(Another) Troubleshooting a Yamaha receiver with DC_PRT (protect shutdown) probl

[NHcoder]

Apologies in advance for the wall of text!

Let me start by just saying I've looked over many similar threads here and it seems that most of them just die out before a resolution is found, so in spite of the number of similar threads, none of them I was able to find actually resulted in a resolution.

I promise to anyone that decides to help, if I decide I've had enough and give up on the project, I'll say so in the thread rather than just disappearing!

I have a Yamaha RX-A1000 receiver that I've owned since new.  It worked flawlessly for ~10 years before it was put in storage for two years.  When it was pulled out of storage, it still worked mostly ok, but one of the HDMI outputs was putting out garbage.  All of the others worked fine.  After using it for another 18 months or so, it suddenly stopped working.  I came home to find it powered off, and attempting to turn it on results in it going into the power protection mode almost instantly.  This happens regardless of what connections are made on the rear panel, including with no connections at all.

I was able to find the service manual online and boot into the diagnostic mode, which indicated a "DC_PRT 0L" message, which the manual describes as

"Power amplifier DC (DC voltage) output is detected.  The voltage at pin 1 (DC_PRT) of IC902 is displayed.  Normal value: 35-68 (Reference voltage  3.3v = 255)"

Several forum posts I saw recommended testing the output transistors.  I assumed this meant the massive pairs of darlington transistors for each output channel (Sanken STD01N/STD01P), which are a real pain in the neck to get to with this unit due to their position combined with the board to board wire routing, but I gave it a shot anyway.  I took apart the the receiver carefully, board by board, to get to those transistors.  Unfortunately due to their location it's essentially impossible to get to them with the receiver assembled and all the boards in place.

Once all the boards were out of the way, I tested the transistors as best I could determine how in circuit with my multimeter.  I checked for:
  - Dead shorts from the gate pin to any other pin
  - Diode forward drop of the N channel pair between pins 1 and 2.
  - Forward drop of the diodes on the P channel pair between pins 4 and 5.
  - Forward drop of the P channel pair between pins 1 and 3.
 
In retrospect I should have probably checked more, but everything here checked out, and I put it all back together.  Honestly, I don't know if these are even the transistors I was supposed to be checking, but it seemed like the right place to start -- because it was the most difficult, probably.

Since reassembling it, it still shuts off with a DC protection error, but now it shows a high value rather than 0.  On the display for P1-1 it shows "DC: 84" at present, but as I watch it, it jumps around between about 79 to 85.

I took pictures of every board front and back with and without my phone camera flash enabled as I was putting them back in after looking them over for obvious damage, without finding any.

This imgur album has all the photos.  The ones taken of the individual boards were just before reassembly, the ones inside the receiver were taken as I took it apart, so I knew which cables went where.

https://imgur.com/a/KnwSsHe

I do basic circuit design and microcontroller stuff, but it's just a hobby, I'm a developer by trade.  I'm at my wits end trying to figure out what went wrong here, which is a shame, because as old as it is this receiver has been great since the day I bought it, and it seems like this should be an "easy" fix -- I just don't know where to go from here.  Please try a dash of ELI5 if you can, clearly when some post says "check the transistors" I may not know which ones are meant or how I should test them.  Mostly I just build things and when I let the smoke out, I throw it away and build again.  I don't usually have to troubleshoot in this way.

So thanks in advance for any help in figuring this out. 

EDIT:

Oh, there are a few closeups of discolored areas of two of the boards.  Just looked like normal heat discoloration to me, but other opinions are obviously welcome.

EDIT2:

Results of onboard diagnostic pages, normal values from the service manual.

  P1-1 - DC: 100 (normal 35-68)
  P1-2 - 130/120 (normal 84-141 / 81-153)
  P1-3 - TM: 120/122 (normal 116-255 for both)
  P1-4 - OUTLVL: 255 (normal .  255 = 3.3v.  Pg. 62)
  P1-5 - LMTCNT: 140 (normal .  255 = 3.3v.  Pg. 63)

[alanep]

Hello NHcoder,

I'll try to help you out. I'm a newbie when it comes to posting to the EEVblog forum but I have had a fair amount of experience with Yamaha amplifiers (& audio amplifiers in general). The version of the RX-A1000 service manual I downloaded was from Hifi Engine (for the page number references).

Before commencing any fault finding, how competent are you with mains powered equipment e.g. this amplifier? There maybe exposed terminals with mains potential inside which could be lethal. Normal practice is to inspect how the mains wiring enters the amplifier & how it is routed e.g. from the service manual, the "Power Board" (schematic page 139, component overlay page 118) definitely has mains power "bities" to watch out for/keep away from. If you haven't done this before, think very carefully whether you wish to proceed...

It looks like it's a 7 channel amplifier. The output power amplifiers sections (page 138 of the service manual) have plus & minus power supplies. With no signal (volume fully down), the power amplifier output should be around 0V DC (or within +/-0.5V) when measured with a multimeter on the DC range. This has to be measured before the output muting/ protection relay i.e. it's no good measuring it on the rear output terminals as they'll be disconnected. The usual spot to measure the output DC voltage is on the centre tap of the emitter resistors (or close to them) - I explain more about this later.

One or more of the power amplifiers has something other than 0V on it's output (i.e. a DC offset) which is generating the DC_PRT error. A DC offset doesn't necessarily have to be a faulty output transistor, it can be anywhere in the power amplifier stage.

The first step is to identify which power amplifier(s) has the DC offset but there maybe some difficulty keeping the amplifier on long enough to identify it. It sounds like it will stay on in diagnostic mode. Still, it's still best to minimise the time the amplifier is left on just in case otherwise there maybe magic smoke...

The emitter resistors mentioned above for each of the 7 power amplifiers are: R1041 (SBR  CB102), R1071 (SR CB103), R1111 (FR CB104) , R1146 (C CB107), R1183 (FL CB108), R1215 (SL CB110), R1250 (SBL CB111). These have two power resistors in a single ceramic package. They are conveniently brought out on the CBnnn connectors (shown in brackets above) mainly for idling current checking. Note that the idling current CBnnn references on page 23, 24 of the service manual don't match the schematics (frustrating). I'm using the schematics CBnnn.

Although their centre tap isn't brought out to the respective CBnnn connector, measuring the voltage on any part of these resistors is "close/good enough" i.e. it will normally only be a few millivolts different to the centre tap. The idea is to measure the DC on the emitter resistor with respect to the signal ground (or output ground) e.g. connect one meter lead to a signal input ground & the other lead to either the first 2 pins of the CBnnn connector (see attached extract for one of the power amplifiers).

You'll need a steady hand when probing the CBnnn connectors with amplifier powered up as there's +/- 56V (as well as mains voltage) inside!! If you're not comfortable, power it down (wait for the DC power supply caps to discharge) and solder a "flying lead" wire onto the rear of the connector & connect the other flying lead end to the multimeter, then power it up & measure it.

Hopefully, you'll identify a power amplifier with a significant DC offset. If so, we'll dig a little deeper.

Alan

[NHcoder]

Awesome, thanks for that response!  I saw those little test headers on the schematic and was wondering what they were for, it didn't occur to me they were for this sort of testing.  I'm comfortable with mains so no worry there, and while the board diagrams don't mention it, they seem to be a perfect fit for JST XH connectors (page 124 -- I'm using the manualslib PDF in which the page #s printed inside and in the PDF itself match), which I have in abundance, so I'll make up a little pigtail to use for testing with the meter and get back to you later today.

Image attached just so we're talking about the same thing.  I agree with the stupid renaming/renumbering of the connection.

[NHcoder]

Hopefully, you'll identify a power amplifier with a significant DC offset. If so, we'll dig a little deeper.

Decided to forego the JST and just use a female "dupont" jumper to check each pin against chassis ground.  6 of the channels show -1.7v, one showed -2.2v.  I mention the polarity just because it might be important relative to chassis, as I (perhaps naively) expected the bias to be positive.  The 0.5v difference doesn't seem huge, but maybe it is indicative of something.  The -2.2v signal is on CB378 (feeds to CB377, which leads to CB111 on the other board).

[alanep]

-1.7V (or -2.2V) is high. One mistake I made was assuming that CB102 to CB111 were the test points but I can see from your video pcbs attachment that they are connectors to additional pcbs video (4), video (5) where the idling current is set and with the real test points. The schematics for video (4), (5) boards are shown on the manual page 142.

CB372 to CB386 on video (4), (5) are indeed the test points. The idling current setting page 23,24 & the component overlays page 124 & their schematic page 142 now agree.

So, you've measured the DC on CB372 to CB386? If possible, I would like you to do a "sanity check" by measuring the DC directly (with respect to signal ground) on one of the emitter resistors (R1071 to R1250). From the component overlay, I think these resistors may have small wires (x 3) protruding from the top. These are an extension of their connection into the PCB i.e. they act as resistor test points & avoid having to probe in the vicinity of their base/the pcb. If not, then probing around the resistor base might be a little tricky... It is a sanity check so only do one channel.

If the emitter voltage is indeed the same as the test point then it is very unusual for all channels to have a DC offset greater than +/- 0.5V i.e. we're missing something.

I should explain how the power amplifier works in stabilising it's DC conditions. I've attached an extract of one of the power amplifiers with a few markups. At DC, we can ignore all capacitors (i.e. as if they don't exist). Q1022 at the input is called a differential pair & will produce an error signal at the collectors if the base voltages don't match (so it's often called an error amplifier). The voltage on the base of the left hand (LH) Q1022 is supplied a reference ground (0V) signal via R1053 & R1056. The DC voltage on the amplifier output is feed to the base of RH Q1022 via R1045 & R1044. The closed loop gain of the amplifier at DC is one (unity) i.e. assuming the amplifier has no faults, whatever voltage is on the LH Q1022 transistor base will be reflected on the amplifier output & back in the RH Q1022 transistor base. Another way of saying it would be, any error between two Q1022 bases voltage will be amplified by the open loop gain & will force the output back in the right direction so that the base voltages match.

If there is a DC offset on all channels, then it's looking like something is wrong with the reference ground (0V).

See how you go...

[NHcoder]

So, you've measured the DC on CB372 to CB386?

Yes, I measured the voltages to chassis ground from all 14 pins of those Video(4) and Video(5) "For Services" (each one labeled CBxxx and the other pin XHI) connectors on each.  I double checked each one because I thought I might have read 1.7mV on a bunch and actually 2.2V on the one odd man out, but it was volts and not millivolts on each one.

If possible, I would like you to do a "sanity check" by measuring the DC directly (with respect to signal ground) on one of the emitter resistors (R1071 to R1250).

I'm searching for them, unfortunately they must be among those printed vertically, which acrobat reader cannot find when doing a text search, so they only turn up in the parts list but nowhere else when I search for them with ctrl-F.  I should also point out that the boards are mirrored in the PDF for some reason.  For example the "MAIN(2) P.C.B. Side A" has R1056 in the upper left corner, but it's in the upper right corner of the PCB.  The PCB diagrams are laid out as though you're viewing them from "underneath" with X-ray vision.  I've attached a screenshot of the PCB from the manual on page 114, as well as a photograph of what I have to work with inside the receiver without disassembling it.  You can also see the "Video" daughter board there with the test points.  I can't access anything below that, including the pins of the actual power transistors and those test point jumpers, without tearing the thing apart again.

Getting these two amplifier boards out requires unplugging a few of the connectors going to them, so that may affect my ability to test them.

This is just to sanity check *me* really.  You can clearly see the markings in the photo for R1114 and R1056 if you zoom in, along with the parts in between, and they're perfectly mirrored from how the board is actually laid out.  Maybe this is normal.

So I'll get back to hunting for R1071 and R1250 on the PCB but if you see equivalent test points on the PCB diagram in a spot I can access them, that would be great!

[NHcoder]

Ok sorry I didn't realize 1071 etc were the big voltage divider things.  I don't have any access to probe those, they are physically below the video a/videob daughter cards plugged into the main a / main b boards as you can see in the photo.  I might be able to get dodgy access to suitable test points on MAIN(1) if I disconnect at least two of the board to board connectors from it -- FUNCTION (1)_CB706 (bottom left of MAIN(1) schematic on page 138) and MAIN (3)_CB106 (top right of the same board).  I haven't looked at where those go yet though so testing may not be possible with them disconnected, just wanted to get a quick update in.

[alanep]

Ok, no point having to dismantle to gain access to the emitter resistors - don't bother with the sanity check.

In my description of power amplifier DC conditions, I forgot to say that the DC voltage on the amplifier output should reflect the DC voltage feed to LH Q1022 via R1053, R1056 by the reference ground (0V). That's why the amplifier output should be at 0V. If not then there's something wrong with the reference ground (0V).

The next step is to check if the reference ground (0V) is elevated or floating (in voltage) with respect to the input ground on any of the amplifiers. For example, looking at previously attached Amplifier DC condition image for the SBR channel, can you measure the voltage on the side of R1056 or R1054 that's connected to the reference ground (0V) & an input ground? If SBR is inaccessible, you could measure any of the other channels e.g. R1085, R1088 for the SR channel etc.

[alanep]

One additional context, "input ground" is any ground that you can connect to that's on the rear panel inputs etc.

[NHcoder]

Ok, no point having to dismantle to gain access to the emitter resistors - don't bother with the sanity check.

Ok.  I'll attach an image to this post showing how the boards are laid out to give some idea of what I can access and what I can't.

In my description of power amplifier DC conditions, I forgot to say that the DC voltage on the amplifier output should reflect the DC voltage feed to LH Q1022 via R1053, R1056 by the reference ground (0V). That's why the amplifier output should be at 0V. If not then there's something wrong with the reference ground (0V).

The next step is to check if the reference ground (0V) is elevated or floating (in voltage) with respect to the input ground on any of the amplifiers. For example, looking at previously attached Amplifier DC condition image for the SBR channel, can you measure the voltage on the side of R1056 or R1054 that's connected to the reference ground (0V) & an input ground?

Access to R1056 is easy enough

One additional context, "input ground" is any ground that you can connect to that's on the rear panel inputs etc.

I was just going to ask if that's what you meant.  I have a solid and repeatable 2.5mV between the subwoofer input and the pin on R1056 that goes to ground, 0.0mV on the rest.  I also have 0.7ohms between that resistor and chassis ground, which isn't much, but it's not zero, just in case it matters.

Thanks for the quick replies so far, hopefully there is a clue in here somewhere.  I'm wondering if I should check some other channels?  I can get to R1123 and R1114 easy enough on that board, and probably all the other resistors on the other board easily as well.

[NHcoder]

Sorry, that was 2.5mv from subwoofer input tip to the resistor.  It's 0 on the input ground.

[alanep]

I'm not sure about these quick replies/chat. It does cause the topic to be elevated each time (& generate forum "noise").

I was really hoping that R1056 had a DC offset or wasn't (eventually) connected to chassis (2.5mV is very small). 0.7 ohm is pretty good (maybe even multimeter lead resistance).

So, a bit more thought must go into it now...

The idea was to understand where the DC offset was being introduced. It's a little difficult not standing in front of the amplifier as you are i.e. a bit like remote control.

By explaining the principles of how the amplifier works with respect to the DC conditions, it might give you a few ideas.

Another suggestion, can you check the voltage across the R1053 +R1056 (or equivalent in another channel). So what you are measuring is voltage between the base of LH Q1022 and the reference ground (0V). You should use a digital multimeter with a 10Mohm input impedance or higher (normal digital multimeter) so that it doesn't load down R1053 (33K). This measurement is effectively across the two resistors (& NOT to signal ground). You might need "flying leads" to the multimeter if they're inaccessible.

[NHcoder]

The replies will stop for a bit now, I need to hit the sack, got an early morning tomorrow clearing snow, but the replies shouldn't bother anyone who's not subbed to the thread.

I'll need to reread the meat of your posts and do some thinking.  To be honest I still haven't really understood how it knows there's an offset in one of the channels, and why it can't tell which channel it is.  I think it's worth mentioning that the last time I turned it on without doing the protection override and let it shut off on its own, it showed a "DC_PRT: 0L" message on the display, indicating a low voltage detected rather than one that was too high -- the manual says the value should be from 35-68, not the 0 I see right now, and not the high 80s like I also saw.  I wonder if it could be the reference voltage it's comparing against that's bad?  I checked the 3.3v lines going to the MAIN (4) and MAIN (5) boards which look like thermal monitors for the heatsink, and they were solid, but I don't know if the reference voltage is coming from the same source.

I couldn't repeat the 0.7ohm measurement, I probably just didn't poke the metal hard enough where I was probing last time.

If you think of anything else feel free to reply, I'll be looking at this again in 14 or so hours.

> The idea was to understand where the DC offset was being introduced. It's a little difficult not standing in front of the amplifier as you are i.e. a bit like remote control.

As a developer & sysadmin, I understand this struggle completely.  Thanks again for trying to help out.

[shakalnokturn]

I admit I've been skimming over this rather than thoroughly reading so sorry if I've missed something on the way.

Have you been able to confirm lack of DC offset on all outputs (measuring at speaker outputs) with protection override?

If you're sure the DC output is within tolerances your last remark on reference voltage may be the way to go.
I recently repaired a Yamaha HTR something that was throwing random Power Supply and Thermal protection errors on startup although everything was in range.
A tiny 3.3V SMD regulator on the small signal board (CPU / HDMI...) was giving a low voltage that was used for protection references.

[alanep]

A little time to gather my thoughts as well.

As shakalnokturn mentioned, recheck the DC offset. Page 23 of the service manual has a table (RHS) with a column "DC Offset Confirmations" & it shows a DC offset of between 0 to 100mV (i.e. 0.1V). 

You might also need to confirm that protection override enables the output muting/protection relays so that the DC offset can be measured at the speaker outputs. If the relays aren't energised, measuring at the speaker outputs would give erroneous results i.e. use the test points instead.

What happens next depends on whether there's any out of spec DC offsets...

[NHcoder]

OK so I went back to the top just to reverify everything.  First, on power up, it is shutting down after a few seconds, engaging the protection circuit.  Starting it up in diagnostic mode it shows "DC_PRT:  0L" as the reason, and not a value of up around 80.  The 0 value is the "normal" one I've seen for the past few months, I only saw the ~80 once when looking at it yesterday.  It's possible taking it apart and putting it together jiggled something, maybe a broken solder joint.

In diagnostic mode you can see the current live value, and right now it's bouncing around between 0 and 6.  The last 4 recorded shutdown reasons say the same, that it was due to "DC:  0".  If the relationship is linear then the value of 6 would represent 6/255 * 3.3 = 0.778mV, but it may well not be linear.  I assume the datasheet for the protection IC would say but I haven't looked yet.

I went through all the readings at the test points on the VIDEO(4) and VIDEO(5) boards again.  These are the JST connectors shown on page 124 and labeled CB372, CB376, CB374, CB378, CB384, CB386, CB382.  I checked the voltage from each pin of the JST connector to both chassis ground and digital ground.  So over all this was 7 channels x 2 pins x 2 grounds = 28 tests.  The results were:



Board Video(4)

CONN	PIN	DIGIGND	CHASSISGND
CB378	1	-2.021v	-1.976v
CB378	2	-2.014v	-1.976v
CB374	1	-1.565v	-1.535v
CB374	2	-1.563v	-1.536v
CB376	1	-1.568v	-1.542v
CB376	2	-1.568v	-1.542v
CB372	1	-1.553v	-1.531v
CB372	2	-1.553v	-1.531v

Board Video(5)

CB384	1	-1.554v	-1.538v
CB384	2	-1.552v	-1.535v
CB386	1	-1.551v	-1.538v
CB386	2	-1.549v	-1.535v
CB382	1	-1.549v	-1.530v
CB382	2	-1.545v	-1.534v

After I collected these values I shut off the receiver and started transcribing the numbers from the excel sheet I'd initially written them in to the forum post and saw the voltage was reading positive now rather than negative, and was slowly rising.  It reached about +2.4v after 5 min when I stopped watching.  This occured on all test points.  I figure that it just had something to do with p-channel FETs being deenergized while the device cooled off and caps drained, but thought it was worth mentioning just in case.

If I turned the receiver off and watched my meter, it only took 20 or so seconds to go from -2v up to +0.6v before continuing to rise at a steadily higher slower rate.  Shutting the mains connection caused it to drop down to -0.6v and sit there slowly rising towards 0 as caps drained or whatever.

Anway, now sure that I really (really, really, really) had tested each and every one of the test point voltages to ground, I moved on.

As shakalnokturn mentioned, recheck the DC offset. Page 23 of the service manual has a table (RHS) with a column "DC Offset Confirmations" & it shows a DC offset of between 0 to 100mV (i.e. 0.1V). 

I have not been testing the voltage between the two test pins, I have been testing the voltage of each one of them with respect to ground.  I just ran through and tested them as indicated in the service manual and got a solid 1.7mV on each one except CB378, which read 0.0mV, an order of magnitude less than what they should be.  I thought maybe my connection with the dupont jumpers was bad so I went ahead and made the JST pigtail I'd originally planned, and tested with that, and got the same results.  I used the pigtail to go through all the voltages in the above table as well and the results were the same, so I can only assume that the 1.7mV reading is correct, as is the 0.0mV reading on CB378 which is also the one with the higher voltage reading above.

This probably isn't a good sign I take it.

You might also need to confirm that protection override enables the output muting/protection relays so that the DC offset can be measured at the speaker outputs. If the relays aren't energised, measuring at the speaker outputs would give erroneous results i.e. use the test points instead.

I hear a relay click when the protection circuity comes on even in diagnostic mode, but there are other options/modes available that cause relays to energize and denergize as I go through them.  I can't tell which ones are energized and which aren't since they're so buried, but I did discover that if I in the "DSP Margin" mode which is the first mode that is activated when you enter diagnostic mode, I can hear the relay click, while switching to the next mode makes it click again, so it would seem I can activate and deactivate them as needed in diagnostic mode.

After discovering this I decided to check the subwoofer again, as well as looking for DC voltages on the speaker outputs in diagnostic mode.  The output grounds are shared together, but not with chassis ground of course.  This prompted me to recheck for DC voltages at the outputs and I did find voltages on all of them when in the default mode you enter in the diagnostic mode, "DSP Margin."

In this mode there is negative DC voltage between all of the 7 outputs and ground that are all the same except for one.  I have -5.8v on all channels except "surround back / bi-amp L" which is about 400mV off, at -6.25v.

It's worth pointing out that all of these odd readings are tracking to the same channel -- Surround Back L.

Switching to the next test mode causes an audible relay click, and all the voltages test at 0.

[NHcoder]

I admit I've been skimming over this rather than thoroughly reading so sorry if I've missed something on the way.

Have you been able to confirm lack of DC offset on all outputs (measuring at speaker outputs) with protection override?

See my reply above, but I measured a DC voltage on *every* output channel, with one higher than the others.

If you're sure the DC output is within tolerances your last remark on reference voltage may be the way to go.
I recently repaired a Yamaha HTR something that was throwing random Power Supply and Thermal protection errors on startup although everything was in range.
A tiny 3.3V SMD regulator on the small signal board (CPU / HDMI...) was giving a low voltage that was used for protection references.

I have my doubts about it being the regulator now after this last round of tests, but I intend to check anyway.

[alanep]

In this mode there is negative DC voltage between all of the 7 outputs and ground that are all the same except for one.  I have -5.8v on all channels except "surround back / bi-amp L" which is about 400mV off, at -6.25v.

It's surprising that on the speaker outputs it's -5.8V and on the test points it's approx -2.0V. Did you measure a test point & a speaker output at the same time/conditions? It probably doesn't matter i.e. I think it has DC offset problem (which still needs to be investigated).

The DC_PRT protect is a single analogue signal representing the sum of all the channels (via R1152, R1120, R1189, R1047, R1258, R1077, R1221). The controller can't discriminate between channels & will just shutdown the amplifier. I wonder if "DC_PRT:  0L", the "0L" actually means overload i.e. beyond the ADC range. If the DC_PRT is really 0V, you would think it would come out of protection (despite the real dc offsets on the outputs).

I'll have to look into the "DC: 0" error & see what it means.

By the way, I did check the schematic for what shakalnokturn mentioned i.e. 3.3V regulator & reference voltages. This amplifier does have something similar (see attachment). I don't think it's the problem but the information is there if you wish to check.

[NHcoder]

In this mode there is negative DC voltage between all of the 7 outputs and ground that are all the same except for one.  I have -5.8v on all channels except "surround back / bi-amp L" which is about 400mV off, at -6.25v.

It's surprising that on the speaker outputs it's -5.8V and on the test points it's approx -2.0V. Did you measure a test point & a speaker output at the same time/conditions? It probably doesn't matter i.e. I think it has DC offset problem (which still needs to be investigated).

Time was within a few minutes, but I just checked the troublemaker (CB378 / Surround Back L) and his neighbor (CB374 / Front L) one after the other. 
-6.53V between center and gnd for Surround Back L.  -2.11V from the test points to gnd.
-6.00V between output and gnd for Front L -1.6V from the test point to gnd.

The DC_PRT protect is a single analogue signal representing the sum of all the channels (via R1152, R1120, R1189, R1047, R1258, R1077, R1221). The controller can't discriminate between channels & will just shutdown the amplifier. I wonder if "DC_PRT:  0L", the "0L" actually means overload i.e. beyond the ADC range. If the DC_PRT is really 0V, you would think it would come out of protection (despite the real dc offsets on the outputs).

I'll have to look into the "DC: 0" error & see what it means.

The service manual on page 61 says it displays "the voltage at 1 pin of IC902, normal value 35 to 68; 3.3v = 255" and the word "overload" only appears once in the document in an unrelated issue.   0L  doesn't appear in the document that I could find so I assumed it was just the value 0 -- I believe it is a 0 (zero) and not an O (the letter) as the dot matrix VFD can display them differently.

By the way, I did check the schematic for what shakalnokturn mentioned i.e. 3.3V regulator & reference voltages. This amplifier does have something similar (see attachment). I don't think it's the problem but the information is there if you wish to check.

Thanks for that.  I had searched for the term "regulator", logged all of them, and looked up all the part numbers to see what voltage they are meant to produce, but haven't tested any of them yet.  The only accessible ones are going to be on the very top board which holds the HDMI ports, Digital (1) and I doubt one of those being out would have anything to do with this.  Below that is the input board with the RCA jacks called FUNCTION(1), then finally at the bottom is the amplified output board MAIN(3).

I checked the mentioned pin of IC902 which is on the top digital board and is easily accessible and got a reading of -191mV between it and chassis ground.  There is 0 ohms resistance between chassis ground and the ground of that chip. 

[alanep]

Wow, there's a fair discrepancy between the test points & the actual outputs (not sure why). Either way I think the DC offset is the problem to investigate.

I'll reiterate the power amplifier at DC is just one big unity gain amplifier. The voltages (when measured with respect to chassis/input ground) of (1) LH Q1022 base, (2) RH Q1022 base & the (3) output (speaker output) should all be identical.

It may be an advantage to have 2 digital multimeters i.e. one continually looking at the output and the other measuring the points I mentioned above. Have a look at the reattached circuit. You can measure the base voltages indirectly by probing the ends of R1053, R1045 (connected to the respective bases).

Re "0L", the closest I could find was page 32 (Self Diagnosis) with "DC_PRT:xxxH". Also on the same page was "PSx PRT:xxxL". Either way it is just a suffix & as you said, not meaning overload.

[shakalnokturn]

I too doubt that the 3.3V could be the problem here as Vref is common to CPU Vcc you'd be experiencing other troubles if it was bad, besides for DC protection to give false low readings Vref would have to be too high.

I'm back to alanep's first impression:

Very unlikely to have a DC offset related to a fault on every channel (although not impossible), this is more likely a ground / amplifier power supply problem. It's possible you may also have a different problem on surround back L that would need addressing later.

Does this model have a cut-out in the bottom frame that would give access to the relay PCB solder side?

Maybe worth doing a continuity check between all grounds that should be common starting from transformer center point to speaker negative terminals, power amp's then check power stage supply voltages.

I'd have to get more involved and download the service manual, that's not happening on my smartphone...

[alanep]

shakalnokturn, to bring you up to speed, have a look at that previous attachment "Amplifier DC conditions.jpg". I was really interested in what I called reference ground (0V) (about posting 7 - don't know how to link). It provides the 0V DC reference for all channels. The unity gain voltage measurement suggestions were a sanity check.

[shakalnokturn]

Corrected that in my previous post, you were indeed the first to have doubts on the ground.
In my books Yamaha are far better than Denon, still dry solders on interconnect can be a problem.

[alanep]

The reference ground (0V) for SBR, SR, FR is via W1042 &  for C, FL, SL, SBL is via W1041 W1043 i.e. completely different connectors (so if it's missing, then it's further back). I think the voltage sanity check is needed just to "tick the box".

PS I hope I don't sound too abrupt. I am an Engineer (retired) i.e. "on the spectrum".

[NHcoder]

I too doubt that the 3.3V could be the problem here as Vref is common to CPU Vcc you'd be experiencing other troubles if it was bad, besides for DC protection to give false low readings Vref would have to be too high.

That was my initial thought too, but when I catalogued all the voltage regulators in the receiver, there are 22 of them and 5 are 3.3v.  Just occurred to me that it could be using a 3.3v zener for the reference as well I suppose, but I didn't check.  I do agree though that a ground issue somewhere seems more likely.

I'm back to NHcoder's first impression:

My first impression was "help, wtf is wrong with this thing?!"   Saw your correction while posting this.

Very unlikely to have a DC offset related to a fault on every channel (although not impossible), this is more likely a ground / amplifier power supply problem. It's possible you may also have a different problem on surround back L that would need addressing later.

Yeah, this is what made me think of the regulator.  Having virtually the same fault across every channel seems really suspect.  I believe I mentioned that one of the HDMI outputs is "dead" as well, it just puts out discolored garbage since bringing the receiver out of storage.  The circuits don't seem related but maybe something failed even worse here and affected the system on a wider level?  Just spitballing.

Does this model have a cut-out in the bottom frame that would give access to the relay PCB solder side?

Unfortunately no, just tiny vent holes.  When I first saw what a pain in the neck it was going to be to get access to the big power transistors to test them, I looked for something like that to probe through before disassembling the whole thing.

Maybe worth doing a continuity check between all grounds that should be common starting from transformer center point to speaker negative terminals, power amp's then check power stage supply voltages.

I don't think I can test that with it assembled, but I'll go for it if nobody has any other suggestions to try before that.  The big transformer is shrouded and appears to be glued to the chassis as well in spots.  It has about a million wires going into it, give or take, and they are short enough that I don't think I'll be able to get direct access even if I remove the mounting screws and try to lift it up with everything attached.

I'd have to get more involved and download the service manual, that's not happening on my smartphone...

I can't attach the one I have as it's larger than the maximum size allowed by the forum, so I uploaded it to google drive.  Not the easiest thing to navigate on a phone for sure, but at least if you want to look at it on the PC you won't have to find it first:  https://drive.google.com/file/d/1n7gMIAvhxn-x_hrzlnGUg5rOaXkB6AOI/view?usp=share_link

Honestly I find it pretty amazing that it was assembled like this at all by humans.  The larger gauge wires running to the amplifier boards appear to be crimped on both ends; it doesn't look like they can simply be unplugged.  I have a ton of spade connectors here and I'll admit I was sorely tempted to cut them all and put spade connectors in when I had it all apart the last time.

The reference ground (0V) for SBR, SR, FR is via W1042 &  for C, FL, SL, SBL is via W1041 i.e. completely different connectors (so if it's missing, then it's further back). I think the voltage sanity check is needed just to "tick the box".

I think I can probe both of these without difficulty.  We're talking about checking continuity/resistance between chassis ground and the "E" marked pins on those two connectors, correct?  I can see on the schematic for MAIN(2) for example, for W1042 that this ties ground together a bunch of emitters and what appear to be filter caps. 

Thanks again to both of you!