Memorex MLT3221 - No Power

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign hi,

i start in the end i got your point i am also don 't care ,one of my friend give it to me.so i am playing to my fun.and i think no one dig very deeply like you and me.
i agree with you,i also think IC8,Q16,Q17,CCFL are good,so at this evening i start to check the controller IC7.this ''IC7'' get the stby command,send command to IC8 to start ,get the feedback from the high voltage,and he send the command to shutdown the high voltage if some thing wrong.

so i start to disconnect the feedback one by one pin 1,2 ,3,4 after disconnect pin 4 the high voltage stay on and the CCFL on.it is work for Several mints.
i must read about this controller.any information will be help,
be continue tomorrow evening
thank for your help

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

Thank you for your explaination of how your HP scope frame is isolated from your probe return connector. I have a TEK scope and have never tried to set up mine the same way.

Your waveforms of the pulses on IC8 pin 11 and pin 14 look VERY GOOD ! They actually look identical to mine while my set is running normally.

At this point, I think it really does not matter what what the other pulses or voltage levels are on IC8 or even IC7. These two pulses are all that is needed on the gates of Q17 and Q16 (along with 400v DC) to start them switching at 60 kHz and feeding this into the input of the final step up transformers to generate approx 3000 vAC to run the CCFLs. I'm not sure if the CCFLs need an initial "strike" voltage first, however.

By forcing +12v DC onto pin 13 of IC8, I believe we are now running IC8 in "unregulated mode", meaning IC7 is no longer actually doing anything we need.

Are DCT1 and DCT2 still getting hot? It would be very helpful if you had some way to SAFELY measure the actual AC voltage going to the CCFLs. If it is reasonable, maby you could try removing DCT1 and DCT2 and see if the lamps come up to full brightness.

If the lamps still remain dim, double check the peak to peak voltages you had the last time across Q17 and Q16. Bacuse of the way you measured them (HP scope ground lead isolated from scope frame) I think they sound reasonable for driving the primaries of the 2 final step-up transformers. It is possible that one of these transformers might have say, some shorted turns, and that is why the lamps are running dim.

It is als a possibility that the lamps them selves may have a problem. That is what I thought was wrong with my set when it had only a dim screen when I forced +12v DC onto IC8 pin 13. So, I am hopeing that yours are OK also.

As far as taking measurements on IC7, I can give you so insights on this chip. It is actually a full blown microcomputer. I spent 7 years at RCA as an engineer writing software for these chips using Assembly Language and debugging in Machine Language. Without some very specialized and expensive equipment, it would be virtually impossible to "reverse engineer" the stored program to figure out what it is actually doing. Even looking at it's input and output ports would have verry little meaning.

I know many of the other users of this form are actually TV repair shops and must repair the set to established standard before they they are allowed to return the repaired set to their customers. I, on the other hand, "got my set out of the trash heap" and therefore have nothing to loose if I should burn it out or even run it with IC7 bypassed and therefore no AC regulation. However, being an engineer and knowing the consequences, I would never leave the set unattended or even plugged in unless I had an eye on it.

I don't know how much "risk" you are willing to take with your set, but please be carefull with whatever you do with your set.

Anyway, let me how you are making out and we will keep working on the problem.

By the way, you took some GREAT scope photos!

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign hi,

this is my measure at ic8:
pin1 go up to 3.5v then down to 1.5v
pin2 go up to 3.5v
pin3 got pules -Img00013 this pules came after CCFL off and it is constant
pin 4 not used
pin5-pin16 the seem your measures
pin 11,14 img11,12,10
on connector 6 pin 1,4 are the seem
for your ask at the measurements Q17,Q16.first of all i working on float place,it means i disconnect the GND electricity from my HP scope.I do not touch any thing connect to GND earth. I am very very careful i also use circuit break, it is cut the electricity if more then 3MA go to the earth. PLS BE CARE
can i ask you pls to measure IC7 pins

thank you for your help

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

Your waveform on IC8 pin 14 looks good to me. Also your frequency of 60kHz looks good.

I took some measurements of IC8 while my TV was running normally. I thought it would be more accurate to make the measurements with my scope, rather than with just a voltmeter. Here is what I have:

Pin1 +5v DC with a very small wave riding on it (similar to output A and B)

Pin2 +4v

Pin3 and Pin4 didn't measure because don't think they are used

Pin5 sawtooth waveform 0v to +3v same frequency as output A and B

Pin6 +4v

Pin7 almost identicle to pin5

Pin8 +7v

Pin9 +6v

Pin10 0v

Pin11 same waveform as we have

Pin 12 0v (ground)

Pin13 +12v

Pin14 same waveform as we have

Pin15 +12v

Pin16 +5v

I could not measure D25,D16,D15,D19 because they are on the back side and I have my power board mounted in the TV with the CCFLs connected.

Also, on connector 4, Pin4 (white wire) has +5v and Pin1 (yellow wire) has +3v.

I think at this point, it would be a good thing to determine on which side of T6 the problem lies. I belive the best way to test this is to hook up your scope exactly the same way that I had mine hooked up (channel 1 onto IC8 pin 11 and channel 2 onto IC8 pin 14) Set scope on alt or chop to see both waveforms. set trigger on positive slope of output A ( pin11)

Then turn set on, let it shut off after 1.5 seconds and apply +12v to pin 13 of IC8. (the same as you did last time).

If you now get the exact same waveforms that I had in my photo, then IC8 and it's associated components should be OK and the problem lies further down stream or in the microcontroller chip IC7. Please keep in mind that the 2 waveforms MUST be 180 dregrees out of phase with each other (Q17 and Q16 can not be on at the same time!).

Since you told me that DCT1 and DCT2 were getting hot the last time, don't leave the +12v on IC8 too long.

Please do this test and post a photo of your of your scope results (if you can).

If the your results are the same as mine, Q17,Q16,and their caps along with the 2 final output transformers SHOULD be lighting the CCFLs at their normal level.

By the way, when you gave me the peak to peak voltage readings across Q17 and Q16, how were you actually able to make the measurements? (hi-voltage probes and an isolation transformer between the TV or scope ?)

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign hi
yes it is the seem at my HP scop as you can see at the pic,i hope i am not mistaken on my calculation the volt button is 2v per square and the time 2usec[micro] per square.
i also checked with frequency counter i get+-60.0khz you can see at the pic
this 2 measures was test at pin 14 of ic8.do not forgot i have only 1.5sec to make all my test measures,and the frequency is change i also see it at the HP scop.
if you operating your p/s can you make few measures for me?with volt meter to the gnd
1 all the pins at ic8
2 at d25,d16,d15,d19
thanks!!!!!! for your help

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

The photo didn't upload on the previous post. Here it is.

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

With my TV running normally, I took a photo on my scope of the output pulses on IC8. The upper waveform is output A (pin 11) and the lower waveform is output B (pin 14).

The scope is set up on 5 microseconds/division and the verticals are set up for 5v/division each.

As you can see, both A and B are about 12v DC (above COLD ground).

They are also 180 degrees out of phase with each other and each has about a 50% duty cycle.

Because one full cycle of A takes 15 microseconds, I calculate that these pulses have a frequency of 66.6kHz

I did't take any measurements across Q16 and Q17 because that would require putting my scope return lead onto the HOT side without an isolation transformer(I don't have one).

The conclusion that I have come to so far is that we need to get your IC8 to be putting out the same waveforms before we can even begin to continue to Q16 and Q17.

I have not yet looked at any of the other pins on IC8 because I need to get a DIPCLIP to do it safely.

So, in the meantime, if you want to look at all of the other signals on IC8, go ahead (you may be able to spot something abnormal and fix it). I will still look at all the other signals on my IC8 and report to what I have on a TV that is working normally.

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk


I was expecting the frequency of the pulses on pins 11 an 14 of IC8 to be about 30 Khz.
12v peak to peak sounds ok because that is it's supply voltage.

It turns out that pin 11 of IC8 already has a diode to ground on it (D24). Pin 14 of IC8 also has a diode to ground on it (D23). It is interesting that your 2 zener diodes caused the CCFLs to oscillate from dim to off, dim to off,... every two seconds.

As far as the 100v ptp on Q17 drain and the 30v ptp on Q16 drain, I am not sure if (or how far from normal) these are.

I have the exact same power board that you do (IPOS 150 Rev 1.8 2006.11.25).

I am going to put it back in my Memorex MLT3221 LCD TV and start looking at all the wavforms on IC8, Q16, and Q17 with my ocilloscope and report to you what I am finding. Then we can compare measurements to find out what is different and then get yours to work the same as mine (and get your TV operational again!)

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign
1 thanks for your help
2 the 2 cap 10u 400v was replaced again-no help
3 at pin 11,14 there is 12v ptp the fq=1:4x1000000=250mhz
4 Q16 drian there 30v ptp at Q17 drian more then 100v ptp
5 TEST- at pin 11,14 i solder 2 diodes zener 5.1v to GND ,the pulse was cut to 6v ptp,the results -the back light was dim. it is light 2sec ,off,then again for 2 sec then off.
at this test i see even if the output for the ccfl is low voltage it is react the seem.
any thought?

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

I had another thought. I know that you told me that EC17 and EC18 were new. However, because these two caps have an extensive history (on this forum) of being bad, maby you should recheck them. Probably, the best way would be to replace them with new ones. I could see how they might be causing the problem that you are having. I'm not sure, but I think part of their job is to absorb any voltage spikes from drain to source of Q17 and Q16 during switching. If either cap can't do this, then the spikes will be applied to the final output step-up transformers an jump from hundreds of volts to thousands of volts. (enough to jump the gaps in DCT1 and DC2 to ground).

So, try replacing EC17 and EC18 before going any further (even though they look ok).

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

It is very good that the DC voltage on Q17 drain is staying at 400v.

DCT1 and DCT2 appear to be there to protect the CCFLs from any voltage spikes greater than 3500v. Because they are getting hot, it looks like they are shunting the spikes to ground (or like you said, the inverter is simply generating too high of an output voltage to the CCFLs).

I didn't want you to burn anything out by forcing the 12v onto pin 13 of IC8, except only briefly to make some test.

Do you have access to an ocilloscope? One thing you could try would be to measure the duty cycle of the pulses on pin 11 and 14 of IC8. I believe that as the width of each pulse becomes greater than 50% duty cycle, then the output of the final step up transformers will rapidly become greater. When this happens, the microcontroller should stop IC8. But of course, by putting 12v on pin 13 of IC8, it can't.

Please don't burn out your board. If you do not have any to measure the pulses duty cycles, let me know and we will think of other tests we can do.

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign

i try 12v at pin13 got dim back light at the Q17 drain have 400v BUT,BUT at the output there is 2 parts DCT1 and DCT2 this 2 parts start to be burning,very hot. i think this 2 parts are protect over high voltage.so maybe my p/s making over high volt for the CCFL. if so i try download the oscillator control datasheet SG3525.
If you have any suggestion or any thought pls

thanks

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

If you put 12v back onto pin 13 of IC8, you said that you now have a back light, but it is dim. Try this again and carefully measure to see if you still have 400v on the drain of Q17 and 200v on the drain of Q16. If you do, then either IC8 is putting out pulses on pins 11 and 14 that are too narrow (less than 50% duty cycle) or Q16 and Q17 (with EC17 and EC18) are not generating a high enough voltage as input to the final CCFL step-up transformers. ( I'm sure this will be seen by the D78F9222CX and it will attempt to shut off the 12v to IC8 - which it can't do because you are giving IC8 12v). However, I,m sure it will complain by lighting up the red led on the power board. I have not yet figured out if it will try to shut down IC8 some other way.

If the pulses on IC8 are too narrow, we will try to figure out why next.

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign hi,
at power on i got picture for 1.5sec then black,no audio or voice.got output p/s 5v,12v,24v.blue light on at the front.i shine a strong flashlight onto the screen- there is noting at all.
i checked the 2 mosfets Q16 Q17 and the 2 x 10u cap at the first ,they are new.
at on ,at the derian of Q17 there is 400v and at Q16 the half 200v,and no change when the back light off.
any suggestion?
thanks for your help

[graywebdesign]

Re: Memorex MLT3221 - No Power

mattmarkgo

I would like for you to do the same test that I asked kobkobk to do. Shine a strong flashlight onto the screen and look for any activity like a menu. Because you said that you have stb=5v,5vstb,12v,and 24v (even though you have a dark screen),your low voltage controller IS WORKING. I don't know why your red light on the front of the tv goes out. For the moment we won't worry about it.

If you get activity on the screen, then we can proceed to find the problem with the high voltage inverter.

Let me know how you make out.

[mattmarkgo]

Re: Memorex MLT3221 - No Power

i check con3 there are all ok...

[mattmarkgo]

Re: Memorex MLT3221 - No Power

thank you guys for help!!!

[graywebdesign]

Re: Memorex MLT3221 - No Power

kobkobk

from what you are describing, it sounds like MOST of your power supply is probably working OK. to verify this, turn on the TV (whih blue light on the front) and shine a strong flashlight onto the screen to see if you can see either a menu or a picture of some sort (it will be hard to see-so look carefully).

If you do get something, your stb,5vstb,5v,12v,and 24v should be OK.

The 400vdc on the big cap also sounds OK and I believe that this is the required input voltage to the input of the 2 inverter switching mosfets Q16 abd Q17. If you can, check that these two transistors are OK as well as EC17 and EC18 (not bulging). Also, check fuse F2 (right next to the big cap) as this protects the above mentioned components.

Let me know how you make out and if you want to, we can keep working on your problem.

[graywebdesign]

Re: Memorex MLT3221 - No Power

Dgtech

You are probably right !!!

[kobkobk]

Re: Memorex MLT3221 - No Power

dear graywebdesign
thanks for your fast replay
yes at power on i got 400vdc at the big cap
12v at pin 13,no change at big cap also 400vdc even the ic D78F9222CX stop
the 12v for the high volt.i also try to check the feedback input at this ic controller for that i need the service menu.
again thanks for the help