Samsung SyncMaster 940T blows inverter fuse

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

The monitor should power on by itself when power is applied,Do as PlainBill suggests use a power strip to power on.
And in a dark room look at the ccfl transformers closely.
Al

The power supply is always-on when the power cable is plugged in and the power switch is flipped on, but the panel and inverter don't activate until I push the front-panel power button. I'm using that to stop things also, and I try to have someone near the power outlet if I'm doing something that seems particularly risky.

I'll have to get someone to help me with this. I'm more interested in not destroying anything than I am in figuring out if I saw a little bit of light from somewhere. If this is causing an arc in a transformer, there should be a safer way to test for it (like measuring the switching frequency or duty cycle). I'm only theorizing that this could be happening based on the current increase of 400mA - if there is something switching power to the transformer on and off, this behavior is likely regulating the amount of current going to the transformer (via duty cycle or switching frequency, I think) as it can't change the voltage. That signal would be coming from the inverter control IC (U201), which would cause the FETs (U202 and U203, labeled 4502C) to switch power from the power supply to the transformers on and off. I would guess, based on this, that either the inverter IC's switching behavior is changing in a bad way, or one/both of the FETs is damaged. Since the thing works at all, I don't think one of the FETs is shorted but I can check that tonight. I think I know what part that is, so I might be able to hook my oscilloscope to whatever pin attaches to the inverter IC and watch the wave. Doing the same to the output pins might be useful also.

Also I have never used a ring tester,Do the tests you've done indicate any thing about the coils?
Al.

The ring tester indicates if any turns in a coil are shorted together. I'm not sure how well these work on the inverter transformers, as the wire gauge is tiny and there are spacers every so often that may impact the ringing behavior (I have no idea). Other inverter transformers I've tested that work fine light 3 LEDs on the high voltage side and none on the low voltage side, which is about what I'm seeing now. As I understand things, a few shorted turns would reduce the resistance of the transformer and would also reduce the output voltage a bit. I already checked the resistance, and I'd need to have some idea of what to look for in the voltage divider on the return from the CCFLs as I don't have any tools that can safely measure 650v (high voltage rating).

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

I checked the low voltage side of the inverter transformers with my oscilloscope and didn't see any noticeable variation in the RMS voltage or frequency on either transformer while current through the inverter was increasing. I also took a look at the return wires from the CCFLs, and they appear to be "or" circuits using the return voltages from the CCFLs attached to each transformer. They're connected together through a couple of resistors later, then that goes back to the control IC through a couple of other surface-mount components. I measured the voltage at the end of R345 after everything comes together, and the voltage and frequency measured there didn't change over time as current through the inverter increased (picture attached). Any other thoughts on what I could check?

[PCBONEZ]

Re: Samsung SyncMaster 940T blows inverter fuse

Wild guess: Did you check the optocoupler?

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

Wild guess: Did you check the optocoupler?

I didn't, as the optocoupler is part of the power supply circuit and I used my adjustable power supply to run the inverter and the problem persisted. If the inverter IC is using the 5v output of the power supply then it's possible that something is wrong there (I haven't tested that yet), though wouldn't watching the 5v output on my oscilloscope for ripple or changes in voltage accomplish about the same thing?

[PCBONEZ]

Re: Samsung SyncMaster 940T blows inverter fuse

You've been chipping at this for a while and I'm working on more than one thing so beg pardon if I get lost now and then.

You cited changing load current and no change in power.
The optocoupler is the feedback to the PSU section from the load and I didn't know you were using an off board PSU.

.

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

You've been chipping at this for a while and I'm working on more than one thing so beg pardon if I get lost now and then.

You cited changing load current and no change in power.
The optocoupler is the feedback to the PSU section from the load and I didn't know you were using an off board PSU.

.

Don't worry about it. I actually have 4 or 5 more similar projects I'm going to need help with (rebuilding the PSU from a switch, a monitor with a white screen (probably dead panel but worth asking about), some PC PSUs that have regulation problems, stuff like that) but I'm trying to stick to one project at a time so I don't get confused

After the recap the ripple in the 13v output from the power supply has been reduced to something like 800mv - 1v. With the inverter on, it actually runs at around 12.5v while the rating on the board says 13v. I used a bench power supply at a couple of voltages and had the same problem every time. I have so far not looked at the 5v output at all, so there may be a lingering problem there causing issues in the inverter IC; the optocoupler should be a relatively easy check (I think) so I can do that tonight. I can also watch the wave on my oscilloscope for oddness.

At this point I'm running out of things to check. I haven't yet checked the diodes on the returns from the CCFLs (likely BAV99 I think, but I can't read a part number off them), but I'll have to do it in-circuit as I don't have a good way to detatch and re-attach surface-mount components reliably. Beyond that I'm out of ideas. I was thinking about attempting to draw a schematic based on visually inspecting the traces on the bottom of the board, but that'll be a bit tricky as Samsung hides some traces underneath components where I can't see them. I might have to get some Chip Quik and desolder the inverter IC to get a good idea of where the traces hiding underneath it go. I also won't know the ratings on the SMD parts as I don't have a parts list, so I'll have no idea if they're still good. I'm starting to wonder if a pair of the "smart tweezers" might be worth the price for this kind of stuff.

[PCBONEZ]

Re: Samsung SyncMaster 940T blows inverter fuse

Might be a good time to wake PlainBill up from his nap.
He knows enough off the top of his head to write a book on these things.
On monitors I'm a novice.

[PlainBill]

Re: Samsung SyncMaster 940T blows inverter fuse

Might be a good time to wake PlainBill up from his nap.
He knows enough off the top of his head to write a book on these things.
On monitors I'm a novice.

HEY!!! I wasn't napping. I was checking the inside of my eyelids for cracks. (I found a couple.)

As far as the 940T, the two ideas that come to mind are 'thermal runaway' and 'magnetic saturation'. Now that those popped into my head, I have to see if they could be related to the problem.

Well, it looks like magnetic saturation may be out. J304 and J307 provide +14V to the two dual FET switches. The other 3 jumpers (J309, J308, and J306?) provide ground. C231 and C232 provide DC decoupling for one end of the primary windings. Unless either of the caps is leaky there is no way magnetic saturation could be an issue.

But thermal runaway is definitely a possibility. When the current starts to rise do either U202 or U203 get warm - or hot? Another thing to try is lifting J304 and inserting an ammeter. Then compare the current to the current flowing through J307. What you hope to see is the current through one of them rising.

PlainBill

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

HEY!!! I wasn't napping. I was checking the inside of my eyelids for cracks. (I found a couple.)

As far as the 940T, the two ideas that come to mind are 'thermal runaway' and 'magnetic saturation'. Now that those popped into my head, I have to see if they could be related to the problem.

Well, it looks like magnetic saturation may be out. J304 and J307 provide +14V to the two dual FET switches. The other 3 jumpers (J309, J308, and J306?) provide ground. C231 and C232 provide DC decoupling for one end of the primary windings. Unless either of the caps is leaky there is no way magnetic saturation could be an issue.

But thermal runaway is definitely a possibility. When the current starts to rise do either U202 or U203 get warm - or hot? Another thing to try is lifting J304 and inserting an ammeter. Then compare the current to the current flowing through J307. What you hope to see is the current through one of them rising.

PlainBill

Well, this was interesting. I pulled both jumpers and put multimeters on both. I have 3 multimeters total, one on each jumper and one across the fuse. I turn on the power and... it behaves just fine. Current steadily decreases for a while (stopped around 1.7A I think). After about 5-10 minutes, current starts going up again and I turn the monitor off at around 2.2A. I checked the transistors, and they were both really hot (like 80 degrees C hot, at least).

If I take my digital multimeter off and leave my analog meter on the other jumper, current through the inverter rises slowly while current through the transistor with the meter attached decreases slowly. Same thing happens if I pull my analog meter instead, only the current change is faster. I swapped meters, and the same thing happened (speed of current change follows the meter, not the jumper). Does this tell you anything useful?

[PlainBill]

Re: Samsung SyncMaster 940T blows inverter fuse

Well, this was interesting. I pulled both jumpers and put multimeters on both. I have 3 multimeters total, one on each jumper and one across the fuse. I turn on the power and... it behaves just fine. Current steadily decreases for a while (stopped around 1.7A I think). After about 5-10 minutes, current starts going up again and I turn the monitor off at around 2.2A. I checked the transistors, and they were both really hot (like 80 degrees C hot, at least).

If I take my digital multimeter off and leave my analog meter on the other jumper, current through the inverter rises slowly while current through the transistor with the meter attached decreases slowly. Same thing happens if I pull my analog meter instead, only the current change is faster. I swapped meters, and the same thing happened (speed of current change follows the meter, not the jumper). Does this tell you anything useful?

Another buzzword - parasitic oscillations? Adding the meter in the supply line changes the behavior?

I started to reread the thread and got bogged down on all the discussion on the replacement caps, so you will have to summarize for me.

Was C345 replaced? What was the replacement? (brand and series?) Recheck the soldering, make sure there are no solder bridges.

What were the other caps on the 12? 14? volt supply replaced with? (brand and series)

Were the caps on the 5V supply replaced?

PlainBill

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

Another buzzword - parasitic oscillations? Adding the meter in the supply line changes the behavior?

I started to reread the thread and got bogged down on all the discussion on the replacement caps, so you will have to summarize for me.

Was C345 replaced? What was the replacement? (brand and series?) Recheck the soldering, make sure there are no solder bridges.

What were the other caps on the 12? 14? volt supply replaced with? (brand and series)

Were the caps on the 5V supply replaced?

PlainBill

Adding the meter changes the behavior until stuff gets hot. Then it does the same old thing. I don't know exactly how hot the chips get, as I can't find the thermometer attachment for my multimeter right now and I'm using it to watch the inverter's input current anyway.

The only work I did was adding a touch of solder to a bunch of points including all transformers and the pins of the CCFL connectors, I turned C106 around to stop the squealing, I replaced C112, C111, C301 with Nichicon HE 680uF 25v instead of CapXon GL 820uF 25v, and I replaced C109 and C110 with Nichicon HE 1000uF 10v instead of CapXon KM 1000uF 10v. I tested a few other power supply components out of circuit, but I put them back (D102, R102).

As far as things I could have but didn't do, I have some Panasonic FK 820uF 25v caps on backorder and I just ordered a pair of 2200uF caps (Panasonic FC 10v, Panasonic FK 25v) for use on another one of the 4 of these monitors that I have. I did not replace anything blue, C107, C123, C113, C345. I have some Panasonic NHG and Rubycon ZL 330uF 25v that I could use for C113 (CapXon KM 330uF 10v), and I just ordered some Rubycon YXF 47uF 50v that might work for C107 (CapXon KM 47uF 50v) and C345 (CapXon ST 33uF 25v). From what I can tell, the CapXon ST series are standard type 5mm caps and since there's plenty of space on the board in that area, a larger cap wouldn't be a problem. 47uF is a lot more than 20% of 33uF though, and I have no idea what that capacitor does. I also have not looked at U301.

I've been over the soldering a few times, though I can take a brush and some MEK or alcohol to the board again to try to break any little bridges that might exist. I used a bench power supply in place of the fuse and the inverter still misbehaves, suggesting that the power supply (13v line, at least) is fine. The supply isn't the best regulated thing ever, idling at 14.8v, jumping up to ~18v when the power-on cycle starts, then it drops to ~12.5v when the backlight switches on. I powered the inverter with 14v and 13v, with no change in behavior. There's also some ripple while the backlight is on, and I think the frequency is about the same as the inverter's switching frequency (~94KHz).

[PlainBill]

Re: Samsung SyncMaster 940T blows inverter fuse

My recommendation is a 'shotgun' approach. If you haven't already done so, replace every capacitor in the 14V supply with good quality low ESR types such as Panasonic FM or FC, or equivalent Nichicon or United Chemicon types. Also replace the electrolytic caps in the inverter area with as low an ESR cap as you can find. Then power it up and observe the current draw. If it starts to rise, replace both dual FETs. My first suspicion is that too high an ESR is allowing oscillations on the power supply line, resulting in the FETs not switching fast enough. The second possibility is that one of the FETs is failing to shut off properly.

PlainBill

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

My recommendation is a 'shotgun' approach. If you haven't already done so, replace every capacitor in the 14V supply with good quality low ESR types such as Panasonic FM or FC, or equivalent Nichicon or United Chemicon types. Also replace the electrolytic caps in the inverter area with as low an ESR cap as you can find. Then power it up and observe the current draw. If it starts to rise, replace both dual FETs. My first suspicion is that too high an ESR is allowing oscillations on the power supply line, resulting in the FETs not switching fast enough. The second possibility is that one of the FETs is failing to shut off properly.

PlainBill

This sounds like it's going to get expensive. I hope this works As I said, I replaced the big 5 already with Nichicon HE. I have Rubycon ZL for C113, and I have Rubycon YXF on the way for C107. I'm not sure what C123 is offhand, it might be a little 0.47uF cap or something, and I don't have a replacement handy for C345 (33uF 25v). Can I use a 47uF cap for C345, or should I go for exact replacements with the lowest possible ESR? I've passed up really low ESR caps in a couple of places (picked FC instead of FM when FM was cheaper, for example), aiming for matching ESR instead. I have 1000uF 10v Rubycon MCZ that I could have used for the 5v filter, but I was advised that they are too low ESR.

What about all the blue caps? I haven't replaced any of them, and I'm not even sure what type of cap they are. My multimeter will measure hundreds of nF capacitance, so I can check some of them against their rating, but I don't have an ESR meter to check them with. Also, what about the primary filter cap? I haven't replaced one of those before (I think this one is a CapXon KM).

The FETs are labeled 4502C, which might be AF4502C from AnaChip as the date codes are of the format listed in the datasheet (attached to an earlier post, on page 4 I think), but the ones I have don't have the AnaChip logo that the datasheet says should be present. 4502C also matches an SPST (I think) from Maxim, but I'm pretty sure that's wrong. The AF4502C is a P & N channel 30v MOSFET, which isn't a type of part I've seen before. I don't see these on Newark, Mouser, DigiKey, eBay, so I'm not sure where I'd get one. I can shop around for a similar part, though I may need a little help.

I may also need some suggestions on how to do the replacement as I don't have a rework station. I can try hot air with my Weller Portasol, or something like Chip Quik which I can order from DigiKey.

[lucky13]

Re: Samsung SyncMaster 940T blows inverter fuse

The inverter would not light only a single CCFL. I used a 19" wide CCFL along with the other lamp from the panel, and tried all 4 connectors. As long as I plugged in two lamps to each transformer, it was happy. All 4 connectors resulted in a nice, bright light. With just one lamp attached, the lamp did not light all the way across, was dim, and the inverter hissed a little.

WAIT. How long did you run your test CCFLs here? If they all run good w/o blowing the fuse, then your problem is your lamp!

BTW, for this model, I use Digikey P12391-ND, 820uF/25V 10*25.

Good luck.

[jetadm123]

Re: Samsung SyncMaster 940T blows inverter fuse

The FETs are labeled 4502C, which might be AF4502C from AnaChip as the date codes are of the format listed in the datasheet (attached to an earlier post, on page 4 I think), but the ones I have don't have the AnaChip logo that the datasheet says should be present. 4502C also matches an SPST (I think) from Maxim, but I'm pretty sure that's wrong. The AF4502C is a P & N channel 30v MOSFET, which isn't a type of part I've seen before. I don't see these on Newark, Mouser, DigiKey, eBay, so I'm not sure where I'd get one. I can shop around for a similar part, though I may need a little help.

I may also need some suggestions on how to do the replacement as I don't have a rework station. I can try hot air with my Weller Portasol, or something like Chip Quik which I can order from DigiKey.

The dual FET you're talking about might be a AM4502C, which is also difficult to find. What is the number of the CCFL controller chip that drives these FET's?

[PlainBill]

Re: Samsung SyncMaster 940T blows inverter fuse

This sounds like it's going to get expensive. I hope this works As I said, I replaced the big 5 already with Nichicon HE. I have Rubycon ZL for C113, and I have Rubycon YXF on the way for C107. I'm not sure what C123 is offhand, it might be a little 0.47uF cap or something, and I don't have a replacement handy for C345 (33uF 25v). Can I use a 47uF cap for C345, or should I go for exact replacements with the lowest possible ESR? I've passed up really low ESR caps in a couple of places (picked FC instead of FM when FM was cheaper, for example), aiming for matching ESR instead. I have 1000uF 10v Rubycon MCZ that I could have used for the 5v filter, but I was advised that they are too low ESR.

What about all the blue caps? I haven't replaced any of them, and I'm not even sure what type of cap they are. My multimeter will measure hundreds of nF capacitance, so I can check some of them against their rating, but I don't have an ESR meter to check them with. Also, what about the primary filter cap? I haven't replaced one of those before (I think this one is a CapXon KM).

The FETs are labeled 4502C, which might be AF4502C from AnaChip as the date codes are of the format listed in the datasheet (attached to an earlier post, on page 4 I think), but the ones I have don't have the AnaChip logo that the datasheet says should be present. 4502C also matches an SPST (I think) from Maxim, but I'm pretty sure that's wrong. The AF4502C is a P & N channel 30v MOSFET, which isn't a type of part I've seen before. I don't see these on Newark, Mouser, DigiKey, eBay, so I'm not sure where I'd get one. I can shop around for a similar part, though I may need a little help.

I may also need some suggestions on how to do the replacement as I don't have a rework station. I can try hot air with my Weller Portasol, or something like Chip Quik which I can order from DigiKey.

Let's go with the easiest first. Short of a hot air rework station, the best way to replace SMT ICs is Chip-Quick, and solder braid. Add a little flux (comes in the kit) to the pins, add Chip-Quik, lift off the IC. Use solder braid to clean up the board, add flux to the pads, position the IC, and solder the IC in place. For me, the hardest part is getting the IC in exactly the right place - my hand shakes slightly.

I'm pretty sure the AF4502C is a 'jelly bean' part. It's a matter of identifying equivalent part numbers; something I'm not comfortable doing.

To me, the most significant point is the extreme amount of ripple you have reported.

And if you have been able to repair the other 3 monitors, there is no reason this one can't go on the shelf as a parts donor. A number of people would be horrified at a 3 of 4 success rate. A baseball batter would be thrilled to hit .750.

PlainBill

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

Let's go with the easiest first. Short of a hot air rework station, the best way to replace SMT ICs is Chip-Quick, and solder braid. Add a little flux (comes in the kit) to the pins, add Chip-Quik, lift off the IC. Use solder braid to clean up the board, add flux to the pads, position the IC, and solder the IC in place. For me, the hardest part is getting the IC in exactly the right place - my hand shakes slightly.

I'm pretty sure the AF4502C is a 'jelly bean' part. It's a matter of identifying equivalent part numbers; something I'm not comfortable doing.

To me, the most significant point is the extreme amount of ripple you have reported.

And if you have been able to repair the other 3 monitors, there is no reason this one can't go on the shelf as a parts donor. A number of people would be horrified at a 3 of 4 success rate. A baseball batter would be thrilled to hit .750.

PlainBill

Thanks for the tips. I've seen this done in videos on Youtube, so with the right supplies it shouldn't be too hard. Since I'm going to have to order that stuff from DigiKey, I might as well buy as many of the replacement parts from them in the same order as I can. Do you have any suggestions on replacement caps for the remaining ones (especially the 33uF 25v and all the blue ones)?

I'll poke around a little over the weekend to see what equivalent parts I can come up with. The important thing is that I correctly identify what kind of chip it is (P-channel, N-channel, both, etc.) and some important specs so I can attempt to match them with other potential replacement parts.

Given the type of load, I had no idea what kind of ripple to expect. If I turn the inverter off, the ripple goes away (I don't have another load capable of handling 2A at 13v handy). I was under the impression that ripple on the 13v output should be controlled by the output filter caps and perhaps also an inductor (PI filter, yes?), so I'm not sure how much of an impact replacing the others will have given that I already replaced the 3 CapXon 820uF 25v filter caps on the 13v output.

As far as repairing others, I regularly fix the newer, cheap Samsung 19" wide monitors just with cap replacements. The LGs I fixed recently were a new adventure into "2 seconds inverter shuts off" repair and CCFL replacement. I have 3 more of these (possibly 4 more, I have one on my desk that I haven't disassembled yet), all with the same problem and I have not been able to fix any of them. Unless you think it would be valuable for me to check everything against all available IP boards (which would be somewhat time-consuming), I'm hoping that successfully diagnosing one of these will yield a solution for all of them, given that they all have the same problem (increase in current draw by the inverter causes the inverter fuse to blow). While a baseball batter might like to hit 75% of pitches, (s)he would probably feel bad if there was a particular pitcher who threw a particular type of pitch that (s)he was completely unable to hit.

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

WAIT. How long did you run your test CCFLs here? If they all run good w/o blowing the fuse, then your problem is your lamp!

BTW, for this model, I use Digikey P12391-ND, 820uF/25V 10*25.

Good luck.

I used only two of the lamps at any time during that test, both connected to the same transformer. The monitor shut off after a few seconds, but it was enough time to verify that the lamps lit properly. If I run with all lamps attached, they all light just fine but current through the inverter rises until the fuse blows. A couple of posts ago I describe what happens with a multimeter attached in series with the inputs to the FETs. I don't have another working monitor handy that I can use to independently test the lamps, but I'll do that if a recap fixes another busted monitor I have.

I was unable to get 10x25 caps to fit in the available space. I used 680uF 25v Nichicon HE 10x20 caps instead, and I have two other options on order (2200uF 25v Panasonic FK 12.5x30 and 3x Panasonic FK 820uF 25v 10x20).

[stevenps]

Re: Samsung SyncMaster 940T blows inverter fuse

The dual FET you're talking about might be a AM4502C, which is also difficult to find. What is the number of the CCFL controller chip that drives these FET's?

BIT3195G (I can't find a datasheet for this).

[PCBONEZ]

Re: Samsung SyncMaster 940T blows inverter fuse

Here's the manufacturer but they aren't so big on data sheets.
Probably have to email and hope they read English.
http://www.bitek.com.tw/e-company.htm