[pvrusb2] Things about noise handling

[isely at isely.net]

A few words about noise...

In an analog electronic pathway, everything said in this thread is 
absolutely true.  An analog signal has effectively zero "hamming 
distance" between the signal level and any source of noise.  If noise 
gets introduced it is impossible to later isolate and remove.  It is why 
analog systems are subject to signal degradation and it is why such 
systems require everything around it to be EMI/RFI "clean".

But in a digital electronic pathway, none of this applies.  Once the 
signal is changed into a stream of bits, the noise has to be strong 
enough to flip bits before it becomes a problem and even in those cases 
if the signal is encoded with forward error correction (e.g. redbook CD 
audio, OTA digital video) then even if bits get flipped the system can 
frequently reverse the damage.  So long as the bit depth (bits/sample) 
is enough to exceed your expected noise floor and so long as the 
sampling frequency is better than 2X the maximum highest frequency of 
the incoming signal (the Shannon limit), there's simply no chance for 
signal loss / degradation.  Once it's digital, noise is generally not 
going to be a problem and if it ever gets bad enough to "be" a problem 
then likely it won't be subtle and the signal destruction will be total 
(e.g. bit stream completely garbled or dirty power causing processing 
elements to just simply crash).

So for a problem like what is discussed here, you have to evaluate what 
part of the audio signal path is analog and what part is digital.  
Because it's only the analog part of the pathway where you really have 
to worry.  And with an HVR-1950 (or an older PVR-USB2), once the analog 
signal reaches the first processing chip it's digital from that point 
all the way to your hard drive (or whatever you are doing with the 
signal).  So the only analog parts here are basically whatever 
electronics might be in front of the chip that is capturing the audio 
(e.g. the cx25840 for HVR-1950).  And I expect it's probably just a few 
passive parts (perhaps low pass filter) and some clamping diodes for ESD 
protection of the input and that's it.  If there really is noise getting 
into the audio - for a line-in input - likely it was already there once 
it reaches the HVR-1950.  Because there's not much of anything beyond 
those RCA jacks that is still analog.  It's not impossible, just 
probably also not likely.

Regarding the RF input, involving things like antenna amplifiers, again 
it's a digital vs analog thing.  If you are capturing older analog TV 
signals, like one of NTSC / PAL / SECAM, then yeah a poor amplifier that 
is injecting or accepting noise isn't going to help at all.  However 
again if your incoming signal is actually a digital bitstream, like ATSC 
here in the USA, then the interference has to be strong enough to 
actually disrupt the bits AND overwhelm the forward error correction - 
and then you're totally hosed.  This is why digital TV signals are 
essentially perfect and seem immune to noise, at least until it gets so 
bad that you lose the signal entirely.  There's generally no middle 
ground here.  So again, if that pathway is digital TV, things like metal 
vs plastic encased antenna amplifiers probably won't really make much 
difference.  It might for analog TV.

  -Mike



On Sun, 21 Apr 2019, Roger wrote:

> > On Sun, Apr 21, 2019 at 08:19:31PM +0000, Gary Buhrmaster wrote:
> >On Sun, Apr 21, 2019 at 9:27 AM Jan Ceuleers <jan.ceuleers at computer.org> wrote:
> >
> >> The audio defect is a clearly audible repetitive "whistling" noise, with
> >> a periodicity of about 1 second. This noise is not present on the audio
> >> signal that goes into the PVR.
> >
> >Just on principal, check the power supplies
> >(swap between the units that work and one
> >of the STB ones).  And also check for ground
> >loops between the PC, the PVR and the
> >STB, which can result in all sorts of audio
> >artifacts (just go over the the audiophile
> >forums and discuss ground issues and
> >whines and whistles).
> 
> I agree with Gary here concerning noise.
> 
> I purchased a Winegard TV preamplifier, for which are only encased in plastic.  
> The other TV preamplifier I had at the time, a Channel Master encased in metal 
> housing.  Soon after purchasing/installing the plastic Winegard TV preamplifier 
> on one of my array of outdoor TV/FM antennas, I encountered significant noise 
> over the radio frequency spectrum heard on my ham/shortwave radios; a 
> buzzing/whining sound.  After a significant repetitive process of elimination 
> of switching-off house circuits/breakers, I finally found the plastic Winegard 
> TV preamplifier to be the culprit, and subsequently/immediately replaced with 
> the likely better metal encased Channel Master preamplifiers.
> 
> Rule of thumb, anything encased in plastic (or unsheilded) can become a noise 
> or interference problem.
> 
> Can also grab an assortment of ferrite cable clips, 5mm & 9mm, for prancing 
> around and clamping to any cable ends (whether RG6, power, audio cable, ...) 
> until you find the source. Sort of futile, but will minimize noise on some 
> cables due to poor SNR manufacturing standards of products, and with hopes the 
> ferrite clips may eventually find the source.
> 
> Try using a (sensitive) shortwave or AM radio for finding any RF noise sources, 
> but sounds more likely your noise source maybe more inline, or power 
> source/feed.  I've also seen high-end or pricey computer power supplies (PSU) 
> cause noise, of which Corsair readily replaced.
> 
> 
> 

-- 

Mike Isely
isely @ isely (dot) net
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8