Kentec QVGA LCD display blank even though proven program

Hello Sergey

Did you check if the backlight is coming correctly. It should be a PWM signal or a logic high connected to the PWM driver

Regards
Amit

There seems to be a logic high on the enable pin of the backlight driver circuit.

Also, the voltage input to the backlight is at 5v, which seems quite low. Why would this be?

You must (better) define your phrase, "Before it turns blank."

If the backlight is itself - switching off (greatly dimming) then Amit's direction is valid.

But - if the backlight's brightness remains (relatively) unchanged - the backlight is (likely) "off the hook."

If yours is an LCD (Not TFT) then a (proper) contrast voltage IS required.   Often when a display behaves as you describe - the contrast voltage is mis-set.   The mis-setting enables a "brief" display - then the improper voltage builds (as caps charge) which vastly decreases the display's contrast.

The screen is a TFT type. Blank means that it is apparently off with no way to tell one way or another. Upon plugging in it will sometimes be all white or display the correct program, both for a very short period of time. Thanks.

There (have) been changes to and a variety of displays used in that firm's "add-on" display implementation.
Being TFT - contrast voltage is NOT an issue.

I'd suggest that you operate the display in a dimly lit room - and try to detect if the display's (backlight) light output changes when it shifts from working to failed.

It is possible that you have a cable and/or connection issue - usually "ohming out" the MCU pins to display pins provides a valid "test/verify."

Ok I will continue to troubleshoot.

Another interesting artifact: when attempting to power the Tiva and screen through a non-usb source 5v, I only get a short blink on the green power LED and nothing works. But when I remove the screen all works as it should. Does that give some clue as to what is happening?

That's interesting - but not (yet) conclusive.

Are you introducing that 5V (properly?)

Is the non-usb source (adequate) to the dual demands of display & your MCU board?

The "improper connection" between MCU board and display has risen much higher on cb1's "Suspect List." (list invalid in 53 states - per el presidente...)

You're a funny person... Thanks for the laughs in this time of frustration.

The supply is most definitely adequate as I've been using it, as well as USB, for several days with the same set up. This all happened when I took my project enclosure apart earlier today. Improper connection is interesting... I'll hit the pins with an ohmeter but I can't image that there has been any damage.

The backlight LED voltage is only 5V, it should be 20 according to the data sheet... Seems like the driver is broken?

Ok so the reason the alternative (non-USB supply) fails is because of built in protection. My guess is that it is short circuit protection that is going off., but I'm not sure.

Edit: probably not a short. But I'm only getting 5v where I'm supposed to see 19V so I figure the LED driver is not working.

That's good detective work. (likely valid in all states)

Now your note of disassembly & re-assembly makes me even more interested in a pinched, shorted or otherwise inadequate connection. I suspect that (usually) that base level 5V is "stepped up" to adequately drive the "series-connected, chain of backlight Leds" w/in your display.

Cannot you beg, borrow or (otherwise) acquire a variable voltage DC supply - current limit it to 20mA (if Leds are connected in ONE series path) and gradually apply increasing voltage to the proper backlight pins - with ALL ELSE disconnected. (we are seeking (just) backlight test/verification.)

I'm reminded that the backlight ground (usually) but not always ties to circuit ground. Check this - my money (not much, though) rides on pinched wire or lost connection - caused by your enclosure adventure...

Your (possibly) suspect Led driver usually is PWM driven - which provides variable brightness.   While you report a 5V "measurement" this may be PWM - which confounds your DVM.   Near to 100% PWM will fully drive the backlight's Led chain - lower % PWM will decrease light output.   I'd (really) ohm-out the: PWM signal from MCU to Led Driver and then "scope" the voltage across the display's backlight connections.

Hello Sergey

In which case you should see the USB port also shorting out. USB ports have generally a built in OC protection that shuts down the port when the 500mA current limit is exceeded.

Regards
Amit

Thanks Amit and CB1

There is only so much I can do from home as I don't have a proper scope or variable supply. It helps to know that USB has protections in place also. I'm still not sure what it is though. The metal case around the LED does not connect to ground, which is suspect. I'm not sure if it ever did, but one one guess that it should.

Access to proper equipment normally is required to resolve such issues.

KISS dictates that you, "Isolate and then properly power the backlight - independent of the backlight driver - and note results." One area of focus at a time yields the fastest (and best) diagnostic results...

It is 100% a backlight issue. I tore the thing apart and can see the correct information displayed but it is very dim. The wiring seems fine but I can't tell. I guess I have to try to power the LEDs up to figure out if it is the driver or the LED string. In any case, I already ordered a new one... thanks guys.

Hello Sergey,

The LED backlight requires current to be sourced (I believe it is 19.2mA). You may want to power the back light separately

Regards
Amit

Once you gain access to proper equipment you may (properly) power (independently) the Led backlight w/in the display. Note the current limit!

The failure of the Led string driver may be caused by an output short (most likely) - this (anywhere) between driver output and its destination Led connector and/or cable/wires.

Well, the power inductor indicated continuity through it. Though I guess that is expected as it is a piece of wire. How can I find a short?

sergey dusheyko said:

How can I find a short?

Measure from LED Driver Output (+) to Output (-.)

Should the short have been strong & long enough - the LED step-up converter IC may have departed...

There is a FET somewhere w/in your driver circuit - it or the converter IC itself are "prime suspects."   (sometimes that FET is w/in the converter IC)

Ah I see. (ic, hehehe)... Anyway, the LED driver output measure 5V and there are no apparent shorts. But this 5V should be 19V.

If a FET is present it should attach to one end of the inductor. The FET's gate should be the recipient of PWM from the micro. (you'll need a scope to note that)

The 5V you note is unlikely (to be kind) to drive a series string of white Leds. Led driver should "step-up" that voltage.
Have you tested the LED driver's output when Disconnected from the display's backlight? While unlikely - excessive current may be drawn w/in the backlight circuit itself - pulling down the driver's (proper) output.