Hardware Support - UNO 4k Hardware Issue - Debugging | vuplus-images.co.uk

Hardware Support UNO 4k Hardware Issue - Debugging

Joined
Mar 18, 2018
Messages
11
Likes
2
Points
3
Location
UK
#1
Hi

I have a Uno 4k that is 'not working'. Symptoms as follows:

(1) Power light comes on
(2) No HDMI output
(3) Power light flashes when buttons on remote are pressed
(4) Does not update when USB stick is inserted with new image (power light stays on and does not start blinking)

I have opened the unit and checked the regulators as I know that these have been an issue on earlier hardware models. I measured the following voltages from the various regulators with a multimeter which seem to be ok for the BROADCOM BCM7252s CPU:

0.9V - U134
1.5V - U103
1.8V - U104 (maybe this should be 2.5V??? but unless I have a datasheet or schematic for anouther receiver I dont know how to verify this)
3.3V - U106
5V - U105

I could not find a datasheet for the BCM7252s but found details on a receiver using the same chip (https://www.syscomtec.com/_produktb...666/BrightSign Hardware Serie1 2 3 Manual.pdf page 38) showing that voltages of 0.9V @10A, 1.5V @3A, 2.5V @4A, 3.3V @4A and 5V@3A are required (design specific but the voltages should be the same for the Uno 4k and current requirements would be a relative indicator).

There are no electrolytic caps inside this unit which I know can be problematic.

I will proceed as follows (and this should preferably have been done before opening the unit but I dont have the cables/parts at hand yet):

(1) check RS232 output from the receiver on the PC
(2) check for composite video output
(3) try to update the flash again with a different USB stick

Can anyone tell me what the pinout is for the composite video output (3.5mm microphone jack with 4 pins) connector?

The Uno 4k is a relatively new model so I doubt many have had issues with the hardware. However, I will keep on updating and hopefully anyone experiencing issues in future will be able to refer back to this.

Please feel free to suggest any debugging steps that you think might be useful.
 
Joined
Mar 18, 2018
Messages
11
Likes
2
Points
3
Location
UK
#3
I have tried connecting to a network with an ethernet cable. The ethernet LEDs do not light up. So I'm starting to believe that this might be corrupt firmware or a hardware issue.
 
Last edited:
Joined
Mar 18, 2018
Messages
11
Likes
2
Points
3
Location
UK
#4
I have now hooked up the serial port. Log below. It seems that there is an issue with the memory.



single board

eMEMC:
M
00000000
ePHY: 00000000BLD:10360 HW:hp
eSHMOO: 12400000 00000082 00000000CPU 01
MCB: FLEX
A
BCM74390011tus=000000f
PRID07252111
v1.10 Curr
RR:0000020300000395(91
BFW
seclib-1.5
BBL v1.0.1emperature
AVS init...024)
AVS init OK

AVS load:select_image: addr_offset: 00051000 part_offset: 00000000 bootStatus: 20333278 [0.3.2.x]
OK
DDR0: !@ ffe0e780 <= ffe0df1
a0f2b00-0 @ f1
AVS load_code offset =00051000
MEMSY
LOADED: 0000
AVS: overtemp mon ON

eMEMC:


single board
ePHY:
MEMSYS-STD
SHMOO 01020100 BLD:10360 HW:hpf0_b1p6 V:1.2.1.0
BCM7439
MCB: FLEX
PRI
AVS start:status=000000ff0
RR:00000203
STB: Current voltage=00000
BBL v1.0.1
PRID0725211100ff
v1.10
RR:00000203Current vol
BFW=00
seclib-1.5
BBL v1.0.1
AVS init...ture=00008c
AVS init OK
AVS load:select_image: addr_offset: 00051000 part_offset: 00000000 bootStatus: 2
AVS FW rev=30333278 [0.3.2.x]
OK
DDR0: !@ ffe
a0f2b00 ffe0df
AVS load_code offset =00051000
LOADED
ME
AVS: overtemp mon ON

single board
00000002
MEMSYS-STDPHY: 00000
AVS FW rev=30333278 [0.3.2.x]
OK
a
DDR0: !@ ffe0e780 <= ffe0df10 MEMSYS-0 @ f1100000
LOADED
AVS: overte
MEMSYS ERROR: 00000001

single board

eMEMC:
MEM
00000000
S
ePHY: 00000000D:10360 HW:hpf0
eSHMOO: 12400000 00000082 00000000CPU 01
MCB: FLEX
AVS
BCM74390011s=000000ff
PRID07252111

v1.10urren
RR:00000203000395(917)
BFW
seclib-1.5
BBL v1.0.1perature=0
AVS init...4)
AVS init OK
PV
AVS load:select_image: addr_offset: 00051000 part_offset: 0000000
AVS FW rev=30333278 [0.3.2.x]
OK
AVS init OK
PV=00000351(849)_image: addr_offs
MV=00000393(915)offset: 00000000
AVS FW rev=30333278 [0.3.2.x]
OK
DDR0: !@ ffe0e780 <= ffe0df10 MEMSYS-0 @ f1100000oad_code offset =00051000
MEMSYS ERROR: 00000001: overtemp mon ON

eMEMC:
00000000board
ePHY: 00000000STD
S
eSHMOO: 12400000 00000082 00000000CPU 011.0
BCM74390011: FLEX
PRID07252111t:status=000
v1.10
RR:00000203
STB
BFWrre
seclib-1.500000
BBL v1.0.1
LOADED
MEMSY
AVS: overtemp mon ON

AVS start:status=000000ff
STB: Current voltage=00000392(914)
temperature=0000926d(37485)
PV=00000350(848)
MV=00000392(914)
AVS FW rev=30333278 [0.3.2.x]
OK
DDR0: !@ ffe0e780 <= ffe0df10 MEMSYS-0 @ f1100000
MEMSYS ERROR: 00000001

eMEMC:
00000000
ePHY: 00000000
eSHMOO: 12400000 00000082 00000000CPU 01
BCM74390011
PRID07252111
v1.10
RR:00000203
BFW
seclib-1.5
BBL v1.0.1
AVS init...
AVS init OK
AVS load:select_image: addr_off
 
Joined
Mar 18, 2018
Messages
11
Likes
2
Points
3
Location
UK
#5
and a google search for "MEMSYS ERROR: 00000001" gives plenty of hits for the dreambox 9xx units :-(
 
Joined
Mar 18, 2018
Messages
11
Likes
2
Points
3
Location
UK
#6
Ok, a follow up. I noticed that pressing down on the memory chips and processor chip resulted in a slightly different console log. Not enough to get things going. My suspicion is that the things are getting really hot in the box and a gently nudge at the wrong time coupled with sufficient cooling effect (contraction) results in the BGA balls getting unseated. This occurs on the pins that is connected to the memory making it look like a memory error (RAM chips). A similar phenomenon occurs with gaming consoles, graphics cards etc.

I took a heat gun and gently reflowed the memory chips (4 x DDR3 devices) and the processor (after removing the heatsink and heatpad) - I won't recommend this but I had nothing to lose. To my surprise the unit sprang back to life when I plugged it in. I'm not sure how long it will last and I am not hopeful.

This is a design issue that needs to be addressed by the likes of vu+ and dreambox. Had they performed sufficient thermal analysis on their designs, they would have known to either use more suitable means of heat transfer (ie. bigger heatsink, more copper on the PCB for thermal relief or using a fan for active cooling). It might only be occurring on a small percentage of units now but overtime - and most annoyingly - when the warranty has expired the problem will become more familiar to the rest of the loyal customer bases. This is something I would expect from a clone not genuine units.

Well, I'm sure someone will soon start selling a service on ebay to add better aftermarket cooling/reflowing on these units now. At least someone will profit from this... but as usual, the customer pays for something he should not have to.
 

Similar threads

Top