[SOLVED] Coloured retrace lines on top of the picture, Sony BE4A chassis KV21T1D

[pcdata76]

I solved the first problem. By using the Capacitance-ESR meter that i've built. i removed and measured all electrolytic caps except the ones at tuner section. Some of them, especially few caps in power supply and one cap in horizontal deflection are way out of tolerance in means of ESR (i mean instead of some 0.x ohms, i measured 1x.x ohms) I replaced them all with new ones by selecting the most low-ESR one in my vault. So in total, i replaced around 20 caps including the slight off-tolerance ones. I powered the TV on but nothing changed in means of retrace lines except some change in geometry possibly caused by the replacement of faulty cap at deflection. Then, i removed EEPROM from the board and put an IC-socket instead. I programmed another new EEPROM using the .bin file which i've found from internet and put it on TV board. I powered the TV on and hopefully, retrace lines were away! Raster was too overscanned due to the default settings on eeprom but i changed it quickly by displaying a pattern on TV using service menu. Now, retrace line problem looks like to be solved.

The second problem, i was hoping that this one should also be solved thanks to the change of all failed electrolytics especially in power supply section, but unfortunately nothing changed. I pushed and distorted the board using a plastic rod in case of a possible cold solder junction but nothing observed. I also re-route the cables inside of the tv to keep them avay enough from deflection yoke and FBT but this one also doesnt changed anything.
Since noisy lines have a frequency of 50 (or 100 hz) i still suspect from power supply. I'll change the main rectifier bridge now in case of a leakage. I measured D601, the 2nd and only diode directly connected to the 50 hz mains together with bridge rectifier, but it was ok (a leakage around 2 microamperes which is OK, and a forward voltage within limits) This problem is more annoying than the retrace lines because they are not stationary at the modes with 60Hz vertical scan rate, which are mostly used on this tv.

 

[retrogear]

D601 provides a temporary startup supply and then D602 supply overtakes it so then D601 should be reverse biased. Any excess AC ripple on anode of D601 could turn it on and introduce a hum bar into the power supply regulation. Check R614 8.2M . It insures a neutral ground. With a meter, measure the DC across C607 then measure AC voltage on anode of D601 with respect to the negative of C607. That AC is a full sine wave. It should measure less than twice your DC reading on C607. Remember D601 is a halfwave rectifier so the DC it produces on the cathode is half the AC on the anode. Of course, don't use your scope in this hot ground area w/o an isolation transformer !!! With a scope you could verify D601 is not getting forward biased but that is dangerous territory to use a scope !!!

Larry G

 

[pcdata76]

Thank you for your quick answer. After taking the measurements, could i verify startup circuit alternatively by disabling it after turning the tv on (i mean for example by mounting a temporary switch between D601 and R604 and breaking the circuit after powering the tv on) I dont think it will harm but i needed to ask again for double checking.
Since i dont have an isolation transformer readily available now, i wont use scope. Thank you for warning.

 

[retrogear]

That's the way it looks in theory but I'm not comfortable telling you to try that. D601 is only tied to one phase of the AC so not sure it would induce 100/120Hz noise but maybe only 50/60Hz.
Also puzzling is how that would create the east/west distortion. May want to wait for Brian's input on this. I have seen a defective degauss coil circuit induce a hum noise but again, that would be 50/60Hz.


Larry G

 

[pcdata76]

If the checks that you've said does not help then lets wait for Brian's input. Is there a safe way to disable and isolate E-W correction circuit (like removal of L804, C827 and maybe R828) to check the lines disappeared or not. Maybe it could help us to identify the problematic section??

 

[retrogear]

Ok back to basics again. If the crawl is being induced in the deflection circuit you should be able to find it with your scope. Put your scope on CN801 pin 1 or 2 to see the hor drive to the yoke which is waveform #17. Get the waveform locked with your trigger then slow the sweep rate down to the vertical rate. Look along the top of the waveform which is the amplitude of the signal. You might see a bow
in the amplitude which is the pincushion correction locked on your scope but watch for a crawl in the amplitude moving from left to right (if 60Hz). If there is crawl that is the problem. Now do the same for pin 6 waveform 12 which is the vertical drive. With vertical signals you will just see a "wobble" in the trace. Do the same for pin 5 waveform 13. The parabolic waveform 13 is actually the east/west correction signal.
With waveforms 12 and 13 you could actually slow your scope even more. As long as you maintain trigger to the waveform, the hum distortion should creep across the screen affecting the amplitude of the signal. Of course, make sure the lines are on the screen while you examine. Also these are high amplitude (and power) signals so be careful. Make sure your scope probe is set to 10 to 1 mode and not direct 1 to 1 so you don't exceed max amplitude input and damage your scope.

PS - there should be a max peak to peak rating written on your scope by the probe connector. That is the 1 to 1 figure. Some scopes are only 50V p-p. In that case, 10 to one sets max at 500V p-p.
Your probe itself might have a max p-p also.

Is your scope dual trace and do you have two probes?

Larry G

 

[pcdata76]

I have a dual trace(with add-subtract modes) 20Mhz analog scope and i have two probes (1x-10x each). Maximum input rating written on the scope near BNC probe connectors is 300V peak. It also has TV-H and TV-V trigger sync modes .

 

[retrogear]

Ok you have a good scope for this work. The crawl on screen is either entering via the video signal or the deflection. Since you have distortion on left and right vs center it must be coming in the deflection.
See if you can spot it.

 

[pcdata76]

That's the way it looks in theory but I'm not comfortable telling you to try that. D601 is only tied to one phase of the AC so not sure it would induce 100/120Hz noise but maybe only 50/60Hz.
Also puzzling is how that would create the east/west distortion. May want to wait for Brian's input on this. I have seen a defective degauss coil circuit induce a hum noise but again, that would be 50/60Hz.


Larry G

Before starting to scoping, I wanted to start from the simplest thing which you've recommended before and found the problem on degauss circuit!! After the disconnection of the connector of degauss coil while TV is operating with parasitic lines on screen, lines disappeared instantly. To be sure, i verified it with and without coil few more times and each time lines are disappeared without coil. Since i dont have any spare PTC, i couldnt verify that either PTC or degauss coil is defective but my guess is on PTC. Instead of minimizing the current to the negligible value after heat-up, it continues to provide some current to the coil and this magnetic field injecting noise to the deflection circuit. There is also a slight picture jitter throughout the screen while coil is in circuit. Tomorrow i'll try to find a replacement from local electronic store, but alternatively i can mount an external switch series with the coil and turn it on whenever i need degaussing.

Thank you again for all of your help.

 

[retrogear]

Well glad it's solved. I was actually thinking about recommending check the degauss circuit because I have seen that cause hum before but as I remember it and also theoretically it only produces one hum bar on the screen since AC current at that point is 50/60Hz. But, so much for theory ...

PS - I see in schematic the degauss coil is represented by two separate coils so maybe each coil produces it's own bar ... interesting

Larry G

 

[pcdata76]

Yes, there are two separate coils. One at the upper back side, one at the lower back side of the tube. Maybe this was the reason of two bars.
ps- i couldnt find a replacement PTC in the nearby electronic components store but i've replaced it with the PTC of a 14" PC monitor which was pending at storage. Bars are not there with coil connected, so i can say that PTC is the source of the fault. I'll buy from online store and replace it.

 

[karesz]

Hi pcdata,
only as some ideas;
For retrace lines; is G2-voltage setting correct?
For hum; maybe the rectifiers have problems with time-heath...
K.

 

[pcdata76]

Hi pcdata,
only as some ideas;
For retrace lines; is G2-voltage setting correct?
For hum; maybe the rectifiers have problems with time-heath...
K.

Hi Karesz,

I solved the problem. Btw, the first thing which i've checked was G2 voltage for retrace lines and it was correct. For hum, i would check bridge rectifier but found problem before on degauss coil PTC and solved it too. Thank you again for response.