repairing switch power supply in samsung plasma

[bountyhunter]

I would not think ESR actually changes at different frequencies .

It does, at least for electrolytics. If I can find a curve, I'll post it but believe me, it is not constant. Life would be a lot simpler if it was. I remember some of the better cap makers used to publish curves of ESR versus frequency.

Here is one for ceramics:

 

[bountyhunter]

still good to understand how the reactance within a capacitor can effect ESR readings though, not just capacitive reactance but ESL reactance.

Network analyzers give poor results measuring ESR, there is a much better way: you drive the cap with a "step change" in current and measure the step voltage you see in the cap's voltage plot. The step voltage is an exact measure of ESR as the step current flows into the cap through it's ESR. This app note describes the method.

**broken link removed**

 

[Preher TV]

It does, at least for electrolytics. If I can find a curve, I'll post it but believe me, it is not constant. Life would be a lot simpler if it was. I remember some of the better cap makers used to publish curves of ESR versus frequency.

I believe you......gives me some good reading to do, I love electronics and find it's easy to be wrong but a lot of fun to keep learning more and correct my thinking on the topics I had wrong....


My thinking was that if Resistance is independent of frequency so should be ESR, Impedence of the capacitor however is dependent upon the frequency and would include the ESR.

If we say ESR changes with frequency then we are saying that ESL reactance and Capacitive reactance of a capacitor are part of it's ESR.

I just thought it was easier for most people to speek of ESR changing with frequency because then they don't have to think, of the reactances within the capacitor causing the ESR to appear high or to think about the impedence of a capacitor which includes ESR ESL reactnce and capacitive reactance.

 

[bountyhunter]

I don't know how Sencore can be claiming a patent on the technique of measuring ESR by measuring the voltage step resulting from a load current step. I published an article some time back using this principle explaining how it works (see attached PDF taken from my article). The internal effects of both ESR and ESL produce measurable voltages in the "load transient test" where load current is stepped abruptly. The ESL produces a narrow voltage spike and the ESR produces a definited DC volage level step change. The amount of load step voltage measures the ESR accurately.

 

[bountyhunter]

My thinking was that if Resistance is independent of frequency so should be ESR, Impedence of the capacitor however is dependent upon the frequency and would include the ESR..

The characteristics of the cap and it's dielectric material affect ESR and they are affected by frequency.

If we say ESR changes with frequency then we are saying that ESL reactance and Capacitive reactance of a capacitor are part of it's ESR.

No, they are certainly not. ESR (at any frequency) is the resistive component of impedance, and it does not include reactance.

 

[Preher TV]

No, they are certainly not. ESR (at any frequency) is the resistive component of impedance, and it does not include reactance.

That's what I thought, But if you look at the ESR vs frequency curve, it looks like it should if you take into account reactance, frequency goes up so does ESL reactance and so does there graph of ESR, frequency goes low the capacitive reactance goes high and so does there graph of ESR it looks like an impedence vs frequency curve would for the same capacitor.....seems ESR and Capacitor Z are interchanged alot to me.

 

[Repair-me]

Right, so getting back to the topic, I’ve relised that the stw26nm60 is actually only a 60, not a 50,. Very small print. My mistake.

Now moving on to the diodes, once the diodes shorted, they obviously blew the stw26nm60. but the other boards where the vs. leads to all have fuses prior to any other components, so diodes connect to the caps, then to other boards with fuses., so if the other boards where damaged.. Wouldn’t the fuses blow or better yet wouldn't there be a short? Cause I connected low voltage to the vs. of the other boards to see if there was any amp drainage. Apart from the charging of the caps, they hardly didn't take no power.

So now im thinking that the caps failed in the psu, and possibly in the other boards. And that + heat caused the failer of the diodes which in turn blew the stw26nm60.

Just don't want to connect new diodes and stw26nm60 and have it blow up in my face.

My second problem is I can seem to find a replacement diode. The diodes are rhr15120 and searching on **broken link removed** I can't find them. Anyone have recommendations?

**broken link removed**
.
**broken link removed**
.
**broken link removed**
.
**broken link removed**

and the pdf of the diode rhr15120

 

[Repair-me]

Just wanted to say thank you for all your help so far. Anyway, so I’m looking for the diode RHR15120 but can't find it, also you guys recommended I get a faster one? So here’s the list, what do you recommend:

**broken link removed**

 

[unclejed613]

That's what I thought, But if you look at the ESR vs frequency curve, it looks like it should if you take into account reactance, frequency goes up so does ESL reactance and so does there graph of ESR, frequency goes low the capacitive reactance goes high and so does there graph of ESR it looks like an impedence vs frequency curve would for the same capacitor.....seems ESR and Capacitor Z are interchanged alot to me.

sad to say it, but a LOT of data is transcribed from engineering notes to catalog pages these days by marketing people who have been given a crash course in "technese", and only have a fuzzy idea what the data is that they're transcribing and reformatting. there are so many terms that are interrelated in complex ways, so they try to simplify it to fit in so many inches of page space, even to the point of lumping a lot of loosely related terms into a single number or curve on a chart. you find spec sheets that use whatever term for a quantity that seems easiest to use, or most commonly used when talking to the engineering department at their company. you will see the related terms ESR, DF (dissipation factor), TanΘ (loss tangent), Q, and "minimum Z", all used to describe the same characteristics of a capacitor. it just depends on the preferred measurement methods used by that company's engineers, and how the marketing techs doing the catalog copy understands the engineer's notes. unfortunately it makes it difficult to look at two competing catalogs side by side and decide which cap you want to buy. ESR is easy, not much math to do, just add X ohms to the reactance of the cap at Y frequency, and you know whether it will work or not. DF, TanΘ, and Q require a bit more math to get results in ohms of ESR. minimum Z includes ESR, ESL, and Xc all lumped together, and is actually more confusing (when i see this used, i tend to "highlight" that section of a catalog with a black marker, since that "spec" is hiding something more often than not)....

i see the same thing in RF components, where one device has insertion loss, while an almost identical device from another manufacturer has "gain" (measured in negative db of course...) while both terms are valid for the same quantity, which one would you buy first, the one with loss or the one with gain?...

 

[bountyhunter]

That's what I thought, But if you look at the ESR vs frequency curve, it looks like it should if you take into account reactance, frequency goes up so does ESL reactance and so does there graph of ESR, frequency goes low the capacitive reactance goes high and so does there graph of ESR it looks like an impedence vs frequency curve would for the same capacitor.....seems ESR and Capacitor Z are interchanged alot to me.

The manufacturers publish all the data: ESR, ESL and impedance versus frequency. It's not mixed up, it's as posted. ESR is related to the cap dielectric material characteristics and it is frequency dependent in some cases.

 

[bountyhunter]

So now im thinking that the caps failed in the psu, and possibly in the other boards. And that + heat caused the failer of the diodes which in turn blew the stw26nm60.

Just don't want to connect new diodes and stw26nm60 and have it blow up in my face.

I have repaired a few (many dozens) of switchers and they rarely fail nicely. The switch transistors have a lot of voltage across them and a slight mis step causes immediate destruction... and the components nearby get clobbered as the transistors junctions fail and pass HV into low voltage circuitry. I would recommend on a blown switch converter to AT LEAST change the switch transistors, snubbers and associated components, output rctifiers and the switcher's driver IC. Of course, the switching caps should always be replaced.

 

[Repair-me]

ok, so im stuck. i need your professional help people. rhr15120 does not exist in australia., "which is where im from" the closest thing i can find is here = **broken link removed** also its been mentioned that i should use a faster one? please advise

 

[unclejed613]

while we're on the subject, i'll include a common misconception about electrolytics.....
i see a lot of posts on a popular audio forum, where people seem to think electrolytics have tons of ESL because of their construction. they surmise this because just about everybody has seen the results of a cap blowing up, with the long spindle of aluminum that results. first of all, this is usually small electrolytics with a case so small that it shoots off rather than rupturing, pulling apart the innards of the cap in this manner. second, they rarely see the construction of larger caps or the methods used to cancel or minimize ESL in the construction. we're all familiar with the aluminum foil/ wax paper cap illustration used in textbooks and even in crystal radio DIY kits, and know that electrolytic construction is similar, with the exception that a chemically treated dielectric is used instead (the chemical action on the aluminum creates a dielectric layer on the aluminum itself). what usually isn't shown is that the ends of the aluminum "coils" are usually crimped, and the wires for both plates terminate at the same end of the "coil". the crimp minimizes ESL by shorting the "turns" together, and the termination at the same end of the plates minimizes ESL by cancellation of mutual inductances, so the only real inductance left should be the lead inductance. of course no cap is perfect, and some "shortcuts" may be taken by manufacturers, causing more ESL. ESR in electrolytics is primarily a function of the formulation and the thickness of the oxide layer on the plates. you may have noticed that higher voltage rated caps have higher ESR. that's because of the thickness of the oxide layer, and the fact that the conductive portion of the plates is also thinner.

 

[Repair-me]

found a place that sells rhrp15120 "even though my diode is labled rhr15120". going to get it and try, also im going to install fans since the psu has 4 outputs for fans and no stupid fans!. cheap go samsung.

 

[Nigel Goodwin]

Fans were normally fitted on only the early Plasma's - later ones were more efficient, and didn't need them.

 

[Repair-me]

Ok so an update- replaced the fuse, transistor, cap, & diode which fixed the vs in the psu. And plugged it in and bang! Repeat of the damage. Ive googled this model and it turns out that allot of people are getting the exact same symptoms as me but there is no current way to fix this.

Sure anyone can say change all the boards. I mean once you replace the x & y & psu. What’s left, nothing? You may as well go buy a new plasma. Im not looking to replace the boards. Im attempting to repair them, so once it’s up and going. Allot of people can follow in my foot steps with this particular model and it's repairs. It’s so famous for being the faultiest Samsung that there are at least 5 of these on eBay faulty per week. Which only indicates a very high failer rate.

 

[Repair-me]

Any future help and investigation into the cause would be appreciated. Think of this as a challenge to the electronic guru.

 

[Repair-me]

Decided this thing is a headache~~~! Seriously I can recommend people to stay away from ps-42s5h. It’s a nightmare of a TV which has so many shortcuts that all you need to do is Google the model to see how many people out there hate this TV.

Not going to bother to fix this one. Time to let go. Moderators please close this thread..

 

[meatynugget]

well Repair-me,.. you done alot of work and it's been a great read, i'm just about to imbark on a PSU of my 50" plasma can't see any blown parts but i'd think once i get the multimeter on there i going to start finding things,.. I be starting tomorrow so i'll try keeping you upto date and would be great if you can keep me in the picture ( if we could only get one!! LOL)

 

[unclejed613]

you may want to pull the switching transformer out to do a ringing test. connect a 100pf cap across one of the secondary windings, feed the primary with a 10khz or so square wave (a function generator with a 50 ohm output is best for this, a 600 ohm or 1k output impedance may not give useful results), and connect an oscope across the secondary with the cap (or one of the other secondaries). you should see spikes or square waves with decaying squiggles following each voltage transition. if the squiggles aren't there, or the waveform isn't there with a reasonable voltage swing, the transformer may have a shorted winding. this would also cause blown fuses and switching transistors.