Need an alternate DC-DC converter recommendation please?

[Tracker]

Hi folks,
My Rohde & Schwarz FSEA 30 has developed a fault, the screen died in front of my eyes 
I have traced it down to a faulty voltage converter on the LCD driver board (made by Hosiden & Philips).
The LCD unit is HLD0921-013020, available used from China on eBay but I really think I can fix this one to be honest.

On the LCD display there is a DC to DC converter (M5291F) which should produce about +12v from a +5v input.
Below is a sample diagram from the datasheet which seems almost exactly the same as the implementation by Hosiden !


The output NPN transistor is being driven hard on with 0.6v at its base from pin 2.
After about 30 seconds the Transistor gets warm, same with the 10uH choke.
Output Voltage is only around 3.8v with an input voltage of 2.8v (it's current limited with a bench PSU to 300mA) the LCD display unit is rated to run off +5v with a max current of 300mA (from its datasheet).
I suspect the 5291 DC-DC converter is faulty... all related resistors and caps  test out just fine.
Do you folk concur ? anything else I can check?

I'm looking for an alternate to the M5291 which I can pick up locally, would anyone have a recommendation ?


Thanks for looking

Cheers  !

Tracker

[DaJMasta]

Are you sure this is the right part and that it's faulty?

LCD drivers of that vintage are going to use LVTTL for screen data (shouldn't need +12V nor +5V), and they're going to use ~1.5kV AC for the cold cathode backlight.  The inverter board is usually easy to spot (two widely spaced output terminals go off to the backlight), and typically runs from +12V.  Any LVTTL translation or decoding that happens would be on the data input connector to the panel, and as mentioned likely doesn't use either voltage, it probably runs 3.3V.

Given the age of the FSEA series, it's very unlikely that the panel uses an LED backlight unless the screen has already been replaced.

[Tracker]

Thanks for the reply.

Actually the backlight is a totally different circuit on this display, it has its own HT supply and that is working just fine, backlight is bright!

The problem lies in a PCB assembly associated with the TFT color display.
Pixels are all 'off' with just a little of the backlight visible.
When I powered on the analyzer the screen came on briefly then after about 5 seconds it just faded away until it was all grey.


See image, I have circled the step-up converter.


According to the TFT specs, it runs on 5V at 300mA.

[fzabkar]

Can you pick up a 12V feed from elsewhere within the instrument? I don't see an Enable pin, so ISTM that the 12V supply is present whenever 5V is present.

[Tracker]

Hi fzabkar,
Yes I guess that's an option, and I might just do that to prove that I only have one faulty DC-DC converter... and the CFL inverter runs off 12v.
But I do want to find a better final solution.

I'm getting more confident the DC-DC chip is at fault, I substituted another NPN transistor and still had the excessive current draw.
However I do measure 50 kHz on the transistor base... so the Chip is trying to work.
I can also put 12v on the smoothing caps (after the diode) and there is no current draw, so this means nothing is pulling the output down.

the Pinout of the IC looks standard enough for a DC-DC device, i wonder if any DC-DC boost chip will do the trick?

 

[fzabkar]

Could the inductor have shorted turns, or could the diode be leaky? Have you tried disconnecting the 12V load and substituting your own dummy load?

BTW, I did search for a substitute but came up empty. I tried searching for PDFs containing "1.17" OR "1.17V", as that is an uncommon feedback voltage. I found a few, but none with 8 pins


Edit:

See page 21 of this document:

http://www.1688eric.com/upload/pdf/2012-9-17/M62503FP.pdf

It lists these equivalent parts:

μPC1100 (NEC)
AN8011S (Panasonic)
FA7611M
FA7622M
M62210
M62211
M62212
M62215
M62216
M6227X
M6229X
*HA16114
*HA16120
*HA16116
*HA16121

Datasheet for Renesas M5291P/FP:

http://static6.arrow.com/aropdfconversion/eee2388f05ccb8854d3d8575ab3c433da9d7e899/3rej03d0841_m5291pds.pdf

Edit 2:

I don't know why those parts are classified under "cross reference", but none of those appear to be a match. Maybe something got lost in the translation.

[DaJMasta]

Is there a blowhole in D1?


If it's switching at 50kHz, there are not too many parts that could be bad, and it's unlikely the controller is one.

[Tracker]

Thanks guys, I really appreciate the input.

@fzabkar
I think the bottom note might explain why the 'equivalents's are not an exact match...
Note2) Electrical characteristics and pin arrangements differ among equivalent.

The inductor... I had the same thought last night, what if it was not generating the required field during the off cycle... but when measured it's 150 uH, correct as per it's marking (151).

and I can't easily disconnect the 12v, but I can prove otherwise that there is no significant load on the converter.
For example, voltage drop is practically zero across a 5cm strip of narrow PCB track, and I can apply my own 12v to the load and there is very low current drawn.

@DaJMasta
No, D1 is fine, it measures good, it was just a price of flux.
better image below... 
I agreed its odd that the converter is generating 50kHz, so it could be ok.
The other possibility is a leaky Transistor I guess, It tests ok ar DC but there might be a problem at 50kHz.

Im going to do some digging on the parametric tables over on Digikey, looking to find a similar device.
and I might just replace both the Convertor and Transistor if I can find equivents.

Cheers!

[fzabkar]

Disconnect the base drive from the transistor and see if there is a C-E leak.

[Tracker]

Thanks for the suggestions, I ended up just ordering a replacement screen from China, all is good now. New screen is excellent, also added LED strip instead of CCFL, looka amazing now.