Hello and thanks from RichFX

RichFX

New Member
I wish to thank BH and several others I've not remembered the names of for the guidance I found here in rescuing my HDR-Fox T2.

My problem was the box suddenly went into the cycle of not completing a boot, no warning.
It took me a while eliminating possibilities, including tidying up a few iffy solder points on the hotter items in the PSU, but this seemed OK as the volts were steady and correct.
However, I verified it worked as a tuner with no leads connected to the HDD, and I also verified the fan worked if given 12V from my bench supply.
It even didn't crash if just the power lead was connected to HDD. But restarting with the SATA lead forced the re-boot cycle.
I flushed it, and upgraded hdf from .12 to .13, but still tuner only.

I bought a new 2TB Seagate (the third HDD for this box), and to start with it was the same, OK with power and I could just hear the HDD but only if no SATA.
However, one of the posts (I think on this forum) suggested plugging in the SATA lead once powered up and working as tuner. Risky, but on a blank drive I've nothing to lose but a drive, no data, and if the box is dead, it's dead.

Bingo.
I was able to format the new HDD, test it, and record a program to it just now, play it back and erase it, as well do some time shifting.
It is now doing a soak test for the next day or so.

The old HDD is now in a self-powered caddy, and the HDR-Fox can read it happily via the USB, so nothing lost.

The only commonality between this crash and one about four years ago on the original HDD from 2012 was I had been doing quite a lot of housekeeping, making space. I assume the directory structure became corrupted in each case by just enough that the cpu could not resolve it. However the caddy tends to fiddle around a bit with this and showed a couple of damaged files on the original HDD, but I've not found any on the most recent HDD (no 2).

I will add a small 12V 40mm Sunon inside the box to stir the air in the direction of the hot bits to reduce local heating.

Thanks again to this community for your help.

Still learning where everything goes on the forum.
 

Black Hole

May contain traces of nut
I bought a new 2TB Seagate (the third HDD for this box), and to start with it was the same, OK with power and I could just hear the HDD but only if no SATA.
However, one of the posts (I think on this forum) suggested plugging in the SATA lead once powered up and working as tuner.
I don't recall any such advice, but if it worked for you that's great.

I can understand a file system might be so screwed up it crashes the HDR, or even a fault in the HDD's SATA interface results in a reboot, but I am struggling to rationalise how a new blank drive persists with those symptoms until initialised or how the (presumably faulty) original HDD then functions perfectly well in a USB caddy.

To me, a main board fault is still indicated.
 
OP
R

RichFX

New Member
You're right, there is a main board fault, but it seems to be intermittent. I'm doing some soak tests and cool-down rests to attempt to pin it down.

The first time it worked properly with the new drive was after I had flexed the board slightly by fitting the SATA lead while it was alive, maybe a coincidence,
The second time I tapped the board near the LDO regulators for the RF modules, and it woke up.
The third time I tried tapping the board near the U52 5V LDO and U24 FET 12V switch for the HDD and no effect, but then a wake up tap worked again nearer those DS9183s. BUT they must be working fine since with no HDD it is always OK. Which makes me think it might even be the SATA connector or a related grain-of-sand component near it, or those tracks going to the hot heat sink device, or the SATA lead.

Next test will be to substitute the Humax HDD lead with a standard SATA lead and power lead from my spare bits.

I measured the PSU performance and it is excellent, 12.5V and 6.7V on full load and very little variation during failed booting; stable while my power monitor showed the variations in mains consumption while it was going round the loop.
 
OP
R

RichFX

New Member
I think I have cured the problems with my particular HDR Fox-T2. Either of two things seemed to be effective, so I have applied both!

1. Additional internal fan
I added a spare 40mm fan I had left over from a project, a nice gentle quiet Noctua and took its 12V from the HDD connector via one of the supplied speed reducer adaptors, and stuck it with DS sticky tape at an angle to simply stir the air, since even an unventilated box this size can easily lose 23Watts. And I put it on soak for a couple of days. This dropped the temperature on the heatsink significantly, but I don't have a measurement for that when the box is closed. However, just once on a start from cold it went through the boot loop for a few times, and then ran OK. So it was not a complete solution.

2. Power supply modification.
After the fan experiment, I changed the 220uF output capacitors for bigger ones (1,000uF on the 5V and 470uF on the 12V, but very untidy over-sized left-overs from the surplus box) for better surge handling. After a week on soak with regular overnight power downs it did not fail to restart.
I traced out the PSU circuit and checked the probable design parameters against the switching-chip data sheet, and bought a set of caps to replace all electrolytics experiencing potentially high peak ripple current, the 120uF/400V, the 1000uF/16V, the 220uF/25V and the 47uF/50V, and I increased the values on some so all the 16V and 25V caps are now 1,500uF/16V. A week with these fitted and still no problems. There is a significant change as well, the peak starting power taken from the mains has increased from about 30W to closer to 36W, suggesting the DC output current is holding up better. It normally runs at about 22W to 23W.

I measured the capacity on the nine-year-old caps I removed; all were between 95-101% of their rated value, impressive for something that has been powered almost 24/7. I assume the new drives just take a higher peak start-up current or for longer, and the higher peak surge rating possible with the total of 4,500uF per supply instead of 2,220 uF has tipped the balance back again to stability.
 

Attachments

  • 20210919_131143_c.jpg
    20210919_131143_c.jpg
    475.8 KB · Views: 4
  • 20210919_143145_c.jpg
    20210919_143145_c.jpg
    460.5 KB · Views: 4

Black Hole

May contain traces of nut
Thanks for your efforts and reporting. I think we had better move this thread somewhere more appropriate (and give it a better title)...
 
Top