Multiplex LCD Driver uses too much power

Thread Starter

nuo9558

Joined Jul 26, 2018
3
I am using four 4-multiplex LCD Drivers (V6118-4) to drive a 120 segment static LCD. This might sound stupid, but I will explain why later.
Basically, the driver comes as a replacement for the old static LCD driver V6108, however, the new driver consumes way too much current compared to the old one, and the circuit board, which uses a charge pump, simply cannot provide enough current for it to function properly.

Back when the circuit board was built, it used three V6108 40-segment Static LCD drivers, those static LCD drivers had a very low power consumption, which is good because the circuit runs on battery and needs to last. But when the new driver came, it had a significant higher power consumption, surprisingly.

I had to use this particular driver for my static LCD despite it being a multiplex driver, because it has the same pin configuration and can avoid modifications to the PCB. If I can't make it consume less, I'd have to build a new power supply, and the entire board will have to be redesigned.

Is there any way to reduce the power consumption of a multiplex LCD driver?

V6108 datasheet: http://pdf.datasheetcatalog.com/datasheets2/33/339105_1.pdf
V6118 datasheet: http://www.emmicroelectronic.com/sites/default/files/public/products/datasheets/v6118_ds.pdf
 
Last edited:

joeyd999

Joined Jun 6, 2011
4,239
I am using four 4-multiplex LCD Drivers (V6118-4) to drive a 120 segment static LCD. This might sound stupid, but I will explain why later.
Basically, the driver comes as a replacement for the old static LCD driver V6108, however, the new driver consumes way too much current compared to the old one, and the circuit board, which uses a charge pump, simply cannot provide enough current for it to function properly.

Back when the circuit board was built, it used three V6108 40-segment Static LCD drivers, those static LCD drivers had a very low power consumption, which is good because the circuit runs on battery and needs to last. But when the new driver came, it had a significant higher power consumption, surprisingly.

I had to use this particular driver for my static LCD despite it being a multiplex driver, because it has the same pin configuration and can avoid modifications to the PCB. If I can't make it consume less, I'd have to build a new power supply, and the entire board will have to be redesigned.

Is there any way to reduce the power consumption of a multiplex LCD driver?

V6108 datasheet: http://pdf.datasheetcatalog.com/datasheets2/33/339105_1.pdf
V6118 datasheet: http://www.emmicroelectronic.com/sites/default/files/public/products/datasheets/v6118_ds.pdf
The problem is the resistor divider on Vlcd to produce the various voltages required by the multplexed LCD. You will not be able to get around this with this part.

Some new PIC micros use a capacitive divider (or higher impedance resistor divider) to minimize LCD drive current. I don't know if such technology is available in the same form factor as your chosen IC.
 

Thread Starter

nuo9558

Joined Jul 26, 2018
3
The problem is the resistor divider on Vlcd to produce the various voltages required by the multplexed LCD. You will not be able to get around this with this part.

Some new PIC micros use a capacitive divider (or higher impedance resistor divider) to minimize LCD drive current. I don't know if such technology is available in the same form factor as your chosen IC.
The problem indeed lies in the resistor divider of the multiplex driver. In fact the datasheet of V6118 says the internal LCD bias levels have a maximum impedance of 25k in the Vlcd divider bridge. With a Vlcd of 5V, this amounts to a current of 200uA constantly passing through the resistor divider and lost in form of heat.

That should be about 600uA if we take all three drivers.

But that's not what I observed.
These are the results I found when I did some current measurement using an external supply as Vlcd:

Code:
VLCD(V)            I(µA)          Calculated Equivalent R (kΩ)
3,2                130                       24,6
4,0                166                       24,1
4,4                180                       24,4
5,0                200                       25,0
5,6                230                       24,3
As you can see, when I measured the total consumption of all three drivers, the current is only 1/3 of what was expected. If the datasheet is correct, the equivalent resistance of three dividers should really be 8.3k instead of the 25k that we see here.
I can't understand how this is possible if the divider resistors are indeed 25k fixed. Is there any chance those resistors are actually variable and is there a way to influence their value?

Thanks

Moderetors note : used code tags to preserve spaces
 
Last edited by a moderator:

joeyd999

Joined Jun 6, 2011
4,239
I can't understand how this is possible if the divider resistors are indeed 25k fixed.
Because the resistors are not 25K.

The datasheet states:

Selection_024.png

It is talking about the maximum output impedance of each of the generated voltage levels, not the value of the individual resistors (which is unstated except for their ratios).

This information is required to know the maximum panel size that can be driven.

Stick to this table to understand the maximum supply currents (recognize that Itotal = Idd + Ilcd):

Selection_025.png
 

Thread Starter

nuo9558

Joined Jul 26, 2018
3
Because the resistors are not 25K.

The datasheet states:

View attachment 157004

It is talking about the maximum output impedance of each of the generated voltage levels, not the value of the individual resistors (which is unstated except for their ratios).

This information is required to know the maximum panel size that can be driven.

Stick to this table to understand the maximum supply currents (recognize that Itotal = Idd + Ilcd):

View attachment 157005
Thank you for your help, we have decided to change the driver.
 
Top