EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: HwAoRrDk on August 09, 2022, 04:43:42 pm
-
I have an old Sony 21" CRT TV, model KV-21X5U, that I recently got out of storage because I want to use it with something that would benefit from being viewed on a CRT.
However, it worked apparently fine for about 30 minutes, then all of a sudden the picture cut out, and now it doesn't want to play ball any more. There was also a slight smell emanating, of the usual 'too-hot electronics' or 'capacitor spilled its guts' odour. :( Its behaviour now is as follows: when power is applied, there's the usual 'fizz' from the CRT, then about 4-5 'clicks', then it will start flashing the power LED in a regular pattern - 2 flashes, pause, repeat, and so forth - and do nothing else.
I found the service manual online, which has full schematics, etc. It also has a table of diagnostic error codes which correspond to the LED flashing. It says that two flashes mean "protection circuit trip".
Am I right in thinking that the 'protection circuit' is a feature found in CRT displays that has something to do with detecting when the high-voltage to the tube is excessively high, so that it does not start emitting x-rays? That is my dim recollection anyway; I don't know much about CRTs and their circuitry.
I'm wondering whether this is something worth pursuing for a repair, if the problem is something related to the HV parts of the TV. I'm not keen on dealing with any of that, given my lack of experience in such matters.
I opened the TV up anyway, just to get a look an see if there's anything immediately noticeable wrong, but I couldn't find anything. No bulging or leaking caps, no blown, charred or otherwise noticeably overheated components. Apart from everything being filthy dusty, everything looked perfect. I did notice there's a large relay on the board (which according to the schematic is used to switch mains power in when it comes out of standby), but I'm not sure if that was what was doing the clicking/ticking noise.
Any suggestions for narrowing down the cause of the problem? Schematic-only version of the service manual attached below if it helps.
-
Just a blind guess: there is always the chance that the flyback HV transformer is defective and leaks to the outside or arcs internally, try to localize the sourc of the 'clicks' or watch for discharges in darkened environment.
Also check the HV capacitors C542/3/8 C553/5 for cracks.
Protection circuit can mean overcurrent protection in the power supply or X-ray protection from excessive HV.
-
HVT +1 unobtainium???
-
Just a blind guess: there is always the chance that the flyback HV transformer is defective and leaks to the outside or arcs internally, try to localize the sourc of the 'clicks' or watch for discharges in darkened environment.
I tried powering-on a few times again, and the rapid clicks are not coming from the relay (although that does clunk on, then off right as the clicks stop). I couldn't narrow it down too much - don't feel like putting my head right in close to the HV stuff! :scared: - but it seems like the clicks are coming from around the large transformer T603.
Also check the HV capacitors C542/3/8 C553/5 for cracks.
I checked them, they all look fine. They're nearly all through-hole dipped epoxy-package parts, and I couldn't see any cracks in the shell. C553 doesn't look like any capacitor I've ever seen, though. It's a tall, thin glass-like looking package, and it's right up against the heatsink for Q533, touching, as if it were some kind of temp sensor. Whatever kind of cap it is, it looks fine too. Edit: on the schematic, it's labelled as 'PT', and there's a legend on another page which says that means a mylar cap.
I think tomorrow I might check the voltage on the +9V and +5V rails from IC604 & 605. I would assume if there's some kind of problem with T603 shorting, its output would be compromised? I already checked the 5V standby output (at IC608), and that's fine.
-
You can also do the good old bulb test: disconnect the FBT from the B+ voltage and load the B+ rail with a 40W bulb.
If it glows, the power supply is delivering and the failure must be in the FBT/control or deflection.
With too much uC intelligence, such modifications also might result in a dagnosis failure and shutdown.
A good idea is to watch the power consumption of the set for irregularities/overload.
-
I measured some more voltages today. The main +5V (at IC604), +9V (at IC605) and +B (at R571). They don't look at all good: +5 = 0.2V, +9 = 0.7V, and +B fluctuating around 6-7V. The input to the 9V regulator was also only 1.7V. All the voltage rails were slow to rise; these figures are the maximum I saw before the main relay clicked off (and the error code starts flashing).
If the problem is a shorting transformer, then I don't fully understand how the main microcontroller is detecting that something is wrong. Although, from looking at the schematic, it seems perhaps it can possibly determine the status of +B through Q571/Q574/Q011 which lead to the micro pin labelled "XRAY PROT", which I guess by the name does tie in with a protection circuit fault.
You can also do the good old bulb test: disconnect the FBT from the B+ voltage and load the B+ rail with a 40W bulb.
I don't actually have any incandescent bulbs in the house! ;D
-
Did you measure the voltages when the set tries to power up or after it has shut down?
Do you have a powermeter to check the power consumption?
Unplug the degauss coil to see whether the set draws excessive current at startup.
Disconnect the supply rails one by one to find out if one of them is overloading the power supply.
-
No further efforts to repair the set?
-
Sorry, I haven't had the time to look at it recently. I plan to tackle it again over the weekend.
By the way, those voltage I previously measured were when it tries to power up, before it shuts down (or, rather goes into standby error state).
By power meter, do you mean mains power? I have a Kill-A-Watt-style mains power meter, but I doubt it'd be too useful. The display on that only updates once per second, and the whole power-on-to-error-shutdown period only takes about 3 seconds, so it'll be "blink and miss it".
I don't know whether I can disconnect any of the supply rails - that'd mean starting de-soldering stuff off the board.
-
I am confused, first I thought the 'fizz from the CRT' was the HV building up, but that's not possible with a B+ voltage of only 6…7V.
Is the fizz the degauss coil? Please disconnect it as already suggested.
Yes, you will have to engage the soldering iron to isolate the assumed overloaded rail.
-
I assume the degaussing coil is connected to the connector marked "DGC" (CN602) on the schematic? I just looked, and that connector does indeed go to a couple of loops of cable around the tube. I will try with and without that and try to measure the current draw.
Although... looking at the schematic, the degaussing coil seems to be directly across the AC mains input (with current limited by a PTC thermistor), before main transformer and even bridge rectifier. So, if that was drawing too much current for some reason, wouldn't it blow the main fuse?
-
The idea is to eliminate any additional current paths and noise sources to only concentrate on the electroncs.
-
I tried measuring the mains current draw using my plug-in watt-meter. With the degaussing coil still connected, it drew about 0.6-0.7A (@ 240V) peak during the repeated 'clicking' noises. Then I unplugged the DGC from the board and repeated. The current draw was essentially the same - around 0.6A.
After repeating this a few times to verify and getting the same results, I had a quick look across the board to see where the PTC thermistor (THP601) that is part of the DGC circuit is. I found it, and discovered that it was very hot. Almost too hot to touch! Is it supposed be doing that? I also really can't comprehend why it is a 3-pin device if it's for some kind of current-limiting.
-
Yes, thats the normal behaviour, the degaussing will only occur during the heating phase.
The PTC needs to cool down for several minutes to re-start the degaussing.
What influence to the fizz did you observe?
-
Yes, thats the normal behaviour, the degaussing will only occur during the heating phase.
The PTC needs to cool down for several minutes to re-start the degaussing.
What influence to the fizz did you observe?
Oh, I see. I guess the repeated on-off in a short period made it much hotter than normal. :-[
The 'fizz' didn't occur with the degaussing coil unplugged.
-
Yes, thats the normal behaviour, the degaussing will only occur during the heating phase.
The PTC needs to cool down for several minutes to re-start the degaussing.
What influence to the fizz did you observe?
Oh, I see. I guess the repeated on-off in a short period made it much hotter than normal. :-[
The 'fizz' didn't occur with the degaussing coil unplugged.
No, the PTC was not hotter than normal. It needs to be hot to have high enough resistance to reduce the current through the degauss coil.
3 leads because sometimes the component is a combination of a PTC and NTC in series. The coil is in parallel with the NTC.
-
It looks like x-ray protection in this set is triggered by excessive B+, rather than excessive HV, so a power supply regulation problem would be my first suspect. Unless you're using an oscilloscope to check the voltages, I doubt you can get an accurate reading before it shuts down since it shuts down the instant it detects excessive voltage. IC603 could be bad.
You could remove the horizontal output transistor to test it, and also to reduce the risk of damage while you're working on it. If it's not a false protection trigger, then every time you turn it on with excessive HV, you risk causing serious damage.
With the HV disabled (by removing the HOT), you could try disabling the protection circuit to give you a chance to check the power supply voltages. Pull the plug quickly if it goes much over 135v. You may need a dummy load on the B+ to get accurate readings, but not usually.
Some sets will also shut down for other reasons such as loss of deflection, or excessive beam current, but not often on a small direct view set like this.
-
We do not have an indication that B+ is present anyway, I still suspect that the power supply is shutting down due to overload.
Please test whether the HOT Q533 is still OK by measuring B-C and B-E with R549 disconnected.
If the HOT has failed there is a good chance that this is a secondary effect, either due to poor control signal or failure of the FBT or damping circuit.
-
Please test whether the HOT Q533 is still OK by measuring B-C and B-E with R549 disconnected.
I de-soldered one leg of R549 and measured Q533:
B-C: 0.32V
B-E: <0.01V
E-B: <0.01V
C-B: 0.39V
C-E: 0.38V
Q533 is either a 2SD2539 (according to schematic) or 2055N (according to parts list), and from datasheets I was able to find of either, it has built-in B-E pull-down resistor, as well as a C-E "damper" (snubber?) diode. I'm guessing this will complicate measurements. However, despite the B-E resistor, I don't think it should measure effectively short-circuit B-E/E-B on diode-check, nor 9 ohms on resistance. According to one datasheet, B-E resistor should be approx. 50 ohms.
So, I'm guessing this transistor has failed?
IC603 could be bad.
Is there any way of testing this IC?
Datasheet describes it as a 135V "error amplifier". I assume its job is to drive the optocoupler to give feedback to the switch-mode controller IC606, so that 135V is maintained on +B.
-
C-E anything below infinity is generally bad if it is the transistor alone.
Desolder the transistor at the emitter and see whether B+ comes up.
-
C-E anything below infinity is generally bad if it is the transistor alone.
Okay, good, that's what I thought - that C-B and C-E should read 'OL' on the meter.
Desolder the transistor at the emitter and see whether B+ comes up.
I'll give that a go. Going to be tricky - will probably have to de-solder the whole part, remove it, bend the emitter lead, then re-solder.
So, one thing I'm confused about: should B+ be 135V, or 200V? I'm sure I saw a label against it on the schematic that said "200V", but I may be mistaken.
-
You're joking
Either the hole is big enough so that the pin can be isolated by sucking off the solder or you remove the entire part.
I'd expect 135V for the B+, any hints on that in the manual?
-
FE1 Chassis fault list if it helps
FE KV25X5U Int no pic with LED flashing 4 times O/C print from pin 14 of IC301 to first link wire.
FE1 Buzzing only on sound from tuner - ok on AV Incorrect sound system selected, the cause being memory corruption. To correct the problem, enter test mode to obtain "TT - -" on screen, & type in "24". This will reset sound system back to PAL I. Turn set off & re-tune, & sound should then return to normal.
FE1 Dark pic The ABL circuit is operating incorrectly. Replace R516 (56K) & R517 (10K or 27K depending on CRT)
FE1 Dead except for st/by LED flashing twice oscillator STR & phototransistor 123 & FOP chip STV
FE1 Int blows LOPTr check line dr tx but was dry joints jungle i/c pins IC301this time
FE1 No power (no standby power supply o/p) Replace IC609 (TOP209) also check C640 (47µF, 25v), D631 (MTZJ-12B), D628 (P6KE200), D629 (µF4005PKG23) and R627 (4.7?,)
FE1 Reverts to st/by at sw on with 2 or 4 flashes from LED 2 flashes indicates an over-current error in line stage, & 4 indicates a frame fault. However, quite often, when 4 flashes occur it can still in fact be an over-current error as well as a frame fault. To prove which is the correct fault, measure pin 52 of the micro (IC001) at switch-on, and if this voltage goes high (anything above 0.6v) then it is in fact over-current. If it stays at 0v then the fault is more than likely within frame stage.
FE1 Set shuts down shortly after start up with standby led flashing 2, 4 or 5 times Check SMD resistors R543 (47K) and R541, R573 and R575 (all 100k)
FE1 Snowy pic & tone on sound IC001 micro faulty
FE1 After about 30 minutes the set goes to standby, the LED flashes and TV starts again. After 1 Minute the TV Displays : Antenna connection failed! Change IC101 of IF on A-Board. Order Nr. 875946647
FE1 Band of interference down left of screen on dark scenes This was cured by adding a 1000pF, 2kV capacitor (part no. 1-161-731-81) in parallel with D539 in the EW diode modulator circuit
FE1 Excessive width Q532 (2SK2251-01) s/c, it is advisable to also check other EW circuit components as a precaution
FE1 Intermittently displays bright red/green/blue raster then shuts down with led flashing 2, 4 or 5 times This was due to the CRT being faulty
FE1 Loss of one colour with teletext and on-screen displays Check the appropriate 2SC2412K transistor Q007 (green), Q008 (blue) or Q009 (red)
FE1 Main psu starts up then almost immediately shuts down This was due to IC603 (SE135N) being faulty
FE1 No picture or dark picture Check R516 (56K, 0.5w) or R517 (27, 0.25w) for high resistance or o/c
FE1 No power although relay operates (main psu inoperative) Check IC606 (STRF6654) for failure, also check resistors R659 (0.33?,3w), R660 (0.39?,3w), fusible resistor R603 (0.1?,0.5w) and capacitor C625 (100µF,25v)
FE1 No remote control operation Check as necessary D203 (MTZJ-5.6B) on main pcb, D907 (MTZJ-9.1A) on pcb H1 or the remote control sensor IC900 (SBX1981-51), part no. 8-742-014-11
FE1 No sound This may be remedied by entering TT24 in the service mode, if fault persists Q202 (2SC2412K) could be leaky
FE1 No teletext just P100 This problem was caused by Q405 (2SC2412K) being s/c
FE1 Pressing the screen format button displays wide, smart, zoom and 4:3 modes but settings have no effect on picture Enter TT35 in the service mode to select a 4:3 CRT
FE1 Purple line down picture Replace IC301 (CXA2060AS), part no. 8-752-082-35
FE1 KV21X5U burst of EHT then shuts down with LED flashing error 2 regulator IC603 SE135N (874992061)
FE1 KV21X5U pic closing in from sides and going off (poor HT regulation) error amp IC603 SE135N (874992061)
FE1 KV25K5U Dead with LED blinks 4x - excess current prot mode LOPTx T511 1145326411 shorted turns
FE1 KV25K5U Dead with LED flashing 4 times Check for 9v at pins 33 & 34 IC301. If present, check R546 47R 1/2w for o/c (135v Line dr supply)
FE1 KV25K5U Error code - 2 flashes 2SD2539 LOPTr leaky & dry joint on line drive transformer. Orig Tr = 10UKP - used BU2520DF in lieu
FE1 KV25K5U lines across screen, frame IC running hot C506 0.01 uF
FE1 KV25K5U No sound or vision, front led flashing Replaced line output transistor, resoldered dry joints on driver transformer. Also check for any other poor joints in this section. Have used a 2SD1878 as replacement, This seems to work ok as the Sony original is very expensive
FE1 KV25K5U not starting up, LED flashes 4 times T511 breaking down at switch on, replace T511 (145326411 for 25", 145326511 for 29")
FE1 KV25K5U St/by only with error code 2x flashing In prot mode due to LOPTr Q533 2SD2539 s/c due to dry joints line driver Tx
FE1 KV25K5U Trips off with a crack - LED flashes 5x CRT faulty
FE1 KV25R2D Reverts to st/by with LED flashing x2 in 2 secs SE135 +B regulator diode leaky, also caused demise of FOP chip STV9379
FE1 KV25X5U comes on then goes off, red LED is on stby IC609 TOP209P u/c
FE1 KV25X5U No frame Lock on OSD Menus and teletext SM cap next to FOP chip had crack across its connection
FE1 KV25X5U Service Mode While the set is in standby Press OSD (i+) then 5 then VOL+ then TV ([])
FE1 KV25X5U To reset EEPROM select chan 59, enter service mode, at TT-- prompt, enter 49 then 24 (for UK) & 08 then sw off at mains
FE1 KV25X5U Tripping with LOPTr 2SD2539 s/c Dry joints line dr trfmr - as orig LOPTr costs £20 - used 2SC4769
FE1 KV25X5U yellow line down screen about 2 inches from RHS IC301 CXA2060BS (875208838)
FE1 KV25X5U etc Enters search mode as soon as its switched on "replace SS902 preset search switch and clean PCB that it is
mounted in"
FE1 KV29K5U dead, in stby C515 220uF 25V s/c
FE1 KV29K5U no remote operation D907 BZX79C9V1 leaky on SIRCS line
FE1 KV29K5U pink band on LHS, only obvious on dark bachground video lead from CN504 on A PCB to CN704 on CRT wrapped around L537 (LINE LIN), tie bottom edge of loom to the hole provided in top right corner of S PCB
FE1 KV29X5U Dead with error code 2 flashing LOPTx HR8520 - use only the top focus control supplied - check for drys on jungle chip
FE1 KV29X5U Dead with error code 2 flashing & no HT R613 1R ceramic o/c & STK HT reg s/c - replacement cured - check for drys on jungle chip
FE1 KV29X5U excessive blue then goes to standby R722 680k on crt base goes high
FE1 KV29X5U Int pic - sound ok or nothing with LED flashing 4 times Dry joint line dr Trfmr T531 (under plastic cradle)
FE1 KV29X5U no remote, front buttons are Ok D908 9.1V zener leaky on H board
FE1 KV29X5U No sigs - ok on scart & OSD OK IC101 TDA9817 (VIF/SIF Demodulator)
FE1 KV29X5U Vol OSD perm displayed & front controls give wrong functs Leaky front control switches - can be cleaned with luck/care
-
I was wrong about the B+ being labelled as "200V" on the schematic. I simply mis-interpreted the schematic - that label was referring to another nearby line.
Anyway, I removed Q533 completely, and then tested while measuring B+. Turning the power on, I no longer get the clicking sound, and B+ measured a solid 135V! :D There was no flashing from the standby LED either.
So, is it likely that this was the sole fault? I can buy a replacement S2055N or 2SD2539 for just a few pounds, so I'm willing to try just putting a new one and seeing if that fixes it.
-
FE1 KV25K5U Error code - 2 flashes 2SD2539 LOPTr leaky & dry joint on line drive transformer. Orig Tr = 10UKP - used BU2520DF in lieu
Thanks, looks like BU2520DF has similar specs and may be another candidate transistor for replacement if I can't get hold of either of the others.
-
The most universal type is the S2000, but you will be more safe keeping to the original type.
Its a matter of will for tinkering and price/availability as well.
BUT its also worth finding out why the HOT had failed as described already before.
Better buy three transistors at once as they will immediately tell you that you forgot checking something...
-
Oh yes, I will buy more than one, just in case. :)
Is it worth checking the horizontal drive transistor (on the other side of the transformer T531), Q535?
What kind of frequency will the horizontal deflection drive signal out of IC301 (the 'jungle' chip) be? Just wondering if I'll be able to scope it using my battery-powered (and thus isolated) 'toy' oscilloscope. It's only good for <100kHz signals. :( The schematic shows what kind of waveform and p-p voltage to expect, but not what frequency/period it should be.
Edit: I de-soldered Q535 and measured it. It's a 2SC2688 NPN. It measured as follows:
B-C: 0.6V
B-E: 0.61V
E-B: OL
C-B: OL
C-E: OL
That all looks good I think, probably nothing wrong with it, so I put it back in.
-
What kind of frequency will the horizontal deflection drive signal out of IC301 (the 'jungle' chip) be?
The drive signal will be horizontal sweep rate. 15.750 kHz for NTSC.
-
When you have the new transistor and solder it in, leave the collector open or interrupt the B+ somewhere else.
Then scope the base signal, compare it to waveform 10 in the service manual.
Then check everything upstream of the collector up to the B+, all the f... diodes and capacitors.
After that you are ready to sacrifice the first HOT.
There is still a chance the FBT is broken.
-
Some Sony sets often had cracked solder connections on the horizontal drive transformer. This would intermittently cause the output transistor to fail. It's a good idea to resolder it just in case.
-
The drive signal will be horizontal sweep rate. 15.750 kHz for NTSC.
This is a PAL set. Although... I suppose that'd only make a difference for the vertical sweep rate? :-//
Then scope the base signal, compare it to waveform 10 in the service manual.
Just wondering, will the form of the base signal depend on whether there is any (or what) video signal being displayed?
Some Sony sets often had cracked solder connections on the horizontal drive transformer. This would intermittently cause the output transistor to fail. It's a good idea to resolder it just in case.
Thanks for the tip, I will inspect those solder joints closely. In fact, will give the whole board a good look over.
-
The drive signal will be horizontal sweep rate. 15.750 kHz for NTSC.
This is a PAL set. Although... I suppose that'd only make a difference for the vertical sweep rate? :-//
. 625/50 horizontal rate is only slightly different. 15.625 kHz. 64 microseconds period.
-
Hello.
Please check bad solder joint on FBT and C531 and C545 for ESR.
-
When you have the new transistor and solder it in, leave the collector open or interrupt the B+ somewhere else.
Then scope the base signal, compare it to waveform 10 in the service manual.
I got a couple of new S2055N transistors. I took the precaution of measuring the new ones too - when measuring B-E/E-B on diode test the meter would still beep for continuity, but measuring resistance they were 40 ohms, which is close to what the built-in B-E resistor is supposed to be (50 ohms), and everything else checks out, so they seem to be good.
Bent the collector pin up and out of the way and soldered one in. Probed the base pin with my scope, as well as measuring the B+. The B+ was a solid 135V, and I got a regular-looking waveform of approx. 15.67kHz on the 'scope (see attached image). Voltage seems a bit low compared to what the service manual says, though. I got 4.7V p-p, and it says 12V.
No magic smoke released or anything, so it seems to be happy so far. Interestingly, about 10 seconds after power-on, it started flashing another, different error code on the LED. This time it was 4 flashes, which apparently means "no vertical sync" according to the service manual.
I gave the board a good look over for bad solder joints. Looking very closely at the horizontal drive transformer (T531), it seemed like there might be a very small hairline crack around one of the pins, but it was very hard to tell. I re-soldered all the joints on it anyway as a precaution.
Anyway, that's all I had time for this evening. Tomorrow I'll solder the transistor back in fully and see how it goes. :)
-
Soldered the transistor back in completely, and... it works!! :D
Picture looks good, no distortion, B+ is still 135V, 'scope shows an actual proper 15.625kHz for horizontal drive (not sure why it was slightly fast previously), mains power consumption is approx. 0.5A (although I haven't hooked the speakers back up yet).
I did have to scrounge around for a while looking for some fresh thermal grease when properly re-fitting the transistor, but managed to find some in a box of old computer stuff.
Thank you everyone for your help, particularly inse. :-+ :clap:
-
Triple thumbs up!!!