Panasonic sc tnpa5081 rebuild

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

Well guy's really bummed out right now....

I received all the new parts rebuilt the board again went over everything and it looks perfect.

Installed it in the TV power it up with the SU/SD disconnected and it powers up perfect no errors.

Recheck the SU/SD they look good to me with no shorts. Reconnect the SU/SD remove the jumper power the TV on and BANG all kinds of shit blown again!

I don't know what to think at this point either the SU/SD are bad or the screen is defective.

Going to take a few days off from this I need to order a few more parts then I will have at it one more time.

If I pickup another TV that uses the same SC board can I install this board it a 42" / 50" in the same series? This is out of a 54" just to prove out that the board is good?

I picked this up cheap and it should be here next week I know it's the wrong board version but it will give me something to look at to compare a few things.
http://www.ebay.com/itm/TNPA5081-/30...vip=true&rt=nc

Thanks for all the help......

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

You're right--I'll have to check that out, Hadn't realised it outputted such a high test-voltage on Diode.--Guess that's why it has no trouble illuminating white-LED's in backlight strips when tested individually....

Using that battered Fluke 23, I tested some (UT-Sourced) IGBTS....

Between C and E they read 2-3 Meg one way, O/C other....

--I only tested two new and a used (OEM) one from a TNPA5066 I was working on, a DG501,--which was OK

Same IGBTs on Beckman (W/S old) read in excess of 20 meg (ohm symbol--no reading, OL)

--IF you are getting a Beckman--Stay with the 'Tech 300' series like, 300, 310, 320 --These I Know for sure use much the same chip and operate in the same way as the ones I have...

Other models are smaller, newer, trendier, (recognised as having a separate On/Off) --and ain't as good IMHO, --They don't have that low-power, 'Insta-Ohms' feature that makes 'em what they are... Really IS 'insta-ohms' while most meters need up to a second to calculate, their little bargraphs going up and down etc, Tech 310 tells you soon as resistor is attached--practically instantly...

These meters do occasionally have issues--Mainly the rubber/silicone/carbon ribbon-strips between mainboard, the LCD Display and the main processor-chip which itself is on a separate PCB fitted behind the LCD and held in a curious plastic frame....
They can lose connection, usual effect is meter wont register any ohms or voltage.

They need cleaning with alcohol--about every 7-8 years Ive found, and the LCD is pretty fragile, Ive a couple I bought for parts with cracked LCD (They still work, but a black halo is encroaching the view)--If you ever see any spares around for the LCD--Let me know.....

Once you've used one for a while--You'll not ever go back to a mere Fluke IMHO Just a damned shame they don't make 'em anymore......

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

If you look at the FGPF4633 IGBT Datasheet it looks like this.

On Characteristics VGE (th) G-E Threshold Voltage IC = 250 A, VCE = VGE 2.4, 3.3, 4.0 V

So it looks like your meter should turn on that IGBT with no problem.

I am beginning to think that if you look at my post #299 where I thought Q608 was the reason Q661 failed after the first repair it was in fact a faulty Q661 and when I replaced it along with the other parts listed I installed a somewhat working Q661 and that is why it worked until it was put it under load.

Q608 was bad and I have also found Q607 & Q490 bad on the board this time. These were removed and tested and they test leaky to me just like Q608. All of these are 2SK3018 MOSFETS (Same as Q647, Q818) which are usually found bad.

I was wondering if anyone has removed and tested these out of circuit.

I have a feeling they are bad on other board and might not be being caught as bad. I don't think your able to really test them in the circuit and they should be removed and tested.

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

Just checked the Beckman's ohm/diode test-voltages....

On all std. Ohms ranges, 0.4V, 200 ohms to 20M ranges (appears to reduce as the resistance range increased, going to 0.2V on 20 Meg range,--prolly due to the internal resistance of the Fluke measuring it.

On Diode-Test (rather surprised) it outputs 5.0V

This is exactly why resistance measurements are so variable across different DMM's

Fluke 79III
.656 VDC MΩ
1.246 VDC Ω
1.245 VDC KΩ

Fluke 87V
2.776 VDC MΩ
7.31 VDC Ω
7.30 VDC KΩ

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

Just checked the Beckman's ohm/diode test-voltages....

On all std. Ohms ranges, 0.4V, 200 ohms to 20M ranges (appears to reduce as the resistance range increased, going to 0.2V on 20 Meg range,--prolly due to the internal resistance of the Fluke measuring it.

On Diode-Test (rather surprised) it outputs 5.0V

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

I have a really old abused (not by me!) Fluke-23....

I'll check a couple with that--See what I get. This may illustrate the drawbacks of DMM generally.
(A reason personally Ive kept and used my Beckman practically daily for over 30 years and never gone wrong seriously, --No frills or daft functions you never use, not autoranging or anything--It just Works--The battery lasts forever--Even if you leave it on for 6 months!)

That one you have thats not turning on in your testing--Is not necessarily bad.

--The Gate Threshold Voltage (Volts at which it starts to turn on) is one of the huge variables between various IGBT's or MOSFETS--Of the Same Type. Check the datasheet for any IGBT/MOSFET and there'll be a range of voltages for V-Th....

G-Th may be as low as a volt--Or as high as 4V...

--If your test-meter doesn't output the amount that tested device needs--It'll read as 'faulty'
With the Beckman only outputting 0.5V on its Ohms ranges--I cant test 'em like that anyway, so don't ever bother.

IGBT's Very Rarely fail O/C--In fact, Cant recall ever having an O/C IGBT Always Leaky/Shorts--not to say it cant happen, just IME not ever seen one.

Tech 310 outputs 1.5V (Will check today) as I recall, in Diode-test mode, again not reliable for turn-on tests in IGBT's....

Trust me that IGBT is BAD!

All the others turn on just fine with the Fluke 79III

I just tried it with the Fluke 87V and it still does not turn on.

Here are the output voltages in diode mode for the meters measured with the DMM connected to another DMM.

Fluke 79III puts out 2.95 VDC
Fluke 87V puts out 7.30 VDC

Please take a look at the picture I posted in post #325 what do you see?

I think that the IGBT's I received from UTSOURCE are fakes. Look at the markings if you go by what I think is the date code it would mean these were manufactured in 2003?

I don't think that these were even made back then?

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

I have a really old abused (not by me!) Fluke-77....

I'll check a couple with that--See what I get. This may illustrate the drawbacks of DMM generally.
(A reason personally Ive kept and used my Beckman practically daily for over 30 years and never gone wrong seriously, --No frills or daft functions you never use, not autoranging or anything--It just Works--The battery lasts forever--Even if you leave it on for 6 months!)

That one you have thats not turning on in your testing--Is not necessarily bad.

--The Gate Threshold Voltage (Volts at which it starts to turn on) is one of the huge variables between various IGBT's or MOSFETS--Of the Same Type. Check the datasheet for any IGBT/MOSFET and there'll be a range of voltages for V-Th....

G-Th may be as low as a volt--Or as high as 4V...

--If your test-meter doesn't output the amount that tested device needs--It'll read as 'faulty'
With the Beckman only outputting 0.5V on its Ohms ranges--I cant test 'em like that anyway, so don't ever bother.

IGBT's Very Rarely fail O/C--In fact, Cant recall ever having an O/C IGBT Always Leaky/Shorts--not to say it cant happen, just IME not ever seen one.

Tech 310 outputs 1.5V (Will check today) as I recall, in Diode-test mode, again not reliable for turn-on tests in IGBT's....

Ok you've talked me in to it I am going to purchase a used Beckman one more meter won't kill me.

I also have one of these on order http://www.greenlee.com/products/DMM...oduct_id=17755

I need to put the Fluke 79III back in the Van I needed it today out on a job and forgot I left it on the Tech bench.

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

I have a really old abused (not by me!) Fluke-23....

I'll check a couple with that--See what I get. This may illustrate the drawbacks of DMM generally.
(A reason personally Ive kept and used my Beckman practically daily for over 30 years and never gone wrong seriously, --No frills or daft functions you never use, not autoranging or anything--It just Works--The battery lasts forever--Even if you leave it on for 6 months!)

That one you have thats not turning on in your testing--Is not necessarily bad.

--The Gate Threshold Voltage (Volts at which it starts to turn on) is one of the huge variables between various IGBT's or MOSFETS--Of the Same Type. Check the datasheet for any IGBT/MOSFET and there'll be a range of voltages for V-Th....

G-Th may be as low as a volt--Or as high as 4V...

--If your test-meter doesn't output the amount that tested device needs--It'll read as 'faulty'
With the Beckman only outputting 0.5V on its Ohms ranges--I cant test 'em like that anyway, so don't ever bother.

IGBT's Very Rarely fail O/C--In fact, Cant recall ever having an O/C IGBT Always Leaky/Shorts--not to say it cant happen, just IME not ever seen one.

Tech 310 outputs 1.5V (Will check today) as I recall, in Diode-test mode, again not reliable for turn-on tests in IGBT's....

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

Finally got some readings of various IGBT's....

I'll only list C-E resistances, Gate should (And does) read Completely O/C with ref to either E or C.
All tests done on 20 Megohm Range, using two Beckman Tech 310, one early (W/S meter) and a later one, Home Meter These use an 'Insta-Ohms' and Low Voltage/power resistance tests. Ive not seen this function as-is in Beckman Tech series on any other meter,--Even mega-bux Fluke etc.

RJP63F4A reads OL with red lead to C and black to E.
---Reads 16 Meg with the leads reversed using a later Beckman Tech 310 and over 20M on early meter (No resistance, just 'OHM' symbol illuminates on display)

The story is much the same for the FGPF4633, OL one way, Very high, --un-measurable the other.

I checked some RJP43F4A, again--Same ball-park readings.

Also checked some RJP56F4 and some RJP63F3A transistors, All much the same.

--One thing I did do was to lick my finger, then 'short' the IGBT to be tested first by holding it between thumb and finger--Object was to remove any residual gate-charge that could have messed readings up...

I tested 5 pieces of each type IGBT. These were all sourced from UT at various times over last year or so. The 20 x RJP63F4A I received this very morning, and immediately scrutinised.

Asthetically, they are rough, meaning tiny chips, scratches and gouges if looked at through a loupe, but test fine, as above. One or two May have witness-marks of having been screwed to a heatsink but no soldering or cutting of leads.

All the casting dents and impressions, symbols etc are identical to a genuine Renesas part, so can only deduce these are surplus stock, thats been loose in a box and shifted and bumped about all over the worst roads in China on back of an Ox Cart....!

Wow Thanks for all the testing.

Here's the deal not having the same DMM on my side it's hard to test for the same measurements that you're getting.

Here is my testing of the RJP63F4A (Q661) this is using a Fluke 79III DMM.
Emitter = + Lead, Collector = - Lead (Reading = 3.138 M Ω and slowly climbing)

Same IGBT Tested with a Fluke 87V
Emitter = + Lead, Collector = - Lead (Reading = 1.247 M Ω and slowly climbing)

I tested 6 of these and they all test pretty much the same.

Now here's the deal one of these I know for sure is bad! When testing in diode mode the IGBT NEVER turns ON!

Example:
Emitter = - Lead
Collector = + Lead (Should get an OL on the DMM if the Gate has not been charged. If you do short Gate & Emitter with your fingers you should now get an OL on the DMM.)

Now move the Positive Lead to the Gate for a second to turn on the IBGT once charged move the Positive lead to the Collector and you should now have a reading)

On the RJP63F4A using the Fluke 79III I get the following readings. (These were ones purchased from UTSOURCE)
#1 = .809 VDC
#2 = .865 VDC
#3 = .876 VDC
#4 = 1.178 VDC
#5 = 1.020 VDC
#6 = OL this IGBT is BAD for SURE! (But using the Ohms testing it shows good or at least the same as the others.)

I feel that these are bad or at least a few were / are BAD. And at this point might be what has led to blowing the repair up several times. I will be using the eBay RJP63F3A in the next repair attempt.

On the RJP63F3A using the Fluke 79III I get the following readings (These were purchased from eBay)
#1 = 1.373 VDC
#2 = 1.186 VDC
#3 = 1.350 VDC
#4 = 1.160 VDC
#5 = 1.330 VDC


Just some side notes to help others and confirm what has been said before.

I ordered some replacement parts from Parts Simple here is the US and here is what I have received. The other parts will be here tomorrow and once I receive those I will be testing the board again.

(The NEW Q621, Q622 {30F125} will be here today and this what I have been waiting for)

I ordered a couple of the D481, D482 Panasonic part # B0FACP000003 {Originals were RFUS20TM4S) the factory replacements that were sent to me were RFN25TM4SW so these should be considered exact replacements.

I have ordered a Peak Atlas DCA75 PRO Semiconductor Tester I have wanted it for a while now and plan on using is to get a base line of a known good IBGT for these repairs moving forward.

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

Finally got some readings of various IGBT's....

I'll only list C-E resistances, Gate should (And does) read Completely O/C with ref to either E or C.
All tests done on 20 Megohm Range, using two Beckman Tech 310, one early (W/S meter) and a later one, Home Meter These use an 'Insta-Ohms' and Low Voltage/power resistance tests. Ive not seen this function as-is in Beckman Tech series on any other meter,--Even mega-bux Fluke etc.

RJP63F4A reads OL with red lead to C and black to E.
---Reads 16 Meg with the leads reversed using a later Beckman Tech 310 and over 20M on early meter (No resistance, just 'OHM' symbol illuminates on display)

The story is much the same for the FGPF4633, OL one way, Very high, --un-measurable the other.

I checked some RJP43F4A, again--Same ball-park readings.

Also checked some RJP56F4 and some RJP63F3A transistors, All much the same.

--One thing I did do was to lick my finger, then 'short' the IGBT to be tested first by holding it between thumb and finger--Object was to remove any residual gate-charge that could have messed readings up...

I tested 5 pieces of each type IGBT. These were all sourced from UT at various times over last year or so. The 20 x RJP63F4A I received this very morning, and immediately scrutinised.

Asthetically, they are rough, meaning tiny chips, scratches and gouges if looked at through a loupe, but test fine, as above. One or two May have witness-marks of having been screwed to a heatsink but no soldering or cutting of leads.

All the casting dents and impressions, symbols etc are identical to a genuine Renesas part, so can only deduce these are surplus stock, thats been loose in a box and shifted and bumped about all over the worst roads in China on back of an Ox Cart....!

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

I didn't get a chance to measure or find some IGBT's today--bit of a crazy day--will see what I can do tomorrow...


45F128 (Toshiba)
330V ⎯ 200A (pulse) ⎯ 26W (@25 degC) TO-220SIS, 1.45 Vsat C-E @ 120A 15V G/E 0.27uS

(Most of the Tosh PDP IGBT's have a similar 200A Pulse rating. The V-Sat C-E is fairly important and can vary widely between the many Tosh types. The Lowest V-Sat C-E is desirable for low dissipation. The I/p is probably most important for PDP due to the extreme currents flying round SUS boards, up to 300A for a 50")

FGPF4633 (Fairchild), 330V 300A (Pulse) -30.5W (@25 degC), TO220SIS 1.55V Vsat C-E @ 70A 15V G-E, 0.26uS
--I would be reasonably happy using this Fairchild part in place of the Tosh....

--Doesn't help that different manufacturers measure at different currents though....
Just sayin'....

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

The 5081 is much the same--no matter what size set it comes from--The Details are in doubling up on certain parts, and the ERC coil/capacitor components specific to the size/type of panel the board was originally fitted in....

Its conceivably possible to take a 42" 'AH' board, change the ERC parts over, add the 'missing' extra IGBT's etc and drive a 54" (or whatever) size panel with it....

For Q621/2 you should be able to sub FGPF4633, I'll be using these in the 5081 I'll be doing after the 5105's that Ive got to do...
--Diode-mode isnt always as revealing as you would like to hope--It only takes a small leak to cause issues with these IGBT's and doesn't always show on Diode-test-mode...

Even the Slightest Doubt--Replace.--This is the motto concerning ANY part of these SC.....

Out of curiosity--I replaced a 5105 with a 5081 in a FHD cheapo Argos special set some months ago, still running fine....

Good info thanks here's some more information I have found not sure if you already have this but it does look like the FGPF4633 is a very close match to the 30F125.

It also looks like the 45F128 I have here is just a higher Collector current rating (DC) of the 30F125.

See page 4 for part Numbering
http://elimex.bg/userfiles/editor/fi...5_GT30G122.pdf

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

with the 30f125 if they only read 200-300ohms in cct they're bad otherwise it's usually in kohms.

adamant as in more often than not they're bad but at 300ohms most meters won't beep as in shorted so people have missed these.

don't have any out of cct to test

Thanks not questioning what your saying just trying to make sure I have this right this time.

I believe that your right and maybe I do have a bad one but just trying to prove that it's bad for sure.

If I had a know good one this would not be a problem but it is my belief that all the ones I received from utsource are bad.

They test good in diode mode but with resistance testing they are all over the place and I have no way of knowing for sure what is right or wrong.

Thanks again for the help

I am not one to give up so one way or another I will get the board repaired.

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

I'll test the FGPF4633 I have--I know not the same device but similar enough to get a ball-park figure for resistance-checks...

If I can find any good 30F125 I'll check those too...

Thanks very much appreciated

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

I'll test the FGPF4633 I have--I know not the same device but similar enough to get a ball-park figure for resistance-checks...

If I can find any good 30F125 I'll check those too...

[tw2005]

Re: Panasonic sc tnpa5081 rebuild

with the 30f125 if they only read 200-300ohms in cct they're bad otherwise it's usually in kohms.

adamant as in more often than not they're bad but at 300ohms most meters won't beep as in shorted so people have missed these.

don't have any out of cct to test

[Alastair E]

Re: Panasonic sc tnpa5081 rebuild

The 5081 is much the same--no matter what size set it comes from--The Details are in doubling up on certain parts, and the ERC coil/capacitor components specific to the size/type of panel the board was originally fitted in....

Its conceivably possible to take a 42" 'AH' board, change the ERC parts over, add the 'missing' extra IGBT's etc and drive a 54" (or whatever) size panel with it....

For Q621/2 you should be able to sub FGPF4633, I'll be using these in the 5081 I'll be doing after the 5105's that Ive got to do...
--Diode-mode isnt always as revealing as you would like to hope--It only takes a small leak to cause issues with these IGBT's and doesn't always show on Diode-test-mode...

Even the Slightest Doubt--Replace.--This is the motto concerning ANY part of these SC.....

Out of curiosity--I replaced a 5105 with a 5081 in a FHD cheapo Argos special set some months ago, still running fine....

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

I never found what the "R" meant but that is the conclusion I reached and have stopped worrying about it ages ago. These parts would have been made for Panasonics needs, pre-cut and bent so it makes sense to have a different part number.

If it were panel I 'd expect the buffers to blow, I don't see how they are ok but the SC keeps blowing once connected to them.

So you think the new buffers are still bad? They measure out good but I guess anything is possible.

Hope to be ready to test it soon I have went over the complete board and found some other interesting bad parts. I will update later on these.

I don't know what I am going to do about the Q621, Q622 (30F125)

The parts test good out of circuit but after reading everything on the forum you seems to be adamant about them being bad even though they test good in diode mode.

The ones I have from utsource look suspect to me and the ones that I replaced the bad ones with the first time were out of a TNPA5764 board that was good. These still test good out of circuit.

If you have some 30F125 on hand could you ohm them out for me please.

I have some on order but who know when they will get here.

I also have a feeling that the Q661 (RJP63F4A) that I also received from utsource were bad I will know more on this for sure later this week when I can do a curve trace on a known good one and the current supply I have on hand.

I see another 50" up for sale and might go and get it today.

Also the board I am working on is from a 54" TV has anyone repaired one of these boards and got it to work?

From reading the forum it looks like all the attempts to repair these have failed. I have a hunch why but will not know for sure just yet until I figure out what to do with the Q621 & Q622

[tw2005]

Re: Panasonic sc tnpa5081 rebuild

Not the first time a manufacture has given bad info....

Well that is kind of what I am thinking the dual diodes I installed were only rated at 300V 20A and if I am reading the schematic right there would be more than 300V on the diode especial at turn on.

Well it's either the diode problem or a bad panel as everything else it testing out good.

Also these are listed as two different part numbers I wonder if one is a R and the other is not as one have the leg bent back and the other froward.

Q16441 B1JAEP000013 TRANSISTOR
Q16452 B1JAEP000012 TRANSISTOR

I never found what the "R" meant but that is the conclusion I reached and have stopped worrying about it ages ago. These parts would have been made for Panasonics needs, pre-cut and bent so it makes sense to have a different part number.

If it were panel I 'd expect the buffers to blow, I don't see how they are ok but the SC keeps blowing once connected to them.

[dolivas27]

Re: Panasonic sc tnpa5081 rebuild

afraid that's bad info on their part. The suffix , FT, FK etc directly relates to the package. Yes , electrically that is the same but there is no way a TSSOP package will populate those locations!

On these later boards they're VSSOP, if they stuck with TSSOP as in the earlier pre - 2010 boards it would be easy as there are heaps of manufacturers.

TSSOP is 1.6mm longer, 2.4mm wider roughly

I would just use whatever 45G128 you have.

The G20,25 guide suggests removal of ss11 and or ss33 TV should turn on with no errors.

I don't understand the diode question. I expect if you went outside the limits in either volts or amps the device would fail.

Personally I think they use higher voltage ratings to improve longevity but they all appear to be in the 20A range. This is out of my area of knowledge

Not the first time a manufacture has given bad info....

Well that is kind of what I am thinking the dual diodes I installed were only rated at 300V 20A and if I am reading the schematic right there would be more than 300V on the diode especial at turn on.

Well it's either the diode problem or a bad panel as everything else it testing out good.

Also these are listed as two different part numbers I wonder if one is a R and the other is not as one have the leg bent back and the other froward.

Q16441 B1JAEP000013 TRANSISTOR
Q16452 B1JAEP000012 TRANSISTOR