Advice for a 16x2 LCD heater

[mcinque]

Hi all, I'm working on a device that use a classic, cheap 16x2 LCD screen. The device will be exposed sometimes a low temperatures (max -5 C) and since at that temp the crystals response is very slow, I need a solution to heat up the LCD screen.

The only commercial solution I've found is a very thin and adhesive transparent film heater that runs at 12V, that must be applied on the rear part of the glass screen.

Unfortunately those films doesn't come in 16x2 size (and cannot be cut due to their working principle) and even if I could use a bigger screen their cost is very high since they are not available as single items.

I was thinking to glue 4X1W resistors to the metal chassis that holds the glass in place and drive them to generate heat, but I don't know if this a good solution (since the glass has a low thermal conduction, I fear that heating only the chassis is not enough).

Do you have any advice about other ways to heat solve my issue? (except changing display technology )

[Rerouter]

nicrome wire looped a few times around the back driven off a constant current source?

[babysitter]

Place the heater resistor somewhere under the LCD, allow air gaps  and let the raising warm air reach it.

In the old times when germanys telco was still the state-owned "Deutsche Bundespost", the public phones had a heater resistor near the dial. The resistor was mounted in a kind of "chimney" which gave some advantage over well, no chimney.

[mcinque]

Thank you for your advices! For both methods, I will need something to control how much heat it will be dissipated, that cuts off the resistors when the temperature rises over a threshold. Since I should use 2W power dissipation with 170mA I guess I can use a PTC to do it. What do you think?

EDIT: I cannot use a PTC, the resistance rising curve is too far from the temperature I need!  (around 18°C, while PTC start from 70°C) 

[dannyf]

A transistor (TO220 for example) controlled by a mcu or a temperature sensor would function like a heater. A TO220 can confortably dissipate 10w and more with TO247/3P.

[holozip]

How about a 16x2 sized PCB with 35um traces arranged in a back and forth pattern? That way it'll be perfectly sized to mount on the back of your LCD display.

3D printers use a similar concept for their heated print beds:
http://reprap.org/wiki/PCB_Heatbed

Maybe you could hack up something like one of these:
http://www.dual-star.com/index2/Rider/heated_grip_kit1.htm

[strangelovemd12]

Is this battery powered?  Assuming you don't need super high efficiencies (4W @ 33% average duty saves ~1kWh/yr) you might try looking for a solution that doesn't overheat at ambient temp, but serves to keep the screen warm enough to function satisfactorily at -5 then run it constantly.

Is there anything inside that you can just overclock when it gets cold?

Generally speaking more small resisters spread around will be more efficient.  Also note that thermal transfer from a cylinder to a plane is terrible.  Maybe square package ceramics?  Probably less practical is making a small board almost solid with SMDs.  I bet it'd coat up with thermal paste real nice.  Might try that for one of those "keeps your coffee warm" electric coasters, actually.

Edit:  holozip came in with a more clever idea while I was typing.  Guess I guessed the printer beds were Peltiers.  Then you could put a fridge under it.  God knows there's enough time for a drink or two while it's printing.

[dannyf]

looking for a solution that doesn't overheat at ambient temp

You just need to put the temperature sensor next to the heater and use negative feedback. The heater would then only turn on if the ambient / heater is cold.

Look for some OCXO design - Elektor has something like that awhile ago.

[strangelovemd12]

dannyf, that should work fine as long as everything is reasonably stable.  As the heat flow gets more complicated and variable, you move towards having to err closer towards always on.  My remark was more about looking for the simple solution.

OP, what are the expected temp fluctuations inside the device due to the other circuitry?  How much/constant of a current does it draw, or is the screen mounted away from the main circuitry? 

[Rufus]

For a similar problem the solution was an array SMT resistors and a temperature sensor on a custom flex circuit.

[dannyf]

that should work fine as long as everything is reasonably stable.

Here is an example of how it could be done.

M1 is the heater (a mosfet in this case), and Q1 is the temperature sensor (=Vbe multiplier) mounted to the heater - it serves as a feedback mechanism.

At 0C, the heater allows Ids of 0.7amp -> dissipating 3.5w. At Q1 heats up, its Vbe goes down, turning off the heater gradually: to eventually 0.3w @ 40c.

Obviously, you will need to adjust the numbers (R2/R3) for your particular implementation but that's the gist of it.

In the real thing, they use kapton cable for more even heating and a comparator for more precise cut-off. But the basic principle is the same.

[mcinque]

strangelovemd12: Yes, it is battery powered, it runs between 12 and 9 volts and drains 150mAh at the maximum voltage. It must stay on for at least 24H (I can use maximum a 2.2Ah LiPo ).

dannyf: thank you for your schematic! I was just simulating the attached scheme before reading your kindly post, it looks similar.

When the PTC reach 25°C has 120R, and toghether with the 270R resistor it  will turn off the FET.

I'd like to shut off the current to the heater when the temperature rises over 25°C, so I think a PTC and a FET should work well.

What do you think about it?

[mcinque]

Generally speaking more small resisters spread around will be more efficient.

Yes, I would use 10x 10Ohm small resistors under the LCD.

[ConKbot]

https://www.sparkfun.com/products/11987 no need for a heater, operating range is -40 to +80C, dont know what characteristics may change, or how sunlight readable you need your LCD to be. 

[dannyf]

sparkfun.com... operating range is -40 to +80C,

I wouldn't bet my life on product spec from sparkfun.

[ConKbot]

sparkfun.com... operating range is -40 to +80C,

I wouldn't bet my life on product spec from sparkfun.

I wouldnt bet my life on a product spec from TI either

Buyer acknowledges and agrees that it is solely responsible for compliance with all legal, regulatory and safety-related requirements
concerning its products, and any use of TI components in its applications, notwithstanding any applications-related information or support
that may be provided by TI. Buyer represents and agrees that it has all the necessary expertise to create and implement safeguards which
anticipate dangerous consequences of failures, monitor failures and their consequences, lessen the likelihood of failures that might cause
harm and take appropriate remedial actions. Buyer will fully indemnify TI and its representatives against any damages arising out of the use
of any TI components in safety-critical applications.
In some cases, TI components may be promoted specifically to facilitate safety-related applications. With such components, TI’s goal is to
help enable customers to design and create their own end-product solutions that meet applicable functional safety standards and
requirements. Nonetheless, such components are subject to these terms.
No TI components are authorized for use in FDA Class III (or similar life-critical medical equipment) unless authorized officers of the parties
have executed a special agreement specifically governing such use.
Only those TI components which TI has specifically designated as military grade or “enhanced plastic” are designed and intended for use in
military/aerospace applications or environments. Buyer acknowledges and agrees that any military or aerospace use of TI components
which have not been so designated is solely at the Buyer's risk, and that Buyer is solely responsible for compliance with all legal and
regulatory requirements in connection with such use.

Either way, its OLED instead of LCD, so its going to operate over a broader temperature range than LCD, but possibly at the expense of being able to read the display with sunlight directly on it.

[mcinque]

https://www.sparkfun.com/products/11987 no need for a heater, operating range is -40 to +80C, dont know what characteristics may change, or how sunlight readable you need your LCD to be.

Thank you for the advice; well... the oled tecnology should be a lot less sensitive to cold with no liquid crystals that "freezes", but is much more expensive (and I should change my code because of new oled library).

Unfortunately the "operating" range doesn't tell us about the display response: the datasheet of this LCD wich is the one I'm using, says "operating range -20 to +70C" but I've tested it with freeze spray and a thermocouple: at 0°C the display takes two seconds to refresh!!

[ConKbot]

Mouser carries some OLED displays too, I dont know how much better their datasheet would be, but they are more expensive than the sparkfun unit.  Depending on production quantities, it could be cheaper than rigging a heater. The company that makes the displays sold on will certainly be more willing to give advice on if it would work than or not than sparkfun though.

Failing that, I'd check out thin kapton heaters (I'd use Omega in the US, dont know about international suppliers)  and see if you can slip them behind the backlight diffuser for the LCD.  If youre feeling daring, straighten out the tabs that hold the bezel on, take the LCD apart and put the heater (which you can get ahesive backed) on the back of the LCD diffuser) and reassemble the stack.  It should work fine if they are using zebra-strips to connect the glass to the pcb. 

[sacherjj]

Have you looked into high temperature range liquid crystal?  I know the display company we met with in China had some medical device LCDs that needed to respond in all temperature ranges.  This may not be an option if your quantity is low.

A quick search turned up a few, such as http://www.lumex.com/images/Lumex_ExtremeTempLCDs.pdf

I think it makes more sense spending money on a display that works in the temp range, than to try to bend the temp range to the display.

[Mark Hennessy]

Some thoughts...

- This thing consumes the best part of 2W. With an ambient temperature of -5C, it ain't gonna be -5C inside the case! If the upper ambient temperature spec allows, a sealed case will keep all of this heat inside nicely, especially if the case is non-metallic. If you can place the voltage regulators near the display, so much the better...

- I would be highly untrusting of the freezer spray plus thermocouple experiment. You'll get a whole range of random temperatures, depending on the thermal mass of the items covered in frost, and the thermocouple will only be in vague contact with part of it. Put the thing in the freezer for a while, monitoring the temperature with a thermocouple thermally bonded to the glass.

- If the LCD really doesn't work, given the above, then it's the wrong technology and heating the damn thing up is just crazy - every way you look at it! Sorry! It might be reasonable if it's a custom LCD with annunciators or similar (like the German telephone), but here it's just text. Other than code (trivial excuse!), why is OLED out? It won't have an appreciable lag at low temperature - it's a diode!

- Alternatively, VFD would work, and they are available in 16by2 compatible versions (meaning no code changes): http://www.noritake-itron.com/subpages/productse/vfmod7000.htm - e.g. GU112x16G-7806A - I've used these a lot, and they have no problems at low temperatures. They take more current than an LCD (without backlight) or O-LED, but no-doubt less that you're proposing to waste with a heater. And they can be switched off by software when not needed.

- LCDs are commonly used in cars, and they work fine below freezing. Yes, they might be slightly "laggy" before the interior heater starts to work, but what's the big deal? After all, they are only telling you non-critical information like the time and radio station you are tuned to. Actually, why is the lag a problem in this application - you're not trying to scroll text or anything? Can a rethink of the UI contribute towards an acceptable compromise?

- An old-school incandescent backlight might be a "nicer" solution, if you can work out the mechanics and optics.

- What if your heater system goes wrong? https://www.eevblog.com/forum/microcontrollers/lcd-display-fault-any-ideas-what-would-cause-this/

Good luck, whatever you go with!

Mark

[cellularmitosis]

mcinque, I have noticed at least two distinct backlight constructions for the common hd44780 boards which have become ubiquitous and cheap recently.  One variant of these boards draws 20mA for the backlight, and another draws 200mA!  If you can find a source for the latter, you might not need a heater at all.

Unfortunately, I can't remember where I got the 200mA variant.  It was either dealExtreme or ebay, but I recall that the display/backlight configuration was red on black.

[Stonent]

I was thinking if there's anywhere in your circuit that you're using a 7805 to drop 12 down to 5, mount that near the LCD to heat it up.

[strangelovemd12]

Dannyf- that should work fine as long as everything is reasonably stable.  I understand how could be implemented, my concern was mostly that a dynamic or conditional (dT thermocouple/dT of screen element) makes it less and less useful.  If the environment is too variable, the case an unpredictable insulator, etc., it is basically just taking a stab in the dark.  Such things could also make this system a huge PITA to tweak to the right values.
Don't get me wrong, it is still a decent solution, and may be the simplest one that is practical in this situation.

Ditto on the freeze spray tests not being the best method.  I second a proper freezer test, though the result might be identical.

I'll try resubmitting another suggestion in verse:  OC below 0C; key to frost free LCD if you ask me.

[mcinque]

I was thinking if there's anywhere in your circuit that you're using a 7805 to drop 12 down to 5, mount that near the LCD to heat it up.

It might be good, but since it must dissipate 1,5W I've put the heatsink out of the case, otherwise in summer I would have heat problems instead cold ones, since I can't use a jumper to disable it when the season is hot. Moreover putting the regulator under the LCD wouldn't allow me to control the heating 

A quick search turned up a few, such as http://www.lumex.com/images/Lumex_ExtremeTempLCDs.pdf

That modules has 15-20W heaters to keep the glass warm.  It's too much

With an ambient temperature of -5C, it ain't gonna be -5C inside the case! If the upper ambient temperature spec allows, a sealed case will keep all of this heat inside nicely, especially if the case is non-metallic

Yes, I use a closed plastic enclosure, and I was thinking the same. But I must test the worst scenario.

Put the thing in the freezer for a while, monitoring the temperature with a thermocouple thermally bonded to the glass.

mmm...  you're right.

- If the LCD really doesn't work, given the above, then it's the wrong technology and heating the damn thing up is just crazy

You're right, unfortunately for me I forgot to consider this LITTLE issue at the start of the project 

Actually, why is the lag a problem in this application - you're not trying to scroll text or anything? Can a rethink of the UI contribute towards an acceptable compromise?

Yes, i'm scrolling text and fast response is needed. Unfortunately I cannot rethink the UI...

- An old-school incandescent backlight might be a "nicer" solution, if you can work out the mechanics and optics.

mmm, i must investigate about this solution.

- What if your heater system goes wrong? https://www.eevblog.com/forum/microcontrollers/lcd-display-fault-any-ideas-what-would-cause-this/

Of course LCD will be cooked

Good luck, whatever you go with!

Thank you for your suggestions and thoughts!

I will try the freezer method and if it fails, I will try the oled 

[DTJ]

Two points:

Have you looked at heaters made by "Minco"

http://www.mod-tronic.com/Minco_Flexible_Heating_Elements.html?gclid=CKn2uarh3LsCFUshpQod2UgAgw


Also resistors are available in TO220 packages.