Tell me more ×
Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. It's 100% free, no registration required.
up vote 3 down vote favorite
1

I am using a TFT display with ssd1963 controller built-in and stm32f4 discovery board, and everything is working well except when writing to the display. There is a significant increase in system noise and it is actually audible on the codec output and also effecting my ADC readings.

All pins are configured correctly and as I mentioned display is working very well. I am driving the display controller from the stm32f4 with 150R resistors because I saw this used in a schematic for a dev board with a TFT. I assumed the resistors were used to terminate the tracks to help with track ringing and impedance, but perhaps they are used to limit current.

Has anyone else seen problems with TFT displays and data pin noise?

Would putting an array of transistor buffers between the micro and display help as long as the buffers where given it's own regulator to keep noise out of the main power supply?

With these kinds of projects making changes like this are not trivial so looking for possible suggestions before busting out the soldering iron

share|improve this question
3  
Sounds like you need some serious decoupling capacitors here and there on the power connections. Do you have a photo and/or circuit diagram of what you built? – jippie Jul 23 '12 at 7:27
Yeah I've tried all the usual tricks with caps and beads but without much improvement. I don't have the schematic handy being on holiday. I have designed loads of digital boards in the past without these issues so usually pretty good at keeping noise low but this is an usual one because I didn't design the development section which is piggybacked on my display adapter board. I wish I knew what the display data inputs were like. Are they using a lot of current or are they highly capacitive? – Bruce Duncan Jul 23 '12 at 16:21
1  
Also make sure that how you are writing to the display in software isn't making the timing of delivery of the audio samples erratic. – Chris Stratton Aug 21 '12 at 13:11
Any luck with those schematics and pictures? – Analog Arsonist Sep 24 '12 at 20:10
I can hardly think of a TFT device that I owned which didn't make noise when updating the display. TFT stands for "tffft! tfffft! tffft!" – Kaz Apr 19 at 6:05

1 Answer

up vote 0 down vote

This is a difficult problem. Normally when you run a wide cable full of single ended signals with 1nS rise time, you can expect high supply/ground currents and high radiated noise. I am not surprised you can hear the video content on your audio.

Every EMC Engineer ought to have a book by Mr Henry Ott in their bookshelf. OK I don't, but I remember what he wrote back in the early 80's and it has served me well.

Let me recall some relevant topics.

Conducted Noise

  • Isolating Analog and Digital ground currents
  • Low ESR caps
  • multi-point grounds

Radiated Noise

  • Common-mode ferrite filters ( lossy noise absorption )
  • Shielding, shield termination and shield transfer impedance
  • ground planes
  • optimal driving impedance
  • controlled impedance cable and terminators
  • interleaved grounds or twisted pair
  • standing wave effects

I recall the biggest radiation on ST506 device FCC testing was always the interface 50 wire ribbon cable.

If you were to pick the top 2 of each list and implement an appropriate fix, I would hope these will eliminate your problems. If you use ribbon cable I would find a flat bar ferrite sleeve to absorb the radiating noise and locate it near the transmitting end or a clamshell sleeve just like on all VGA cables. The ferrite beads help on differential noise but not CM noise radiation. I think this fix is a must have. Laptops will usually have a flat ferrite layer somewhere over the LCD ribbon cable.

I probably missed a few topics but I don't believe buffers will help your issue unless they are ECL differential.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.