Driving LCD... Dot clock is a sine wave?

[cj555]

 Hi all,

I am looking at a wireless drive-through communication controller. It came off the wall and  hit the concrete floor. There was a small daughterboard that was dislodged but there seems to be something else wrong too.  Sadly I broke the ZIF latch on the LCD (7 inch LG graphical LCD with LED backlight)  and I'm wondering if I may have damaged some logic in the driver circuitry by the flat flex  moving around with the power on.

Unfortunately the only data sheet that I found for the LCD doesn't match the pin out. But the only line that I can find that looks anything like a dot clock (32 MHz or something) is a nice sine wave . 

 There is a BlackFin DSP and a Lattice micro of some sort, I haven't figured out yet which is driving the display.   I've checked the passives that I can find going to the LCD signal lines and haven't found any obvious irregularities.

 So… Any good ideas out there? The dot clock should be a square wave, correct?   Would a blown output driver on an ic somehow be able to make a sine wave? Can supply more details or post pics if it helps.
Many TIA !

[ataradov]

32 MHz is a pretty high frequency. How do you observe it? Make sure that you don't have 20 MHz bandwidth limit turned on.

[cj555]

Oh yeah! Hadn't thought of that.

BW limit was indeed on... However, turning it off doesn't affect what I'm seeing  .  Granted, I am using cheap Chinese eBay probes... That say they're 100 MHz on 10x. My scope is a Tek TDS 1002B, 60 MHz, 1 GS.

 So 36 MHz (the actual clock speed) should still be within my capabilities, theoretically.... Right? Even if not 100% accurate due to Nyquist?

Would attach a pic but having several issues due to trying to do it from my iPhone...


Nb, I just noticed my error in the first post... Dot clock should be a *square* wave... Will correct that.

[ataradov]

So 36 MHz (the actual clock speed) should still be within my capabilities, theoretically.... Right?

Not really.

What you are observing is a first harmonic of that square wave. The second one is at 36*3 = 108 MHz, and the third one is at 36*5 = 180 MHz.  You will not see much of that.

[cj555]

Oh yes, of course, silly me!  Have been out of this for too long, I guess.

Now to figure out what is actually wrong...
 Backlight comes on. Reasonable looking data on a bunch of the pins. The LCD did display for a bit when some things for wiggled around, but now nothing. That's what made me think I may have blown something.  But it didn't display the "leaky liquid" type of physical damage. Thankfully I caught that on video so I'll review that again and maybe post a frame or two.

Come to think of it, I need to check the reset signal.

I'll keep at it...

[Dachpappe]

As you mentioned the missing clip on the connector, I made the experience that a missing clip can lead to serious connection problems on these connectors. Maybe you can reseat the clip and see if it works then or measure if all signal lines are reaching the display-board?

[Rasz]

wireless drive-through communication controller.
7 inch LG graphical LCD with LED backlight

those things often use commodity parts, whats the model of lcd? is it a normal color lcd or special Monochrome /with controller build in type?

[macboy]

So 36 MHz (the actual clock speed) should still be within my capabilities, theoretically.... Right?

Not really.

What you are observing is a first harmonic of that square wave. The second one is at 36*3 = 108 MHz, and the third one is at 36*5 = 180 MHz.  You will not see much of that.

This ^^^
And consider the "actual" impedance of your probes at the frequencies of interest. They are 10 Mohm at DC only. They also have a tip capacitance of something like 15 pF (or more, for cheap 100 MHz probes). At 36 MHz, those 15 pF become < 300 Ohm. At the 3rd harmonic  of 108 MHz, the impedance that is loading the signal you are probing is < 100 Ohm. Consider the effects that has - the harmonics are loaded even more than the fundamental, attenuating them more.  For high frequency work I have a pair of Tek "Low Z" passive probes. These have a impedance of only 1 Kohm (not 10 Mohm) but input capacitance around 1 pF. So at high frequencies (above just 20 MHz), the "low impedance" probes have a higher impedance than "high impedance" probes. The ones I have are supposedly good to 3 GHz. No High Z passive probe comes close. To do better, you need at least an active FET probe which combines low capacitance with high-ish resistance (100 K +up).