Thank you to our Sponsors
Anybody knows which capacitive touch interface chip the Beo6 screen is using?
Often those chips are directly located on the flex cable that connects to the digitizer.
once you have the type you can try to find its data sheet and try to interface with it using an Arduino or whatever. Typically it’s just I2C along with reset, interrupt and power supply.
if that’s not working it might be worth replacing the CPT chip which might got defective due to mechanical or electrical stress.
Anybody knows which capacitive touch interface chip the Beo6 screen is using?
Often those chips are directly located on the flex cable that connects to the digitizer.once you have the type you can try to find its data sheet and try to interface with it using an Arduino or whatever. Typically it’s just I2C along with reset, interrupt and power supply.
if that’s not working it might be worth replacing the CPT chip which might got defective due to mechanical or electrical stress.
It is this IC.
https://www.analog.com/media/en/technical-documentation/data-sheets/AD7147.pdf
/Weebyx
It is this IC.
https://www.analog.com/media/en/technical-documentation/data-sheets/AD7147.pdf
/Weebyx
okay, thanks. Good to know.
Luckily still available. So if that’s the common problem here it can be easily replaced.
It would just need a little further investigation to find out if that’s the root cause or not.
Other possibilities could be a broken flex cable or a damaged I2C port on the host processor.
It is this IC.
https://www.analog.com/media/en/technical-documentation/data-sheets/AD7147.pdf
/Weebyx
No, that seems unlikely! The chip has only 13 sensor inputs, so at best you'd get a 6x7 matrix of touchscreen points and that's way too low for an acceptable touchscreen resolution. I'm quite sure that the AD7147 is the chip that provides the wheel functionality on the keypad. Just look at the "Typical Application Circuits" section at the end of the datasheet.
It is this IC.
https://www.analog.com/media/en/technical-documentation/data-sheets/AD7147.pdf
/Weebyx
No, that seems unlikely! The chip has only 13 sensor inputs, so at best you'd get a 6x7 matrix of touchscreen points and that's way too low for an acceptable touchscreen resolution. I'm quite sure that the AD7147 is the chip that provides the wheel functionality on the keypad. Just look at the "Typical Application Circuits" section at the end of the datasheet.
None the less, this is the IC that is connected to the P400 connector, which is connected to the screen digitiser :) I am looking at the PCB right now, also the cable that comes from the touchscreen, has 13 connections that maps directly to the 7147.
/Weebyx
No, that seems unlikely! The chip has only 13 sensor inputs, so at best you'd get a 6x7 matrix of touchscreen points and that's way too low for an acceptable touchscreen resolution. I'm quite sure that the AD7147 is the chip that provides the wheel functionality on the keypad. Just look at the "Typical Application Circuits" section at the end of the datasheet.
The number of cap sense inputs does not equal to the maximal possible touch points. Those values are not 1/0 but more like an analogue reading from multiple cells that the controller will then use for calculating the exact position. When placing your finger on the screen the value of multiple cells will change.
The click wheel most likely uses an optical encoder which is located on the interface board in the "ball".
The number of cap sense inputs does not equal to the maximal possible touch points. Those values are not 1/0 but more like an analogue reading from multiple cells that the controller will then use for calculating the exact position. When placing your finger on the screen the value of multiple cells will change.
Good point. But it still seems an odd thing to do, repurposing this chip for touchscreen use when you have proven dedicated chips for that purpose. Synaptics seems to be a controller that is found quite often in high-end electronics for such applications since the early 2000s.
But well, the point is moot since Weebyx confirmed that the chip is directly connected to one FFC that goes to the screen and therefore it can have no other function than touchscreen decoding.
But - and that's a good thing - it does open the door for troubleshooting and even hacking the function of this touchscreen input. You'd need a good signal analyzer or digital scope for that though.
I find it hard to believe that the chip could have failed completely and I suspect a systematic problem just like a dead capacitor, bad solder joint or similar, which might interfere with the ongoing autocalibration procedure. The bias pin and the capacitor connected to it seems to have a crucial function for measurement and is worth looking at.
I'd be interested to know the CPU model B&O used for the Beo6. Is it still the MC9328MXs from Freescale that they used in Beo5 or is it a later model? (The CPU is just above the large FFC connector that goes to the screen (that cable is for the RGB signal for the screen).
Is pin 8 on AD7147 connected to the FFC that goes to the digitizer? That pin 8 (ACshield) provides noise immunity for the measurement signal.
And my final question, did you find a model number on the LCD panel itself. This might be helpful for finding the specs and/or a potential replacement screen. Beo5 used a Samsung screen, LTV250QV but there'a absolutely no information available on the net about the Beo6 screen maker or model.
The best way to start is just taking an oscilloscope and try to see if there is clock & data activity and a stable power supply.
Due to the good availability of the chip the next step would be replacing it carefully. While dead ceramic caps are very unlikely one additional thing could be replacing the single aluminum one. Even there I think a fault is pretty unlikely.
If that does not help chance are low breathing new life in the touchscreen.
I can't help here further as I neither have a Beo6 nor a broken screen.
The best way to start is just taking an oscilloscope and try to see if there is clock & data activity and a stable power supply.
Due to the good availability of the chip the next step would be replacing it carefully. While dead ceramic caps are very unlikely one additional thing could be replacing the single aluminum one. Even there I think a fault is pretty unlikely.
I fully agree with the first part. Measuring various pins on the chip with a scope can give you a good idea where the problem might come from.
I strongly disagree with replacing - or even ordering - the chip unless you have experience in SMT soldering and the right equipment for it. On Mouser it says that this is a 4mm CSP package (chip scale package) and without underheat and a temperature-controlled hot air pencil I see little chance of successfully getting a new chip mounted. Don't even think of just yanking it with an off-the-shelf uncontrolled soldering iron!
Replacing a 2-pin capacitor with a normal soldering iron and a high quality solder wick may be fine when you have some experience, but not a CSP package.
The second thing to try once you see that the digital signals are fine is to use a plain piece of flex cable and connect it to the digitizer zif-socket on the motherboard. This flex cable will act like a capacitive sensor and by gliding your fingers across that insulated FFC you might see a response from the touch controller. If so it would be a strong sign that the touch controller is detecting the change in capacitance and responding accordingly in which case you'd need to take a look at the ITO touch layer - or maybe the connection from the FFC to the ITO layers.
I strongly disagree with replacing - or even ordering - the chip unless you have experience in SMT soldering and the right equipment for it. On Mouser it says that this is a 4mm CSP package (chip scale package) and without underheat and a temperature-controlled hot air pencil I see little chance of successfully getting a new chip mounted. Don't even think of just yanking it with an off-the-shelf uncontrolled soldering iron!
Replacing a 2-pin capacitor with a normal soldering iron and a high quality solder wick may be fine when you have some experience, but not a CSP package.The second thing to try once you see that the digital signals are fine is to use a plain piece of flex cable and connect it to the digitizer zif-socket on the motherboard. This flex cable will act like a capacitive sensor and by gliding your fingers across that insulated FFC you might see a response from the touch controller. If so it would be a strong sign that the touch controller is detecting the change in capacitance and responding accordingly in which case you'd need to take a look at the ITO touch layer - or maybe the connection from the FFC to the ITO layers.
Sorry, but I really don’t see where this should be a CSP package. Have a look at the datasheet. It’s a standard QFN with EP and 0.5 mm pitch.
I agree that you should only try to replace it if you know what you are doing. I’m working with embedded systems every day so might have forget to mention that disclaimer... If you are trained doing such rework it is certainly much easier and quicker than any other possibilities after confirming power supply and host activity. You can also probe the interrupt pin that should go low on every change.
In don’t understand the second part of your reply. The open end of the flex cable (which connects to the MCU board) only carries power supply and digital signals (either I2C or SPI according to the datasheet). Touching there won’t have any impact on the readings from the CPT panel. The other end of the cable usually is well sealed and hidden behind the frame so no way touching there either. You can debug caps sensors with oscilloscopes or spectrum analyzers...
You must be logged in as a BeoWorld member to post replies to forum threads.
Register to access the forums, reference material, and exclusive member benefits — including our monthly prize draw.
Become a Member