|
As this forum is rarely used anymore, we've locked it. Feel free to browse and read. Questions? Please reach out to us directly. Cheers! |
|
 |
|
|
| Author |
Message |
paw
Joined: 08 Mar 2006
Posts: 1176
Location: Arvada, CO
|
| Posted: Tue Dec 16, 2008 4:36 pm Post subject: |
|
|
| kschmit2 wrote: | The Extron formula only applies to equipment in the middle of a chain (switching, amplifiers, splitters), not to the last device in the chain (display).
Unfortunately the calculator on csgnetwork is also based on the Extron formula. |
In addition, Extron's formula was developed for commerical/production situations. Where there maybe multiple devices in the middle of the chain. Each device will reduce the signal. These reductions are additive. So, a higher bandwidth is required in each device to not adversely affect the signal after it passes through all of them. In a HT, there should be fewer devices in the chain. Therefore, the bandwidth requirement for each device can be lower.
_________________
Aubrey
|
|
| Back to top |
|
 |
Gary M.
Guest
|
| Posted: Tue Dec 16, 2008 5:19 pm Post subject: |
|
|
so what is the estimated bandwidth needed for 1920x800p at 60hz ?
-Gary
|
|
| Back to top |
|
|
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Tue Dec 16, 2008 5:29 pm Post subject: |
|
|
| Nashou66 wrote: | Yes he mentioned that the higher flat gain bandwidth is more important. At least 3 times the required bandwith is all that is needed for proper head room, so a 800 MHz chip at a gain of 2 is enough i would think. I use the EL5166 op amp but I think the OPA695 will do better after looking at the flat gain bandwidth of 300, even the original CLC449 has the same spec at 200 which is not bad. I have seen a close up pic of his VNB's and notice he doesn not change the feedback or gain resistors to alter the gain of the chip, so i wonder if he is still using the CLC449? Hmmm  I have been searching for chips that wont use any new feedback resistor values and only the EL6166 is close I think the OPA695 might work also with out a resistor change. Its the decoupling caps that make the difference i would think. Once again I go Off topic...
Athanasios |
Yes, I've found that for 1080P it's best to go after flatness. And you would also need to look for the best 'rise' and 'fall' (timings/speed) when looking at things with a scope. The challenge with the scope is the final section (grid/cathode) of the neck boards. It's not easy measuring with a scope that section of the video chain, so you would have to come up with a good method or use FET probes.
The Intersil 5166 is a very good low noise chip, but may not be the best for every circuit, even if it shows an improvement visually, it may not do so well when measuring things with a scope. IC's are very finicky in a video chain. It all depends on the board design and what's before and what type of circuit follows the chip. Not necessarily true with cap coupling, but in the case of the marquee, which is almost completely direct coupled out to the tube. And that makes for the best video chain. The direct coupling creates a challenge getting the finicky devices to respect each other along the chain.
The 695 is a really good chip, but when using in Marquees, you'll need to make some radical changes in the associated circuit.
I don't use the 449. I replace every OP Amp in the chain, to include some tansistors.
Not sure what Greg does. And to be honest with you, If he's not presenting a valid means to support his testing, he can't really claim higher bandwidth. Almost any of the older video chains that's in all of our CRT projectors could benefit from the latest lower noise chips. And just because you're seeing a better image does not mean increased bandwidth.
It's a real task to get better bandwidth from the boards that are being used. And it takes a lot of work and some good equipment to sweep out and confirm the increase.
You would more likely have to go though about 10 chips for each stage when dealing with direct coupling stages and chains. And that requires months on end of work, and in my case--years. And the ones that actual work best, are not the higher bandwidth chips.
So when shooting for a substantial increase from present designs, you'll need the equipment to sweep test and check your changes, to include the parts and patients to bring it all together.
|
|
| Back to top |
|
|
Nashou66
Joined: 12 Jan 2007
Posts: 16171
Location: West Seneca NY
|
| Posted: Tue Dec 16, 2008 6:44 pm Post subject: |
|
|
| mp20748 wrote: | | Nashou66 wrote: | Yes he mentioned that the higher flat gain bandwidth is more important. At least 3 times the required bandwith is all that is needed for proper head room, so a 800 MHz chip at a gain of 2 is enough i would think. I use the EL5166 op amp but I think the OPA695 will do better after looking at the flat gain bandwidth of 300, even the original CLC449 has the same spec at 200 which is not bad. I have seen a close up pic of his VNB's and notice he doesn not change the feedback or gain resistors to alter the gain of the chip, so i wonder if he is still using the CLC449? Hmmm I have been searching for chips that wont use any new feedback resistor values and only the EL6166 is close I think the OPA695 might work also with out a resistor change. Its the decoupling caps that make the difference i would think. Once again I go Off topic...
Athanasios |
Yes, I've found that for 1080P it's best to go after flatness. And you would also need to look for the best 'rise' and 'fall' (timings/speed) when looking at things with a scope. The challenge with the scope is the final section (grid/cathode) of the neck boards. It's not easy measuring with a scope that section of the video chain, so you would have to come up with a good method or use FET probes.
I agree!! Still learning the Scope , can you give me some pointers? I have tried to solder wire to the tes points and do it that way but your not getting the best reading through long wire runs than you would right at the chip or resitor output.
The Intersil 5166 is a very good low noise chip, but may not be the best for every circuit, even if it shows an improvement visually, it may not do so well when measuring things with a scope. IC's are very finicky in a video chain. It all depends on the board design and what's before and what type of circuit follows the chip. Not necessarily true with cap coupling, but in the case of the marquee, which is almost completely direct coupled out to the tube. And that makes for the best video chain. The direct coupling creates a challenge getting the finicky devices to respect each other along the chain.
I have used it all my locations up to now but have been thinking of trying them only in the one part, the second RGB op amp on the Vim, I want to try the OPA695 Ahead of it or the unity gain OPA693(400 BGF @.2db) and remove the FB and G Resistors from the board for a gain or 2. I got the THS3201's to somewhat work but they seam to "Bright", those are hard to tame had to change the feedback gain and the dc restore resistors, but still have a hard time. Plus those work best with higher operating voltage than the Vim can give it. Gave up on those for now, Jarmo did massive testing with those but it looked like a rats nest on the Vim !! I also want to try a few more the LMH6703 somewhere maybe the second RGB Op amp on the Vim.
The 695 is a really good chip, but when using in Marquees, you'll need to make some radical changes in the associated circuit.
Hmmm 
I don't use the 449. I replace every OP Amp in the chain, to include some tansistors.
Not sure what Greg does. And to be honest with you, If he's not presenting a valid means to support his testing, he can't really claim higher bandwidth. Almost any of the older video chains that's in all of our CRT projectors could benefit from the latest lower noise chips. And just because you're seeing a better image does not mean increased bandwidth.
Yepp, I cant get the one on one off to resolve like yours do but The low end details have come out more from the
El5166 on the VNB's But I think that location is not the right spot for that chip, but still better than the CLC449. I think it benefit more from a faster rise and fall time chip with better heat characteristics and noise numbers.
It's a real task to get better bandwidth from the boards that are being used. And it takes a lot of work and some good equipment to sweep out and confirm the increase.
You would more likely have to go though about 10 chips for each stage when dealing with direct coupling stages and chains. And that requires months on end of work, and in my case--years. And the ones that actual work best, are not the higher bandwidth chips.
So when shooting for a substantial increase from present designs, you'll need the equipment to sweep test and check your changes, to include the parts and patients to bring it all together. |
Thanks Mike for the tips, I wish i did have a frequency generator to test, doing the visual testing gets very tiresome, and its not good for the boards or PJ to keep popping them in and out a million times.
Athanasios
_________________
Don't blame your underwear for your crooked ass~ unknown Greek philosopher
"Republicans believe every day is the Fourth of July, but the Democrats believe every day is April 15." --- President Reagan
One Smart Dog!!!
Marquee High Performance Bellows now shipping!!
Marquee Modifications and Performance Enhancement
Marquee C-element and Bellow removal
|
|
| Back to top |
|
|
VideoGrabber
Joined: 09 Apr 2006
Posts: 933
Location: Michigan
|
| Posted: Tue Dec 16, 2008 9:58 pm Post subject: |
|
|
AR commented:
> Driving trians backwards can be hazardous. <
Didn't you mean driving "sniart"s backwards? 
It's fascinating how many times this question resurfaces, how many ways it gets answered (and misanswered, with bogus misinformation), and how short peoples' memories are. There's some really good discussion of this over at the A forum, that has everything you'd ever need to know, that I'm too tired to dig up again and regurgitate.
Gary asked:
> so what is the estimated bandwidth needed for 1920x800p at 60hz ? <
What for? I'm wondering if you realize just how meaningless that question itself is? "What's the top speed my car needs to be able to reach?" Again, for what?
I'll shock the sh!t out of you with an answer to your question that you won't believe, but is 100% true. 60 MHz. I'll leave that answer as ambiguous as your question. Give you something to think about.
_________________
- Tim
|
|
| Back to top |
|
|
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Wed Dec 17, 2008 12:13 am Post subject: |
|
|
| Nashou66 wrote: |
I agree!! Still learning the Scope , can you give me some pointers? I have tried to solder wire to the tes points and do it that way but your not getting the best reading through long wire runs than you would right at the chip or resitor output. |
You'll more likely need a set of good probes for standard measuring everywhere except the finals of the neck boards. You'll also need to make sure the trim pot on the probe is calibrated using the square wave generator on the scope. You'll need to set that pot so that the square wave does not have slanted or distorted lines. All lines must be straight.
| Quote: | | I have used it all my locations up to now but have been thinking of trying them only in the one part, the second RGB op amp on the Vim, I want to try the OPA695 Ahead of it or the unity gain OPA693(400 BGF @.2db) and remove the FB and G Resistors from the board for a gain or 2. I got the THS3201's to somewhat work but they seam to "Bright", those are hard to tame had to change the feedback gain and the dc restore resistors, but still have a hard time. Plus those work best with higher operating voltage than the Vim can give it. Gave up on those for now, Jarmo did massive testing with those but it looked like a rats nest on the Vim !! I also want to try a few more the LMH6703 somewhere maybe the second RGB Op amp on the Vim. |
The 6703 is not a good chip on the Marquee. I'll not get into more on the chips, but a 693 will do well somewhere else in the chain, but whenever you use a 695, you'll have to play around with taming it. It's like keeping a tiger in a cardboard cage.
The 3201 is oscillating when that happens.
The right combination is key. But before that, you'll have to substantially lower the noise floor on the rails.
| Quote: | | Yepp, I cant get the one on one off to resolve like yours do but The low end details have come out more from the El5166 on the VNB's But I think that location is not the right spot for that chip, but still better than the CLC449. I think it benefit more from a faster rise and fall time chip with better heat characteristics and noise numbers. |
Yep!
You don't need a special generator for the scope sweeping. Almost any test pattern CD, DVD, PC program would have a pattern on it that would give you the square wave test that you'll need.
Last edited by mp20748 on Wed Dec 17, 2008 12:50 am; edited 2 times in total
|
|
| Back to top |
|
|
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Wed Dec 17, 2008 12:32 am Post subject: |
|
|
| VideoGrabber wrote: | | I'll shock the sh!t out of you with an answer to your question that you won't believe, but is 100% true. 60 MHz. I'll leave that answer as ambiguous as your question. Give you something to think about. |
You can always learn a lot from this guy here.
Now you see why I don't trow around bandwidth figures when I claim 1080P performance. The actual bandwidth needed is really not as high as some think. And most of the bandwidth calculators are for passive devices or deviced that'll be used in either connecting the different pieces of equipment together, or in-line switching.
The video chain is a bit different, and that's why the specs are all over the place. For instance, the Barco 909 claims 180mhz...
If that's true, how can they make that happen considering they are basically using the same boards that were being used in their Barco Graphics 808 that came out over decade ago. And with that same video chain that also has the Analog Device AD 835 (250mhz) in the chain. And not just that chip, it also has a newer version neck board that has the worst path for best bandwidth...
Ever hear of the OPA688?
Well.. they put one of those in the neck board chain to clamp both white and black levels. A good idea, but the wrong chip for best performance. The chip that should have been used was the Harris/Intersil HFA1130.
When I modified the neck boards on a 909, I had to make substantial changes to get them to where they needed to be. The HFA1130 was only one of the changes.
|
|
| Back to top |
|
|
AnalogRocks
Forum Moderator
Joined: 08 Mar 2006
Posts: 26706
Location: Toronto, Ontario, Canada
TV/Projector: Sony 1252Q, AMPRO 4000G
|
| Posted: Wed Dec 17, 2008 12:38 am Post subject: |
|
|
| VideoGrabber wrote: | AR commented:
> Driving trians backwards can be hazardous. <
Didn't you mean driving "sniart"s backwards?
It's fascinating how many times this question resurfaces, how many ways it gets answered (and misanswered, with bogus misinformation), and how short peoples' memories are. There's some really good discussion of this over at the A forum, that has everything you'd ever need to know, that I'm too tired to dig up again and regurgitate.
Gary asked:
> so what is the estimated bandwidth needed for 1920x800p at 60hz ? <
What for? I'm wondering if you realize just how meaningless that question itself is? "What's the top speed my car needs to be able to reach?" Again, for what?
I'll shock the sh!t out of you with an answer to your question that you won't believe, but is 100% true. 60 MHz. I'll leave that answer as ambiguous as your question. Give you something to think about. |
Ok getting back to my origional question for a second. I was asking for someone who asked me. They want to switch COmponent through a Denon recever that was rated 100Mhz. Over at the A site I saw 183MHz.
'Grabber you say 60MHz. Of course you don't specify switch or projector.....
So I'll ask this way. In your opinion (everybody ) will the component section of the Denon be able to switch 1080P/60 without degrading the signal so much as to make you not want to watch it.
Also his source was Dish at 1080P so already compressed.
Opinions?
_________________
Tech support for nothing
CRT.
HD done right!
|
|
| Back to top |
|
|
VideoGrabber
Joined: 09 Apr 2006
Posts: 933
Location: Michigan
|
| Posted: Wed Dec 17, 2008 1:38 am Post subject: |
|
|
Jeremy, you've been specific enough. So I be will as well.
> They want to switch Component through a Denon recever that was rated 100Mhz. <
and
> will the component section of the Denon be able to switch 1080P/60 without degrading the signal so much as to make you not want to watch it. His source was Dish. <
The answer is not only will HD sources at 1080p be "watchable", there will be minimal degradation, probably only detectable by measurement. This assumes that the Denon rating is at the -3 dB point, which is the standard reference point. AND that you're going from the source, through the Denon switching, and directly into the PJ or PJ driver/transcoder. I.e., not a chain of other BW-limiting components. And lastly, that you're not trying to display a 1080p desktop from your PC.
There's not a lot of extra headroom at 100 MHz (-3 dB) for 1080p, so I wouldn't want to stack many pieces of gear, but the Denon is adequate here, w/o problems.
_________________
- Tim
|
|
| Back to top |
|
|
AnalogRocks
Forum Moderator
Joined: 08 Mar 2006
Posts: 26706
Location: Toronto, Ontario, Canada
TV/Projector: Sony 1252Q, AMPRO 4000G
|
| Posted: Wed Dec 17, 2008 1:43 am Post subject: |
|
|
| VideoGrabber wrote: | Jeremy, you've been specific enough. So I be will as well.
> They want to switch Component through a Denon recever that was rated 100Mhz. <
and
> will the component section of the Denon be able to switch 1080P/60 without degrading the signal so much as to make you not want to watch it. His source was Dish. <
The answer is not only will HD sources at 1080p be "watchable", there will be minimal degradation, probably only detectable by measurement. This assumes that the Denon rating is at the -3 dB point, which is the standard reference point. AND that you're going from the source, through the Denon switching, and directly into the PJ or PJ driver/transcoder. I.e., not a chain of other BW-limiting components. And lastly, that you're not trying to display a 1080p desktop from your PC.
There's not a lot of extra headroom at 100 MHz (-3 dB) for 1080p, so I wouldn't want to stack many pieces of gear, but the Denon is adequate here, w/o problems. |
Oh yeah, he also wants to run this up component to Cat5 cables. But that's besides the point.
_________________
Tech support for nothing
CRT.
HD done right!
|
|
| Back to top |
|
|
VideoGrabber
Joined: 09 Apr 2006
Posts: 933
Location: Michigan
|
| Posted: Wed Dec 17, 2008 2:20 am Post subject: |
|
|
The problem with bandwidth specs for switchers is that the common (-3 dB) reference point isn't what you're really concerned with. A switch is a device that takes a number of input signals, and selects one of them. You want (and general expect) that Output = Input.
Which would you rather know?
a) what's the maximum frequency where there will be virtually no degradation? (loss <= -0.1 dB, or "maximally flat")
b) what's the frequency where noticable degradation is guaranteed to have already occurred? (loss @ -3 dB)
Most folks would pick (a). But (b) is higher. Guess which spec they publish?
Knowing (b), how can you figure out what (a) is? Answer... you can't. 
_________________
- Tim
|
|
| Back to top |
|
|
VideoGrabber
Joined: 09 Apr 2006
Posts: 933
Location: Michigan
|
| Posted: Wed Dec 17, 2008 2:24 am Post subject: |
|
|
> ...he also wants to run this up component to Cat5 cables. <
Didn't I catch you at the Comedy Club last Saturday?
_________________
- Tim
|
|
| Back to top |
|
|
AnalogRocks
Forum Moderator
Joined: 08 Mar 2006
Posts: 26706
Location: Toronto, Ontario, Canada
TV/Projector: Sony 1252Q, AMPRO 4000G
|
| Posted: Wed Dec 17, 2008 3:19 am Post subject: |
|
|
| VideoGrabber wrote: | > ...he also wants to run this up component to Cat5 cables. <
Didn't I catch you at the Comedy Club last Saturday? |
Please return the waitress to her original uprite position, don't forget to tip your server and please, try the veal.
_________________
Tech support for nothing
CRT.
HD done right!
|
|
| Back to top |
|
|
Gary M.
Guest
|
| Posted: Wed Dec 17, 2008 3:30 am Post subject: |
|
|
I personally wouldn't run analog video thru anything with less than 350mhz, I have used alot of analog devices and the higher the bandwidth the better, this is not scientific it is just observation and severe pickiness
-Gary
|
|
| Back to top |
|
|
Spanky Ham
Joined: 22 Mar 2006
Posts: 5643
Location: Comedy Central
|
| Posted: Mon Dec 22, 2008 12:30 am Post subject: |
|
|
As Tim said, this has been discussed over at AVS. Look for Chiem's thread about this.
I seem to remember Scott giving the formula for pixel clock in one of those threads.
|
|
| Back to top |
|
|
Stonefool
Joined: 24 Dec 2006
Posts: 253
Location: Sweden
|
| Posted: Sat Dec 27, 2008 1:18 am Post subject: |
|
|
Huh, I don't get it.
Even if you need to press your eyeball to the screen to se if the lines are seperatin' ... the PJ can handle it, it seem.
And if, even when your eyeball is pressed to the screen and it hurts like you wouldn't believe, the lines overlapping, but the image looks just way cool, stellar, and wow, from far far a screenwidth away ... then, what the hell does the friggin bandwidth mean?
Cheers.
_________________
Trying to get everything to work.
|
|
| Back to top |
|
|
beun
Joined: 28 Jun 2006
Posts: 676
|
| Posted: Sat Dec 27, 2008 2:29 am Post subject: |
|
|
Another much more pessimistic way to look at this:
The EBU defines the following standard: EBU System S4 1920x1080/p/50 (this is equal to SMPTE 274 System 3 for 1920x1080/p/60) with a sample rate of 148.5MHz.
Acoording to the Nyquist theory with an ideal (read unobtainable) low pass filter, this system will have information up to 148.5/2MHz=74.25Mhz.
So for components in the chain it makes sense to double this bandwidth to around 150MHz each. This makes it possible to put several component in the display chain without degrading the source bandwidth. Anything more is overkill.
_________________
ESL-0.5 Full Range Electrostatic Loudspeakers | RTC2200 Component to VGA Converter with Gamma Boost
|
|
| Back to top |
|
|
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Sat Dec 27, 2008 9:31 am Post subject: |
|
|
| beun wrote: | Another much more pessimistic way to look at this:
The EBU defines the following standard: EBU System S4 1920x1080/p/50 (this is equal to SMPTE 274 System 3 for 1920x1080/p/60) with a sample rate of 148.5MHz.
Acoording to the Nyquist theory with an ideal (read unobtainable) low pass filter, this system will have information up to 148.5/2MHz=74.25Mhz.
So for components in the chain it makes sense to double this bandwidth to around 150MHz each. This makes it possible to put several component in the display chain without degrading the source bandwidth. Anything more is overkill. |
Bingo!!!
|
|
| Back to top |
|
|
WTS
Joined: 08 Mar 2006
Posts: 1276
Location: Calgary
|
| Posted: Sat Dec 27, 2008 8:37 pm Post subject: |
|
|
Would that be -3db point or .1db flat?
_________________
Thanks
Walter
|
|
| Back to top |
|
|
beun
Joined: 28 Jun 2006
Posts: 676
|
| Posted: Sun Dec 28, 2008 4:53 am Post subject: |
|
|
The 150MHz could easily be -1dB using modern high speed opamps, the bandwidth of the 74.25MHz for the source bandwidth is a bit more difficult to quantify. The closer you want to go to the 74.25Mhz the steeper the filter has to be and the more ringing you will get. So for safety the 74.25MHz is most likely -3dB.
Kim
_________________
ESL-0.5 Full Range Electrostatic Loudspeakers | RTC2200 Component to VGA Converter with Gamma Boost
|
|
| Back to top |
|
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
|
Forum powered by phpBB © phpBB Group
|
|
