Author Topic: Media Player: Power supply dead short (Read 8295 times)

NivagSwerdna · « **on:** January 29, 2016, 11:53:28 pm »

I have been given a Minix X8H Plus media player and told it was broken but no details of why (or the circumstances it was broken)... a bit weird but anyway...

I didn't want to look a gift horse in the mouth so...

I had hoped for a simply supply side fault or a dodgy switch or connector but....

The power supply is a 5V 2A wall wart, checked with the DMM and shows 5V OK

I attached the power supply and nothing happened; I measured between the centre plug (whilst in the player) and ground (on the player) and get 0V. Using my bench supply just gives current limit as the media player appears a dead short.

I was wondering if it is the socket on the player but it looks OK.

As the 5V enters the board it goes to a Fuse which checks out OK

There is then a combo of a diode, two ferrite beads and a big-ish SMD capacitor

I'm thinking that this is un-repairable? What would cause the short if it wasn't a catatrophic failure of the board somewhere deep in its circuit, a surge to the capacitor? Could that cause a short? (Tricky to get at since it is surface mount).

Should I bin it?

Thanks in advance

Shock · « **Reply #1 on:** January 30, 2016, 04:11:25 am »

Shorts can be characterized by a low resistance path but some times it occurs under heat, voltage, current conditions which can make it harder to troubleshoot. As a general rule though voltage drops and resistances across the short when measured are in the low mV and mOhm range.

When they are a little more tricky to find you may need to current limit the circuit and look for warm components indicating drawing plenty of current and dissipating that as heat. Then there is the divide can conquer technique, where you isolate part of the circuit. With semiconductors and IC's you have to look at input vs output, remove and recheck or sometimes test out of circuit.

It's repairable and will most likely be a voltage regulator, capacitor, diode. They can fail from age as well, especially capacitors after a few years.

NivagSwerdna · « **Reply #2 on:** January 30, 2016, 06:07:33 pm »

Thanks.

It's a new-ish device, the previous owner seems reluctant to discuss how it failed... I think they probably plugged in a higher voltage wall-wart. They supplied me with the units original 5V supply and that checks out OK.

In the picture is a surface mount electrolytic. Could that create the short?

rrinker · « **Reply #3 on:** January 30, 2016, 06:20:59 pm »

Quote from: NivagSwerdna on January 30, 2016, 06:07:33 pm

Thanks.

It's a new-ish device, the previous owner seems reluctant to discuss how it failed... I think they probably plugged in a higher voltage wall-wart. They supplied me with the units original 5V supply and that checks out OK.

In the picture is a surface mount electrolytic. Could that create the short?

If subject to overvoltage enough to kill it - it could fail shorted. If it's directly across the + and - on the input - well....

NivagSwerdna · « **Reply #4 on:** January 30, 2016, 10:27:12 pm »

Thanks for the suggestions... I have a hot air combo thing on order, so I'll wait until I have that before trying to get that electrolytic off... Fingers crossed it has shorted. I assume it is intended as a filter capacitor.

Please could you help identify explain the components for me... I know I have an electrolytic, an inductor and a semiconductor but I'm not familiar with SMD labels..

NivagSwerdna · « **Reply #5 on:** January 30, 2016, 10:49:17 pm »

I'm going to guess... just to show I've made some effort...

100uF Electrolytic, G4 is form factor, 255V?

6.8uH SMD Inductor

Voltage regulator of some kind?

station240 · « **Reply #6 on:** January 31, 2016, 12:45:33 pm »

The mystery IC is a 6 pin sot23 package, I wasn't able to figure out what it is however.

Given it's connection to the inductor, I'd say it's a switching regulator (buck mode) to supply the main 3.3V rail. A linear regulator that small wouldn't be able to power everything.

If you can't figure out what the regulator is, then trace the 3.3V rail back from some other the other ICs.

NivagSwerdna · « **Reply #7 on:** January 31, 2016, 06:02:58 pm »

The mystery IC is very near the inductor so the switching regulator theory sounds good.

I did a bit of googling and found a few of these but they seem to have 5 pins rather than 6 pins. I guess it doesn't matter... without a spec I wouldn't be able to replace this anyway?

NivagSwerdna · « **Reply #8 on:** February 07, 2016, 09:56:52 pm »

Hm... well I unsoldered the electrolytic and the short still exists... I even un-soldered the small fuse to isolate the jack and there is still a short across the capacitor pads... I unsoldered the diode and there is still a short... which is a bit weird since I assume the diode was a protection diode so would seperate out the two sides...

I even went as far as unsoldering the inductor and the mystery six legged voltage regulator... and still a short across the capacitor.

cvanc · « **Reply #9 on:** February 08, 2016, 01:19:38 am »

...Could the power connector itself be internally damaged, causing the short?

NivagSwerdna · « **Reply #10 on:** February 09, 2016, 12:35:58 am »

Quote from: cvanc on February 08, 2016, 01:19:38 am

...Could the power connector itself be internally damaged, causing the short?

That's what I was hoping for. But I removed the fuse F1 and the connector no longer has a short between the central pin and the outer so the problem is beyond the fuse.

The next component a long seems to the electrolytic, presumeably operating as a filter capacitor, but I removed that and still get a short across its pads.

I guess I might be following the path the wrong way... there seems to be a zener? and another regulator in the other direction

(Ignore the bodged diode, that's a protection? diode that I removed and replaced very badly,, interestingly that appears to be shorted too.)

Very challenging!

poot36 · « **Reply #11 on:** February 09, 2016, 05:07:27 am »

I am going to suspect one of the ceramic caps (eg C3 or the two bigger ones next to it). Here is how I would approach finding the problem component. Get a current limited 5V supply that will not shut down with a short on the output and feed that in after the fuse. You can then use your finger to find the component that is getting hot (be careful! I have burned myself on one of these caps before when troubleshooting a laptop with a similar fault).

NivagSwerdna · « **Reply #12 on:** February 09, 2016, 08:28:00 am »

Quote from: poot36 on February 09, 2016, 05:07:27 am

Here is how I would approach finding the problem component. Get a current limited 5V supply that will not shut down with a short on the output and feed that in after the fuse. You can then use your finger to find the component that is getting hot (be careful! I have burned myself on one of these caps before when troubleshooting a laptop with a similar fault).

What short of current do you think? I don't what to fuse the board. 250mA? 500mA? 1000mA?

mich41 · « **Reply #13 on:** February 09, 2016, 05:49:47 pm »

Start with low current and if you really can't find anything hot anywhere on either side of the board, ramp it up. If you limit to 0.5A and 5V you can't exceed 2.5W, which should be safe to PCB and components surrounding the shorted one.

If you removed F1, D1, EC1 and U2, all that remains is C3, C4, C14, C15 ... and ... this block of 9 vias between F1 and L4. Which means there can be other things powered from 5V somewhere else on the board. Can you find any?

NivagSwerdna · « **Reply #14 on:** February 09, 2016, 11:44:41 pm »

It's not going hugely well

Firstly I wired a bit of wire across where F1 lives (and checked for shorts in the socket again)...

Then I attached the power socket to +5V 500mA from my power supply and turned on, it quickly went to "CC" with the voltage display dropping to less than one volt. I'm assuming it was passing current. (I have a EL302D TTi dual supply)... nothing got warm. I even cranked it up a bit... nothing got warm.

I decided that I needed to remove a few of the more obvious components and keep trying...

I removed C3, C4, C14 and C15. No joy... still a short across EC1

(Incidentally R7 has been removed; it was a bit of collateral damage from the first time I used hot air and obviously had the fan too high.

)

In desperation I removed FB1 and FB2

There is no longer a short across EC1, this is indicative of the short being further into the board. It looks like FB1, FB2, EC1, C3 and F1 and the socket are good. The short is on the board side of FB1 and FB2. $:-\$

I measured a few things and added this as an image. Lots of things are at ground potential marked "0" on my diagram. If FB1 and FB2 the short will move towards to power jack.

I hadn't previously noticed the 3x3 grid of vias, I need to follow this down and through....

Looks like I'm no nearer finding the short.

poot36 · « **Reply #15 on:** February 10, 2016, 04:51:46 am »

I would redo the limited current into the short test again but this time check the entire board for anything getting warm. I would ramp it up to 1A so you get 5 watts of heat. You should be able to feel that. If it does not heat up try 2A as that is what the power supply is rated for so the board should handle it. Just hope it is not a chip that needs to be programed to work.

mich41 · « **Reply #16 on:** February 10, 2016, 08:07:52 am »

If you do the math, 0.5A at <1V is <0.5W. You'll probably need something like 0.5W-1W, depending on the size of the component. IOW - moar current

One more thing you can do: since now there is a measurable difference between 5V rail and ground, get your DMM and look for things which have the same potential as the 5V input. Try other voltage converters and ceramic capacitors near bigger chips (maybe not the chips themselves, to avoid shorting two of their pins together while the board is under power).

NivagSwerdna · « **Reply #17 on:** February 10, 2016, 08:27:49 am »

I've been up to 2A (which is the limit for my supply) and still can't detect anything warm. At 2A the power supply is indicating 0.73V 2.00A.

It looks like whatever has failed has very very low resistance. So not getting much I^2R effect.

Tricky

In fact 0.73/2=0.36 Ohms.... that's almost zero?

NivagSwerdna · « **Reply #18 on:** February 10, 2016, 10:07:39 pm »

Another long day and then home to tinker...

I decided that I wasn't getting very far in the 'North East' corner of the board so decided that I needed to look for other regulators on the board with the hope that the fault was more in that direction.

I removed the big heatsink and fired up my very rubbish USB 'microscope' and started wandering over the board trying to find some usual suspects...

... however when I found myself in the 'South East' corner of the board I spotted some things that really don't look right.

Firstly, behind the pushbutton switch there seems to have been an area of conflict! Looks like the PCB is cracked and there is a hole disappearing downwards...

There are three components without designations in a box labelled A. One of the components seems to be connected to a via and that via is very close to the deformation.

Secondly C180 isn't where is should be, it has wandered. (This is not an area I have been to so these components have not been at the mercy of my hot air/soldering iron).

When I have time I will remove C180.

Not sure about the damage around the "A" area. $:-\$

poot36 · « **Reply #19 on:** February 11, 2016, 06:38:29 am »

It looks like a lot of current has flowen through that via. Does the solder on the moved cap look like it melted? If so the cap is short as well (or very close to it). I hope you can trace where the via goes or else the chance of repairing this is very slim. Keep in mind that the via could have shorted to an inner trace and caused more problems and that carbonized pcb can be conductive. PS: try and take your photos in at least 640 x 480 or even better 800 x 600, 320 x 240 is just to small.

NivagSwerdna · « **Reply #20 on:** February 11, 2016, 08:18:08 am »

Quote from: poot36 on February 11, 2016, 06:38:29 am

PS: try and take your photos in at least 640 x 480 or even better 800 x 600, 320 x 240 is just to small.

That's a limitation of my rubbish USB microscope. I'll see what I can do.

Yeap the burnt area around the via looks like it might be an unrepairable fault. It looks like a many layer board, all sorts of things could have gone wrong on the lower layers.

C180 looks like it is touching; I'll remove that at my next session.

NivagSwerdna · « **Reply #21 on:** February 11, 2016, 08:49:47 am »

Here's a shot from this morning. Interestingly since the current has been off and the board has had the night to rest it looks like the board damage has subsided? (Doesn't mean there isn't an internal short of course)

poot36 · « **Reply #22 on:** February 11, 2016, 09:50:21 am »

Hmm most USB microscopes are based on webcams. I would check the driver and see if there is any settings for a higher resolution using a generic webcam viewing software. I do know my Microsoft webcam III defaults to 320 x 240 for its built-in picture taking button and as far as I can tell you cannot change that but you can use the generic webcam control software to set it to 640 x 480 and take a picture with that resolution from the computer.

rob77 · « **Reply #23 on:** February 11, 2016, 10:11:39 am »

that sot23-6 seems to be a MP1470 - 2Amp buck converter.

http://www.monolithicpower.com/DesktopModules/DocumentManage/API/Document/getDocument?id=418

most probably the whole device is 3V3 powered and it's coming from that buck converter - so your issue might be that chip

NivagSwerdna · « **Reply #24 on:** February 12, 2016, 08:49:24 am »

"sot23-6" being U2. Yeap. I've removed this from the board. If it is a MP1470 that would work well as removing it definitely isolates the two sides of the circuit.

I removed C180 (the wandering capacitor) and the power demand changed significantly, it is now passing 2.0A at 2.73V so the resistance of the short has significantly increased...

... unfortunately there is now definitely a component that gets hot and that is the main processor.

Finally I removed Q6 the FET around 'Ground Zero' but no change.

When I have time I will repopulate the area in the North East corner as this seems to have been a red herring.

Looks like the actual behaviour here is that an over-voltage applied to the power jack routes directly (possibly when the previous owner pressed the power button to activate the FET) and that over voltage went to the main processor that could not take it and now has an internal short. The high current through the processor subsequently heated up the trace sufficiently to melt the trace and C180 to such an extent that C180 wandered and created a shorter short.

I will re-populate everything and test once more.

I guess I should try and remove the processor to finally prove that is where the final short exists

Sadly, nearing the end of the road


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Media Player: Power supply dead short (Read 8295 times)

Share me