Repair of an AVR, (virtually) no sound

[zzattack]

With the holiday period I want to tackle the repair of a huge pioneer AVR with an issue. The problem doesn't seem too bad but the complex design makes a scattergun approach infeasible. There's literally over 30 boards inside with multiple dozens of connectors and flatcables. This makes troubleshooting/following paths very difficult, as most stuff is inaccessible while assembled. So probing in circuit requires disassembly, soldering some wires, and then reassembling first. I've fixed receivers before by blindly measuring until I find shorted FETs, burnt up diodes or fuses, but I'd like to learn doing this is a more organized and structured manner. I'd like some hints on how to proceed.

Now, for the symptoms:
 - when I got the device, it did not turn on and immediately fell into a protection mode
 - according to user manual, the specific accompanying LED pattern indicates the problem had to do with the SMPS assy
 - that error could be reset, after which receiver turns on for ~5s, then gets stuck in the same protection mode again
I took the device apart all the way to the bottom. Nothing too bad aside from dirt and dust, possibly 1 slanted ribbon cable (highly doubt this was an issue). Since I could not find any problem with any of the diodes, FETs or fuses I reassembled it and for whatever reason, the receiver now turns and remains on. I don't have an explanation for this. I don't think it's likely I solved a short circuit issue due to some conductive remains anywhere.

Either way, unit now turns on and the next issue exhibits: I get virtually no sound output. Symptoms:
 - when changing the volume, with every tick, a slight pop is heard over the hooked up speakers (tested all outputs)
 - there's no kind of hissing or background noise, even at max volume
 - digital inputs are recognized correctly from optical, coaxial and HDMI sources (but again, no sound output). Muting/unmuting the input causes the same small pop on the output speaker.
 - analog inputs are heard over the output, but very quietly only; the volume changes with source volume, but turning the receivers' volume know does absolutely nothing (aside from the pop with every tick)
 - nothing discernible on pre-outs, there's a slight change in noise on the scope when muting/unmuting, but it does not resemble the analog input
 - there's no measurable DC over the outputs

I'm not sure what I can conclude about these symptoms, and what would be a good starting point to dig deeper. Any hints are appreciated.
I've got a service manual available for most of the machine, aside from the SMPS. Probably because that part was built by Bang & Olufsen. Analog path diagram is attached.

[Chris56000]

Hi!

Could you tell us the model number or provide a link to a service manual please?

I can give you some fault finding and repair hints once I've studied it!

Chris Williams

[Chris56000]

Hi!

This is an insanely complex system relying on at least six main microcomputer systems to all talk to each other for the unit to operate, and from the sound of it one of the Audio subsystem micros isn't enabling the audio paths through to the ICE power amplifier assemblies!

Open Manual RRV3708 at page 46 and enter the "Service Mode" by means of the instructions at the beginning of page 46, and press the up down keys until the screen headed "VERSION" is shown on the unit's display screen.

Are all the entries on the "Version" page showing a software number?

If they are, then all of the microcomputers should be operational, if any of the lines on the Version page or blank, zeros or question–marks, this indicates this particular i.c. isn't operational or communicating with the rest of it!

If the version page shows OK as described on page 46, enter the Control Panel Mode as described on page 52 of Manual RRV3708 and carry out an "ALL RESET" to clear any stored errors in the system.

Does this help with the fault symptoms?

Chris Williams

[Chris56000]

PS!

I can drill down to components–level diagnosis with this from the information you provided for me but we need to go this in logical stages as you quite correctly remark in your first query!

I'll deal with each further step in sequence depending on the answers to your fault–finding tests made before in the previous steps!

Chris Williams

[zzattack]

Thanks for looking into this. The version info of each component is showing, so there's at least communication between all of the processors. I had also already reset the error count and other statistics. After a few hours of toying around, I reentered this calibration menu and confirmed all fault counters are still at 0.

[zzattack]

Looking at the analog audio block diagram, it seems to me that the issue would lie between the "multi ch. in assy" and the "preout assy". The preouts are the first external location I can probe easily and the multi ch. in assy is the last on the path where I can inject a signal. By probing the audio bridge assy #3 and #4 it should become clear whether there's a problem with the volume assy or whether it's earlier on the path. Most suspicious IC seems to be IC3351 or IC3267, since those are the only ones that are unconditionally included on any path, be it analog or digital, for all channels (whereas volume assy uses 1 ic per channel).

Does my reasoning make sense? And clearly preouts are outputs, so maybe that's enough of an indication that it will be a stupid idea to do so, but would I be able to test individual amplifier channels by injecting a signal on those outputs?


Edit, update after doing some tests: there seems to be no audio signal on the bridge assy 2 or 3, I think, unless I need <100mV/div to even see it. I can inject (with the 'input' side of the bridge not plugged in) a tone on these bridge boards, and this is audible on the hooked up speaker. Turning the volume knob makes no difference in the volume of the tone, but changing the input amplitude does. I need a higher amplitude on bridge board 2 than I do on board 3 to reach the same level of volume. Roughly 5Vpp results in a very clearly audible tone. If I inject a tone on the bridge 4 assy, I need to use a much much lower amplitude. About 10mVpp results in a well amplified tone.

On the one hand, this seems to indicate a problem with the volume assy. On the other hand, I would have expected the tone to be easily found on bridge assy 2/3 if that were the case. Which of my conclusions is wrong?

[Chris56000]

Hi!

Ah – it looks like everything's OK after the "Audio Bridge 4 Assembly" and the loss is occurring either in the volume control assembly assembly "I" (AWX8737) or elsewhere in the signal path from Audio Bridge Assemblies 2 and 3, etc., prior to it!

Can you trace input signals on the pins of CN3901 and CN3902?

Chris Williams

[shakalnokturn]

Follow it through with Chris, you'll get a thorower diagnosis of the problem.
From my experience with AVR's the starting point is to run through the numerous cracked solders.
In your case it sounds like a digital volume control IC failed due to bad power to one of its rails.
If such is the case, you'll need to find why power failed in the first place.

[Chris56000]

PS!

Here is the datasheet for the TC94A32BFG Electronic Volume Control ICs – please don't confuse it with the TC9432 which is a much larger Digital Servo IC Processor & DA Converter!

The Volume Control ICs are all powered in parallel from the Common Analogue ±15V power supplies at CN3901 pins 14, 15, 16 and 17, whilst the two data (MVOLDAT) and clock (MVOLCLK) lines control all six ICs together on pins 4 & 5 on CN3903.

The volume control data is then latched onto the TC94A32BFG ICs by Master Control Strobe pulses coming in on pins 2 and 3 of CN3903 whenever the volume control is physically moved to change the volume.

However, four of the ICs for the rear channels, sub–bass/subwoofer, etc., have a Strobe Control line separately pinned (MVOL STB2) on CN3903 pin 2, whilst the Front and Centre Control ICs are controlled by Line MVOL STB1 on CN3903 pin 3.

Are you getting data on the MVOL DAT pin of CN3903 that changes when you rotate the volume knob, and are both MVOL STB1 and MVOL STB2 lines  showing positive–going Strobe Pulses when you physically move the Volume knob?

MVOL CLK on Pin 5 of CN3903 in receipt of a continuous train of clock pulses?

Chris Williams

PS!

I appreciate how aggravating it is to get at test points, connectors, etc., etc., on this, the best way is to furnish yourself with a few metres of colour–coded ribbon cable, strip off lengths with numbers of cores according to the number of ways of each board connector you want to test, then solder the ends (carefully!) to the connector pins of the board you're investigating, after removing the assembly, then refitting the assembly back for test purposes afterwards, bringing out the piece of ribbon cable towards the top of sides of the unit!

The Service Manuals do mention extension leads, etc., but unless you're lucky enough to find 'em listed on fleabay, etc., the chances of getting them from Pioneer are probably nil!

Finally, this unit seems to have been made entirely from that lead–free muck according to the Main Service Manual RRV3711 – did you check all the connector pins on the PCB assemblies for cracked or dry joints with a magnifying glass whilst you had the unit dismantled?

[zzattack]

The Volume Control ICs are all powered in parallel from the Common Analogue ±15V power supplies at CN3901 pins 14, 15, 16 and 17, whilst the two data (MVOLDAT) and clock (MVOLCLK) lines control all six ICs together on pins 4 & 5 on CN3903.

The -15V rail is present, but +15V sits at around +0.45V. Likely culprit! I'll try to find out what the source of that +15V should be.

* Are you getting data on the MVOL DAT pin of CN3903 that changes when you rotate the volume knob
*  MVOL CLK on Pin 5 of CN3903 in receipt of a continuous train of clock pulses?
* and are both MVOL STB1 and MVOL STB2 lines  showing positive–going Strobe Pulses when you physically move the Volume knob?

The clock is not continuous but seems to sit normally at 0V. are definitely pulses on it when I turn the knob, and they sync up with pulses on the DAT line. See attached image. Voltage on those signals seems rather low for a digital signal at only 1VDC? Strobes on MVOL STB1/2 indeed show positive-going pulses when I turn the knob, but only if this results in a change of volume. I.e. if it's already at max volume and I turn it up, there's no strobe. Analogous for volume down when already at the lowest.

I appreciate how aggravating it is to get at test points, connectors, etc., etc., on this, the best way is to furnish yourself with a few metres of colour–coded ribbon cable, strip off lengths with numbers of cores according to the number of ways of each board connector you want to test, then solder the ends (carefully!) to the connector pins of the board you're investigating, after removing the assembly, then refitting the assembly back for test purposes afterwards, bringing out the piece of ribbon cable towards the top of sides of the unit!

Yep, that's how I'm currently doing it, no biggie. I'm not very time constrained so not too much risk of getting impatient
The HDMI board completely blocks all access to everything underneath it which is precisely where we're looking now. Luckily, not connecting the 4 ribbon cables seems to take out a bunch of digital signals and the front panel display, but none of the analog stuff seems affected. So with the minimal available surplus length on the connectors there, I can position the board so that the analog boards are somewhat accessible.

Finally, this unit seems to have been made entirely from that lead–free muck according to the Main Service Manual RRV3711 – did you check all the connector pins on the PCB assemblies for cracked or dry joints with a magnifying glass whilst you had the unit dismantled?

Only the boards that I took out I did look at under the microscope. They seem to be in fairly good shape. No bad joints found at all. However, all of the horizontally stacked boards for all the analog audio stuff remained seated. Also, when attaching wires, I notice the leaded solder I use mixes relatively easily with the old stuff that's already there. That seems like a good sign.

[zzattack]

Possibly final update: ok, the missing +15V certainly was an issue. I took out the 78M15 and supplied +15V limited at 500mA (for which the 78M15 seems rated) from the bench. Seems good now, draw is ~275mA and volume controls work as expected. DC load confirms the regulator gives out early, significantly dropping output voltage when >50mA is drawn. Cheap fix!
Thanks Chris for your help, appreciate the guiding hand!

[Chris56000]

Hi!

Glad you located your fault!

The +15V and –15V come from IC4311 and IC4312 on the Analogue Regulator PCB by the way!

Pleased to be of assistance, and enjoy your listening!

Chris Williams