problem is my V supply is drifting lower, and I cant use it in a V divider (at least I dont see how).
Thus I dont think your solutions would work for me.

Is there any other way to get a const V if my V supply is drifting lower.
Its not critical I set it to 0.615V it can be set to any value actually.

Regards
arrow

 

an opamp running on a moving power supply has somehting called PSRR. That is its ability to reject (not see on its output) the fact that it's supply is changing. If you have ripple for example on a power supply rail, the op amp is designed to reject this. Now this only works up to a limit. Obviously if the supply drops below what the opamp requires, it will cease to function but above that, you can get rejection. Why wont the opamp work? Or is your whole system powered by something like a battery or capacitor that is steadily dropping? If you can do what you need to before it drops too far, there is no issue.

What is your power source and why is it dropping?

 

Hi

My power supply is a discharging capacitor.
I need to setup the MAX6433 to allow the capacitor to charge up, reach a certain V and then switch on, discharging the cap through the useful part of the circuit.
I dont know if an op amp on the capacitor will work. I will need to produce a constant V off it- how can I do that?

Thank you
arrow

 

Well a discharging capacitor eh.. How fast is it discharging. A linear (low drop out regulator) sounds like a good fit for your application. Do you know that inside the linear regulators is an opamp, reference and pass transistor? Efficiency is the only real reason I can think of that makes a case for NOT using an LDO. My point with the op amp was that in order for its output to be stable (0.615V) it does NOT have to have terribly stable power supplies. The opamp will maintain its output the best it can unitl the supply drops so low that it cannot function. With that said, you can get opamps that work down to 1.2V supply rails and alll the way up past 18V. 3V -18V is not uncommon for a suitable supply rail on an opamp.

If the dishcarge is slow(relatively speaking, you need to specify) the op amp will reject this well (in otherwords your output wont move much from 0.615V). If the discharge is fast, the opamp will be able to reject that less causing a similar movement in the output (the magnitude of which, still might not be offensive though!)

If you cannot use an opamp, you can get a programmable voltage reference from a number of manufacturers but guess what is driving that output? An opamp. Your options:

1) Linear low dropout regulator. Inside: opamp drives a pass transistor. opamp still sees a falling supply input but PSRR rects it.

2) Programmable reference. Inside: a reference diode/Vbe with an _opamp_ gaining it up in a programmable way. Same situation as above.

3) You put your own opamp down on the board. Same situation again, PSRR rejects supply falling until a minimum is reached where it stops working.

4) discrete transistor buffer of a voltage divided zener. This is very similar to the programmable reference except your buffer is cheaper and you may or may not want gain. Cheaper parts but supply sag rejection will be worse.

5) A fixed current source into a resistor. The current source, if done correctly can have very high rejection of the falling supply voltage (this is the same as saying it has very good voltage compliance) then buffer the voltage with your favorite method (opamp, emitter follower whatever.)

6) another more exotic thing I can think of is build a high efficiency buck converter. It will accept a wide input range of voltages meaning you could design its input to work from say 2V to 18V or so.. feed it your sagging supply voltage and it will efficiently regulate its output at 0.615V. This will work but I think is major overkill for what you need. It also brings alot more switching noise into your circuit which is never desirable.


I think all 6 options can be made to work with a falling supply voltage to regulate 0.615V output. Which one do you want? You likely wont avoid an opamp somewhere in the circuit. Of course if your capacitor is fully discharged, nothing in the universe can run with ZERO power. So plan your recharge cycle accordingly.