Sign in   |  Join   |  Help

BeoLab 9 Overhaul, Crackling

rated by 0 users
This post has 15 Replies | 3 Followers

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member
Menahem Yachad Posted: Wed, Feb 24 2021 2:08 PM

This is the first time I've opened a BL9, and I'm very impressed. A customer in Tel Aviv imported a pair of these from Poland, and they had a number of problems. Orange and Red LED's when not supposed to, Poor quality sound, and one of them has a crackling sound - it's NOT the tweeter!.

I have already done my standard procedure - replacing all the SMD electrolytics on the Crossover PCB's, and those faulty electrolytics on the ICE PSU. All solder joints resoldered on ALL PCB's - there were many cracked joints. Discussed in expanded detail later.

Both BL9's reassembled, and all problems are solved, except for ONE - the intermittent crackling sound on one BL9. The other one is fine.

I've read through all the previous BL9 posts which may have indicated a cause for the crackling, but no luck. I did find many posts with BL9 PSU failures, which were solved here, and those members may be able to obtain some help by reading this post carefully.

I am very fortunate that I have another perfectly functional BL9 to use a a swapping source. So, I am now at the stage of swapping individual PCB's between the BL9 chassis, in order to isolate the faulty PCB. I will then troubleshoot that PCB. Sending the BL9 to a "Service Center" is not an option - the problem will be solved here.

Here's my large table open and occupied for organized swapping of the PCB's.


Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

So, I will discuss now my work already done on this ICE PSU PCB.

I saw that other members found a blown fuse, and hoped that a replacement will solve the problem. Wrong.

 

Firstly, it must be understood that this PSU is a Switch-Mode (SMPS) PSU. 

It is NOT a Linear PSU, like the Pre-2000's models sold by B&O.

A SMPS operates at very high frequencies, and reserve voltages are well over 300Volts.

That means that the high frequencies cause very high breakdown rates on the passive components (such as resistors and capacitors), and also the Semiconductors can overheat, if the heat by-product is not managed correctly.

 

Secondly, when a fuse blows in a SMPS, that means that (beyond all doubt) there is a real and catastrophic problem, usually in the Primary (High-voltage) section of the PSU. 

Throwing more fuses at the PSU is guaranteed to cause only MORE incremental damage.

 

Before I started work on the PSU PCB, I did a visual inspection to identify any visual clues for defects.

There were no bulged capacitors (but no-bulge is NO guarantee that a capacitor is functional or not), and I was impressed to observe B&O's choice in the components  - Panasonic, Vishay, and ChemiCon - all very high quality.

 

The only component of which I didn't approve, was the RIFA paper "X2" 220nF supppression capacitor C111 275VAC, especially since the other "X2" cap C113 is a super-high-quality Vishay 339 rated at 310VAC.

In Europe, 310VAC rating is more "forgiving" than a 275VAC rating. Why would B&O have chosen to use the Rifa in one X2 position, instead of installing 2 identical Vishay caps? Maybe lots of Rifa caps in stock? 

These Rifa X2 caps are notorious for exploding and causing all types of damage. After repairing hundreds of Tektronix oscilloscope PSU's, ALL with Rifa X2 explosive damage, I automatically junk these Rifa paper caps whenever I see them installed. They have not stood the test of time.

So, replacing the Rifa paper cap (with the identical Vishay 339 as C113) would be my first recommendation, when looking at a blown fuse. And of course, check for any other collateral damage.

 

The other visual damage was the cold-solder (cracked) joints on the PCB underside.

Very common on SMPS PSU's, especially with today's unreliable Lead-free solder.

So, I resoldered all the solder joints with Kester 63/37 fully-leaded solder.


Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

I now started measuring all suspect passive components.

The first suspect was R114, a Carbon Composition 1/2W resistor. Carbon Comp resistors do not age well - they absorb humidity, and depart from their specified resistances, sometimes drastically.

This 470R measured over 600 ohms. Predictable, and I replaced it with an new Carbon Comp resistor. 

Carbon Comp resistors are installed for a very good reason - they handle high-frequency pulses better than any other resistor types, and it would be foolish to replace it with any other type.

Here is the new Carbon resistor and new Vishay 220nF "X2" capacitor installed.

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

The next suspects were the electrolytic capacitors.

Normally in SMPS's, the small capacitors fail at a much higher rate than the large units.

So it was here.

The large Vishay, Panasonic FC and Chemicon electrolytics checked perfectly OK on my ESR meter, and there was no need to replace them. Excellent quality, and super-reliable.

 

Only these electrolytics needed replacing (circled in red):

22uF/50V C514 and C516 

100uF/25V C425, C429 and C611.

I used my gold-standard for SMPS's - Nichicon UPW and UHE.

 

Subsequent testing proved the SMPS completely functional.


Dennis
Top 150 Contributor
Denmark
Posts 614
OFFLINE
Gold Member
Dennis replied on Wed, Feb 24 2021 10:20 PM

Following this with great interest, as it is nice seeing someone repair some of the newer B&O products - especially as I guess you don't have any schematics to help you? I'm sure this thread will become useful for others in the near future. 

Keep up the good work. Smile

/Dennis

Hyponx
Not Ranked
Posts 20
OFFLINE
Gold Member
Hyponx replied on Wed, Feb 24 2021 11:10 PM

Thanks for posting! I own a pair of Beolab 9's myself and definitely appreciate seeing some of the issues I may expect down the road.

Beobuddy
Top 25 Contributor
Utrecht, The Netherlands
Posts 3,863
ONLINE
Founder
Beobuddy replied on Thu, Feb 25 2021 8:58 AM

Before you start to replace all the capacitors, start with identifying the crackling sound. Where does it come from. Just from one driver, or does al the drivers contain that intermittent crackling sound? Crackling noises are never caused by degraded electrolytic capacitors.
From all the defective BL9's I've had, none of them had issues caused by these. 

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

Beobuddy:
Crackling noises are never caused by degraded electrolytic capacitors.

From all the defective BL9's I've had, none of them had issues caused by these. 

 

All the electrolytic capacitors have already been replaced.

That's the first order of work - to replace ALL confirmed defective parts.

The crackling is the ONLY problem remaining, therefore you are correct that the capacitors are NOT the cause of the crackling.

 

The crackling comes from the tweeter.

The tweeters were already swapped in order to determine whether the problem was from the actual tweeter. Not so.

 

So the first PCB to swap will be the X-over.

My instinct is pushing me in the direction of a cracked SMD solder joint somewhere, but time and further checking will determine whether I am correct.

 

Have you actually solved any crackling problems on the BL9's?

Beobuddy
Top 25 Contributor
Utrecht, The Netherlands
Posts 3,863
ONLINE
Founder
Beobuddy replied on Thu, Feb 25 2021 11:40 AM

I’ve solved crackling noises on BL4, 4000(both mk1 and 2), BL1, 3, 5, 7 and 9.

Causes are diverse. 

Beobuddy
Top 25 Contributor
Utrecht, The Netherlands
Posts 3,863
ONLINE
Founder
Beobuddy replied on Thu, Feb 25 2021 11:49 AM

Menahem Yachad:

That's the first order of work - to replace ALL confirmed defective parts.

You do not know this, unless you desolder all capacitors for a reliable measurement. Even then a measured valid value isn’t a guarantee that it works as it should.

 

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

Beobuddy:
You do not know this

On the contrary!

Checking ESR in-circuit is a perfectly reliable method of determining whether the capacitor is functional or not (except in a tank circuit).

 

For capacitance, you are correct. Capacitance MUST be checked out-of-circuit.

However, after ESR has confirmed a bad cap, the capacitance is of academic value only.

 

Can you expand on the actual causes of crackling problems in your past experience with BL9's?

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

PCB1, also described as the Crossover PCB, is the faulty PCB.

Next step is to isolate the faulty components.

 

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

Checked the Inputs and outputs of the STK amp.

Pin 11 on the STK403-130 is the input - shows the crackling, so I'll be reverse engineering the circuit backwards from the preamp to the STK to find the culprit.

The schematics would be helpful here....


Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

Scoped all the inputs and outputs of the MC33078 Dual and MC33079 Quad OpAmps.

Without a scope, this troubleshooting would be absolutely impossible.

 

What I was looking for is exactly this: OpAmp input Clean, OpAmp output Distorted.

Found this exactly on IC204 MC33079 - Pins 12 and 13 Input - Clean, Pin 14 Output - Distorted.

 

This is now the area of interest, as shown on the photo. This IC is under the metal shield on PCB1 (Crossover).

Either the OpAmp itself is defective, or the surrounding components of Pin 14 are defective, and are introducing distortion onto that rail.

Currently diagnosing.


Beobuddy
Top 25 Contributor
Utrecht, The Netherlands
Posts 3,863
ONLINE
Founder
Beobuddy replied on Sun, Mar 14 2021 5:25 PM

Most likely that it wil be caused by the 33079. With some coldspray you probably can influence the amount of distortion.

Menahem Yachad
Top 75 Contributor
Jerusalem, Israel
Posts 1,226
OFFLINE
Bronze Member

Yes, I would very much like to believe that it is so simple - just replace the 33079. 

Experience has taught me otherwise.

So, I had already used Freeze Spray, and that caused no change at all.

In the old days (before SMD), it would have been very simple - just desolder the legs of the parts around here, and in 20 minutes, the defective part is identified.

But here, I've had to cut the PCB traces from Pin 14, which go to the surrounding parts, and I have not been able to cut every trace, simply because they are so small, that there is the real danger of not being able to repair them afterwards.

So if you look at the above picture, to the left of the red circle are 8x MMBF4393 FET's.

The Drain pin on the right-side set of 4 FET's is connected directly to Pin 14, therefore also shows the distortion. I have no means of cutting that connection, so my next step is to do DC voltage checks on all those pins, to see if there is one pin, slightly non-uniform.

A real journey with these micro-parts, and being unable to desolder and lift legs.

Page 1 of 1 (16 items) | RSS
Beoworld Security Certificate

SSL