G3, Bad FlyBack and then weird vertical lines without video

[daveyk021]

Hello,

I have never seen a CRT/TV do this before (I used to work in TV repair, but that was in the late 70s through 1983, so I am a tad rusty). I replaced the flyback and horz output to get this iMac to come on. The unblanking has vertical strips in it, missing video. This one had worked almost perfectly until the first time I turned it off (it was having display size jittering, and some flashing, so I assume the flyback was failing). I doubt the CRT got damaged while sitting on the floor of my storage room for a week or so. I have no idea if this could be the video out of the computer board or the display board. I checked all display board caps before replacing the flyback and horz output. If I changed the OS resolution down to 800x600, the blank vertical lines are still there. Could be a yoke, but the mouse passes through the blank line, like they still there. If I move a window, the window doesn't jump across the lines. So it kind of looks like a really weird un-blanking issue. I cannot find schematics for this thing anywhere.

The original flyback was: 6174Z-1017D. The flyback from the parts unit (can not be sure it is 100%) is a 6174Z-1003D. From what little I have found on-line, the 6174Z-1003D is used to replace a 6174Z-1017D.

When the original flyback (6174Z-1017D) was working no lines. With the used 6174Z-1003D on the board, there are lines that are effected only by the screen control. I ordered a new 6174Z-1003D from EBAY for $41. Not a huge loss if it doesn't fix the problem. I don't know how the flyback could be causing this. Oh and if I resize the image via the monitor settings in the control panel, the lines move with the re-sizing, so that rules out the CRT.

 

[hjohnson]

https://www.applerepairmanuals.com/index.php (access with caution, avoid downloading "PDF" software)
(or use archive.org and search for "imac g3 apple service manual")

There's many iMac G3 models. These Apple manuals provide basic board-level descriptions and parts. Won't provide component level servicing info. Apple did that at the factory. Otherwise you'd need a Sams / Photofact and those aren't online and may not exist for the G3's.

YOu've focused on the flyback, but possibly the problem is the horizontal driving transistor. With other Macs, and with CRT TV's, it can be important to replace the horizontal driver with the *same brand* as well as same-model transistor. Older books on fixing 9-inch Macs call this out, and my TV repair buddies note it also. IT's entirely possible, a variation in the flyback model could cause problems. It's all about inductance and capacitance and reactance - that horizontal waveform is *very critical*.

Maybe, what's going on is the return on the sawtooth waveform has bad timing, and is happening too soon - before the full image is scanned? So you are getting some fold-back? If you can tweak the horizontal coil - with all due caution, remember your TV repair history! - and see some changes, that may be informative.

In any event, I think this is more of a CRT TV problem, than a digital side problem. Review your CRT TV history!!!

Web search for me on "imac CRT flyback horizontal repairs"

Mac Classic

Tips and tricks to help keep your Mac OS 9/Early Mac OS computers usable today. Software downloads, guides, hacks and news.

mac-classic.com

iMac Flyback Transformer

iMac Flyback Transformer

macos9lives.com

Testing of Flyback (LOPT) Transformers

regards, Herb Johnson
(saw your VCFed-MidAt post)

 

[sentrytv]

Hello,

I have never seen a CRT/TV do this before (I used to work in TV repair, but that was in the late 70s through 1983, so I am a tad rusty). I replaced the flyback and horz output to get this iMac to come on. The unblanking has vertical strips in it, missing video. This one had worked almost perfectly until the first time I turned it off (it was having display size jittering, and some flashing, so I assume the flyback was failing). I doubt the CRT got damaged while sitting on the floor of my storage room for a week or so. I have no idea if this could be the video out of the computer board or the display board. I checked all display board caps before replacing the flyback and horz output. If I changed the OS resolution down to 800x600, the blank vertical lines are still there. Could be a yoke, but the mouse passes through the blank line, like they still there. If I move a window, the window doesn't jump across the lines. So it kind of looks like a really weird un-blanking issue. I cannot find schematics for this thing anywhere.

The original flyback was: 6174Z-1017D. The flyback from the parts unit (can not be sure it is 100%) is a 6174Z-1003D. From what little I have found on-line, the 6174Z-1003D is used to replace a 6174Z-1017D.

When the original flyback (6174Z-1017D) was working no lines. With the used 6174Z-1003D on the board, there are lines that are effected only by the screen control. I ordered a new 6174Z-1003D from EBAY for $41. Not a huge loss if it doesn't fix the problem. I don't know how the flyback could be causing this. Oh and if I resize the image via the monitor settings in the control panel, the lines move with the re-sizing, so that rules out the CRT.

Just an FYI, if you do a video you may be able to catch a frame that has that black sideways bar out of frame.

But as for the problem, it looks to me as if it’s a damper issue, which means that the horizontal, output transistor could be wrong. Or damper diode, which usually is inside the transistor.
Look up the original transistor number and replace it with the NTE equivalent.
The original part number would be preferred.
I had a TV shop years ago, so I still have a lot of parts from televisions.
Otherwise, it’s a bad capacitor in the horizontal output circuit.
Did anybody try “ringing” the fly back to make sure that the fly back is definitely bad?
You could put the original flyback in but risk blowing the horizontal output transistor.

 

[daveyk021]

https://www.applerepairmanuals.com/index.php (access with caution, avoid downloading "PDF" software)
(or use archive.org and search for "imac g3 apple service manual")

There's many iMac G3 models. These Apple manuals provide basic board-level descriptions and parts. Won't provide component level servicing info. Apple did that at the factory. Otherwise you'd need a Sams / Photofact and those aren't online and may not exist for the G3's.

YOu've focused on the flyback, but possibly the problem is the horizontal driving transistor. With other Macs, and with CRT TV's, it can be important to replace the horizontal driver with the *same brand* as well as same-model transistor. Older books on fixing 9-inch Macs call this out, and my TV repair buddies note it also. IT's entirely possible, a variation in the flyback model could cause problems. It's all about inductance and capacitance and reactance - that horizontal waveform is *very critical*.

Maybe, what's going on is the return on the sawtooth waveform has bad timing, and is happening too soon - before the full image is scanned? So you are getting some fold-back? If you can tweak the horizontal coil - with all due caution, remember your TV repair history! - and see some changes, that may be informative.

In any event, I think this is more of a CRT TV problem, than a digital side problem. Review your CRT TV history!!!

Web search for me on "imac CRT flyback horizontal repairs"

Mac Classic

Tips and tricks to help keep your Mac OS 9/Early Mac OS computers usable today. Software downloads, guides, hacks and news.

mac-classic.com

iMac Flyback Transformer

iMac Flyback Transformer

macos9lives.com

Testing of Flyback (LOPT) Transformers

regards, Herb Johnson
(saw your VCFed-MidAt post)

Sorry I can not get my reply to not show as a quote. I tried with the original horz output from the donner and from the bad one. There isn't a single fricken adjustment on that board. The Service Manual is useless "replace board". I'll have to look, the horizontal width coil may me heat-shrunk (that might have fooled me). Th Primary of the original flyback rings great (with a coil ringer kit I made several years ago), the secondary does not. It at least kind of works with the doner flyback. I don't know how the 1003D is used for the 1017D without any adjustments on the board. Fold-back of the sawtooth sounds right, I'm just not sure what to do about it.

I did search and search the web for this issue. All I really found was "If it's dead; replace the flyback". I saw one reference to the vertical line(s) but no response. I will check out the links you gave.

For now, I must head to Virginia for family. I'll be back Saturday afternoon. (sigh)

Dave

 

[daveyk021]

Just an FYI, if you do a video you may be able to catch a frame that has that black sideways bar out of frame.

But as for the problem, it looks to me as if it’s a damper issue, which means that the horizontal, output transistor could be wrong. Or damper diode, which usually is inside the transistor.
Look up the original transistor number and replace it with the NTE equivalent.
The original part number would be preferred.
I had a TV shop years ago, so I still have a lot of parts from televisions.
Otherwise, it’s a bad capacitor in the horizontal output circuit.
Did anybody try “ringing” the fly back to make sure that the fly back is definitely bad?
You could put the original flyback in but risk blowing the horizontal output transistor.

Dead with original flyback. Primary rings good, secondaries do not. I used the horizontal output from the donner and the original one, same issue. Not a single fricken adjustment on the board which includes the horz width coil). The doner board did have a width adjustment switch, this one does not.

I have to leave for holiday very shortly but can get back to it on Sunday.

Thanks for the information and ideas.

Dave

 

[hjohnson]

First: to Davey. This won't immediately help, but as a favor please identify the exact iMac model, there were many. It's a pain to identify iMacs, I grant. The value of the Apple "service manuals" is to identify boards in use with given models, so start there. How about, the Apple part number on the CPU board, and the same part number on the video driver whatever board? Why bother? maybe someone can offer actual parts from a truly matching video board. Have a photo of the video board around in case it comes down to "here's the board I have, is this like yours?"

I'm reluctant to repeat this, but exact match or known-to-match parts are needed to repair these analog flyback circuits. I believe that's universal to CRT transistor (not vacuum tube) TV and monitor circuits. So 1990's TV repair books may be the best bet for further info. By the time Mac owners needed to dig into these CRT's for component repairs, they had zero skills with any CRT's. "Will I die if I touch the CRT?" is among the ask-for-help posts about iMac CRT repairs. This thread is already more informative than most threads on the subject since the SE/30 Mac., repairs of which are very well documented.

Regards Herb

 

[Hugo Holden]

I find a lot of stuff posted online about VDU repair to be pretty poor myself.

TV's and VDU's operate on the same physical science principles, especially the horizontal output stage and flyback transformer. The principle is energy recovery scanning, invented by Blumlein and perfected by Otto Schade at RCA labs in the 1940's.

The first thing to distinguish if there is a defect in the screen image; is it in the video information controlling the CRT's beam current (or an auxiliary circuit like an abnormal blanking pulse which alters the CRT's beam current) or is the defect in the scanning raster ?

The lines moving with the resizing does rule out the CRT, that is correct, but it does not definitely distinguishing between a raster scan fault and a beam current issue. But there is one clue that does:

The vertical lines have very sharp edges, this normally doesn't happen with a raster scan fault.

Also, if there was anything wrong with the horizontal output transistor (HOT), the drive to it, the energy recovery diode (damper diode) or the auxiliary voltage from the flyback transformer you would not likely have a basically normal scan otherwise, with reasonable linearity and obviously normal EHT. (the EHT must be close to normal because the scan size is inversely proportional to the square root of the EHT voltage, and the size looks about right)

The vertical lines you are seeing are happening because along the H scan starting at the left after flyback, the beam current is getting uniformly low, equally in the three CRT guns in that zone. The horizontal blanking pulse is abnormal. The first bar is little wider than the second, which suggests there is a decaying oscillation in the origin of the pulse, so initially a normal blanking pulse is likely there, followed by a few decaying oscillations that should not be there, and you are only seeing the first two, likely there is a third and 4th not visible because the amplitude fell too low to initiate blanking.

Depending of the circuit design, the H blanking pulse could be applied at the CRT's g1 grids as a negative going voltage. Or, it could be injected into the three video amplifier stages, to take the CRT's cathode voltage positive during blanking time.

The same basic problem can put vertical bars across the entire screen depending on the width and amplitude of the aberrant blanking pulsed (see attached). Most veterans of the TV industry know about this dilemma. It can happen in varying degrees in monochrome or color sets.

Even without the schematic, it is a simple matter to scope the CRT's cathode voltages and the G1 grid voltage and firstly identify where the blanking pulse is being injected and trace it back to its source. It is derived from the flyback transformer windings or the collector circuit of the horizontal output transistor, often with a capacitive divider and coupled by diodes into the video amp circuits or CRT grid to achieve the beam blanking. Some sets used a blanking amplifier transistor too so as to acquire a nice rectangular blanking pulse.

There are some borderline conditions where increased load on the flyback secondary can result on additional pulses after the main flyback pulse appearing on the HOT's collector waveform, due to the leakage inductance between the windings, but normally though, this results in a fairly grossly abnormal raster scan and low EHT. So that is probably unlikely the cause, especially since, by the look of things the auxiliary voltages from the flyback transformer are normal or there would be other defects.

The protocol for diagnosing & repairing it is straightforward. You need to scope the H output transistor's collector voltage initially. You need to buy a x 100 2kV rated scope probe (protect the input jFets in your scope) and set the scope on 50v/cm. Have a look at that collector waveform first. Then after that scope the CRT's cathode and grid voltages to initially identify where the defective pulse is being injected and trace that back to its source. If you can acquire these recordings we can get to the bottom of it.

(I think you are right, it is a blanking/un-blanking issue, but don't forget though, if you are starting out with the premise that you are dealing with a problem ultimately involving the CRT beam current, that includes then the entire video amplifier chain, so it is prudent to check with the scope that the video signal feeding the VDU is normal, I think in this case it is likely to be)

Even in the old days, CRT blanking anomalies were common, they often turned up on magazines. The story being like Marty McFly saying to his jailbird uncle Joey, when he was a baby in the crib "better get used to those bars kid"

 

[daveyk021]

I find a lot of stuff posted online about VDU repair to be pretty poor myself.

TV's and VDU's operate on the same physical science principles, especially the horizontal output stage and flyback transformer. The principle is energy recovery scanning, invented by Blumlein and perfected by Otto Schade at RCA labs in the 1940's.

The first thing to distinguish if there is a defect in the screen image; is it in the video information controlling the CRT's beam current (or an auxiliary circuit like an abnormal blanking pulse which alters the CRT's beam current) or is the defect in the scanning raster ?

The lines moving with the resizing does rule out the CRT, that is correct, but it does not definitely distinguishing between a raster scan fault and a beam current issue. But there is one clue that does:

The vertical lines have very sharp edges, this normally doesn't happen with a raster scan fault.

Also, if there was anything wrong with the horizontal output transistor (HOT), the drive to it, the energy recovery diode (damper diode) or the auxiliary voltage from the flyback transformer you would not likely have a basically normal scan otherwise, with reasonable linearity and obviously normal EHT. (the EHT must be close to normal because the scan size is inversely proportional to the square root of the EHT voltage, and the size looks about right)

The vertical lines you are seeing are happening because along the H scan starting at the left after flyback, the beam current is getting uniformly low, equally in the three CRT guns in that zone. The horizontal blanking pulse is abnormal. The first bar is little wider than the second, which suggests there is a decaying oscillation in the origin of the pulse, so initially a normal blanking pulse is likely there, followed by a few decaying oscillations that should not be there, and you are only seeing the first two, likely there is a third and 4th not visible because the amplitude fell too low to initiate blanking.

Depending of the circuit design, the H blanking pulse could be applied at the CRT's g1 grids as a negative going voltage. Or, it could be injected into the three video amplifier stages, to take the CRT's cathode voltage positive during blanking time.

The same basic problem can put vertical bars across the entire screen depending on the width and amplitude of the aberrant blanking pulsed (see attached). Most veterans of the TV industry know about this dilemma. It can happen in varying degrees in monochrome or color sets.

Even without the schematic, it is a simple matter to scope the CRT's cathode voltages and the G1 grid voltage and firstly identify where the blanking pulse is being injected and trace it back to its source. It is derived from the flyback transformer windings or the collector circuit of the horizontal output transistor, often with a capacitive divider and coupled by diodes into the video amp circuits or CRT grid to achieve the beam blanking. Some sets used a blanking amplifier transistor too so as to acquire a nice rectangular blanking pulse.

There are some borderline conditions where increased load on the flyback secondary can result on additional pulses after the main flyback pulse appearing on the HOT's collector waveform, due to the leakage inductance between the windings, but normally though, this results in a fairly grossly abnormal raster scan and low EHT. So that is probably unlikely the cause, especially since, by the look of things the auxiliary voltages from the flyback transformer are normal or there would be other defects.

The protocol for diagnosing & repairing it is straightforward. You need to scope the H output transistor's collector voltage initially. You need to buy a x 100 2kV rated scope probe (protect the input jFets in your scope) and set the scope on 50v/cm. Have a look at that collector waveform first. Then after that scope the CRT's cathode and grid voltages to initially identify where the defective pulse is being injected and trace that back to its source. If you can acquire these recordings we can get to the bottom of it.

(I think you are right, it is a blanking/un-blanking issue, but don't forget though, if you are starting out with the premise that you are dealing with a problem ultimately involving the CRT beam current, that includes then the entire video amplifier chain, so it is prudent to check with the scope that the video signal feeding the VDU is normal, I think in this case it is likely to be)

Even in the old days, CRT blanking anomalies were common, they often turned up on magazines. The story being like Marty McFly saying to his jailbird uncle Joey, when he was a baby in the crib "better get used to those bars kid"

Wow, thank you for that! Was a great read and advice.


Today: Installed brand new "exact same flyback" from ebay and it smoked. Put in the old working donner. replace horz driver again, and damping diode. Scoped and scoped. Yes, I need a 100:1 probe. Replaced yoke, and that helped a little. I now only get the little squishy vertical line if way over-driven. Yoke and flyback came from same parts donner.

I will now see if I can scope the G1 grids and see if I can identify the un-blanking. Being able to trace it back is another things. Holly crap, this CRT board is complex, especially the horizontal section. That is because width/centering and what not can be controlled by software.

BTW, all caps have low ESR and I even scoped most of them looking for noise.

Back to the bench, and again thank you!

 

[Hugo Holden]

If the flyback transformer overheated, it implies something is overloading it, or the drive to the horizontal output transistor is incorrect or the transistor is shorted, but those two latter things are not the case because you would not have had the raster scan you had before you changed it. This suggests there is a problem on one of the auxiliary circuits that the transformer feeds. Still, that does not explain why you had a scan (albeit with the blanking defect) before. Unless another fault got introduced in the parts changeover.

One thing to consider in making repairs; don't go changing any part like the flyback or yoke without any evidence it has a problem. You can simply add in other issues. For example if you made some error changing the flyback transformer, and that caused it to smoke, you may have introduced damage elsewhere and now have two superimposed faults, not one.

When you have a fault that you can see on the CRT's face, like those vertical bars, focus on finding the cause of that with the scope and leave everything else alone until you have a reason to replace something or it could create a downhill spiral.

 

[daveyk021]

Okay, yes, the unblanking has ringing in it. This explains why it worked good with its original flyback (until it failed). The original was a - 1017D, which seems unavailable. The donner one is a -1003D. It's PCB had a lot more complex un-blanking circuit, at least a lot more parts to get rid of the ringing. Now, I don't know what to do about it. Before the un-blanking amplifier transistor I tried adding some caps to ground. I could narrow the black band, but not eliminate it. Adding more just made the jail bar bigger.

So, you got me on the right track. Now to figure out what to do about it......

 

[Hugo Holden]

The next move would be to start an exhaustive search for the schematic.

If it cannot be found, the only thing one can do is to sit down initially with a pencil & paper and start to draw the schematic out. Once that is done, then a neat copy of it helps with a drawing program.

You would not likely need to draw out the whole thing, just the blanking circuitry and blanking amplifier most likely.

It sound like a daunting task to have to do it. I have done it for an entire VDU because there was no manual being a mil spec model. I started with pencil & paper, drew out everything and basically made a service manual for it. I did the same thing for the IBM5155 psu too and some other gear like printer interfaces and lathe motor speed controllers where there is no manufacturer data. The process is aided by looking up the data sheets on the active devices. It pays to make the diagrams as tidy and orderly as possible.

www.worldphaco.com/uploads/The_1987_Vintage_Avionics_Conrac_Video_monitor..pdf

 

[daveyk021]

During my experimentation, I blew out the initial rectification (to ground) diode for the flyback unblanking pulse. A 1N4148 gave me jail bars all across the screen. So I pulled on from the donner board that appeared to be in the approximate same location (donner board newer and double-sided with some SMT). Okay back to where it was with a slightly narrower bar. Then adding a 220pf ceramic from the base to ground of the transistor that creates the un-blanking pulse fixed it! Now if I turn the screens too far up, the jail bar comes back (along with re-trace, of course). So the ringing is not cured, it is just suppressed enough to work, I wonder what will happen if it gets really hot? Through the control panel, I can make the picture as big as the CRT Mask and it looks good except the convergence needs fixed. I do think changing the yoke to the one that went with this flyback helped quite a bit too. This video pcb just doesn't have the circuitry to tune the flyback or suppress its ringing. Looking at the board that this flyback came from, that unblanking circuit is a lot more extensive.

I would have never looked this way if it wasn't for your lesson; thank you. It is now "fixed", biting and holding tong at just the right angle.

 

[Hugo Holden]

During my experimentation, I blew out the initial rectification (to ground) diode for the flyback unblanking pulse. A 1N4148 gave me jail bars all across the screen. So I pulled on from the donner board that appeared to be in the approximate same location (donner board newer and double-sided with some SMT). Okay back to where it was with a slightly narrower bar. Then adding a 220pf ceramic from the base to ground of the transistor that creates the un-blanking pulse fixed it! Now if I turn the screens too far up, the jail bar comes back (along with re-trace, of course). So the ringing is not cured, it is just suppressed enough to work, I wonder what will happen if it gets really hot? Through the control panel, I can make the picture as big as the CRT Mask and it looks good except the convergence needs fixed. I do think changing the yoke to the one that went with this flyback helped quite a bit too. This video pcb just doesn't have the circuitry to tune the flyback or suppress its ringing. Looking at the board that this flyback came from, that unblanking circuit is a lot more extensive.

I would have never looked this way if it wasn't for your lesson; thank you. It is now "fixed", biting and holding tong at just the right angle.

Normally there is no ringing of any significance on the terminals of the flyback transformer primary, except for the flyback pulse itself, which represents a half cycle of un-damped oscillation where the flyback is operating in a resonant (un-damped mode) and this allows the polarity of the magnetic field in the flyback and yoke to reverse, placing the beam on the left side if the screen to start a new scan.

During scan time on the R hand half of the screen, there is either a constant voltage switched across the primary by the horizontal output transistor (HOT), so that a linear current develops, and on the L side the damper diode conducts clamping the voltage to a constant level, so that there is a near linear return of current to the power supply to scan the left side of the screen. The point being though, the voltages on the flyback's terminals on the primary windings are generally constant during scan time.

In some cases, depending on the design of the flyback, there can be enough leakage inductance between a primary and secondary winding, such that when the voltage clamping on the primary (due to the HOT and damper diode) occurs, this results in the leakage inductance appearing in the secondary (If you short out a primary winding on a transformer, or switch a constant DC potential across it amounts to the same thing from the AC perspective, the leakage inductance appears in series with the secondary). This can leave a series of decaying oscillations, during active scan time, immediately after flyback on a secondary winding on the flyback transformer. The usual winding to have these oscillations is the EHT winding, but most people don't know it is there because the EHT winding output prior to the EHT rectifier never gets scoped for obvious reasons.

On the other hand, in most VDU's, the H blanking pulse is taken from the primary winding of collector circuit of the HOT, which is free from such oscillations. But there are always design variations.

The actual "tuning of the flyback" is a combination of the self and lumped tuning capacitance on the primary and it affects the frequency and amplitude of the half cycle of the main flyback pulse on the primary, and is not related in any way to the jail bars problem because the primary voltage which occurs when the transformer is heavily damped with constant voltages applied across the primary during scan time has no oscillations.

In any case we can deduce two things about the problem in your VDU, firstly the blanking signal is likely taken from a secondary winding on the flyback transformer, because it has oscillations in it (and there are no oscillations in the scan current, the raster is ok so the primary and yoke are free from oscillations) , and the blanking circuit that processes that signal, something has gone off with its bias conditions, and it is not ignoring those oscillations after the main flyback pulse. Generally the blanking amplifier circuit should just slice off the flyback pulse during H retrace. It should be obvious what was going on scoping the blanking circuit.

 

[daveyk021]

Generally the blanking amplifier circuit should just slice off the flyback pulse during H retrace. It should be obvious what was going on scoping the blanking circuit.

I was wondering about that. There is another transistor in parallel with the one that produces the blanking pulse. It looks like it ties in with the programmable horizontal driver IC (controls width, centering, etc.). I am wondering if it is supposed to be turning on just as the other one is turning off? I will have to look in to that. In some instruments that I work on, an FET would be used to sharply terminate a pulse. This is another BJT. This did not appear to be a problem before this flyback with the different part number, or it never reared its head before. I report back if anything is new.

BTW, I left it run all days with towels over it, so it was pretty hot. A skinny blank line has started to appear. The width of the line was most prominent at the top and the bottom, not fully there in the middle.

 

[daveyk021]

and it is not ignoring those oscillations after the main flyback pulse. Generally the blanking amplifier circuit should just slice off the flyback pulse during H retrace. It should be obvious what was going on scoping the blanking circuit.

I was playing around again. Instead of a board jumped I wanted to put two 1N4148 in series, passing only the positive portion of the pulse. I put them together wrong. The unblanking pulse is now perfect going to the neck boards. I can crank the screens all the way up and not a hint of jail bars.

Forgive the crudeness of this drawing. The diodes are in-place of jumper J59.

The scope image shows the unblanking pulse now going to the neck board. The fix was accidental serendipity.

 

[Hugo Holden]

I was playing around again. Instead of a board jumped I wanted to put two 1N4148 in series, passing only the positive portion of the pulse. I put them together wrong. The unblanking pulse is now perfect going to the neck boards. I can crank the screens all the way up and not a hint of jail bars.

Forgive the crudeness of this drawing. The diodes are in-place of jumper J59.

The scope image shows the unblanking pulse now going to the neck board. The fix was accidental serendipity.

If the diode polarities are as you have drawn it, one 1N4148 diode is conducting in the forward direction, while the other is going into zener breakdown, acting like a 90 to 100V zener, so only the peak of flyback getting through to the blanking transistor. It could be though that the H blanking pulse is now a little narrow, but it may not show up as a defect.

The real question though is what caused the real fault and why the apparent signal amplitude feeding the blanking transistor appears way too high in level.

Some faults, and this could be one of them, can be fixed with a "kludge" like this, but it is not very satisfying, not to have found the root cause of the problem.

Though, I have done this sort of thing. I was once repairing a time-base corrector, based on many programmable devices and a myriad of logic and analog IC's and no manufacturer data at all. An aberrant pulse had appeared in the blanking period fouling up the output video's sync. Presumably one of the FPGA's had malfunctioned. I hunted around the board with a scope until I found a near identical pulse on an IC pin and used it to phase cancel the aberrant pulse in the video output, by inverting it and mixing it in with a diode & resistor. It was "cheating" but it worked, somewhat like the trick Captain Kirk taught Spock.