| Author |
Message |
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Thu Feb 25, 2016 9:29 am Post subject: |
|
|
| redfox001 wrote: | problem is using a 1080p@60 pattern will give you 75MHz or something like that at most. So if you give that to a spectrum analyser the limit will be the video resolution and not the equipment.
So how to set a vga card to get high frequency harmonics? Perhaps a pattern generator goes to 200MHz analogue bandwidth? That could work. So I do need a pattern generator for this. A PC dow not go so high in pixel resolution something like 2048x1536 that could give an analogue bandwidth of 100MHz or so but -3db etc. |
Look for a Chroma or other good pattern generator on Ebay. The last one I saw was going for under $100.00
Stay away from the Burst test pattern. Looking at a pixel on the scope yields far much more. You'll learn to also look at noise and distortions this way. very important for improving video quality
|
|
| Back to top |
|
 |
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Thu Feb 25, 2016 11:31 am Post subject: |
|
|
I use PC as test pattern source, on the bench I use either the VGA card, or an "ECL pulse generator" this later can output a 300MHz square wave at 2Vpp out (terminated to 50 Ohm), and has a rise time of 700ps, it was a bargain deal for 35USD, usualy they are in the 100ish USD range.
_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Thu Feb 25, 2016 2:32 pm Post subject: |
|
|
That makes sense nice pulse generator.
I done my best to do some more measurements. This time hf probe well compensated and short leads. This is the beginning of the driver. Seems my former measurements where totaly wrong because of the leads. Question is whether these are right? Perhaps I am facing the roll off from the probe capacitance here?
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
El Duderino
Joined: 23 Jan 2011
Posts: 4653
Location: Portland, OR
|
| Posted: Thu Feb 25, 2016 7:24 pm Post subject: |
|
|
|
I think your probing is much improved and approaching about the best you can reasonably do. As far as why the rolloff and if it's your probes, your scope, or the circuit under test, its likely a combination of all three IMO. To whittle it down, you may need other instrumentation and other known references to suss that out.
|
|
| Back to top |
|
|
the dude
Joined: 11 Sep 2013
Posts: 179
|
| Posted: Thu Feb 25, 2016 9:06 pm Post subject: |
|
|
I got much bettter results on the same circuit with the same scope and a simple short ground wire. So i think it is the probe.
Maybe its because the ground spring has only contact with the ground ring at the end, if there is any space between the turns and no ground ring under it a incuductor may be formed anyway.
|
|
| Back to top |
|
|
the dude
Joined: 11 Sep 2013
Posts: 179
|
| Posted: Thu Feb 25, 2016 9:13 pm Post subject: |
|
|
|
Connect the probe to the calibrator. You should not only check for a perfect square wave at 1ms/div, also verify the amplitude at different time bases. It should be 0.4V at should not change when switching to faster time bases, only the square wave may become a sine like waveform.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Thu Feb 25, 2016 10:37 pm Post subject: |
|
|
Ok is 0.4V all scales but strange is when I go to 20ns per div it takes two div to go from 0 to 0.4V. Since rise time should be 1.4ns I wonder why it takes so long?
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
El Duderino
Joined: 23 Jan 2011
Posts: 4653
Location: Portland, OR
|
| Posted: Thu Feb 25, 2016 11:23 pm Post subject: |
|
|
| redfox001 wrote: | | Ok is 0.4V all scales but strange is when I go to 20ns per div it takes two div to go from 0 to 0.4V. Since rise time should be 1.4ns I wonder why it takes so long? |
Your probe has a 1.4nS rise-time (10-90%) rating. The signal your scope calibrator is generating isn't near that fast.
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Fri Feb 26, 2016 6:09 am Post subject: |
|
|
High frequency passive probes usualy have more than 1 adjustments for compensation. I have a Tek P6106 250MHz that has four adjustments, and even my 250MHz testec has two. While my Tek P6105 (100MHz) has only one. Also interesting that the P6105/6 probes are modular, they have replaceable cables and the probes themselves are also identical, the only difference is the compensation box, one is rated to 100MHz the other is to 250MHz. But even the 250MHz one don't come close to my FET probe in performance.
So if your probe has only one adjustment it most likely won to to high freq...
_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Fri Feb 26, 2016 8:49 am Post subject: |
|
|
Ok but two other measurements I did. I directly connected the coax from the switcher that is 50 Ohm terminated to coax on the scope 50 Ohm terminated and I directly connected the 75 Ohm terminated coax from the moome to the 50 Ohm terminated scope. Like this.
50 Ohm switcher
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
75 Ohm moome
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Fri Feb 26, 2016 8:58 am Post subject: |
|
|
It is fun to see the ringing. The first pic is done on the moome with longer wires and the second pic is on the driver but with the direct connection. Clearly i got rid of ringing what is left must be the imperfect construction of the rectangular wave. At least it has become symmetric. 
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
mp20748
Joined: 12 Sep 2006
Posts: 5689
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Fri Feb 26, 2016 9:52 am Post subject: |
|
|
|
what is the source signal coming from on the last two?
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sat Feb 27, 2016 9:02 am Post subject: |
|
|
3 pixel test patter if I remember well.
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Sat Feb 27, 2016 10:57 am Post subject: |
|
|
Back to the protection diodes...
I just happend to look at the Ampro schematics again just for the fun and I spotted an interesting method they realized in their CRT amplifier circuit. Here is my theory behind it: in the CRT the spark is always coming from a positive potential. Either from G2 static build up (~1kV) or from anode charge build up (~30kV). So on Cathode output the ground diode is needless on G1 output the negative diode is needless. I don't know what was the failure rate of those CRT amplifiers, but they also used Sanyo video transistors, so Even according to Mike those are more sensitive to arcs.
| Description: |
|
| Filesize: |
163.54 KB |
| Viewed: |
9218 Time(s) |
|
_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sat Feb 27, 2016 1:56 pm Post subject: |
|
|
I see it is a 400MHz sanyo too. Where those Ampro's high bandwidth?
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sat Feb 27, 2016 2:06 pm Post subject: |
|
|
I did some simple capacity measurements on lug tubes between cathode and g1 and some are 2.8 pf others are 3.2 pf so luck with the tubes helps a little too?
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Feb 28, 2016 8:45 am Post subject: |
|
|
I found a seos training manual and it says video bandwidth 75MH that is 150MPixel -3db finally a honest number. It is the -3db at 1080p that we are observing. It is a808s manual. So they say the 808s has 150 MPix and the 909 180 MPix -3db.
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Feb 28, 2016 10:53 am Post subject: |
|
|
I wonder if we could make this modification on the 909 neckboard.
Both on ebay and how difficult can it be to make a small print that fits the heatsinc with the right parts?
In the topic is a schematic of the internals of the vph with most vallues. Can we make a small print with these parts and the sanyo's?
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Feb 28, 2016 1:53 pm Post subject: |
|
|
Ok on with the peaking on the lug tube. Original there is 180nH. I tried 470nH but to much interference and streaking or something. I tried lower resistance that makes it a little better. I tried 330nH but still streaking and interference. So I tried 120nH and that seems to work ok. Not sure if it is better. I have to wait for my fet probe and p6136 to know more. I will try measuring with the fet on the output.
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
Difficult to make a picture lot of black bars over the screen.
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
Last edited by redfox001 on Sun Feb 28, 2016 6:12 pm; edited 1 time in total
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Feb 28, 2016 2:43 pm Post subject: |
|
|
Original 180nH
 Untitled by Radio Head, on Flickr
120nH looks a little better here
 Untitled by Radio Head, on Flickr
 Untitled by Radio Head, on Flickr
_________________
701s->runco933->8500ultra->cinemax->9500mp->919 splitpack + cinemax
Last edited by redfox001 on Sun Feb 28, 2016 6:11 pm; edited 1 time in total
|
|
| Back to top |
|
|
|
|
