Start Up Fails When HDD Connected

Hey guys, I've just discovered another 5 LDOs on the underside of the PCB! What is it with Humax and LDOs??

View attachment 5825

View attachment 5824
They're a good way to isolate circuits for saving power or testing or en/disabling 'features.'

I used to place the extra-low voltage switching regulators (fed from a 5V or 3V3 general power plane), right under the middle of each big fat FPGA for the core voltage, and their power startup could be sequenced as required and (depending on the chip type) sometimes even turned off for testing
 
That doesn't seem particularly applicable for the HDR-FOX. What I think is (a) reduced impedance on the power distribution, and (b) reduced dissipation in each regulator eliminating the requirement for expensive heatsinking.
 
Woody17 kindly sent me the duff 4u7 cap he removed from his PVR for me to check. As expected - replacing it did fix his PVR, after all - it is one sick puppy. I checked it with various instruments and initially found the results varied quite a bit, depending on the particular instrument, the range selected, etc. Its value was too low for the Peak Atlas ESR60 to measure but on my Peak Atlas LCR40 it now consistently reads just over 1n. 1nF is 0.001uF - so nowhere near 4.7uF.

My repaired PVR has been working fine for around 10 days now; no problems with the replacement SPDIF transmitter - and we can now actually see the display.

I was wondering, as we have a prospective solution to the problem stated in the thread’s title (certainly a “first-thing-to-try” solution) is there a way to signpost the casual visitor to the site facing a similar problem (as I was at the start of November) to my post #122?
 

Attachments

  • Woody17 cap 1.JPG
    Woody17 cap 1.JPG
    328.6 KB · Views: 16
  • Woody17 cap 2.JPG
    Woody17 cap 2.JPG
    315.1 KB · Views: 16
  • Woody17 cap 3.JPG
    Woody17 cap 3.JPG
    314.6 KB · Views: 16
see if replacing that cap fixes it today, on plus side if it does I now have a 1tb drive lol

did you all have same symtoms, for months before odd recording failed, then odd time youd turn on get reboots, but usually pressing power button few times got it on, then eventually nothing got it on, but leaving the reboots for 1-30mins u eventually heard hard drive power up then you could get it on? also failed recordings became more frequent
 
Well just completed doing this

Before
1640443484160.png

After
1640443520721.png

Is anyone elses bottom pcb like this
1640443547925.png

I had to clean it all off lol

So far first boot was instant, so hopefully its all fixed
 
Last week I inherited another T2, same age as mine - both bought just before 2012 Olympics, this one with the Humax seal still intact, but rarely used since about five years ago.
It also had developed intermittent failures to wake up, requiring on/off switch.

So I put the Picoscope on the two power rails to watch them during boot, with and without the original 500GB hard drive. They were always OK, never dropping more than about 0.2V even at peak mains demand with HDD connected.
I looked at the control signal going to the integrator around that SM 4.7uF and it can be noisy, suggesting startup might be erratic elsewhere (beyond me to search this out if the factory had known and applied a bodge fix).
This C was not integrating as it should so I tagged a leaded one to the pads, leaving the old SM in place.

This has restored normal operation for the last 48 hours now, and the original power supply is untouched.

Thank you Newcoppiceman for digging deeper than I was prepared to do!
 
Is anyone elses bottom pcb like this
Yes, although not quite as bad. I'm usually fanatical about removing flux residue from boards I work on but resisted the temptation this time as I wanted to minimise the work to the board in case I disturbed/damaged yet more things (I'd already blown-up a Mosfet). You didn't need to remove the main board to replace the cap, of course, so maybe it was down to simple curiosity.
 
if i had the right tools, to be able to get into remove capactior easily I wouldnt have, but only had bigger cutters, but used the below method


So a for few screws was worth the effort, well worked first time everytime, no missed recordings last night

Had the box from 2010, so a 30p capacitor + the £3 quids worth I changed prior, isnt too bad :)
 
if i had the right tools, to be able to get into remove capactior easily I wouldnt have, but only had bigger cutters, but used the below method
That approach is brutal! Damage to the board itself (eg lifted tracks) is the main risk when undertaking a job like this and readers are strongly advised to see my post #102 (and others thereabouts - I joined the thread at #67) for a more board-friendly solution. Nevertheless, congratulations on successfully applying the fix. Like you, I started with the power board, as this is the natural first suspect - but this fault was more subtle and I was fortunate to identify that it was (in my case) temperature-related, which opened the door to tracking it down. It's clear failure of that cap can now be considered a "stock fault"; to the wider readership, have we collated a list of any others somewhere?
 
have we collated a list of any others somewhere?
I think this is the first to be pinned down, other that those mentioned in HDR-FOX Commissioning, Disassembly, Repair (click)

I was holding fire until able to confirm your findings, but the evidence is stacking up. Congratulations on finding a fix – even if people are unable to apply the fix themselves, it will enable a repairer to keep these boxes going a bit longer.

I can't work out how an out-of-spec capacitor is responsible for my 'scope observations, but then again a lot of the exact circuit remains undetermined, and if the SoC is involved that's a black box anyway. To those who suggested dried out capacitors might be at fault at the outset: it seems you were right, and at least that makes it fixable.
 
Hi I am a new member. I was looking through this thread looking for a solution to the continuous looping fault at START SYSTEM on my Humax Fox T2 Freeview box with the 500GB HD. The unit had been working perfectly every day for many many years. I've forgotten when I bought it. Anyway I came upon Newcoppiceman's fault description i.e. Remove HD power plug and it re-boots perfectly with no HD facility. I then measured the pins on the U52 5V regulator IC with a voltmeter and found that with the HD power plug connected, pin 1 CTL on the IC was at near zero volts. Every time the POST looped there was a small blip or pulse of 0.7V. This I think is the system trying to start. Both the 5V and the 12V on the plug were missing during this phase. Of course when I removed the HD power plug both the 12 and 5 volts became present. After a bit of thought and reading through the postings here I began to think that maybe a capacitor around this area may indeed be at fault by being too slow to charge (a guess). I then opted for the 4.7uF 35V electrolytic Newcoppiceman solution. I twisted it off with a pair of pliers placed over the can after checking the polarity. The legs on the cap break easily doing this as the solder is so soft. Then I sucked up the remnants of leads on the pads with a hand solder pump. I soldered in a conventional 4.7uF 50V capacitor with little right angle bends on the legs to get a good strong joint with the pads. I re-assembled and 'lo and behold' success! One less piece of electronic junk in landfill. Many thanks for your postings. I think this is the third unit with this fault that worked with the 4.7uF 35V capacitor replacement. I think you will need to be specific with the symptoms for this remedy to work.
 
I then measured the pins on the U52 5V regulator IC with a voltmeter and found that with the HD power plug connected, pin 1 CTL on the IC was at near zero volts. Every time the POST looped there was a small blip or pulse of 0.7V. This I think is the system trying to start. Both the 5V and the 12V on the plug were missing during this phase. ... Then I sucked up the remnants of leads on the pads with a hand solder pump. ... One less piece of electronic junk in landfill.
Another satisfied customer - congratulations!

The attached shows what's really going on with the HDD supplies (top: 12V; bottom: 5V) - in this case it all comes good after the dud cap has warmed up enough (HDD supplies switch off at end as this was a "mains-on" start).

In one of my posts I cautioned against using solder pumps on boards like this because of the risk of solder splatter and attendant shorts/intermittentcies. Solderwick/mop is much safer (and more satisfying to use!)

To keep your PVR from landfill longer, consider investing in a spare remote - you might be able to extend a remote's life once its buttons start becoming unreliable by dismantling and cleaning, but I've just bought a couple of brand new Humax originals on eBay for £23.95 delivered. The PVR is pretty useless without a working remote, after all.
 

Attachments

  • 04 Faulty PVR, HDD 12V+5V, xxs sweep, bad startup.jpg
    04 Faulty PVR, HDD 12V+5V, xxs sweep, bad startup.jpg
    177.6 KB · Views: 25
Another satisfied customer - congratulations!

The attached shows what's really going on with the HDD supplies (top: 12V; bottom: 5V) - in this case it all comes good after the dud cap has warmed up enough (HDD supplies switch off at end as this was a "mains-on" start).

In one of my posts I cautioned against using solder pumps on boards like this because of the risk of solder splatter and attendant shorts/intermittentcies. Solderwick/mop is much safer (and more satisfying to use!)

To keep your PVR from landfill longer, consider investing in a spare remote - you might be able to extend a remote's life once its buttons start becoming unreliable by dismantling and cleaning, but I've just bought a couple of brand new Humax originals on eBay for £23.95 delivered. The PVR is pretty useless without a working remote, after all.
Yes after I sucked off the solder and cap leg pieces from the pads there were a couple of solder shards lying loose over the area which I had to pick up with sticky tape. A solder pump is not the tool for the job on surface mount boards.
 
Back
Top