AVS Forum banner

How to DIY ; TX-SR576/TX-SR606/etc HDMI daughterboard repair (for dummies?)

544K views 1K replies 451 participants last post by  Worf 
#1 · (Edited)
--------------------------------- PART 1 ----------------------------------
Did your ONKYO work great for a year or more and then start exhibiting difficulty locking or remaining locked on your HDMI stream ? Do you need to warm it up like a 1957 television and it keeps getting longer ? Some of you instead lose component video sources or On Screen Display from its HDMI out.
Well my unit finally became unbearable and after a few hours of time well wasted on the internet, the uncertain deed was attempted. This post is an effort to enlighten those who may be interested too. I figured that eventually as these AVRs age beyond warranty, some of you may want to also take matters into your own hands.
This guide is based on my own '576, but should benefit other models in the design family with HDMI related failures provided that you extrapolate the concepts to fit your application.

UPDATE 2015 - serafis describes multiple successful TX-SR606 repairs plus prevention > http://www.avsforum.com/forum/90-receivers-amps-processors/1246078-how-diy-tx-sr576-tx-sr606-etc-hdmi-daughterboard-repair-dummies-31.html#post31142105
...one of so many people who took time to help others - and why this thread gained momentum.
Before proceeding make sure you ruled out other causes/options like in this sticky > http://www.avsforum.com/forum/90-receivers-amps-processors/1652514-onkyo-acknowledges-failed-units-extending-warrranties-until-2018-a.html
 
See less See more
#4 ·
ping runnin' : Most likely* able to solve the common degenerative 'issues' ; won't improve something that never worked, just to be clear. ( *first attempt, need more feedback )


ping maxcooper : I can't be 100% on which model/build dates are applicable, but you are surely kind for expressing that.
 
#5 ·
What led you to key on the caps?
 
#6 ·
------------------------- PART 2 -----------------------------

OK, so you have tried PART 1 and it's working - congratulations - but don't be silly and put that cover back on.....we won't be needing that anymore, unless you want to perpetuate the grief. Ridonkulous, you're thinking, where else are you going to pile those DVD cases
. Fine, just as long as you complete the repair by adding active ventilation/circulation !. I'm not going to elaborate on what other resourceful forum members have already put out there....just remember you've already crossed that line and have warmed up the iron. I'm currently wrestling with an elegant solution for a PART 3...


in the meantime ;


*Here is the re-capped board as viewed from the top. No improvement after these initial two were changed

*I added a CPU cooling fan connector (previously enjoyed) to an unpopulated capacitor pad ('606 and others won't be empty). It supplies around 9-10 volts (thru the 2 1N4007 diodes) which should quiet the eventual internal fan
Attachment 174320






*This is the bottom view and worth noting is that after replacing the two uppermost caps, all malfunctioning ceased (tested after every second cap). They reside in an area directly above the smokin' hot transistor/heatsink seen in the image that follows this - coincidence ???

*optional: a small dab of RTV silicone, colour of your choice, will keep these new caps secure during any rough rides
Attachment 174321






*While apart, I added a 'previously enjoyed' heatsink to the Bridge Rectifier ('cause it burnt my finger). Although silkscreened on the PCB, this likely was omitted on the '576 due to the absent processor/components on the HDMI board.

*This area is just begging for circulation IMO
Attachment 174322





*I came across the following manufacturer's suggested design circuit example. These boards use this verbatim as the initial supply for the regulators we are ultimately targeting. The 470 and 220uF caps involved are not on the hot underside and as such are a lower priority on the replacement wish-list.
Attachment 174323




If you've read this far, great, but this has become more FYI than DIY.

Many have observed how temperature affects the faulty ONKYO but I can elaborate. While my unit would be on the verge of instability with artifacts in the video, I'd barely cool* the chip SiI9185' (3to1 HDMI switch) and the signal disappeared until it warmed. Same conditions, cooling* neighbouring chip SiI9013' (HDMI receiver) and the signal became flawless temporarily. IOW change a chip's tolerance to noisy power by means of temperature. (Initially I feared faulty chips!) This helps explain the commonly reported wait times after first powering on.

( *with an inverted can of 'electronics duster' - yes, a third use ;-)


[ to be continued....bullet-proof or bust ]



 
#13 ·
Thanks for the post and the research done. I'm looking forward to reading more on this before I attempt it myself.

How much of a difference in the syncing have you observed after change the caps?


------------------------- PART 2 -----------------------------

OK, so you have tried PART 1 and it's working - congratulations - but don't be silly and put that cover back on.....we won't be needing that anymore, unless you want to perpetuate the grief. Ridonkulous, you're thinking, where else are you going to pile those DVD cases
. Fine, just as long as you complete the repair by adding active ventilation/circulation !. I'm not going to elaborate on what other resourceful forum members have already put out there....just remember you've already crossed that line and have warmed up the iron. I'm currently wrestling with an elegant solution for a PART 3...
 
#7 ·

Quote:
Originally Posted by Chu Gai /forum/post/18557085


What led you to key on the caps?

Logical question Chu Gai. Within the novel I read on the subject ( actually a series of 1328 short stories, aka posts ) there were 2 or 3 references to 8-10 capacitors on repair invoices. One vic actually quoted a long ONKYO part number...g00gled it to one Euro site which added (something)4V100M or so and that finally focused me. No single post, but the sum of the parts.


( I'll try to fill missing info/questions regularly )
 
#8 ·
It all boils down to heat. What's your elevation where you live in Alberta? I've had Rotel gear that would not work here in NM at 5,000 ft (1,525 m) but worked perfectly when sent back to Rotel in New Jersey (or wherever they are). I don't think the manufacturers test at high altitude for proper cooling.
 
#11 ·
Penngray, for the purposes of dissipating heat in our electronics, the boiling of water instead has more of an effect on my macaroni and cheese....which by the way, I could almost cook on top of this onkyo some days. Water does boil at a measureably lower temperature at higher altitudes.

OTOH, as petew's point is valid in principle (barometric pressure+relative humidity=mass), but it won't make or break an epic fail design. PIXAR is currently producing an animation for this thread
but in the meantime they sent me this sneak peak; Attachment 174398




>>>>>>>> TX-SR5xx/6xx/xx layout, cross-section looking left2right
 
#14 ·

Quote:
Originally Posted by petew /forum/post/18564519


Forgot about humidity. As my drunk friend Marvin likes to say:

"It's dryer'n a popcorn fart in New Mexico"

lol, sounds like your friend knows how to compensate for the dryness.

Quote:
Originally Posted by lummy /forum/post/18566477


Thanks for the post........How much of a difference in the syncing have you observed after change the caps?
lummy, this OiNK' now behaves like I remember. Typically 5 seconds from cold power up or input switching until picture. I exclusively run my PC (ATI HD4650) on a 47PFL3603D Philips. At it's worst it wouldn't sync cold and once 'warm' it had varying amounts of visual noise and audio snaps when it wasn't losing it altogether. Hopefully yours goes well!
 
#15 ·
--------------------------------------------PART 2.5----------------------------------------------

UPDATE; December 2011: Although the following exerpt is an insight towards the adding of auxilliary capacitors, there are more examples littered throughout this thread. I soon describe how to determine if your situation requires this additional measure.

Quote:
Originally Posted by Macks1974 /forum/post/18443480


Work Performed

Replaced C8072, C8096, C8602, C8609 and added 4 TF caps on HDMI board per Onkyo tech bulletin #10012702. Checked operation with a Blu-Ray player and a 1080p TV monitor.

- - - - -{in regards to a TX-sr606 with handshaking issues }




*top/bottom of daughterboard; TX-SR576 version
Attachment 174496




I'm taking the liberty of assuming C8602 should have read C8607 - as in the case of my version of board, I couldn't locate the former. Of the four listed above, two apply to mine. As it happens I had only replaced C8607('C8602'), tested;unchanged, then replaced two not listed(C8162+C278), tested;good! then finally replaced one listed + one not, testing;redundant. There should never be any regrets if a few extras get replaced IMO

.

*UPDATE; December 2011/aka 19 months later...Unit repaired with repurposed parts as described - powered 24/7 - still flawless (AFAIK
running @1080P , non-HDCP, various audio formats*

*Please note that my contributions contain much speculation and therefore posts by others who own your specific model (search each page for it, ie "606") should supercede the former*
 
#16 ·
Good stuff CZ_the_DAY!
 
#18 ·
I tried the approach offered by CZ on my 606. I replaced all 10 100 uF Caps with 105 degree 10V components. My results are that I have OSD using 480P output and sound from both composite and component sources while displaying OSD. However, I have no video or audio from any sources HDMI/Component/Svideo/composite. When I remove OSD, the sound from the selected source (in the case of my component and composite sources) also stops leaving the black screen. I have also noticed that when cold, there is a great deal of cracking and popping in the audio that I can hear that seems to dissapate when warmed up.

I know many of these results are consistent with other posters. Just wanted to see if there was any feedback before I go about replacing the rest of the caps on the board haphazardly. Here are some shots of my board in case anyone was wondering what the 606 difference were.


Might also be worth noting that I am only using TV audio through HDMI, not speakers attached to receiver

 
#19 ·
navtek; good to have your feedback! It's just unfortunate that there was no happy ending.

-Could you clarify what if any improvement this repair made

-Also, if possible, carefully measure the main supply to the board (5 wire ribbon*) both without and with a viable signal being output. Mine varies from 11 to 10+ Vdc respectively. I have been accepting these values because it has been behaving anyways. (*connector silkscreened as 'gnd/12v/gnd/12v/gnd')

-I can't tell from the pics if your new caps have polarity markings; are they installed PC ? (polarity correct :)

PM me if you prefer


[ Updated! Our common experiences, as our units change temperature, really make me think there is a similar mechanism involved> Degenerative Capacitor Syndrome

As my posts are kept as fresh as possible (which would explain the incontinuities ;p ), we have learned that more parts are involved than in my case. It comes down to the luck of the draw. Under duress, some components outshine their life expectency and some fall short. Focus on what is under fire first (the underside of the board), replace; if all seems well then improve its living conditions from now on. ]
 
#20 ·
CZ,

First of all, thanks for your interest in this effort. I think if we can build up some data, most everyone will be able solve their problems with these boards.


1. As far as I can tell, there is no difference at all. All the symptoms are the same as before the repair was attempted. So, the good news is that I guess I did no harm
. As I said before, I am able to display the OSD in 480P only and hear audio from the selected source (non HDMI only) in the background (with crackling and popping until warmed up). I haven't tried the canned air test but maybe this would be a good idea.


2. I will measure the power supply feeds tonight and post results. The only question I have on this is that I don't see how filtering issues will be determined using a DVM, wouldn't I need an O-scope to derive how much noise is on the DC signal? Of course, I don't have access to one so that presents its own issues.


3. I would expect the voltages to read 10-11 when under load so that seems like a good thing. If you have a blown IC, maybe that would read a full 12V open. Concur? Maybe that was what you were going for anyway



4. I couldn't really tell from the diagram you posted, but it seems that the 470 uF caps are in LC networks with the 470 inductors which I would think would handle the AC filtering aspect. Maybe I'm way off here. Just thinking out loud. In any case I have measured the coils and they seem to be good. Perhaps this could rule out the transformer.


Anyway, I'll carefully check the power input (five wire cable) and report back this evening.


PS: My caps did not have Polarity indicators on the leads so hopefully that means there was no difference. However, there were different lead lengths on the components.

I could not determine the polarities of the sockets, maybe you could enlighten me? Thanks again!
 
#21 ·

Quote:
Originally Posted by navtek /forum/post/18646440


CZ,

...I don't see how filtering issues will be determined using a DVM, wouldn't I need an O-scope....

Absolutely, would prefer to scope this but if DC was lower than expected, we might suffer unpredictable operation.
Quote:
...If you have a blown IC, maybe that would read a full 12V open. Concur? Maybe that was what you were going for anyway
...

I was going for excessive leakage via caps/chips/etc pulling down the 12v. We'd need to be sure it's healthy first.
Quote:
...it seems that the 470 uF caps are in LC networks with the 470 inductors which I would think would handle the AC filtering aspect...

Since the 3 DC-DC circuits operate at 300khz, there is great opportunity for introducing noise to the downstream regulators.
Quote:
....there were different lead lengths on the components. I could not determine the polarities of the sockets, maybe you could enlighten me?...

After looking at my new caps; +++=long lead. The board's printed cap outlines will have a bevelled end; that is +++. At the moment I dont know what effect using a non-polarized cap has. Try and sort this aspect out before proceeding. (crossing fingers + drinking beer for luck)


I am happy to assist anyone willing to further this effort...
 
#22 ·

Quote:
Originally Posted by CZ_the_DAY /forum/post/18647435


Absolutely, would prefer to scope this but if DC was lower than expected, we might suffer unpredictable operation.


I was going for excessive leakage via caps/chips/etc pulling down the 12v. We'd need to be sure it's healthy first.


Since the 3 DC-DC circuits operate at 300khz, there is great opportunity for introducing noise to the downstream regulators.


After looking at my new caps; +++=long lead. The printed cap outline will have a bevelled end; that is +++. At the moment I dont know what effect using a non-polarized cap has. Try and sort this aspect out before proceeding. (my fingers are crossed)


I will likely edit this later, too rushed! But I am happy to assist anyone willing to further this effort...

OK, test results from voltage across the 5 wire connector.


In standby, both + leads measure .67 volts.

Power on, no display both + leads measure 12.45 V.

Power on, OSD in 480P on screen, + leads measure 12.46 V.


very consistent eh?



One more observation, I used the ground connector on the side of the board to measure GND. BTW, the 5V test point measured 3.9 which thought was kind of low for TTL. But maybe this is normal.
 
#23 ·

Quote:
Originally Posted by navtek /forum/post/18648148


OK, test results from voltage across the 5 wire connector.

In standby, both + leads measure .67 volts.

Power on, no display both + leads measure 12.45 V.

Power on, OSD in 480P on screen, + leads measure 12.46 V.....5V test point measured 3.9 which thought was kind of low for TTL. But maybe this is normal.

You have 2 ( leftmost edge ) of the 5 primary ICs 'active' while the AVR is performing that function - and not fully either - whereas my scenario below uses 4 briskly.

My results referenced to ground plane near power connector;

STANDBY @ 12vIN=.56v

No HDMI active @ 12vIN=11.4v

1080p HDMI IN/OUT w/LPCM audio @ 12vIN=10.4v


Seeing your voltages has me questioning my supply. It has just a bridge rectifier and 10,000uF cap, but does feed areas besides the HDMI board.


I was curious as to the source of the new caps ? Any more info on their type or orientation board-wise ?
 
#24 ·
the test point area you mentioned, I believe will work. That is where I measured mine. I would be curious to see what yours is especially since your overall voltages are lower.


Is there a stripe on your caps? since my leads have been cut, I can no longer use that as a guide. I think we are moving in the right direction. Thanks!
 
#25 ·
navtek, the 576 is stripped of the functions in that corner of the board, hence the lack of testing on my part.


I read that some capacitors are marked only by lead length. If those were non-polarized aka bipolar, they should be inherently larger than same spec polars or at least have an 'NP' marking. Perhaps the printing on each cap is the 'stripe'(?) If you cannot locate a datasheet from the supplier, then the only recourse might be to remove one and try to measure DC leakage in both directions. The test need not be destructive; the in circuit exposure could be eventually :-(


. . . . . . . . . . . . . . . R . . . . C . . . . . .

5 to 7V Pos > -----/\\/\\/\\/----] [--------
. . . . . . . . . . . . . .1k-ohm . . . . . . . . . . . . . . .

Apply DC only long enough to measure steady voltage across R and if lower (near zero) in one direction, assume cap good and prefers that orientation. No difference would indicate bipolar or failed.


Hope remains!


[I'm tempted to remove this soon...way too specific to your predicament]
 
#26 ·
OK, I've verified that the stripe on my caps means -. As it turns out, I had two installed reverse polarity. Shame on me! So, although I might need to replace them anyway, I went ahead and reinstalled correctly, just to test.


Two things.


1. Some improvement! In the "through" setting under hardware settings on the OSD, I am able to fairly consistently get component and composite video over HDMI. TV says it is coming in at 480i. Image quality appears stable.


2. After a considerable amount of warmup time, audio for component and composite also kicks in. These are preceded with a long period of silence...crackling...stuttering audio...solid audio.


OK, now the real question, why is the temperature factor so huge? A TTL logic board operating with 5V and 12V doesn't generate much heat on its own. Could it be that we are looking in the wrong place entirely? What if the real problem lies in a faulty power supply sending very dirty voltages that caps are struggling to filter and possibly blowing HDMI boards. This would explain the repeat failures of some users and the temperature operational differences. Just trying to think big picture! Thoughts?
 
#27 ·
(Heh, I was just checking for an update)

I may have underestimated the amount of variables (flaws if you may) that any give ONKYO might be stricken with.


To sum up the temperature issue; the ICs seem to become more (or less - refer to end of "PART 2") tolerant to noise with heat. The individual regs may also improve with heat.


There must be a deficiency in one or more replaced caps if you experienced an improvement; not to say there isn't more noise to filter - as you suspect.


My complete turnaround on my first attempt (10 buck investment) may have been a best-case-scenario in which I am leading most to false expectations. But if one simply makes the same mistake by replacing the same board into the same 'oven', history will repeat itself. (refer to beginning of PART 2)


If you truly duplicated my repair (granted, more caps) without any 'complications' we would focus forward - perhaps on those 470uFs - otherwise the tiny decoupling caps serving the two leftmost chips (the analog to HDMI path, AFAIK). Oh, for the opportunity to scope, eh?


Won't produce heat? My 'TTL' 4 core AMD CPU runs on 1.3 volts but has a huge heatsink! Although 3.5Ghz vs
 
Top