That is an indication of a weak signal - I had all the symptoms, and the CC fixed it using a pre-amp.
(The digital device can normally fill in any small holes, if it has enough to work with orginally. If it gets more spurious signal with the weak desired signal -fan hash, e.g., , it will manifest itself more strongly)
Check the "db" of the incoming signal to make sure you have the required signal strength.
Sometimes making a programme unwatchable. We have managed to
One other item - each splitter degrades the signal by 3 db minimum (cuts it in half) just to split - so you get one-fourth of the input out of each leg of a two-way splitter - so minimize the splitters
(and to minimize noise - if you connect to two input devices with input coax -dvd, vcr, cable box, etc, - you need to break the metal shield on the coax at the input to one of the devices or the resultant loop of coax shield-chassis-coax shield/ chassis-coax shield will act like a loop antenna for nearby interference)
That is correct. You should have stopped there!
If it were 1/4 to each leg, that would be a 6 dB drop, but it is
3 dB as you stated originally. That is half the power to each leg. (In fact it is probably more like 3.5 dB, as there is approximately 0.5 dB loss in the hybrid itself.)That is not true. Do *not* break the outer shield on a coax cable. If you do, then you *will* get an antenna instead of a transmission line.
On Sun, 11 Jun 2006 13:22:48 -0800, snipped-for-privacy@apaflo.com (Floyd L. Davidson) Gave us:
I think that he was thinking of an audio patch cord there.
No, I did mean 3 db for the fact that the device is in the line trying to separate a signal, and
the original signal is split in half because it is a two-way splitter (or in thirds because it is a three-way splitter)
thus giving 6 db between input and either output leg on a two way splitter.
Why?
First, if the splitter was only 3 db, as you say, it would be a perfect device - each leg output being half the input. Sorry, 100% efficient devices are not possible, especially in RF. And there would be no isolation and no impedance matching.
Second, the equations predict a "cost" of 3 db to split a signal in a perfect splitter - either from impedance mismatch externally to the fed devices, or due to the parts needed to match impedance internally so as to have matched external impedance. Passive splitters are theoretically at best 50% efficient (check the equations - we had to do them in lab) - that means half the power is lost in the split, not half the power goes out each leg.
A detailed lay explanation can be found at
Having worked as the cable grounding specialist for the largest ballistic missle warning system, with its multimegawatt transmitters and "many db" amplifiers all around, I can assure you that ALL of our coax that could create a loop due to shield path was made without the shield in place on ONE end only, and that I have PERSONALLY removed coax with shields connected at both ends that were put in by otherwise competent techs, coax that was seriously degrading portions of that system, and replaced those leads with coax with the shield grounded at ONE end in order to restore the system.
I DID mean to make the poster's shields end up in a "spider" grounded system, using the system input coax as the mother lead, and breaking the end loop made by coax shields and chassis-power grounds at one point in the coax shield.
OK, perhaps he has two-prong rather than three-prong power plugs, and he has floating chassis grounds rather than fixed, and he could get by with having a weak loop if his digital signal was strong enough for the discriminator to weed out too-sharp rise-time noise and only pass pro forma pulses - but it sure sounds like he does not have a strong enough signal-noise ratio for his (crap?) discriminator, and he needs whatever he can get.
fwiw...
On Sun, 11 Jun 2006 19:18:46 -0500, "hob" Gave us:
snippitude
So now you are saying that he needs a new/better tuner?
Bit-error-rate... that is the term you are looking for.
Not too bad, but you failed to actually *read* it. It does
*not* say the trans-hybrid loss will be 3 dB. What it actually says is that it will be "less than a 3dB loss". As I noted, it is usually about 0.5 dB. That is a *long* ways from 3 dB.Typical insertion loss for a hybrid is between 3.5 and 4.5 dB.
Me thinks you are blowing blue smoke. What you describe might work for certain applications, but it *won't* work in the one under discussion.
And I would be *very* surprised if it can be made to work with "multimegawatt transmitters".
It appears that you don't understand what ground loops are or how they work, not to mention having not a clue what a transmission line does or does not do.
That is a really dumb idea. I realize that a lot of people use that for audio, where the "transmission line" is very very very short compared to the wavelength of the signal. It isn't the best way to do it there either, but it actually can work. But doing that at RF is absurd.
Now I'm *positive* that you are blowing blue smoke. That's meaningless verbiage.
Nothing.
"TimPerry" wrote in news:j8idnSZc88CjohHZnZ2dnUVZ snipped-for-privacy@adelphia.com:
Got an in-line amplifer already, thanks.
Need an inline spliiter, I'm afraid - these digital boxes only give one output, so you need more than one digibox if you want to record a program whilst watching another.
Yep, any TV that's capable of receiving and displaying a TV picture.
"hob" wrote in news: snipped-for-privacy@comcast.com:
Signal strength is usually ok.
(snip)
speaking of not reading it -
see page 1 - insertion loss table
see page 4, last paragraph - "A signal applied to port A will be routed to port S, minus a 3dB loss in the internal resistor;" (you will note why that resistor has to be in there on page 2, and the effect of mismatch on page 5 - note the 6db loss INSIDE the splitter for mismatch)
I knew the loss was there before I read it - and it is the same after I read it.
It seems you failed to understand what you read: You lose the 3 db internally to match impedance, or you lose 3 db or more in the impedance mismatch. Either way, it costs you 3db loss to insert it into the line. (Having two outputs side by side is called a parallel ciruit, and their impedance is half of what was there before the splitter was inserted. And maximum power transfer ONLY occurs at matched impedance.)
That out-of-context quote of yours refers to the mismatch effect on insertion loss - read what it said just before that sentence. It is talking about the effect of shorting out one of the ports - "THE POWER LOSS WITHIN THE POWER SPLITTER WOULD BE... -6DB of the orginal signal power [power in- at the input port]" And IF you short out port B output, in a real splitter, the LOSS at port A output (not the signal) is less than 3 db because some of the signal power that was to go to port B (that cannot leave because of the short) is reflected into port A.
As I noted, it
No, in a real passive filter, it is closer to 3.5 db. The internal loss is well known and accepted.
that is along way from .5 db you had just claimed. Apparently you are unfamiliar with the term "insertion loss"
Special RF in those RG-59/58/6 coax lines of the poster? His Rf doesn't follow the laws of physics?
So why don't you enlighten us and tell us WHY shield breaks and "spiders" won't isolate the signal from RFI/EMI? There are a couple of authors on rf grounding systems I will pass it on to.
"multimegawatt transmitters".
perhaps that is why they paid me to do it as a civilian specialist, and not you
( the actual size and db of the equipment was classified, fwiw )
Well, I was the BMEWs site grounding engineering tech for two years and have no need to blow smoke - and I have a good deal of experience working with L band and X-band grounds.
- but since you apparently have no real world experience, try a book on grounding and look up "spider" systems.
Well, apparently you have zip experience working with RFI/EMI grounding systems or antennas, or you wouldn't say such stupid things. To make it very simple for you hobbyists - Ever see a TV antenna collector/reflector rod? How long is it and what is that signal's wavelength? And the length of loop made by equipment and connectors? And what is that signal's wavelength?
And who uses coax for audio or worries about transmission line length vs wavelength in audio cables? Monster cable buyers? That statement is so far out of reality it says you are clueless in this field
your 60 cycle hum is not in the same field as RFI
That statement clearly and unequivocably says that you have absolutely no clue how grounding is done inside equipment and that you have no real professional experience whatsoever.
Not in the computer sense per se, but perhaps in the digital signal to ratio sense -
What I am saying is that since digital "amplifiers" do not amplify a signal as in an analog fashion, passing all the signal, rather they recreate the signal at a higher level, isolating at each stage, they are susceptible to creating extra bad bits in noisy signals. That is, while an analog amp will give you most of what comes in, a digital output is only reproduced timewise from what comes in. ( Removes the need for good filters and such.)
A digital amplifier's weakness, however, is that it is looking for a bit to reproduce - and it only "recognizes" signals having a certain risetime -it discriminates against signals rising too fast for the amp ( it can't "trigger") and it discriminates against signals that rise too slowly (the "gate" set by the clock closes before the voltage level rises enough to trigger an output).
If you look at a real digital signal, you will see a series of pulses and noise (non-signal) - if the noise is well below the level of the pulses, the pulses drive the amplifier and low noise has little effect - the bits are pretty much repeated and thus cleanly reproduced at a higher level. If the level of the pulses are weak and thus are in the noise, then there are too many false output bits created for a coherent signal. It does not discriminate properly between noise and signal.
( A "better" tuner would fix the problem only because the word better implies one that does not have the problem -) a "better at handling low signal to noise ratio" tuner might fix the problem, but normally you get what they give you and expect you to get the signal up to a certain strength and above a certain signal to noise ratio
(if that ability to discriminate in high noise is what is called bit-error-rate in digital signals, then that is the term.)
fwiw
On Mon, 12 Jun 2006 02:57:45 -0500, "hob" Gave us:
Nearly ALL audio RCA type patch cables are a pair of coaxial runs.
I cannot think off any that are not. Can you? Give the model and brand name here:
On Mon, 12 Jun 2006 02:57:45 -0500, "hob" Gave us:
No shit, Dip Tracy.
On Mon, 12 Jun 2006 03:30:34 -0500, "hob" Gave us:
Wrong. It is not a term that has ANYTHING to do with computers.
It is a COMMON term used in MPEG2/digital audio video transmission jargon.
ANY compressed, digital, A/V signal that uses FEC for correction of bad data uses THIS TERM to refer to signal quality.
You keep forgetting about the FEC. There is enough of it such that even missing segments of the signal can be replaced. Up to generally about ten percent.
Now that you've read it, I have to explain it to you???
Okay, you didn't read that one apparently:
insertion loss as follows: Number of Theoretical Output Ports Insertion Loss (dB) 2 3.0
....
Not 6 dB, just 3 dB.
But, lets stop the foolishness right here, and do a definitive reality check. Here are a few URLs all of which lead to the technical specifications for a 2-way splitter. Every single one of them says the insertion loss is ~3.5 dB, not 6 dB.
Monster Cable TGHZ2RF GHertz 2-Way Splitter dB Loss peer output is: 5MHz-1GHz: -3.4 dB
Allen Tel Coaxial Splitter - 2 Way 3.5dB insertion loss
Here is a short and simple explanation:
INSERTION/THRU-LOSS
When a 2-way splitter is inserted in the line, the signal level at the output ports will be reduced by approximately 3.5 dB (4.0 dB). This loss is referred to as thru-loss or insertion loss. For calculation purposes it is best to utilize 4.0 dB as the actual loss. If a 4-way splitter had been used than resulting loss would be 6.5 dB (7.0 dB).
Hence, we might disagree on exactly what mistake you made in reading the material in the reference you presented (or I suppose you could claim it is simply wrong), but one thing we don't need to argue further about is just what the insertion loss of a 2-way splitter is: it *is* 3.5 dB.
So lets see what you had to say, and figure out where the error is.
That is how it works as a *combiner*.
However, a combiner is nothing other than a splitter where each of two ports (that are isolated from each other, hence no signal goes from Port A above to Port B) split each signal between two other ports (Port S and the resistor in this instance).
A signal applied to either Port A or Port B does exactly the same thing, which is output at Port S with a 3 dB loss. The other half of the power goes to the resistor (which is actually nothing other than a dummy load placed on one of a splitter's two output ports). Note that a signal input to Port A does appear at Port B (or at least not to the degree that it actually is isolated).
And of course that is identical to inputting a signal into Port S, which is then split between Port A and Port B equally, with
*nothing* going to the resistor because that port is isolated by more than 20 dB (assuming all ports are correctly terminated).That of course is not correct. As that document clearly states, there is a 3 dB loss, not 6. When a signal is input to Port S, it is split half and half between Port A and Port B, and *no* signal appears across the resistor, which is isolated from Port S.
If you would *read* what it says instead of looking for snippets of text that you believe support what you think it should say, you might do better an understanding how a hybrid works.
Nothing in that URL says that. And there are no 2-way splitters being marketed that list in their technical specifications an insertion loss of greater than 4 dB.
What do you think happens if the load impedance on each of these ports is doubled???? (Hint: NOTHING!)
Go back an look at what it is saying.
Given that we *know* the insertion loss is actually 3.5 dB, not
6 dB, I'll leave it to you to figure out the mistake you made in analyzing that paragraph.If you do accept that your analysis *can't* be correct, but cannot figure out what it should be, I'll be happy to explain it in different terms that the reference we are using. You cited it, but I'd say that it is terribly confusing, and is not helping you to understand this at all.
It is 3.5 dB in virtually *every* commercially available splitter you are going to find on the market!
I have no idea what you mean by "a real passive filter". Real passive filters can have almost any insertion loss, depending on how they are designed.
I said the 0.5 dB was hybrid loss. Along with the 3 dB loss from the two way split, a typical 2-way splitter has an insertion loss of 3.5 dB. It happens that some (not the kind used for RF at TV frequencies) may have as much as 1 dB hybrid loss, which would then make for an insertion loss of 4.5 dB.
Because it is RF, what you describe *WILL* *NOT* *WORK*. Go blow that smoke up someone else's rear. You don't know what you are talking about with grounding and transmission lines any more than you understand hybrid splitters.
If you have a broken shield on a coaxial transmission line... you don't have a transmission line! It's called an antenna.
If your purpose is *not* a transmission line, but instead is a shielded cable (for example with DC leads inside the shield), there is a huge difference. That difference might be expressed as the simple fact that a transmission line keeps the RF inside, while a shielded wire tries to keep it outside. That is way over simplified, but...
Giggle snort!
Choke laugh, cough!
In two years I'm sure you learned all about how to brew coffee. I've seen people who had 10 times your experience who *still* didn't understand it!
[snip]Please go to a nice library and find almost any decent book on transmission line theory. Try learning how it works first, then mouth off.
[snip]Think of the telecommunications industry...
Or perhaps that I actually do know about it, and can uncategorically state that you are babbling nonsense.
Here's the best phrase of all:
"to weed out too-sharp rise-time noise and only pass pro forma pulses"
Reminds me of a guy who picked up a box of round headed rivets used to put together metal shelving to store electronics parts, who assumed it was some sort of electronic gizmo, and said "What's this?". To the absolute delight of all the techs watching, it was a facilities mechanic who instantly piped up with "Oh, those are bi-directional diodes!" The new supervisor sat the box down, said "Oh...", and continued on his way.
Didn't they kick you upstairs into management too? To keep you from touching anything critical?
Eh? Are you going to run off at the mouth with *another* one????
Bit-error-rate is a term used for digital communications systems. It actually has nothing to do with computers. I have no idea what you mean by "digital signal to ratio sense" either.
"Bit error rate" is the standard measure of quality of service for a digital data circuit.
Wow, you are not particularly articulate, but that is almost correct. It won't add "extra bad bits", but it will generate errored bits.
Oh, if only that were true... An analog amplifier can and
*will* add signals to the output which are not present at the input.The difference between analog and digital is that analog _noise_ is cumulative, while digital _errors_ are cumulative. Any noise, no matter what the Signal-to-Noise Ratio, at the input is guaranteed going to be included in the output of an analog amplifier. A digital amplifier will not have any noise in the output if the SNR is greater than a critical value. It happens that the critical value for a digital amplifier is one at which an analog amplifier is almost useless. Of course if the SNR is low enough that errors do occur, they are cumulative (which is where error correction comes into use).
Here is a chart, showing BER (Bit Error Rate) and thresholds for SNR required to achieve them, for a binary transmission system:
Error Rate S/N dB Error Rate S/N dB
10e-2 13.5 10e-7 20.3 10e-3 16.0 10e-8 21.0 10e-4 17.5 10e-9 21.6 10e-5 18.7 10e-10 22.0 10e-6 19.6 1o3-11 22.2I'm not sure what you mean by "reproduced timewise", but trust that "good filters and such" *are* essential (to keeping the SNR below the threshold referenced above).
Eh? Risetime may or may not have anything to do with how a digital signal is decoded. It depends on just what parameter is modulated. For example, typical v.90 modems are in fact sensitive to phase hits, on the other hand a typically ISDN or T1 is not.
That does not describe "a real digital signal". You are describing a voltage level encoded digital signal. T1 and ISDN are two examples of types of digital tranmission systems than have pulses as you describe (while a v.90 modem does not).
But lets do note that at the SNR where that signal will become
100% error free, an analog system would be absolutely useless due to excessive noise! Typically, for example, a digital transmission system can function with an SNR of 12 dB and provide a channel equal to an analog system requiring a 60 dB SNR.The tuner of course is *not* looking at "the level of pulses" in any way shape or form. It is receiveing a modulated RF signal and *generating* pulses!
See above, relative to a 12 dB SNR being sufficient for a digital system to provide the same quality as an analog transmission system that has a 60 dB SNR.
Worthless. You should *ask* what this is about, not try to fabricate what you think it could be.
-- Floyd L. Davidson Ukpeagvik (Barrow, Alaska) snipped-for-privacy@apaflo.com
you really don't know what an insertion loss is, do you? You apparently think its a measure of what you get out of either port relative to the input port - (try the price in signal for inserting that splitter in the line.)
or anything about power transfer and impedance matching?
you think you can insert a parallel line in series into a line, create an impedance mismatch, and still get 100% efficient transfer. You really need to take a basic course in electronics.
they don't have 6db insertion loss - they have 3 db insertion loss - and the halving of the signal in a two-way splitter makes for the 6 db drop in strength at either output port from the input strength
nuff said - if you can't get the basics of electronics and don't even know what the terms mean, you sure won't get anything technical - especially RF grounding
( reread the article giv> >> "hob" wrote:
they don't have 6db insertion loss - they have 3 db insertion loss - and the halving of the signal in a two-way splitter makes for the 6 db drop in strength at either output port from the input strength
Utter parsing bullshit - "it won't add ...bits, it will generate...bits"
and the difference between additional unwanted bits and "errored bits" out of a functioning digital ampilifier is what, pray tell?
Apparently you are saying that given sufficient noise, the sensing-repeating part of the amplifier won't add "extra bad bits", that is, it won't add pulses to the output stream that are not in the input?
Talk about grasping.
Read - oh, I forgot, you havbe a problem reading... I said it gives you back most og what comes in, NOT that is did not add signals (just as digital does).
I have no probelm with that because that is what I said (re:timewise only)
Amplifiers discriminate based on the risetime of the digital pulse.
It depends on just what parameter is
Was not talking about phase here,. just rise time discrimination
"noise" can be voltage or phase or whatever masks the intended signal
Apparently you have never designed or worked on digital amplifiers -all amplifers see the input, and one common way to get a clean digital signal is to "clip off the tops" of the pulses making up the bits above the noise and passing them on (think a diode biased to stop everything near the noise and pass everything above the noise - designers have been using that method for nearly 40 years)
don't care about an analog system here - it's digital.
What you hobbyists call things is not necessarily what engineers call things or what they really are -
(And you who think I should ask about your narrow parsed definitions for your convenience apparently think RCA audio cables are coaxial cables, and you think insertion loss is total loss.)
I can think of no RCA patch cables that are coaxial conductors.
coaxial cables are concentric conductors (co-axial conductors , hence the name) that use fields to transfer energy near the conductors. Both the outer and inner axial metal parts of a coaxial cable are ("same-direction") conductors.
RCA audio cables are shielded conductors that use the center conductor to carry energy. The outer metal part is not intended to carry the signal, just to shield the conductor and sometimes provide a ground path.
( Some people undoubtedly sell coax for audio hookup and some fools buy it, but some people also buy "super-pure" copper to carry a signal to a speaker that has more inherent distortion than a thousand miles of romex.)
Yes, coax will conduct audio signal in its center conductor if you want to waste your money, but an RCA audio cable is not coax and doesn't work like it.
On Mon, 12 Jun 2006 17:55:09 -0500, "hob" Gave us:
Tell us... oh DIPSHITTO...
What is QAM 256?
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.