Red Lines on VFD - DVD Player Stopped Working

[paul_g_787]

I have a weird issue here.

I have a HDMI DVD player I have had since 2008. It is a Silvercrest brand player (not sure of model until I pull it out the cabinet) which I got from Lidl in 2008. Up until last night is has been a great player!

I was watching a movie last night and all of a sudden the HDMI signal just dropped. I looked at the players display and it had 'frozen'. The time counter has just stopped but the pause icon was not illuminated. 

I tried turning it off with the remote but it would not respond. So I unplugged it from the mains and left it for 1 minute before connecting it back up.

When it connected back up the display looked like this (see picture). In this state the unit is dead. it will not respond or do anything at all.   

The display is getting very toasty warm in this state so I powered it down right away to avoid further damage.

Now I have done some research and Youtube has taught me that these lines are the Cathode filaments of the Vaccum Fluorescent Display.

Now in normal operation they don't glow red so my instinct says that perhaps this might be a power supply issue?? Maybe bad capacitors?

Has anyone encountered a similar issue before? And if so what might be the cause here.

[Gyro]

Yes, that's a severe overvoltage on the VFD filaments. The worry is that the other supply rails from the PSU have gone up by a similar amount, which would almost certainly be fatal to the ICs.

I don't think such a fault would be down to a capacitor problem, they tend to be more subtly, causing unreliable operation, at least initially. This looks like a sudden loss of PSU regulation, maybe a semiconductor failure in the feedback loop.

It's hard to say more without knowing what has happened to the other supply rails. The player is clearly well out of warranty, but I don't know how confident / competent you feel about working on mains powered equipment.

[paul_g_787]

Yes, that's a severe overvoltage on the VFD filaments. The worry is that the other supply rails from the PSU have gone up by a similar amount, which would almost certainly be fatal to the ICs.

I don't think such a fault would be down to a capacitor problem, they tend to be more subtly, causing unreliable operation, at least initially. This looks like a sudden loss of PSU regulation, maybe a semiconductor failure in the feedback loop.

It's hard to say more without knowing what has happened to the other supply rails. The player is clearly well out of warranty, but I don't know how confident / competent you feel about working on mains powered equipment.

I am hoping it has not fried the ICs 

Looks like I will be pulling the player out. I will take some photos of the inside.

I guess a good first step is to see if the power supply can be disconnected and I will test the voltages of the outputs rails.


Yes I am more than confident/competent at working on mains powered equipment and well aware of the potential dangers.

[Gyro]

That looks like a good first step (isolate and measure) then. Best case would be a fault just on the VFD filament supply, but the fact that all other functionality ceased at the same time seems a bad omen.

[paul_g_787]

That looks like a good first step (isolate and measure) then. Best case would be a fault just on the VFD filament supply, but the fact that all other functionality ceased at the same time seems a bad omen.

I know

I really hope it is something simple. Would be a shame as this player is great.

Will post back in the next few days with my findings and photos.

There was no magic smoke or anything with a funky smell. But even so could be serious.

I am assuming that the VFD filaments would run on a high negative DC voltage?

I have never really worked with VFDs before. To me it seems like the work just like a valve.

[Gyro]

The VFD filaments normally run on some form of AC - traditionally from the mains transformer, but in these SMPS days probably from a none rectified winding on the SMPS(?) or an H-bridge driver.

The VFD is a valve in all meaningful ways. The filaments are directly heated cathodes so, if fed from DC, the brightness of the display would vary from one end to the other. The filament is also biassed negative (by something like -25V) relative to the display segments and grids. This allows these to work at roughly logic levels.

One random thought... The filament bridge drive might be controlled by outputs from the micro, but even stopping the bridge completely wouldn't affect the temperature of the filaments so it still comes back to PSU.

Good luck with the detective work, hopefully we can provide help after measurements and photos.

[shakalnokturn]

Since it seems to have lost regulation if you do power the PSU disconnected from the rest for troubleshooting beware that it doesn't pop an electrolytic at your face.

[commongrounder]

I’ve seen this on units that came through my service facility.  Typical switching power supplies in these use one of the output supply rails (usually the +5v or +12v) to drive the regulation feedback loop to the “hot” side of the supply through an option-isolator. The problems arise when the caps go bad on that rail. The voltage drops, which drops the feedback, the hot side drive increases in an attempt to bring it back to normal. The result is all of the other rails go completely over voltage. Now, well designed power supplies have zener clamps on each output to prevent damage to the load devices. The question is what happened to your unit? My guess is recapping the power supply will restore proper voltages, and you can see if the chips survived.  Good luck!

[TheBay]

I'd say there is a good chance it's capacitor related in the SMPSU, if you're lucky you may actually end up with a brighter VFD after this fault as the VFD cathode filaments have been run for a little bit with too much voltage.

[paul_g_787]

Ok this is all good stuff! Thank you.

I will see if the power supply can be disconnected and check the output voltages.

I have an ESR meter. I will check all the power supply capacitors and see what I find.

[Gyro]

Note that if an electrolytic cap has suddenly gone short circuit due to overvoltage, it might look as if it just has a very low ESR. It's always worth checking on a DMM DC ohms range too - or check that the related PSU output rail isn't shorted beforehand.

[paul_g_787]

Ah yes good call. Will bear that in mind, thanks!

[DaveG8KBV]

I sort of wonder if the initial symptom, of the HDMI signal quitting and the display freezing (but no red lines?) indicates a shorted decoupling cap on a digital board somewhere.  Or failed unobtanium IC. 

But concur with the comment that pulling the main regulated output of a cheap multi rail SMPS could cause the other rails to run high.  But, I've not personally seen VFD filaments run direct from the main SMPS.  More usually from an onboard inverter that also provides the -25(ish) voltage.   But, there again, I rarely work on "consumer" stuff.

You may find, when disconnecting the PSU from the rest of it, that the main rail (+5 or +3.3) is fine, but the others are now low.   That would indicate the PSU may be OK.  (Check by loading the main rail to something like half an Amp, and see if the others come back, so long as that main rail doesn't collapse under any load, however small, in that case the PSU is likely duff.

SMD Tant' caps are prone to suddenly going short cct, especially those placed across a power rail.  (Don't ask how I know..)  They don't tolerate surges too well either.

Even if you can't fix it, you'll learn a load about fault finding, that will do you well in the future.

Good luck.

Dave 'kbv.

[paul_g_787]

OK so here is my player. I cannot see a model no. on it anywhere mind you.

[paul_g_787]

Here is the inside.

First thing I see is a swollen capacitor!

[paul_g_787]

This is the power supply board.

I will check the output voltages of this board next and post back.

[commongrounder]

Here is the inside.

First thing I see is a swollen capacitor!

Yup!

[paul_g_787]

On this connector I have the following measurements.

Between the two pins marked AC3.5V I have 6.0V DC and 0V AC 

All the following measurements taken with respect to pin marked GND. I also took the measurements with the probes reversed and have the same readings.

-24V Pin I have -45V DC  and 0VAC - This is not good at all!
5V I have 4.99V DC and 0V AC - Perfect here
5V-ST I have 4.99V DC and 0V AC - This seems good I think
STBY I have 6.0V DC and 2.25V AC - Is this showing a high ripple current??

[paul_g_787]

And on this other connector I have the following measurements.

Both 5V pins 4.99V DC and 0V AC 

+12 V Pin is at +24V DC and 0V AC 
-12 V Pin is at -24V DC and 0V AC 

[paul_g_787]

I notice on the primary, R5 is a bit toasty looking.

It appears to measure 10K
It is a bit hard to tell on the colour bands though.

I also tested the diodes here for good measure and they are all fine.

[TheBay]

No point checking voltages until the faulty capacitors are replaced.

[paul_g_787]

Just checked all the semiconductors on the secondary.

There are no direct shorts on any of these which is a good sign!!

Yes, my next step is to change the capacitor that is obviously faulty. And I will check the ESR of all the others with my MESR-100 meter.

[paul_g_787]

the swollen capacitor is a 2200µF 10V and is measuring 6Ω ESR. So definitely dead by the look of it. I will replace it.

The one next to it is 1000µF 10 V and has a 0.5Ω ESR which my meter says is dead. I will change this also.

Those are the two blue capacitors in the middle. I will change these right away and see what happens.

The other capacitors with red dots on are right at the top threshold of their ESR but I think they are still OK. I will leave these in for now.

The capacitors I haven't marked have good ESR.

[shakalnokturn]

It looks very similar to some players made by Peekton, then again a lot of these cheap players are similar regardless of the name on them.

[paul_g_787]

OK, so I have found some capacitors in my spare parts to replace the two faulty ones. These have been soldered in.

I also removed R5 and had a look in the bright light. It is definitely Brown / Black / Something?? 

It measures 9.996K so I am convinced it must be 10K as this is so close. I have soldered in a new 10k 0.25W resistor. The original was 0.125W. Not sure what caused it to burn up though..

I am going to plug it in now and see if there are any explosions.

If there are no explosions I will measure the output voltages again.