LCD bias supply

homosapien

Hi all,

I am working with a LCD display that requires a negative voltage for bias (~ -20v, 5ma). I am using a Propeller Proto Board, so I have regulated 3 and 5 volts available. Any one know of a quick and easy way to get the needed -20v?

I have looked at the MAX749, but it seems a tad expensive ($8 + shipping) and still needs external components.

There is a design by Duanne Becker here:

http://home.myfairpoint.net/~snowleop/gdisp3/negboost.bmp

that seems to be what I am looking for, but I would like to understand it a little better before actually using it. My questions would be;

1) I believe the 555 is just creating a square wave for the rest of the circuit - can I replace that part of the circuit with a square wave output from a prop pin (I 'think' the transistor would buffer the prop pin from issues)?

2) Does anyone have any good references as to how the non-555 part of the circuit works (I assume it is creating inductive flyback to create the (-)voltage - how does one figure the rest of the components to produce the desired (-)voltage/current? Yes, I know the schematic has these values, but how were they arrived at? I am not really looking for an exact mathematical description, maybe just a layman's idea of how it works....

Thanks for looking.

Leon

Use an output to generate a square wave and use a diode pump circuit to generate the negative voltage. It might be difficult to get -20V, though. A simple switching supply with an inductor might be better. You don't need the 555, anyway.

kwinn

Attachment not found.

You should be able to get away with just using this part of the circuit provided you are close to the frequency and duty cycle the 555 produced. Not really sure it is a square wave though so you may have to do a bit of experimenting.

When the 3906 is conducting the current through the inductor increases and produces a magnetic field. When the 3906 stops conducting the magnetic field collapses and induces a voltage across the inductor. If the inductor had no resistance the power out of the inductor would be equal to the power put into it. If the average input power was 100mW (5V x 20mA) the output power would also be 100mW (20V x 5mA). Keep in mind these are the averages. The actual currents through the inductor has higher peaks.

homosapien

Is the part of the circuit below the transistor actually a Boost Converter (http://en.wikipedia.org/wiki/Boost_converter), but with the diode reversed to 'boost' to a negative voltage? I am quickly throwing together the 555 part of the design to see what the actual output of the 555 timer is....

Tracy Allen

Observe that the 555 circuit in negboost.bmp specifies a CMOS 555. That is important, because it uses the symmetrical feedback connection from the output, not the DIScharge pin. Something like the LMC555 (with the "C") is good. The operating frequency will be several hundred kHz with the 100pf capacitor and 20kΩ feedback resistor specified in the negboost.bmp file.

Current builds up in the inductor when the base of the '3906 switching transistor is at the low Vss level. Current builds up in the inductor at a rate that is determined by the inductance and the supply voltage, in this case approximately
amps per second = 5 Volts / 100 microHenries
(ignoring some other less important factors)
So it takes 20 microseconds for the current to build up to one amp. The pulse that turns on the transistor has be set so that the current does not build up too high, both from the standpoint of what the 2N3906 can supply, and also from what the inductor can support in terms of heat and magnetic flux.
That negboost.bmp circuit has no feedback and the voltage on the output can potentially build up to high, even destructive, levels if the load is light, as would be an LCD bias.