Hi I'm thinking of using a 320x240 graphical touch screen for a new project. Price is a consideration. In NZ we can get a 'Truly' MFT-G320240DPEW brand for around $US50 in low quantity, which initially looks a bit expensive but compared to a smaller LCD at say $15 and a keyboard interface at around $15 the difference narrows. Can anyone educate me on what I should be looking out for when making a buy decision on these Graphic touch screens? Thanks, Alistair.
1 - What's it going to take to drive the LCD?
2 - Is an inverter required? Included in the price?
3 - Are the viewing angle and maximum ambient light for readability acceptable for your application?
4 - Does the LCD meet your temperature requirements?
5 - Is it going to be available long term?
I've been using 10.4" TFT LCD with a resistive touch screen (of the same size) and hopefully some of my pain can save you some of your own :-)
They take quite a lot of current to run - mine is over an amp to run the inverter. There can be odd voltage requirements too. The old QVGA screen my new one will replace is a 5V screen but the inverter needs 12V and also a -18V is required to control brightness or contrast (or something, I'm getting rid of it so I haven't looked too closely at the existing screen). Check the datasheet closely. Some of the cheaper oriental screens' datasheets are quite "creative" and don't always tell the full story.
The first few quotes for our LCD and touch screen packages came through without the inverter cost and it was a while before we cottoned on. Figure on adding another $US15- 20 for an inverter. They are the major current draw but I've found the 12V to be more efficient than the 5V, and cheaper, but you might not have 12V available. Some smaller screens (I saw a nice Hitachi one) use LED back lights and can run without an inverter but still are a bit power hungry. OLED screens for small displays are nice (and getting bigger all the time) and use much less power but don't come cheap.
If it has to be viewed in sunlight then get a transflective screen. They cost more but ordinary transmissive screens are normally pretty poor in sunlight. Most screens I've seen have pretty respectable viewing angles and it would appear that technology is such that even the cheap basic LCDs are pretty good. However, you might want to consider a screen with a very poor viewing angle if your user input is sensitive - like an ATM - to defeat snooping eyes.
Check the data sheet and be aware of cheating. In low temperatures then a decent picture will appear eventually but the response time can be shocking.
This is very important. Some of the better priced screens I've been offered are a little dicey in terms of future supply. This is fine for me because my setup is very flexible and I can use any old screen but if you're setting up for a big production run then it could be a BIG problem. I've heard that some of the cheaper chinese suppliers sometimes dramatically change their product between production runs and the screen you get in one delivery might have a completely different spec to the one you get the following week. But they are cheap and I can adapt my design a few days to accommodate that - not everyone has that luxury.
In addition to Lewin's excellent comments:
Consider the touchscreen design as well. I've used capacitive screens in the past which work regardless of how dirty they are but you have to touch them with skin - no gloves. However, one of them gave me a serious punch in the nose by failing minutes before an important demonstration and its replacement went out of it's mind and lost its calibration data at a tradeshow in front of customers. Never again. I switched to resistive after that and never looked back - reliability with my 3M resistive touchscreen and the controller in my Sharp79524 has been first class.
The resistive screens are much cheaper, especially if you are using a microcontroller with touchscreen support, but even without controllers are cheap and you are not tied to the touchscreen manufacturer's controller, as with capacitive. A slight down side is that they are not actually "touch" screens but are really "slight pressure" screens which seems to have foxed some of our dimmer salesmen but most normal people work that out pretty quickly.
Thank you so much for your response on this matter has been very helpful. I was not aware of inverters need with these things. Both units I've been looking at are the resistance type, so that is desired in any case for my job. Cheers, Alistair.
No problem. The inverter is a prickly problem on several fronts so spend some time on it. Not only are they power hungry, they are also high voltage which may well have implications on your standards and certifications and certainly on your labelling. They can also generate a lot of electrical noise so be careful where you position it.
Their sizes seems to vary wildly as well and I've had some nice neat little units and some great sprawling goliaths so it pays to shop around.
With the touch screen then I'm pretty sure resistive is the way to go unless your screen is likely to get serious abuse. As I say, I use a controller built into my processor but I've heard that the Tsharcs from the Hampshire company are very good controller ICs and cheap too. The support is supposed to be first class as well.
Of course, what I really want is a reliable way of detecting two touches at the same time but no one has come up with a good way of doing it yet :-(
We've not had problems with that, probably been lucky, but have a Xenon flash going so tend to stick things in metal boxes, screwed tight.
Thats my experience too. Most units require 12V. For a QVGA we're using 12V at 300 mA (can't remember the size, can't get to one to measure 'cos the floor is being relaid). Another product based on similar electronics uses a smaller QVGA display with white LED backlight but have had complaints about low brightness.
I'm using a LH75410 and would be interested in pointers to guidance or code to control a resistive panel. Currently we are using a grey-scale QVGA screen without pouch panel, plus a scanned keyboard but would like to move to touch panel in the future. We wired the scan returns to the touchpanel interface (in GPIO mode) in an attempt to simplify the change.
The 12V ones do seem to be more efficient than the 5V but you need an extra supply so space may become an issue.
Sharp do an outstanding application note which got me up and running very quickly. This is the direct link:
but it may be a redirector thingy so if not try
and the app note is the second one in the list. Very useful and gives a good background on resistive technology in general. This is very good if you want to know about calibration
. I begun having a go at a fancier method using linear regression but my maths classes were too long ago and I soon suffered an intelligence deficit:-(
I used the polled method with a slight modification which meant that the polling function returned immediately while the ADC was busy performing the conversion. This meant I could poll as often as I wanted without degrading performance. Before I did this then moving my finger across the screen quickly dramatically slowed down the system for some reason. The ADC probably converts more quickly when there is no touch. It was a head scratcher for a while. However, it is blindingly quick to respond now and there is no delay whatsoever - in fact I probably sample 50-100 times a second while the system is idle. I guess there's no point going faster than the refresh rate of the screen or you can't display the touch anyway.
That's some good foresight. With that app note then you should have no trouble getting up and running in a very short space of time. The calibration is more of a pig but if you're brighter than me (and nine out of ten gold fish are) then you'll be fine.
We already had 18 V available. Still, the backlight is the single largest power draw on the system on average.
This is very useful, but it looks familiar. I guess a skimmed it when originally doing the design.
It will probably pay for me to look at all the app notes and errata - must be 18 months since I did that.
I have lost many brain cells over the years but started out with a Math(s) masters, so there is some hope for me.
Its good to know what can be acheived.
Well the board layout guy has tried to thwart me at every stage by adding pullups and EMC filters and an inconvenient connector, but I try... foresight seems to be actively discouraged just about everywhere I've worked [grumble].
Not what I meant... OP needs to consider what is required to DRIVE the LCD, not what is required to POWER it. Discrete controllers are moderately rare and ~fun to interface. OP will probably find his choice of micro being driven by the LCD and this can balloon into a horde of interlocking issues - availability of compatible [SD]RAM, difficulty of PCB layout due to the need to keep trace capacitance constant across ~100 SDRAM controller interface lines, ......
From the tone of the OP it sounded like he already had his LCD controller decided and was just looking for a decision on the LCD and touch screen. I see what you mean by drive now.
On that subject controller to LCD interfacing, the cabling should be considered as well because there are some exotic and hard to make connectors about. My Logic PD kit has a rather funky little PCB which attaches directly to the little D-shaped connector (whose name I forget) on the Sharp LCD and allows the touchscreen ribbon to connect straight in. The inverter connects to it too but is dangling loose covered in heatshrink - a very tidy package all in all.
Thanks again for the heads-up on backlight. My previously preferred unit @ $NZ75 had no specs for the backlight except for a wee note in the power connection diagram of 100v backlight supply. Whoowee I would have been trapped on that one. The alternative which costs $NZ125 uses a LED backlight and uses much less overall power. Hmm, the $NZ15 2 line display is starting to look more attractive for this particular product. The usual formula is the end user pays four times what we pay for components. Alistair.