"Abusing" the wrong ICs to generate +/- voltage

[__stdcall]

Hey, I want to drive an LCD panel. I think I have everything I need except that I bought the wrong ICs to generate its power. 

The LCD requires Min 5.60 - Max 5.90v (+ and -). I bought a MAX660; not correctly reading that it doubles the voltage. Furthermore I have a few LM317 and LM337 (Not knowing you need negative voltage to regulate negative voltage).

Sigh.. So after some experimentation I've found out the following:

Driving the LM317 without resistors gives me a voltage of 5.77 at 6.25 input voltage.
Driving the MAX660 with the output voltage of the LM317 with 3 (instead of 4) 200nf capacitors at C1+-, C2+-, V- GND gives me -5.725v.

I guess using both ICs this way is out of spec, but in my naive understanding I now have +- voltage that fits the displays spec.

Can I use this setup temporarily to already tinker with the display until I actually get a proper solution, or should I not even try - as it might break the display, or might give me some other issues I need to figure out?

New ICs will take a while to arrive and I'm a bit inpatient, so a temporary solution would be fine for me. Next time I'll be more careful for sure. 

Thanks!

[Andy Chee]

Absolute maximum rating of MAX660 is 6V, so you are really close!

Absolute maximum rating of LM317 is 40V.  I'm not sure how you are connecting without resistors to get 5.77 volts.

[Ian.M]

The LM317 must be in dropout, with much of its control circuit non-functional as you've got nothing to sink the typical 10uA bias current from ADJ.  Unfortunately its dropout voltage with ADJ open is likely to be highly unpredictable and a small change in input voltage, load current or a moderate change in temperature could cause its output voltage to fluctuate beyond the acceptable ranges for the MAX660 and the LCD.

If you've got a higher supply voltage available IMHO your best option would be to set the LM317 up with its usual resistors carefully adjusted to get 5.70V out giving you 0.1V margin on the min. LCD voltage, but remaining 0.3V under the abs.max voltage for the MAX660.   That's not *good* as MAX660 correct operation above 5.5V isn't guaranteed, but if it works at all it should remain working until you can design and build a better power supply.

The next fly in the ointment is your display's unknown current consumption from each power rail, and the possibility it may have power sequencing requirements this simple circuit cannot meet.   Without knowing its specs, we cant comment on how high the risk of damage is. 

[Kleinstein]

One could set up the LM317 to create some 2.9 V and than use the max660 to double the voltage. Depending on the load the voltage may drop and one may have to adjust the voltage of the LM317 a little higher. This is still a bit dangerrous, as with a lower load the voltage may go higher than 5.9 V.

The voltage range for the LCD is rather small - usually they are a bit more foregiving. Sometimes they also need more like a negative supply in addition to the supply that also handles the logic signals.
A LCD only makes sense if there is something to send data to it, usually a µC of some kind. The µC may also provide an added square wave to dirve a diode based charge pump for a little extra voltage (either negative or positive).

[Buriedcode]

What voltages do you need exactly? It isn't clear what your input voltage is, or what polarity your output voltages are to be.  Does the display require +5.6V? or -5.9V?  And are these to be regulated and adjustable?

[Terry Bites]

Buy the right ones.