Denon AVR-X2000 mainboard caps bulging

[re-atari]

I have been using a Denon AVR-X2000 receiver in my living room for several years now, works perfectly, am very pleased with it. Recently I acquired another one, advertised as defective with flashing red led a few seconds after switching on. Picked it up to try repair and use in hobby room. After switching on, the on/off indicator led indeed initially flashes green for about 2 seconds, during which the display is visible, and then starts flashing red (at 2Hz) with a switched off display.

I'm afraid the service manual is too large to upload here (55Mb), it can be found at https://www.manualslib.com/manual/10...-Avr-E400.html

After opening the receiver up I straight away noticed the tops of both large electrolytic caps c4052 and c4054 (ELNA 12000uF 71V, marked 'For Audio' on the casing) are bulging. The breach lines are still intact, so no gunk oozing out, the two tops are just domed. I have never seen this phenomenon yet on any amp or receiver I have worked on, and don't see a reason why the two caps would be domed either. Looking at the service manual page 127, they filter the voltage for the power amps, get this voltage fed from a large linear transformer and bridge rectifier, so are not part of a SMPS. On my working AVR-X2000 the tops of these two caps are completely flat, no bulging or domed tops at all.
Could these two bulging caps be the single cause of the receiver shutting down? And I'm curious, what could have caused this bulging? Massive overheating, eg. if the receiver had been built into a tight cabinet without enough space for cooling or running it at max. volume for a long time at a party?
I'm a bit reluctant to order replacements just yet, as caps like these are quite expensive, and there might be other gremlins lurking beneath the surface.

I already did some mesasurements after dismantling the PCB's. On the power amp PCB all output transistors and resistors are OK, no shorts detected. I thought of running the receiver with the power connector to the power amp board disconnected, to see if it will do a normal start up then, but as I measured no shorts on it, I don't really see the point in doing so.

On the standby PSU board I did find a defective cap, a 100nF 275V AC X2 made by Carli (c4143), that is located directly after the 230V input socket (see service manual page 121). Its capacitance had dramatically decreased, my meter showed it now was down to only 15nF. I happened to have a proper replacement available, so this one is already replaced. As I expected, this didn't solve the problem, though. The standby PSU outputs a solid 5,2V standby voltage, so seems to work OK. It is built around a TOP258MG, which is known to fail easily, but it appears to be OK here.

I measured the bridge rectifier and all diodes on the various boards, no shorts detected there. The five voltage regulators (two 7805's, 7808, 7905 and 7908) showed no shorts either. Sadly they are just about impossible to reach to do measurements with the receiver switched on, so no way of telling if the regulators output the correct voltages. The internals of the receiver are quite cramped overall, which makes it hard to measure the voltages going to the various parts of the receiver.

Another thing I noticed, is that on the HDMI board the 5,2V standby voltage is filtered by two 470uF 6,3V SMD electrolytic caps in parallel (c3789 and c3802). Strange enough, they are not the same type. Where c3789 is a standard run off the mill one, with no manufacturer's marking visible, c3802 looks to be a special one made by ELNA. This is exactly the same on my working AVR-X2000. I don't see a reason for this, looking at the service manual page 142 both are just filter caps for the 5,2V standby voltage. Can one of the experts here shed some light on why Denon took this particular route?

I can upload photos if needed.

Thanx in advance for any info!

[stj]

Re: Denon AVR-X2000 mainboard caps bulging

are the metal tops domed, or are you looking at a plastic disc?

[re-atari]

Re: Denon AVR-X2000 mainboard caps bulging

The caps have no plastic discs on top, I can see the breach lines in the top of the metal can.
Sadly, the attached photo doesn't give a good image of the tops bulging. In reality it's clearly visible, and you can feel the domes as well when you rub your indexfinger over the top of the cans.

[momaka]

Re: Denon AVR-X2000 mainboard caps bulging

Was the bad Carli X2 cap placed across the incoming AC power lines for the 5VSB supply?

If yes, I suspect the amp might have seen a power surge on the line while it was on. Since the large Elna caps are connected to a linear (line transformer) power supply, it is possible that if there was a surge indeed, it damaged the output caps. I mean, I've seen one-offs with cheapo cap brands, where one of those large filter caps would open/bulge/go bad on a linear PSU like this. But since both of your caps have failed, I'm leaning more towards power surge. If you're in the USA or in a country that provides 2-phase residential AC with a center-tapped Neutral (most houses in the USA), then a bad Neutral connection could surely cause something like this.

Either that, or the unit was severely overheated for most of its life... thought I think this case is a lot less likely to be possible, because if the amp is to get so hot as to cook the caps, imagine how hot the transistor heatsinks would have gotten. Surely an output transistor or one of those 78xx linear regulators would have fried way before then. And you would have noticed this on the solder side in form of darkened / burned PCB, despite the PCB using a dark brown color (which IMO is the worst when it comes to troubleshooting.)

[re-atari]

Re: Denon AVR-X2000 mainboard caps bulging

Hi Momaka, many thanx for your insights, it's always a pleasure to read your posts on Badcaps!

The X2 cap is indeed placed across the AC input of the standby PSU. I have attached page 121 of the service manual, showing its schematic. I wanted to encircle the culprit cap (c4143) in the schematic, but sadly after the last Nitro Pro update I cannot edit pdf's anymore...

I'm in 230V territory (EU), our wall sockets carry 3 wires, brown (phase), blue (neutral) and yellow/green (GND). One strange thing is that the receiver's AC input socket does not have a GND pin, only phase and neutral are connected. I don't know how the previous owner had the receiver hooked up, he possibly had it earthed through the tuner's coax lead. Cable radio/TV is standard in my country, so just about every home has a coax wall socket.

The internals of the receiver don't appear to have been excessively and prolongedly overheated. There's no visible massive discoloration on the PCB's or on the cooling heatsinks. Besides that, I would expect the cooling paste on the output transistors to be dried out then as well, and that's not the case. The regulators are bolted to the steel bottom halve of the receiver housing for cooling, and the steel isn't discolored there either.
There is a tiny spot on the mainboard PCB that looks slightly discolored. It's the area where r4025, r4026, r4027 are situated (see service manual pag. 127). They are three 1,1K resistors that are connected in series immediately after the positive lead of ELNA cap c4052. The resistors all measure in spec, as do r4029 (10K), d4007 (1SS133) and zd4002 (ZJ6.8B, a 6,8V zener) that come after them. These components produce the Mute+7V signal. The discoloration may very well have been caused by a possible power surge, which is what you suspect has happened.

Curiosity already got the better of me, so I disassembled my working AVR-X2000 to swap boards. This showed that only the mainboard of the failing receiver is at fault causing it to not start up, all other boards are OK. I took my investigation one step further and desoldered the two bulging large ELNA caps. This wasn't easy, as the bottom of the cans was stuck to the mainboard PCB with some kind of white silicone glue, but in the end I got them out undamaged. I was astonished to find both caps measured completely in spec: 11 mF (or 11.000 uF) on my M4070 LCR meter and an ESR of 0,018Ohm on my MESR-100!

I had been thinking of measuring the voltages that the five 78XX and 79XX regulators put out beforehand, just to exclude obvious and easy to fix faults there. Decided doing these measurements wouldn't be easy, though, as the regulators are quite hard to reach. And now I had come this far in my fault search, I really didn't want to inadvertently make a short circuit with my test leads and potentially turn an only partially defective receiver into a complete total loss. And as none of the regulators showed a short between IN, OUT and GND, I suspected they would be OK, so let them be for the moment and started desoldering the large caps. Sadly this didn't identify the culprit :-(

That leaves the million dollar question which component(s) on the mainboard could be preventing the receiver from starting up. I did some measurements on the various components on both mainboard PCB's and compared the readings, but got no differences. And as it looks like the two large ELNA caps are not the reason the receiver didn't start up after all, I'm now no step further. Very awkward.

I agree with you, that the material Denon used for making the PCB's (Phenol) sucks big time. I mean, pertinax has already been around for more than 40 years, why would a serious manufacturer even think of using Phenol.
Denon did use pertinax for the CPU board, probably because it's a double sided PCB, whereas all other boards are single sided. Must probably have something to do with cutting cost corners...

[RJARRRPCGP]

Re: Denon AVR-X2000 mainboard caps bulging

The large caps, look like bad caps to me!

[re-atari]

Re: Denon AVR-X2000 mainboard caps bulging

I really wonder if and in what way the ELNA caps could be failing. Both measured completely in spec, 11.000 uF on my M4070 LCR meter and an ESR of 0,018 Ohm on my MESR-100. Could it be they fail when under load? The receiver shuts down 2 sec. after being switched on, no loudspeakers attached or input signal fed into it. So no real load present. I'm stumped, but then again don't have an education in electronics.

I now realise there's one thing (well two, to be more precise) I didn't measure yet, and that's the inductance and voltage output by the large linear transformer. Maybe it took a hit by a power surge and has some windings shorted. I will do those measurements with my LCR meter and multimeter asap and post the results here. It's very convenient that I can compare the readings with measurements on the transformer in the working receiver.

[momaka]

Re: Denon AVR-X2000 mainboard caps bulging

The X2 cap is indeed placed across the AC input of the standby PSU. I have attached page 121 of the service manual, showing its schematic. I wanted to encircle the culprit cap (c4143) in the schematic, but sadly after the last Nitro Pro update I cannot edit pdf's anymore...

No worries, I found it easily.

That cap is for EMI/RFI filtering/compliance only, so it's not that important to the operation of the amplifier.

For this cap to loose capacitance / go nearly open-circuit like that, one of the two must have happened: either there was indeed a voltage surge on the line/phase or the cap is very old. After all, it is normal for metal film caps to loose capacitance over time. However, this amplifier isn't that old (2013-2014 "vintage", according to online sources), so I think this rules out age-related failure.

I'm in 230V territory (EU), our wall sockets carry 3 wires, brown (phase), blue (neutral) and yellow/green (GND). One strange thing is that the receiver's AC input socket does not have a GND pin, only phase and neutral are connected.

Right.
A lot of audio equipment is not grounded and uses only a 2-prong cable for phase/line and neutral so that way there is less likely chance of creating a ground-loop, which can sometimes result in a lot of hum created from the amp/speakers due to 50/60 Hz line noise getting picked up on the ground look and getting amplified.

That aside, the real question is whether the house in which this amp failed had only a single-phase or 3-phase power. If it's 3-phase, there is a possibility to have a high line voltage from a failed neutral connection (depending if the building uses Delta or WYE connection.)

The internals of the receiver don't appear to have been excessively and prolongedly overheated. There's no visible massive discoloration on the PCB's or on the cooling heatsinks. Besides that, I would expect the cooling paste on the output transistors to be dried out then as well, and that's not the case. The regulators are bolted to the steel bottom halve of the receiver housing for cooling, and the steel isn't discolored there either.

Exactly, I don't think heat was the reason why those Elna caps bulged.

There is a tiny spot on the mainboard PCB that looks slightly discolored. It's the area where r4025, r4026, r4027 are situated (see service manual pag. 127). They are three 1,1K resistors that are connected in series immediately after the positive lead of ELNA cap c4052. The resistors all measure in spec, as do r4029 (10K), d4007 (1SS133) and zd4002 (ZJ6.8B, a 6,8V zener) that come after them. These components produce the Mute+7V signal. The discoloration may very well have been caused by a possible power surge, which is what you suspect has happened.

Yes, it could be that those components got a little overheated from the power surge.

Of course, it could also be due to poor design. In particular, I see this sort of cost-cutting technique done on a lot of amplifiers. Instead of adding separate transformer taps from the main transformer to get those lower voltages needed for the pre-amp and logic circuits, they just use several resistors and a linear regulator to drop down the voltage from the main amplifier output stage voltages, since the pre-amp and logic circuits usually don't draw a lot of current. Of course, over a large voltage drop, the power dissipation can get up to several Watts high. So it's actually not too uncommon to see this section in a lot of amplifiers run a little "toasty" if they used a cheap design like this.

Since your components (the resistors and Zener diodes) are in spec, I don't imagine anything was damaged in that section. Of course, it wouldn't hurt to measure the voltage with the amp plugged in to check.

Curiosity already got the better of me, so I disassembled my working AVR-X2000 to swap boards. This showed that only the mainboard of the failing receiver is at fault causing it to not start up, all other boards are OK.

So you got the broken receiver to work with the main board from the working AVR-X2000?

I took my investigation one step further and desoldered the two bulging large ELNA caps. This wasn't easy, as the bottom of the cans was stuck to the mainboard PCB with some kind of white silicone glue, but in the end I got them out undamaged. I was astonished to find both caps measured completely in spec: 11 mF (or 11.000 uF) on my M4070 LCR meter and an ESR of 0,018Ohm on my MESR-100!

Ah, I've seen that a few times, but it's rare. Usually only happens to good caps, because they are built better.

I suspect the surge was high enough to over-volt the caps way past their surge rating, but perhaps it didn't last long enough to destroy the caps, so only a little pressure built up inside them from Hydrogen production, as the caps were over-volted.

I had been thinking of measuring the voltages that the five 78XX and 79XX regulators put out beforehand, just to exclude obvious and easy to fix faults there. Decided doing these measurements wouldn't be easy, though, as the regulators are quite hard to reach.

Well, if you got the receiver to work with the main board from the good one and the regulators are not located on that board, then I don't think the issue would be with the regulators.

That said, if you do want to check voltages in a tight spot where it's trecheros to stick your hand in with the power On, then what you can do is solder some extension wires to the input and output of those regulators and measure that way.

I agree with you, that the material Denon used for making the PCB's (Phenol) sucks big time. I mean, pertinax has already been around for more than 40 years, why would a serious manufacturer even think of using Phenol.
Denon did use pertinax for the CPU board, probably because it's a double sided PCB, whereas all other boards are single sided. Must probably have something to do with cutting cost corners...

Yeah, Phenol board are cheaper, so it is done for cost cutting. What I was referring to, however, was the black/brown die they used to coat the top of the PCB, instead of leaving it uncoated or using some lighter color.

I really wonder if and in what way the ELNA caps could be failing. Both measured completely in spec, 11.000 uF on my M4070 LCR meter and an ESR of 0,018 Ohm on my MESR-100. Could it be they fail when under load? The receiver shuts down 2 sec. after being switched on, no loudspeakers attached or input signal fed into it. So no real load present. I'm stumped, but then again don't have an education in electronics.

If they measure OK for capacitance and ESR both in spec, they likely are OK.
As mentioned earlier, probably the power surge didn't last long enough to damage them, so they only got a little "gassy" inside from the over-voltage, but the vent and other seals didn't get compromised.

Over time, though, that pressure will find its way out and may even force electrolyte leakage from the bung on the bottom. So you can probably continue to use the caps for testing. But once the main issue is found and fixed, then it would be best to replace those caps for better long term reliability.

I now realise there's one thing (well two, to be more precise) I didn't measure yet, and that's the inductance and voltage output by the large linear transformer. Maybe it took a hit by a power surge and has some windings shorted. I will do those measurements with my LCR meter and multimeter asap and post the results here. It's very convenient that I can compare the readings with measurements on the transformer in the working receiver.

Yes, if you can, get readings with both the working and non-working board of the voltages from the main transformer (both AC and rectified DC would be good to see.) This might give us a better clue to what's going on.

Hi Momaka, many thanx for your insights, it's always a pleasure to read your posts on Badcaps!

Thanks, I appreciate the compliment.