WXforum.net

Weather Station Hardware => Blitzortung => Topic started by: Nova on November 01, 2014, 05:33:57 PM

Title: Reducing Interference Part 2
Post by: Nova on November 01, 2014, 05:33:57 PM
Image 1543 shows 1000 Hz 1mV peak-peak interference caused the the system display dimmer.  This noise goes away when the display brightness is set to 100%. 

Has anyone been able to eliminate this noise without increasing display brightness to maximum?

Will full brightness shorten the display lifetime?

Thanks,
Joe
Title: Re: Reducing Interference Part 2
Post by: W3DRM on November 01, 2014, 05:43:55 PM
Joe,

This is a known issue with the LCD used in the Blitzortung RED systems. The recommendation from the developers is to set up your display settings so the display is at full-brightness and/or so it shuts-off after some period of time. I have mine set as shown below:

(http://carsonvalleyweather.com/images/wxmisc/2014-11-01_1439-LCD-Settings.png)

It shuts off after 60 seconds. There is a noticeable difference in signal noise between it being on and off.
Title: Re: Reducing Interference Part 2
Post by: miraculon on November 01, 2014, 08:03:31 PM
Quote
Has anyone been able to eliminate this noise without increasing display brightness to maximum?

Will full brightness shorten the display lifetime?

I have mine set to 100% full time. The controller is out in the garage, and I have a Raspberry Pi cam aimed at it, so I want it on all the time. It is an LED backlight, so I expect a long life out of it.

(http://www.miraculon.net/garagepi.jpg)

I have not dug into the issue, but I suspect that some combination of a lower frequency and slower rise/fall times on the PWM might help. Also, if the grounding was poor for the return path of the PWM there might be a common-impedance cause for the noise. (ground ripple getting into the ADCs)

Greg H.

Title: Re: Reducing Interference Part 2
Post by: W3DRM on November 01, 2014, 08:47:18 PM
It's the PWM that is the cause of the hash noise. From what I have seen on my station, it's really not that much but, in an effort to keep all noise at the lowest possible level, I simply have it set to time-out after 60 seconds. I see about 5mv with the LCD backlight on and about 2mv with it off. That is looking at the Signals page with the 5v sampling Enabled.