No commercial links

Sorry, I had to close comments due to permanent spam. Too much cleanup work.

2016-01-26

Sony micro HiFi system CMT-G2NiP - not starting - defect voltage regulator toko 73400 and shorted capacitor

I took this micro receiver home on new years eve. It belongs to my friend's neighbor. Coming home at 2:30am, I could not resist to check it out. The device wouldn't start. A click, standby red and then nothing.




After some poking around I found that a voltage regulator on the processor/main board was getting very hot. It is a Toko 73400 and it did not produce any reasonable voltage at its output. I measured the output line against ground and it was short. Then I lifted the output pin and measured again. The short was on the IC. The plus line had no short. At least that was what I thought I had measured then...



Ordered 10pcs for little money from Aliexpress. Almost a month later, the chips arrived.

The chip had not been soldered to the pad properly. I applied solder to the pad with the iron and reflowed a new chip with hot air.


With great expectations I started the device. Nothing. Damned.

Measured again. There was still a short! It couldn't be the new regulator. More likely, I screwed up my first measurement. I started unplugging the many connectors one after another. It turned out that one connector was linking M-GND to GND. That shorted the regulator output to GND. There was a short between M-GND and the +7V line on the main board. M-GND had no connection to GND unless this connector was plugged in. I poked around to find something and after a while I hit the large black e-cap you see on the left side. It had the same reading against M-GND on both pins. This sucker did not look suspicious at all. No buldge. The ESR reading was good, because it had a short!
Ahhh, measurements...sometimes they are your enemy :-/

It was cheap crap of a brand I had never seen before. Shame on you, Sony! I replaced it with a decent Nippon Chemicon. This fault will never appear again!



The receiver is working all right again. Fixed for 3€ parts. In a repair shop, they would have swapped the whole processor board for a lot more bucks, if at all.


I am not entirely sure whether my first assessment of the regulator defect was right. Well, not much time and money wasted and a new one can't hurt, as it had to withstand a short. Now I have a batch of 9 adjustable, robust 1A, 1.3-12V low drop voltage regulators left. They might come in handy sometime.

2016-01-07

Philips 37PFL9604/H12 - distorted colors / solarization - AS15F chip defect on TCON board

Same procedure as last year, Miss Sophie!

A colleague asked me to fix his TV. From the symptoms he described, I was pretty sure this would be another AS15F chip case. Philips TVs of the generation 2008/2009 will all break after about 6 years it seems. The AUO TCON boards have this built-in bug.

This image is typical for a mild AS15 failure. Shimmering magenta colors where they shouldn't be.




This is what it should look like. The boot image is actually an excellent test for gamma failures. I wish it would stay longer for testing. Any aberration in the blue gradients to other colors is a clear sign of gamma problems (the magenta stripes of the moiré patterns don't exist in reality).



Let's have a look into the machine:


The 9er series has nice features like big speakers and built-in WLAN. The back cover is removable without loosing the stand screws (that's extremely rare), and the TCON is not buried under the stand. Nice one, very serviceable.

So here it is, the blue trouble child. The AS15 is in the upper right corner.


On the heater plate with 160°C and with the hot air gun at 420°C, the chip comes off easily.


The first attempt was almost successful. The image didn't look 100% right. It was difficult to pin down its faults. Some dark areas in some scenes where too bright. Other scenes looked perfect. In other words: some gamma value had to be wrong. How can this be? It is not unusual that new AS15 chips don't work perfectly. I swapped the chip again and this time it looked good.

Here are the VGAMMA measurements of the first chip and the second chip in brackets. As you can see, the first chip isn't completely broken, but he second is much more continuous, even though there are a few jumps in it.

v1: 15.17 (15.2)
v2: 15.09 (15.12)
v3: 9.5 (10.24)
v4: 8.5 (9.55)
v5: 8.4 (9.24)
v6: 5.9 (7.08)
v7: 5.85 (7.0)
v8: 5 (6.25)
v9: 5 (6.2)
v10: 3.75 (4.6)
v11: 3.57 (4.4)
v12: 3.13 (3.9)
v13: 0.280 (0.333)
v14: 0.235 (0.271)

It is a mystery to me how some of the chips work in such a distorted way. I always handle them the same, therefore I am pretty sure it isn't my fault. Some work, some just don't.

Et voilĂ . Everything is back to normal. Frau Johansson was not available today, so I kindly asked Miss Evangeline Lilly for a test.



This TV is able to produce very dark blacks. I think it has a special filter layer in the panel. The orange and yellow tones are a little weak, but I am a spoiled plasma viewer :-)

The sound is the best I have ever heard in any TV thanks to the proper two-way stereo speakers, which live in their own, closed compartment.

Due to high-quality parts everywhere I could not find any suspicious capacitor. All made in Japan and good for many more years.

The higher series Philips from 2008-2010 are just very good machines.