|
|
|
View previous topic :: View next topic |
Author |
Message |
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Mar 02, 2014 11:25 am Post subject: |
|
|
CasetheCorvetteman wrote: | Its worth a try mate, its no long term solution but it may reveal some things.
I have done this in the past with some domestic grade TVs that only had composite inputs, by connecting composite sync to the composite video input, then wiring the RGB lines to the OSD section of the TV and blanking the composite signal, you get an RGB grade image on a regular TV. Thats not what your purpose here will be, but it may just show a difference in there somewhere that could help figure out the real issue. |
That's one wicked mod
Actually thinking one step further, the neckboards expecting 0,7V signal as input, so I should connect the signal right to it, that would pretty much split the things apart...
_________________ 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 |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Mar 02, 2014 11:42 am Post subject: |
|
|
gjaky wrote: | CasetheCorvetteman wrote: | Its worth a try mate, its no long term solution but it may reveal some things.
I have done this in the past with some domestic grade TVs that only had composite inputs, by connecting composite sync to the composite video input, then wiring the RGB lines to the OSD section of the TV and blanking the composite signal, you get an RGB grade image on a regular TV. Thats not what your purpose here will be, but it may just show a difference in there somewhere that could help figure out the real issue. |
That's one wicked mod
Actually thinking one step further, the neckboards expecting 0,7V signal as input, so I should connect the signal right to it, that would pretty much split the things apart... |
This was quite easy to do, the sad thing is the neckboards are the weakest link. However while visual inspecting these boards looked clean, but now I am wondering if there is some thin grime layer on it that mess up the high freqency response, for example the CRT socket itself looks rather yellow than anything else...
_________________ 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 |
|
|
CasetheCorvetteman
Joined: 09 Nov 2008 Posts: 6319 Location: Australia
|
Link Posted: Sun Mar 02, 2014 12:44 pm Post subject: |
|
|
Was it from a smoking household? Wash those boards in the dish washer or with some sort of soap to get rid of that sh*t. I dont expect it to help much, but they will look nice
_________________ Barco CineMAX, 120" OZ Theatre Majestic 16:9 screen, H/K AVR 7.1...
RUNCO DTV991 LC ( NEC XG 852 LC ) 100" 4:3 screen, H/K AVR 5.1...
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Mar 02, 2014 12:56 pm Post subject: |
|
|
CasetheCorvetteman wrote: | Was it from a smoking household? Wash those boards in the dish washer or with some sort of soap to get rid of that sh*t. I dont expect it to help much, but they will look nice |
From the smell of it, yes. Now I reassembled the machine... I put 20+ hours in it already, yet only saw a single movie, so it's time for some enjoyment now, then will get back to this later.
_________________ 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 |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Wed May 14, 2014 8:45 pm Post subject: |
|
|
So got not one but two neckboards from Blockbullie, tried one, and guess what? Nothing's changed! It's the same crap! This is getting more mysterious!
_________________ 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 |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Mon Jun 09, 2014 7:53 pm Post subject: |
|
|
Ok, figure this out! Built a test power supply to test neckboards on the bench. I've tested two neckboards (one which I bought as spare, the other was originally in my machine) and it turns out both of them had the same problem.
The Cathode amplifier ICs are weakened: they are working but distorting the signal quite badly, and have very slow fall time, rise time is OK -> poor bandwidth.
The discoloration/overpeaking issue was caused in the Cathode side switched peaking network, it seems the switching transistors were leaking somehow, after removing them the neckboard measures good.
As for my luck: I have 5 pieces of poorly working neckboards with the same problem, this couldn't be a coincidence, now I am quite surprized no one else have experienced a similar issue. Mind you all video ICs have the "7G3" batch code.
_________________ 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
Last edited by gjaky on Fri Apr 22, 2016 10:37 am; edited 1 time in total
|
|
Back to top |
|
|
CasetheCorvetteman
Joined: 09 Nov 2008 Posts: 6319 Location: Australia
|
Link Posted: Tue Jun 10, 2014 7:38 am Post subject: |
|
|
Funnily enough, my XG 852 has over 8,000 hours on it and ive never ever had a single issue with it at all, its such an incredible machine!!
_________________ Barco CineMAX, 120" OZ Theatre Majestic 16:9 screen, H/K AVR 7.1...
RUNCO DTV991 LC ( NEC XG 852 LC ) 100" 4:3 screen, H/K AVR 5.1...
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Tue Jun 10, 2014 7:12 pm Post subject: |
|
|
CasetheCorvetteman wrote: | Funnily enough, my XG 852 has over 8,000 hours on it and ive never ever had a single issue with it at all, its such an incredible machine!! |
It's more like matter of how it was used rather than how much. Mine has some 2700 hours, but even the tube toe-in fixing crews are busted, the tubes on the other hand are kept well at least...
_________________ 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 |
|
|
CasetheCorvetteman
Joined: 09 Nov 2008 Posts: 6319 Location: Australia
|
Link Posted: Tue Jun 10, 2014 9:31 pm Post subject: |
|
|
Yeah so are all mine, i replaced a hell of alot of screws due to that. Original red and green were fried black, yet blue had almost no wear. The KELVIN settings were way down on the blue and way up on the others.
I doubt mine was treated any better than yours!!
_________________ Barco CineMAX, 120" OZ Theatre Majestic 16:9 screen, H/K AVR 7.1...
RUNCO DTV991 LC ( NEC XG 852 LC ) 100" 4:3 screen, H/K AVR 5.1...
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Sep 14, 2014 7:47 am Post subject: |
|
|
Some progress in this problem. I hooked my new HP 1120A 500MHz FET probe to a neckboard and measured all gain stages for speed.
The board under test is a board where both videopacks were changed already, one is a VPA15H, the other is the originally used VPJ13S.
-The first stage is built around the Sony CXA1709P preamp IC, its datasheet says it has 2ns rise time when the output is 4Vpp, in this circuit I measured 3,5ns rise time at 2Vpp output (the input signal had 1,3ns rise time), this is notably worse than the advertised value. There is no explanation for that, the signal goes in then comes out, very few external components are used.
-The second stage is on the daughter boards, the discretely built preamplifier, it seems to be good, as it almost didn't affect the signal speed.
-The last stage is the videopack, I measured 12ns rise time at 55Vpp on their output, the signal was nice, without odd distortions, both the VPJ13S and the VPA15H looked quite similarly, but the VPJ13S looked slightly better. The 12ns rise time is roughly equal to 30MHz bandwidth and this is in conjunction what I see on the projected picture.
In the next step I separated the videopack and the output of the preamplifier on the XG neckboard and connected the preamp's output to a PG xtra's neckboard (with VPA13) and this way I could measure 4ns rise time at 55Vpp on its output, that value is quite good, since the feeding signal wasn't much faster than that.
What is the conclusion? I don't know...
What is noteworthy, I swapped the videopacks many times and tried all ICs in all combinations I had here (VPA15H, VPA13, VPJ13S) without the tiniest improvement on the XG boards, also cut the trace just after the output of the videopacks, to ensure nothing is loading the output (no success), still the VPA13 on the PG xtra board works much better.
The sony preamp ic performs poorer than its specs. What is interesting here that the videopacks and the sony ic are on the same PCB and they perform poorer than expected, not to mention the odd problem with the switched peaking network, while the discrete preamplifiers are on a separate PCB and they are working fine. So could be something in the PCB material, caused by heat stress? but this is still a quite steep assumption, since I have 5 pieces of poor performing boards but they came from two different part of the world.
_________________ 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: 2251 Location: The Netherlands
|
Link Posted: Sun Sep 14, 2014 10:45 am Post subject: |
|
|
So this is very interesting. So you are saying that the same VPA13 has a fast performance on the pg neckboard and a bad performance on the xg neckboard. I get the feeling you are close to solving a deep issue here. You checked the electronics and there is no difference there? Are you absolutely certain? If not something is causing a stray capacitance degrading both the sony and the video packs. You suspect the PCB. I would think it must be the PCB layout than? Perhaps something with the groundplane? I check if I can find anything on this issue.
Could you explain a little better how you measure the risetimes? What kind of signal do you use? Testpictures?
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
Back to top |
|
|
redfox001
Joined: 16 Mar 2009 Posts: 2251 Location: The Netherlands
|
Link Posted: Sun Sep 14, 2014 11:10 am Post subject: |
|
|
I am reading something that pcb traces should be close to each other to decrease inductance.
Are there bad low frequency resistors in the board?
Quote: |
current feedback amplifiers do not like to have
capacitances on their inputs. To that end, ground planes should be cut back from the input
pins
|
Transmission line problems?
http://www.analog.com/library/analogdialogue/archives/43-09/edch%2012%20pc%20issues.pdf
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Sep 14, 2014 11:50 am Post subject: |
|
|
It seems I am the only one who experienced this problem with an XG yet, Casethecorvetteman assured that his XG852 isn't doing anything like mine, and his machine also has the same neckboards, MikeEby posted screenshots a while back doing 1080P72Hz on his XG1351 (same model as mine), more over Curt reported that he also never experienced problems like these. So I'd rule out the possibility of a design flaw, more like something have happened to these boards.
I have bought two spare neckboards in this forum and turned out they are exactly suffering from the same problems as my original boards.
These boards are expecting 0,7V input signal terminated at 75 Ohm, so I can directly hook them on the RGBHV cable. they need a clamp pulse that can be the H sync signal if it has (+) polarity and needs a contrast voltge which is set as per the SM (~2,7VDC). I send a (luma) multiburst signal to the board from the computer (the VGA card has a rise time about 1,3ns) and simply measure the transition time between 10% to 90% value of the rising edge at the VPA's output pin, the output is only loaded by the FET probe, which has a tip capacitance less than 1pF when the dividing factor is 100:1, while the CRT's gun acts as a capacitance of about 10-12pF, I've loaded the output slightly with additional capacitance on the XG neckboard but it didn't change anything.
I can't be sure about the VPJ13S performance, since all of them were in those boards that performing poorly, so theoretically all of them can be bad, but I have here VPA15H and VPA13 ICs, which were tested in other board before, and they are also doing the low bandwidth when I put them in the XG boards.
In theory the tubes could have some odd parasitic capacitance meaning the guns acting up much bigger capacitance, but without the tubes the boards still performing poorly... so not the tubes are the problem.
Everything is working fine in the projector. Video gain is fine, brightness, contrast and G2s are at normal levels, neckboards are set to proper black levels, only the bandwidth is missing.
_________________ 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: 2251 Location: The Netherlands
|
Link Posted: Sun Sep 14, 2014 12:52 pm Post subject: |
|
|
Ok than time to get another set of neckboards You said you tested 5 but 3 will be from the same machine so the statistics might be a bit low.
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sun Sep 14, 2014 3:45 pm Post subject: |
|
|
redfox001 wrote: | Ok than time to get another set of neckboards You said you tested 5 but 3 will be from the same machine so the statistics might be a bit low. |
Hm.. let's see if I get any luckier..
_________________ 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: 2251 Location: The Netherlands
|
Link Posted: Sat Oct 11, 2014 11:45 am Post subject: |
|
|
gjaky wrote: | redfox001 wrote: | Ok than time to get another set of neckboards You said you tested 5 but 3 will be from the same machine so the statistics might be a bit low. |
Hm.. let's see if I get any luckier.. |
To get back on this thread. It might be that the good boards are the exception.
Is it not possible to make a complete new board and transfer the components?
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sat Oct 11, 2014 1:20 pm Post subject: |
|
|
redfox001 wrote: | gjaky wrote: | redfox001 wrote: | Ok than time to get another set of neckboards You said you tested 5 but 3 will be from the same machine so the statistics might be a bit low. |
Hm.. let's see if I get any luckier.. |
To get back on this thread. It might be that the good boards are the exception.
Is it not possible to make a complete new board and transfer the components? |
Yes, I also have been thinking about his as well. So far I decided to rebuild the preamlifier section with the sony IC, that only needs a few parts anyway, and see if that gets any better, if yes really the board is the issue. I already removed the needed items I just have to put them on a new circuit board. Just didn't have time to figure out if it would be useful to build it on a test pcb (like this:link), or I should design and fabricate a test pcb for it, I guess the later one would be the better.
_________________ 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: 2251 Location: The Netherlands
|
Link Posted: Sat Dec 27, 2014 8:37 am Post subject: |
|
|
Would be nice to make a PG inspired neckboard for the XG. With that mod you would have an exceptional machine. Are you any further now? Perhaps it is interesting to mention in this thread that a thread showed up where MikeEby's machine does not have a ver high bandwidth. Actually lower than the PG. It appeals to the conclusion that the neckboard design is a problematic one.
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
Back to top |
|
|
gjaky
Joined: 05 Jun 2010 Posts: 2790 Location: Budapest, Hungary
|
Link Posted: Sat Dec 27, 2014 10:14 am Post subject: |
|
|
redfox001 wrote: | Would be nice to make a PG inspired neckboard for the XG. With that mod you would have an exceptional machine. Are you any further now? Perhaps it is interesting to mention in this thread that a thread showed up where MikeEby's machine does not have a ver high bandwidth. Actually lower than the PG. It appeals to the conclusion that the neckboard design is a problematic one. |
There were three runs of XG series:
...0: XG 750, XG 1100, XG1350; all AC machines
...1: XG 751 (AC), XG 1101(LC), XG 1351(LC)
...2: XG852(AC), 1352(LC)
The difference between the ..0 and ...1 series is the biggest. In fact, the XG ...0 series inherited almost entirely the whole VGA signal path of the PG xtra, only the first preamp IC is slightly better in the XG (CXA1779P instead of CXA1209P). So it shouldn't be any worse than the PG xtra.
The signal path in the PG xtra is the following: first stage a 100MHz input preamp ic (CXA1209P,used at unity gain), on the second stage there is one LM1201 (bw:200MHz) for each colour amplifying the 0,7V signal to 5,5V, and the neckboard is a hybrid push-pull configuration: on the positive side there is a VPA13 ic (130MHz), while the negative side amplifier is constructed of discrete transistors, consisting from transistors, that have ft: 5-700MHz.
In the XG...1 and ...2 the reception ic is the CXA1779P (150MHz) used at unity gain, there is no second stage in the path, all further amplification is done on the neckboards. The neckboards therefore are more complex: the Neckboard consist of three gain stages, there is one reception ic (CXA1709P, 250MHz) amplifying the signal from 0,7V to 2,2V, this ic has symmetrical outputs (+/- side), the second gainstage constructed from discrete transistors each of these preamps are on a separate daughterboards on the neckboards (so two daughterboards are on a neckboard), this stage amplifies the signal from 2,2V to 5,5V, then the signal goes to the output ics, VPJ13S (130MHz) for each side (+/-).
-I've tested that the CXA1779P has very little impact on the signal bandwidth (compared to the CXA1209P). These aren't the bottleneck.
-The neckboard expects 0,7V signal terminated on 75Ohm, so I could simply connected the RGB signal to it from the computer, the machine produced the same poor bandwidth, so the problem is surely on the neckboard or the tube itself. (Note the neckboard needs power from the PSU!)
-I built a PSU that can supply all the needed voltages to run the neckboards on my desk, this way the XG's own PSU can be ruled out as suspected problem cause (see later for prove).
-For measuring purposes I invested in a 500MHz active FET probe (with 1:10, 1:100 dividers) for my 275MHz oscilloscope, these allowing me to take measurements with the least intervention to the signals. This probe has less than 1pF tip capacitance.
-When I was measuring the poor-performing XG neckboards, I found that both the VPJ13S and CXA1709P ICs putting out poorer signals than it was suggested on their datasheets, these two ICs are on the same PCB by the way. On the other hand as I said the second gainstage is populated on two separate PCBs soldered to the neckboard's base PCB, these preamps essentially passing (and amplifying) the signal without measurable bandwidth loss. The neckboards are producing the symptoms without any tubes attached to them, the tubes as the cause are ruled out his way. The measured bandwidth is around 30MHz, and this is in conjunction to what I can see on test patterns too!
-Since the biggest bandwidth loss happened at the last stage with the VPJ13s I spent quite some time with swappping those. I have here good working VPA13 (from PG xtra) ic, VPA15H (same as VPA13 but 150MHz, from Sony 1292 neckboards) and VPJ13s from the spare XG neckboards (which also performed poorly), by simply swapping the output ICs the performance was even poorer with both VPA13 and VPA15H. Note, the VPJ13 is the newest generation of these ICs so probably it has the best performance right away. There was no real difference between the VPJ13s.
-I have here a fully unmolested original PG xtra neckboard here as well. I hooked the input to the XG neckboard to only use its pleamplifier section, but I connected the PG xtra neckboard as the final stage to the output of the preamps, eliminating the VPJ13s. This way the output only lost as much bandwidth as it was expected from the VPA13. This way my PSU proved to be fast enough to serve the circuits, and turned out that the VPA13 on the XG's neckboard doesn't serve half as well than on a separate board (even with an interconnection cable), while the peaking components are not different.
-So far all of my efforts to obtain more cheap XG neckboards failed. And I have to save money to order from a set from Curt, but when I asked Curt if he can test the neckboards for bandwidth before shipping he said no. So I have hard time spending $120+taxes on something that could be just as bad as I alredy have here.
-And yes I find the evidence on XG bandwidth performance also lacking as no one is willing to make screenshots But from pure technical standpoint the NEC XG construction is indeed very good (not perfect, but very good), and permits to achieve high bandwidths -at least on paper.
Back to Mike Eby's screenshots: All of his screenshots were taken az 1080P 72Hz, while the 9PG+'s is at 1080i96Hz (aka. 1080P 48Hz), Mike Eby's worse screenshot is still very good for a stock machine, I think a stock Marquee isn't better either, let alone a barco, especially a Cine 8.
_________________ 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: 2251 Location: The Netherlands
|
Link Posted: Sat Dec 27, 2014 11:54 am Post subject: |
|
|
This is a cliffhanger
You connect the vp13 on the PG board and it performs as expected but on the XG board it performs worse while another chip on that board performs good if I understand correct. So it is not the board material? But it is still likely something specific to these boards....No oxidation? little cracks? I assume you plugged the vp13 form the PG on the XG too? So not this particular chip?
Board not available we will never know the solution! I keep my bets on screenshots showing a bit to high bandwidth.
_________________ 701s->runco933->8500ultra->hd1->hd350->vw100->cinemax+919sp+3x919+9500mp->cinemax+919sp(modded)+kuro600a
|
|
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
|
|