Videokarma.org

Go Back   Videokarma.org TV - Video - Vintage Television & Radio Forums > Vintage TV & Radio Tech Forum

We appreciate your help

in keeping this site going.
Reply
 
Thread Tools Display Modes
  #1  
Old 03-06-2013, 06:11 PM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
Cool Knox Studio 40.ND Character Generator Trouble

Hello again, all. Hopefully this post is not too far off topic. It's not really "vintage" repair, per se.

By suggestion, I recently picked up a Knox Video Studio 40.ND video character generator. Because you won't find much of anything about it online, I'll tell you a little bit about it. This unit can do not only full page titles in color and with a few different fonts, but also crawling and rolling titles. A solid background color can be displayed or the background made transparent so that incoming video will show through. My example was made in 1997.

Two video inputs are available, for either composite or Y/C video signals. A switch at the back lets you choose which type of input signal you will use.

The character generation part of things seems to work just fine.



It's when you display video from either of the two inputs that it becomes obvious something is terribly wrong.



I have more captures of the output here.

When I first got the unit, there was essentially no color at all when inputting a video signal. Only later did I notice that one object, a particularly bright red record album sleeve, had something of a "beating" or "pulsating" green color halo around it.

Finally I started to look at the chroma handling circuitry inside this unit. A Harris/Intersil CA3126E chroma processor IC is used. From what I could see, all voltages and the color clock signal (3.579 MHz) looked to be OK. So I turned my attention to the input end of things when I noticed that putting my finger on the chroma input pin of the IC would result in dramatic color changes. And then I found what I thought was the problem...a capacitor in series with the chroma input. Substituting a good capacitor resulted in the correct colors appearing. Well, that was an easy enough fix.

When I soldered the new capacitor in, things were bad once again. I tried another capacitor just to be sure I hadn't damaged the first while soldering it in. There was no improvement.

I looked again at the clock frequencies going to and from the CA3126E chip. On pin 6 (VCO input) I got a reading that differed somewhat from pin 7#(VCO output, 3.579 MHz) and pin 8 (carrier output, 3.579 MHz). I've forgotten the reading from pin 6, but can get it again. I want to say it was around 3.35 MHz.

Removing the CA3126E results in loss of all color for both input video and generated characters. Luminance output continues to function.

Knox Video was of absolutely no help. (Frankly, I find it inexcusable for any company selling products to a professional market to not support them for as long as they are in business.) They ignored my e-mail and when I called, told me that they had thrown everything away concerning this unit while cleaning their office.

I'm stumped. I want to suspect the IC at this point, but I don't feel that I've got enough proof to do so. I'm kind of starting to reach the point where I feel like slinging it into the "stuff for recycling" bin because I'm getting nowhere at this point.

There's a video here showing the defect in all of its glory. Note how the color is unstable at times, especially with regard to that router.

Last edited by uxwbill; 03-06-2013 at 06:19 PM. Reason: forgot the self test image
Reply With Quote
  #2  
Old 03-07-2013, 11:30 AM
Chip Chester Chip Chester is offline
VideoKarma Member
 
Join Date: Apr 2012
Posts: 759
What's the timing relationship between the camera and the CG? Are they both referenced to black burst, or some kind of sync?

It's very likely the CG has a ref or black burst input... try sending the camera video into the video input and then looping it out to the ref input, or vice-versa. (Assuming it has looping inputs on either the vid in or the ref in.)

Chip
Reply With Quote
  #3  
Old 03-07-2013, 05:11 PM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
Quote:
Originally Posted by Chip Chester View Post
What's the timing relationship between the camera and the CG? Are they both referenced to black burst, or some kind of sync?
Not so far. If that is required, I wouldn't have known to do it. Just one camera has been hooked up at any given time. (The application for which I'd be using this--provided I get it working--isn't genlocked or time base corrected, though it could be if it needs to be. I know this isn't any way to do things, but I've gotten away with it thus far.)

Quote:
Originally Posted by Chip Chester View Post
It's very likely the CG has a ref or black burst input... try sending the camera video into the video input and then looping it out to the ref input, or vice-versa. (Assuming it has looping inputs on either the vid in or the ref in.)

Chip
I don't recall seeing anything like that on it, but I will take a look again tonight.
Reply With Quote
  #4  
Old 03-09-2013, 12:18 AM
NJRoadfan NJRoadfan is offline
VideoKarma Member
 
Join Date: Jul 2011
Posts: 123
Post a photo of the rear panel jacks. I don't recall this unit being genlockable, but I could be wrong. FWIW, I used one of these in HS. Even that unit in a genlocked studio had a slight hue shift on outputted video, but nothing like this! Somewhere along the line, something is throwing the chorma signal out of whack. I just find it weird that the chroma shift doesn't evenly fill the entire frame during the first part of the sample clip.

EDIT: I got a shot of the rear, no reference input. Its too bad you don't have a vectorscope handy, that would aid in diagnosing and adjusting color.

I would check if the unit works properly if fed an S-Video signal. If so, that points to the problem being in the comb filter portion of the unit.

Last edited by NJRoadfan; 03-09-2013 at 12:49 AM.
Reply With Quote
  #5  
Old 03-09-2013, 08:14 AM
Chip Chester Chip Chester is offline
VideoKarma Member
 
Join Date: Apr 2012
Posts: 759
I wonder if it might have an internal/external choice in some menu. Since it puts out a image with nothing else attached, we know it has some internal reference capability. It may not auto-switch to correctly handle external sources... or it may be set for PAL or something like that, too...

Of course, it could be busted, as you thought, too...

Chip
Reply With Quote
Audiokarma
  #6  
Old 03-09-2013, 03:57 PM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
Quote:
Originally Posted by NJRoadfan View Post
I just find it weird that the chroma shift doesn't evenly fill the entire frame during the first part of the sample clip.
I find that extremely odd as well. Even in a view where the camera doesn't move, the picture color is not always stable. Since replacing the cap, it has "snapped" to the correct colors for just a moment or two.

Quote:
Originally Posted by NJRoadfan View Post
EDIT: I got a shot of the rear, no reference input. Its too bad you don't have a vectorscope handy, that would aid in diagnosing and adjusting color.
I did get your message about the vectorscope plug-in for VirtualDub. I just haven't tried to use it. I don't know how much it can help as I already know there's a serious problem. I just don't know why, or more importantly, what should be done about it. (That and my copy of VirtualDub seems to be uncooperative at the moment in that it will crash-by-sudden-disappearance.)

I do plan to try the S-Video input as soon as I figure out where those cables are.

Quote:
Originally Posted by Chip Chester View Post
I wonder if it might have an internal/external choice in some menu. Since it puts out a image with nothing else attached, we know it has some internal reference capability.
As best I can tell, the unit's program (which seems to be quite rudimentary) only handles character generation...though I definitely don't have and could not get any documentation. What I know about it is simply "reverse engineered" from pictures of the Knox keyboard and experimentation. I have not found any evidence of an onboard menu.

Quote:
Originally Posted by Chip Chester View Post
It may not auto-switch to correctly handle external sources... or it may be set for PAL or something like that, too...
When I first got the unit, I noticed that the information label on the bottom said "Studio 40.ND PAL". This made me pretty sure it began life as a PAL unit. Yet I know the unit was used here in the US and why the previous owner would be using PAL equipment instead of NTSC is beyond me.

I really did think this might be the reason for the lack of picture related color, but it doesn't explain the shift in this unit's behavior. Since I haven't ever tried it, I'm also not sure what would happen if one were to display a PAL signal on an NTSC TV. I have not tried feeding the unit a PAL signal to see what would happen.

I couldn't imagine that Knox Video would have designed the unit such that it had different circuit boards for each television standard. After poking around, it seems the video board can be switched between PAL and NTSC by way of a jumper. A Hitachi HD440072 is used for sync signal generation. It is to be a multi system capable part (for PAL, NTSC and SECAM).

There's also no explanation for why the unit clearly seems to be putting out an NTSC color signal...the best I can say is that I don't think it's set up for PAL operation any longer.

I was really hoping someone might have some documentation hiding in a forgotten file somewhere.
Reply With Quote
  #7  
Old 03-09-2013, 06:19 PM
Chip Chester Chip Chester is offline
VideoKarma Member
 
Join Date: Apr 2012
Posts: 759
Found a couple of links you may have run across:

http://www.videomaker.com/article/17...&gbv=1&ct=clnk

http://www.shopping-advices.com/feat...generator.html

Do you have the "instructional overlay" outlining basic operation? If not, perhaps one of the auction sellers could make you a copy before sending it away. The video technologies article says the manual is only 5 1/2 pages long...

Also, there are two adjustment pots (with hex nuts) tucked up high on the back panel... you may have a go at those... after marking their current position fairly accurately. Best to do once you get video scope going.

And with regard to PAL... they (or you) may have modified the unit's main video board output for NTSC, but the video input you're using might still be PAL... which would probably look squirrely. The fact that it snaps into correctness on occasion leads me to believe it's a sync problem.

Chip
Reply With Quote
  #8  
Old 03-09-2013, 06:32 PM
NJRoadfan NJRoadfan is offline
VideoKarma Member
 
Join Date: Jul 2011
Posts: 123
PAL on NTSC TVs results in no color at all. Plus the display will roll due to the difference in refresh rate (50hz vs. 60hz). The output of that unit is definitely NTSC (60hz, 525 lines, color sub carrier at 3.58Mhz) as you are getting a stable color picture. The only PAL format video that was 60Hz with a 3.58Mhz color sub carrier was PAL-M used in Brazil. Viewed on NTSC TVs, it would produce a black and white picture.

If an S-Video cable doesn't resolve the problem, I would start tapping the chroma signal throughout the unit. The first place to check would be at the input pin of the Harris chip. We can likely assume that the problem isn't after the text is overlayed into the video signal as that displays correct and stable color. If the chroma video is correct at the input pin of the Harris chip than the IC is likely bad. If not, its somewhere in the circuit leading up to that chip.
Reply With Quote
  #9  
Old 03-10-2013, 12:20 AM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
Quote:
Originally Posted by Chip Chester View Post
Found a couple of links you may have run across:
I'd run across one, but not the other. I also found some auctions for the earlier beige model with the built in keyboard. (The pictures were readable enough to figure out what the major function keys were by process of elimination.) The only info I had was the sticker on top of the unit itself, which isn't helpful if you don't have the specially labeled keyboard. (The Studio 40.ND uses a separate, electrically compatible with the IBM AT, keyboard.)

I thought sure that someone would have scanned the documentation or at least the schematic for this unit. Apart from people who are selling them, not much information ever came up. Makes me wonder how many of them Knox Video actually sold...

Of the sellers I asked about copying that information from their unit, none responded.

Quote:
Originally Posted by Chip Chester View Post
Also, there are two adjustment pots (with hex nuts) tucked up high on the back panel... you may have a go at those... after marking their current position fairly accurately. Best to do once you get video scope going.
Those are some kind of video centering/position adjuster.

Quote:
Originally Posted by NJRoadfan View Post
If an S-Video cable doesn't resolve the problem, I would start tapping the chroma signal throughout the unit. The first place to check would be at the input pin of the Harris chip.
I finally managed to find an S-video cable after much looking around. Funny thing is, I know I had some Apple Desktop Bus cables hanging around and now they seem to be gone.

I fed the unit video through its S-Video input and there was really no improvement at all. Colors were still wrong, though I didn't notice the "beating" effect nearly as much. I also found that it was possible to get a correct rendition of color bars by tweaking the trimmer in the 3.579 MHz oscillator circuit for the chroma processor. Yet this was very unstable and normal video gained two rainbow colored bands running through it. At no point could I get correct color from anything other than the color bars.

So then I turned my attention to picking up the chroma signal from the input pin to the CA3126E after removing it from its socket. I moved the S-Video cable to the main output, and cut into it so I could tap the chroma line while leaving luminance in place. Then I took the chroma line and touched it to the input pin on the IC socket.

Not having an S-Video capable monitor, I set a Sony DCR-TRV27 Handycam to accept video input. I hooked up another Handycam for video input to the Knox unit.

I didn't get a very strong rendition of color when trying to pick up the signal at the IC socket, but I figured that maybe the resistor in series with the chroma input was attenuating the signal too much for the Handycam to really pick it up. So I moved backwards in the circuit and picked the chroma signal up before the resistor and cap that are normally in series with the chroma input.

At that point I had good, if a little bright and washed out, color on the Handycam display. Test video here.



I think I'm finally comfortable with the conclusion that the IC has gone bad.
Reply With Quote
  #10  
Old 04-03-2013, 11:21 PM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
Unhappy

I'm back and don't have good news to report. Replacing the CA3126E chroma processor didn't help at all. Everything chroma-wise is just as messed up as it ever was.

Near as I can tell, the chroma signal makes it to the processor IC in good shape (see my previous post for an example). I don't know what's happening to it after or during its time in the IC...and I really can't imagine the replacement IC is also bad!

The 3.579545 MHz oscillator's output seems as though it goes onward from the chroma processor to a National Semiconductor LM1886N/LM1889N pair. I believe these two ICs are responsible for delivering the character generator output to the final video mix, which is handled by a couple of GT4123A video mixer ICs from Gennum. (I haven't verified that this is fact the exact signal pathway, it's just an educated guess. The PCB is multi-layer and that makes tracing it more difficult.)

The most recent thing I've done is to replace all of the passive devices around the chroma IC for which I have exact replacements. No change has resulted. I didn't expect it to...this was, admittedly, "shotgunning" and wishful thinking.

I've had no luck finding an owner's manual or service data for this unit, and at this point I think I need one or ideally both to go further (although I have a few ideas I'm going to check out in the near future). If you have some of this information for this unit, PLEASE contact me!

Otherwise...I'm nearing the point where I feel I've spent more time on this thing than it's worth. (So far what keeps me going is the fact that I really don't want to let this thing "win".)
Reply With Quote
Audiokarma
  #11  
Old 04-05-2013, 10:12 AM
NJRoadfan NJRoadfan is offline
VideoKarma Member
 
Join Date: Jul 2011
Posts: 123
It looks like you are getting a phase shift in the chroma part of the signal. Either something is messing with the input signal before it hits the GT4123A, or the internal electronics are not locking onto the video input correctly (the hardware's reference clock is out of phase with the input video). Usually the latter can be fixed with a "chroma phase/hue" potentiometer if the hardware is working right. Somewhere in that unit there is circuity to sync (genlock) the unit to external video sources. Otherwise it can't overlay graphics onto video! That Hitachi chip likely does it and I would focus on components in that area if fiddling with potentiometers doesn't have any effect. The datasheet for the HD44072 has a sample circuit for a color camera application (figure 3) that is likely very similar, if not identical, to what the Knox uses.

Sample application for that Hitachi chip. Might be helpful in tracing the circuit: http://www.epanorama.net/sff/Compute...0Converter.pdf

Last edited by NJRoadfan; 04-05-2013 at 11:51 AM.
Reply With Quote
  #12  
Old 04-08-2013, 02:39 AM
uxwbill uxwbill is offline
VideoKarma Member
 
Join Date: Jun 2012
Location: Illinois
Posts: 52
I'm afraid I have finally reached the point where enough is enough and I don't care any more. I really thought I'd be able to crack it eventually.

It hasn't happened and there is a point when you just have to say "this cost $x and my time is worth more than that". I have reached that point.

Perhaps my attitude is bad and wrong (or I just need to vent). I've spent many hours looking, testing and trying to figure out how the blocks of circuitry are meant to work. Some parts of the design just don't make any sense to me. It's a task I'm VERY tired of!

Knox Video was of no help at all, though they did say that attempting any adjustment of the pots on the board would not be a good idea. They also mentioned that the adjustments interact with one another.

Maybe someday the documentation/schematics will surface and perhaps the unit will still be around. I know of one more place I can think to ask.
Reply With Quote
  #13  
Old 04-08-2013, 10:54 AM
NJRoadfan NJRoadfan is offline
VideoKarma Member
 
Join Date: Jul 2011
Posts: 123
It can be frustrating testing parts without knowing how it works. Even more so when you can't even see the board itself (I'm just guessing here based on symptoms and chips present)! I can see them not wanting you fiddling with pots, particularly if they aren't labeled. I wouldn't be surprised if its something simple/stupid... it always seems to be the case!
Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 07:39 AM.



Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
©Copyright 2012 VideoKarma.org, All rights reserved.