tiistai 26. toukokuuta 2020

TV repair


We've had Samsung LE37M86BC LCD full-HD TV for more than 10 years now, and it has served us surprisingly well. If things had gone as originally planned, it would be obsolete already as DVB-T transmissions were supposed to be replaced with DVB-T2 (which this TV doesn't support) few months ago now, but there was some serious disagreements between companies doing the signal transmissions and that was postponed - so far indefinitely.

At the moment this was already serving as secondary TV, as we got a new one as our primary TV, but since this is still used, I was not happy when it lost TV signal and started rebooting itself every 30 seconds or so just recently. HDMI still worked - except for 10 seconds every 30 or so when it was rebooting. Not exactly fun to watch.

Due to its age the first assumption obviously was capacitors so I brought it to work and opened it up. This thing is built to be surprisingly maintainable, which was nice.


There are two main PCBs there (lower middle and right here), and two more under metal sheets (left and upper middle). Latter ones are very likely both LCD and backlight drivers so nothing too interesting there.

Date codes of components put manufacturing date to late 2007 or early 2008, we I'd guess age of this TV would be some 12 years, with quite active use. Pretty good, considering how badly electronics typically seems to age these days.


Immediately I noticed this capacitor that was obviously goner, at upper right corner of the power supply board. This was to be replaced.

Note on thermal management here. This TV is passively cooled, with cold air coming in from bottom, passing above there boards and coming out from top. Whoever designed this power board did otherwise nice job with layout but this specific capacitor is right next to heat sink and airflow is effectively blocked with those two large inductors. Note that other caps are near heat sink too, but at least they have some air flow going, and they weren't showing bad aging so far (even when measuring voltages)

So I replaced this one (1000u 35v 105 degrees C) with one I happened to have handy (reserved for some other repair, I think, I don't keep this type of caps typically handy). Unfortunately this wasn't enough to fix the TV.

Time to focus on main PCB. Since none of those caps showed obvious signs of pending rupture (unfortunately common with SMD caps, it seems), I went with next plan - measure the voltages. With one hand very strictly behind my back - the powered-up power board is uncomfortably near to main board I was trying to probe.

Nothing immediately obvious showed up, but one capacitor was kinda suspicious - its voltage read at 1.3 volts. It might be okay, but my hunch said it wasn't - that seemed to be way too low for typical electronics. But I also noticed few capacitors underneath that large metal shield next to card slot that might be bad too.

That shield needed to come out.  Unfortunately it was soldered in. Fortunately the solder was very, very easy to pull out. Way too easily for it to be unleaded solder, in my opinion. Slightly suspicious... But hey, don't look in the mouth of gift horse; the shield might have as well had been a real PITA to pull off!

The metal shield apparently functions as combined heat sink and EMI shielding, as main application processor and (I guess) display processor were underneath, as well as two more large capacitors. Voltages of these were okay but showed some ripple, so I chose to replace them too (due to their location underneath the shield - likely very hot location), along with the one I found suspicious before. They are marked red in picture below, and shield location is highlighted with blue here.


If that shield was easy to take off, these things were completely the opposite. Even with two soldering irons set to 370 C they stubbornly refused to come off easily. Hot air wasn't an option due to nearby other components - those would come off first and I absolutely, positively didn't want that. Eventually they gave and I managed to lift them off without damaging copper underneath.

Solder in new caps (all were 1000u 10v, a type which I do keep handy) and TV is working again. The two under the shield still show same ripple, but at least they're fresh now. And voltage of last one - the one that read 1.3v or so before - rose to more correct sounding 1.8v. I think that was the main culprit here.

All together this took me some hour or two of active work and several more waiting (while doing other things) to see what TV does after changes. I'd say this was well worth the effort - with this the TV will likely serve us for several more years, and by then the DVB-T2 situation might be clarified so we can re-evaluate whether this thing is past its prime or still able to serve as media display still.


keskiviikko 13. toukokuuta 2020

ATV electric conversion: final part (or not)

This project has been on hold for a while, mainly due to me being incredibly busy with other things. But finally I found time to put this together and test it. But not without one final modification which I a bit dreaded from get go.

The rear sprocket I originally had had 54 teeth, and combined with motor's 11 teeth this gives relatively low down gearing. I feared that this might cause issues getting the ATV moving as motor might not be able to generate enough torque to get moving.

It certainly wasn't powerful enough to move me, even on perfectly flat terrain, but with kid's weight this wasn't too bad an issue (as we tested outside) but I wasn't too happy with it either. This ratio also gave me theoretical maximum speed of 30 km/h, which I deemed a bit too fast for comfort, even if it could not realistically be reached in real world.

So I ordered new chain and rear sprocket, this time with 68 teeth. This dropped max speed to 20-ish km/h (according to my notes, it's been a while since doing the calculations) and gave nice boost to zero-speed torque. It isn't exactly powerhouse still but it should move nicely in slightly uneven terrain too.

Time for test drive!


Well, it worked. For a while. Then, as battery was getting near-empty, kid tried to drive on upward slope on wet-ish mud, overloading the engine and eventually burning it out. So my fears about gear ratio were completely justified, but mitigation with larger rear sprocket was not enough.

Back to the drawing board.