Anyone ever see an LCD display behave like this before?

[iaresee]

Slow, scrolling refresh is the only way to describe it I can think of so I made a clip. Pictures tell 1000s of words:

http://dl.dropbox.com/u/870088/rocktron-all-access-display-issue.mov

The unit was dropped off by the post office today. I let it sit in the house for 8 hours before powering it up, so it'd warm up. It's an aluminum shell so it's cool to the touch, but it should be room temperature by now. Argh.

[G. Hoffman]

I've no idea what's wrong, but if that's a new unit and it is still doing it tomorrow (not that it should matter), I'd say it is time to get an Return Authorization from Rocktron.


Gabriel

[iaresee]

Unfortunately it's used. I did fire off an email to Rocktron and while I was at it I had a peek inside the LCD is on a separate board, not the very large main board. Here's hoping it's an issue that can be fixed with an LCD board swap because that looks fairly easy to get in and out of the controller.

[cloudscapes]

isn't that a LED display? or mayeb a VFD?

anyways, looks like whatever multiplexer is being used to cycle through the digits is acting up.

[iaresee]

isn't that a LED display? or mayeb a VFD?

Opps. Yup. It's not LCD. It's a mulit-segment display ala VFD or LED.

anyways, looks like whatever multiplexer is being used to cycle through the digits is acting up.

That's what I was originally thinking, but see how the edges of the display blocks fade? That's weird. With a misbehaving multiplexer I'd expect characters sent to the display to continue to display as it cycled through the positions across the screen -- not fade away like that.

I'm wondering if there's something wrong with the power to the display. Filter cap gone bad maybe.

Good news is Rocktron tech support is very responsive. They can't seem to view that video, but they're asking all the right questions and seem to want to help even though the unit is used.

I opened it up last night and the display is on its own board, but unfortunately the ribbon cables joining it to the main board aren't socketed. Oh well. Hopefully a fix won't involve removing the main board because that looks like a beastly job. It's basically held in and suspended by all the switches. Lots and lots of things to undo!

[iaresee]

Sweet! Rocktron gave me the schematic and they too suspect a filter cap on the power rail for the display. I'll check it tonight with the scope.

[G. Hoffman]

Sweet! Rocktron gave me the schematic and they too suspect a filter cap on the power rail for the display. I'll check it tonight with the scope.

OK, now that is some pretty damn good customer service! 

I mean, giving that away, for a used unit?  OK, granted, it is of limited use without the code for the MCU, but still.  Pretty cool of them.


Gabriel

[iaresee]

Sweet! Rocktron gave me the schematic and they too suspect a filter cap on the power rail for the display. I'll check it tonight with the scope.

OK, now that is some pretty damn good customer service! 

I mean, giving that away, for a used unit?  OK, granted, it is of limited use without the code for the MCU, but still.  Pretty cool of them.


Gabriel

I know! They didn't have to give me the time of day, but the guy was incredibly helpful. And the schematics! Nice. You're right: you need the code, but it's all there in an eeprom...

Anyhow, I meant to bring the darn thing in to work with me today to check it and forgot. Tomorrow though. It looks like there's a ladder of caps on the main VCC line for filtering and then on the offshot that feeds the LED display there's another filter cap. The unit is fed AC, not DC, and given that everything else is functioning fine I'm guessing it's that filter cap on the line that goes to the display. If there was that much ripple on the processor's VCC in I doubt it'd be functional.

[iaresee]

Fixed!

I replaced the two big power supply filter caps (it takes AC from the adapter and has two AC -> DC conversion sections on its internal power board) and it is all good. But here's the weird thing. I started out probing with just a DMM. I figured the change happens slowly enough that I should be able to see it on ye olde DMM. But when I probe: nothing. It showed a steady 4.96 V across all points I poked at on the power rail that feeds the LED display.

Weird.

So I hauled it into work and sure enough I could see ripple on the line. What???

I tried the DMM again: no ripple.

And then a colleague happened to notice that the glow from the display on the desk was bright when I used the DMM. I couldn't see it because of where I was sitting in relation to the unit while probing.

Propped it up on end, probed with the DMM and...display is fine. Bright and vibrant. No ripples.

Very, very weird.

[JKowalski]

Maybe your DMM poking was fixing a messed up connection? Maybe the caps were fine after all?


And remember you won't see any ripple if you use your DMM on DC settings, since it will measure RMS.

[iaresee]

Maybe your DMM poking was fixing a messed up connection? Maybe the caps were fine after all?

Thought about that...but it fixed the display when I probed at different spots.

And remember you won't see any ripple if you use your DMM on DC settings, since it will measure RMS.

Ha! Got me there: I had it on DC. I thought I'd see it drop and climb on DC. Doesn't explain the display working when probed though.