|
|
Post by gjaky on Dec 1, 2015 15:28:25 GMT -5
One other question can you somehow relate the 2ns measurement to this picture? Is that close to the 380MHz? Here at home my pulse generator has a self rise time of 700ps, my oscilloscope has a self risetime of 1.3ns, and my FET probe has a rise time of 750ps. Compenstaing the 2ns value (differences of square-roots) would give 320MHz for the circuit under test. I think this is quite close, since in this measurement the oscilloscope had lower bandwidth than what was predicted for the circuit, small errors in reading give huge difference in calculation (calculating with 1.9ns would give 390MHz, but calculating with 2.1ns would give on the other hand 270MHz). Also there is some fluctuations in the frequency plot, and the amplification of the harmonics therefore depend on the actualy measuring frequency too. Still, there is a quite broad range where the the response within 1dB, and then there is a broader range that fits in a 2dB margin, I don't think this would be much visible on the screen, especially 1:1 and 2:2 etc. patterns also don't show "every" frequency on screen |
|
|
|
Post by stridsvognen on Dec 1, 2015 16:01:43 GMT -5
|
|
|
|
Post by gjaky on Dec 2, 2015 11:14:26 GMT -5
Thanks, it is a nice instrument for that money, but in my opinion (also my findings seem to support this) a plain PC with an nVidia VGA card serve just as good as test pattern generator. Otherwise my pulse generator has better specs (0.7ns rise time vs. 1.1 ns) for bench tests, and it only cost me 30€  |
|
|
|
Post by gjaky on Dec 16, 2015 16:02:08 GMT -5
So far I've made two faint attempts to test the XG-VNB board in projector. Sunday I dared to put the board in the XG. I got visible scanlines looking into the lens and very faint menu display. Engaging the AKB however did not bring the cure for this, it just lighted up raster insanely. This is indicatees a problem with the current sensing circuit on the VNB (which I taylored to be same to the original XG circuit). But as it turned out I did not study this circuit close enough, because the working point of this circuit depending on the actual Cathode drive too which is significantly smaller on the VNB, that's why the auto kelvin balance circuit turned up the G2 so much... It seems I have to tune this sensing circuit manually. Good thing the spot burn protection works at least. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 17, 2015 11:43:05 GMT -5
I once put an opamp on a 909 neckboard with a different impedance and that shifted the blacklevels a lot too. I did not dare to do the adjustments.
|
|
|
|
Post by gjaky on Dec 30, 2015 13:41:36 GMT -5
Nowadays I find very hard this project... If I'd knew it will be this hard I wouldn't even begin I have a PG xtra chassis here to play, as I thought it is quite close to the XG for this project. I was wrong, however I managed to get all the auxiliary signals from the boards to drive the VNB but I had to spend a considerable time to check the schematics for the needed signals, and still the AKB not even working with the VNB... But in the end I got some picture out of the VNB in the PG xtra, but there are still problems... Pictures first:    So you can see some kind of picture coming out of the red tube (on which the Electrohome VNB was applied), but there is some kind of oscillation on the picture too, it seems this noise is introduced by the +/-85V PS unit which I got from Scott. I'll have to look to that stuff too, hopefully it only needs some snubbing... In worse case the VNB is oscillating, but the space is tight so I can't really access the VNB for measuring it while it is in the projector. Maybe I should get red the tube out of the chassis, but then the HV lead will be short etc... problems everywhere  Also it seems I can't omit the CLC449 stage from the VNB, because now simply there is no enough drive coming out of the neckboard, moreover maybe I'll even need a level shifter stage before the AD834 too. |
|
nashou
Unmoderated Off Topic
Tech in Training.....
Posts: 1,239 
|
Post by nashou on Dec 30, 2015 17:37:39 GMT -5
Try to lower the G2 to the red, when too high a G2 is applied the image gets that noise like that.
Nashou
|
|
|
|
Post by gjaky on Dec 31, 2015 2:33:50 GMT -5
I can set G2 manually (G2 pot) so the picture totally disappear from the tube, and in fact when I increase G2 these odd running dots are the first which become visible.
I also noticed some of this noise is visible on the blue tube too, but this noise is only there if the VNB is powered up, this is why I said it's either the power supply or the VNB itself which generates this.
Blanking and spot kill signals are there and they seem to be effective, the dots are even there without signal applied.
As I checked it is now sure the 85V PSU needs some snubbing, I just don't know if it will help.
I think I should pull out the red tube out of chassis...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 31, 2015 5:26:30 GMT -5
Think you are the first person that got a picture from a Marquee neckboard in a Nec. Bit wobbling but hé the first wheel was square  |
|
|
|
Post by gjaky on Jan 9, 2016 15:32:17 GMT -5
In the past days I was really desperate about this project. But it seems today I reached a major milestone. -As I wrote before first I focused on Scott's prototype PSU, and figured that it entirely lacks mains RF filtering, so I put a nice CLCL filter network on the mains input and also experimented out a snubber network parallel with the rectifier diodes on the secondary side. -Pulled out the red tube out of the chassis to be able to measure on bench. -Cleaned up my bench -Cut long cables to connect the VNB with the chassis -Finally, added "XG like" OSD driving circuit to the PG xtra to be able to use AKB circuitry. And tadam: Nothing! I got no picture at all just a full lit up raster with retrace lines... Since the tube was already on the bench I could figure out thecontrast voltage is too low, despite what is writen on the PG xtra's SM. I also figured out, that for the testing I don't even need to be able to set the contrast for the neckboard, so I simply wired to "full drive". Also I had an overlooked resistor on VIDEO OUT board, which blocked all effective video output... And tadam (again): Got picture with OSD, and working AKB! And even the picture noise went away eventually! No awful retrace lines and fully lit raster anymore! Also, I got plenty tube drive this way, so my original assumption still holds: no need for second CLC449 stage on the VNB. Let's see some pictures then: Bench test 1  Bench test 2  Hey, we have drive!  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 9, 2016 18:41:06 GMT -5
I would not advice anyone reading this to try this at home |
|
|
|
Post by mastertech on Jan 9, 2016 19:06:10 GMT -5
Ahhhh, I can just feel the radiation coming off that tube. |
|
|
|
Post by gjaky on Jan 10, 2016 2:49:07 GMT -5
I would not advice anyone reading this to try this at home Sometimes I wish I'd never try this either... |
|
|
|
Post by Casethecorvetteman on Jan 10, 2016 5:34:20 GMT -5
Do you reckon youll be able to rig in P19LUG tubes with these neckboards?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 10, 2016 7:06:30 GMT -5
He but now you got it working with the PG you might not need to remove the red from the XG to get it working there too?
|
|
