Looking for low-end uC families with LCD controllers

Hi all,

I have an application that needs a very low-end, low-cost micro that can drive an LCD (maybe 100 segments total). So far I've found AVRs and PICs. Does anybody know of any other families I should be looking at? 8051 would be nice, but I'll look at anything that can get the job done. Thanks.

Mike

Reply to
Mike Silva
Loading thread data ...

MSP430, but you pay quite a bit extra for micro's in general with built-in LCD controllers, for very low end we just use any micro's I/O pins and a handful of resistors and some software to drive the mux LCD glass directly.

formatting link

Reply to
steve

The ATmega169P is very very very cheap.

Reply to
larwe

I've seen the technique, but I didn't realize there'd be a significant cost penalty vs. doing it with I/O pins.

Mike

Reply to
Mike Silva

We got the Atmega169PV-8MU (MLF-64) for $2.5. But we want a smaller LCD-less chip (MLF 32) for $1.5.

and transistors

The only problem is generating the proper AC voltages with varing battery voltage.

For us, it's the package size penalty.

Reply to
linnix

not compared to an 88 it isn't

Reply to
steve

You mean a mega88? About $0.25 difference in large quantity.

Reply to
larwe

it's not that the LCD controller itself is expensive, it's because the non-LCD micro selection is about 100x greater, so you are more able to find something that just meets your needs, as opposed to buying a lot of stuff you don't use

Reply to
steve

for $2.5 you can buy a nice 32 bit ARM with 20K of RAM

resistors

no transistors, just two resistors per common, see the reference article

in those cases we read the voltage and add a "contrast" cycle to adjust the AC voltage. But it depends on how much your voltage varies, modern LCD fluids are getting so good that many of the PIC micros with LCD controllers don't have a built in contrast controllers (using a charge pumps) any more

Reply to
steve

Right, that certainly makes sense.

So, if I wanted to explore the GP I/O approach, I see some questions arising:

1) The articles I've seen show resistive dividers to get the 1/2Vcom level. That's a killer for a battery-powered device. How does one get the 1/2Vcom level without wasting current in resistors?

2) Related to #1, how does one get the 3-3.5V swing on the LCD from a

1.5-1.8V battery?

3) One poster mentioned needing transistors as well as resistors. Where do the transistors come in?

I know I didn't specify that my device would be battery-powered in the OP. That fact alone may argue for the built-in LCD controller, I'm not yet sure. I'm still _very_ early in the exploration phase.

Reply to
Mike Silva

Must be fairly small quantities. You can get under US$1 for the first, and well under US$0.75 for the second goal.

Reply to
larwe

There are a number of 80C51 cored devices that target Energy Metering. These usually have segment drive sufficent for meter-display apps.

Analog Devices, Austria Microsystems, Teridian, Winbond (telephony) all have Metering/LCD drives with 80C51 cores

Not 80C51, but low end, are EM Microelectronics, etc

-jg

Reply to
Jim Granville

that must be really large quantities, how large is large?

Reply to
steve

You can duty cycle it rather than voltage dividions.

Transistor boosters for constant 3V supply from 2V to 3V battery (CR3032).

See above.

You don't have to. Battery is expected for LCD devices.

For steve: Should compare to 168 (16K), just to be fair. AC voltage above 3.3V will kill the LCD over the long run. AC voltage below 2.7V will kill your eyes (trying to read it).

For larwe: No way we are getting 169 for $1, even for 10K.

Reply to
linnix

Bit of an apples and oranges thing. The customer I'm thinking of uses very small quantities of mega88 (maybe 15k/year). The other part, probably 250-300K/year, maybe more.

Reply to
larwe

For that, you should really look into mixed-signal ASIC (w/ uC core). It might not get much cheaper than $1, but you can put everything else in it, i.e. op-amp, charge pumps, etc.

Reply to
linnix

Not worth it. The application consists basically of the LCD, some switches, the micro, and a radio. NREs and foundry costs would be more than what we currently pay for the components. And we'd have to redo the software.

Reply to
larwe

depends entirely on the fluid, I've used coin cells to drive low mux LCD's directly (3 to 2 volt range)

formatting link

Reply to
steve

and on the glass and on the manufacturer

That's for static LCD.

For mux-ed LCD, cross couplings between segments and leakages of glass are problems.

Reply to
linnix

LOL! I have customers in the scientific instrumentation market for whom 200 units per year is cause for celebration and holiday bonuses all around! The large/small quantity decision is generally related to the DigiKey quantity 25 price!

One of the nice things about niche markets like that is that the customer doesn't really care whether I specify 256K flash part and use only 120K. Selling price is a several times parts cost to pay off the R&D involved, so a few bucks here and there on parts costs to make the programmer comfortable is no problem.

Mark Borgerson

Reply to
Mark Borgerson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.