Dali Subwoofer Repair and Learning Excercise

[phoenixdigital]

Hi All,

I have busted Dali Sub I got from a friend which was just making a loud single frequency hum through the woofer when turned on. I've previously fixed a Jamo sub which had goo which had become conductive and burn't traces on the board. This time it looks to be a little more challenging.

I removed the plate amp and saw that a resistor on the mainboard had exploded and destroyed a few components around it. So I managed to find a service manual online for an older model and the circuit looks to be pretty much the same as the one I have.

So I replaced all the components around it as per image below



However it still has the same hum through the woofer.

I was pretty much ready to give up on it but thought "This could be a great excuse to finally buy an oscilliscope" So after a few days of trawling this sub and youtube I settled on the Siglent SDS1104X-E as a good starter.

I did learn circuit design back in my electrical engineering days at Uni a few decades ago so I'm moderately competent with the basics understanding of circuits. I haven't really needed this knowledge in my day job as I steered more towards the other CSE and Cyber areas in my job. So I'm not a novice but I'm definitely not an expert.

I have a few questions for when the oscilliscope arrives. This is my understanding of the circuit so far.



Oscilliscope Test

• Verify all voltages marked above with red stars are DC and not oscillating.
• Check all audio signal paths and try to find out where the hum is being introduced. See green stars on above image.

Questions

• What does that section do that fried? It seems related to turning on the main woofer amp circuit.
• What does that section to the right of the fried area do? I can see it's connected to a "Relay2" which looks to cut the signal to the woofer.
• The 4558 opamps seem to have V-1, V+1 and V+2 going into them. I can't find reference to them anywhere except on the right circuit bottom where my note is. Is that where they are connected to?
• Is there anything else I could check? Looking through the forum with sub related questions it could be a problem with one of bridge rectifiers B.D2 or BD1 on the circuit or bad caps associated with these which are feeding DC to various parts of the board..

Any help would be appreciated.

This is more of a learning excercise than a recovery of a dead sub. However if I can bring it back to life that would be a massive win.

[phoenixdigital]

Some additional pics. Just then I did extract the input board (left of circuit diagram) and it appears one of the diodes has failed. That can be the next replacement before the oscilliscope arrives.

Some more pics of the boards and their locations on the circuit diagram attached.

Edit: Also just did a diode check on the bridge rectifier B.D2 on right circuit (part number RB152). It appeared to be borked as well. The same part number of BD1 (left circuit) measured with a diode check correctly so it was good to have a reference part to see if they were the same.

Ordered a ZD1 and B.D2 as well as a bunch of other matching caps (just in case) from Mouser. Will await their arrival. Maybe I didn't "need" that oscilliscope after all.

[phoenixdigital]

OK so I took out ZD1 and BD2 from the board and tested out of circuit with the diode check on the multimeter. Both seemed to test OK.

I'm guessing the capacitor in parallel to ZD1 causes problems with the diode check. Does that sound right? I'll be replacing it regardless as the board under it was quite dark.

As for BD2 even though it checks fine out of circuit I'll replace it and it's associated capacitors. Interestingly multimeter diode checking BD1 in circuit tested fine but for BD2 in circuit the -ve side of the bridge rectifier didn't work but the +ve side did.

Maybe I'll end up using that oscilloscope after all.

[NiHaoMike]

Might be easier to just replace the amplifier with a new board, they're really cheap nowadays.

[phoenixdigital]

Might be easier to just replace the amplifier with a new board, they're really cheap nowadays.

Thanks for the reply. They are definitely cheap in the USA but in Australia not so much after shipping.

Regardless as I said this is more of a learning exercise to understand the circuit and repair process. I've already got a sub that is more than adequate and getting this working would be bonus but not essential.

I've watched a bunch of vids on opamps and have started to get a bit of an understanding of more of the circuit. There are some crazy multiple feedback loops in a chain of opamps on the on/off/auto side of the left circuit which go way above my head.

After studying the circuit a bit more as well I noticed there are two relays.

• R1 - Seems to be related to the section of the circuit that fried. When triggered it opens up the power to B.D3 which provides power to the main woofer amp section
• R2 - To the right of the fried section which when triggered closes the woofer circuit to ground.

Not sure why they would need two relays. Maybe to ensure that one starts before the other so there is no "pop" as the woofer amplifier or opamps are powered up.

Also I suspect the section that fried (B.D2 area) also provides the V+2 for the opamps on the left. The circuit diagram doesn't really show this but I can't see any reason why that connector would send that voltage back from the right circuit to the left circuit. I'll have to trace on the circuit board to see where the opamp power rails come from.

Interestingly I noticed that the R157a I replaced has already started to show heating issues but still measures fine. It's possible that the potential failure of B.D2 rectifier is sending too much current through that resistor.

I'll keep plugging away at it and post my results even if it's just me posting to myself   Hopefully in the future someone else might find this useful watching me bumble on through (or give up). If this is not the place for me to post this happy to stop.

[phoenixdigital]

After a bit of reading on how opamps work I think the circuit works like this. If I've got anything horribly wrong I'd love for someone to school me.

The high pass filter kind of confused me as why would you want to pass high frequencies to a subwooofer. That is until I plugged the numbers into an online calculator for a "Sallen-Key High-pass Filter Design Tool"

http://sim.okawa-denshi.jp/en/OPseikiHikeisan.htm

Which gave 3Hz 28Hz. I'm assuming this is to stop popping as it turns on and also protect the speaker coils from DC signals since it can only really reproduce 20Hz and above.

Regardless when the oscilliscope arrives I'll be checking all the voltage rails on the opamps to see if they are stable and also check the outputs to see if I can spot where the hum is being introduced.

Again this is mainly a learning experience. I know I could just buy a new amp but this is more interesting.

Edit: I think I got my capacitor values wrong in the high filter calculator and the cutoff might be 28Hz which probably tapers off slow enough to capture 20Hz and up.

Edit Edit: It also had me wondering how the amp stays on for a few minutes when set to AUTO even when the audio signal stops. This lead me to lookup opamp timers and I can see a few components related to this in the auto section of the U5 opamp. Very cool. I updated the diagram

Edit Edit Edit: This analysis has obviously gone above and beyond identifying the core hum problem but I'm finding it pretty satifying spotting various functions that match up with examples online.

Class A - Darlington Transistor Configurations - https://www.electronics-tutorials.ws/amplifier/amp_5.html

Class B - https://www.electronics-tutorials.ws/amplifier/amp_6.html

Timer Opamp - https://www.hobbyprojects.com/operational_amplifier/opamp_as_a_timer.html

The attached diagram is getting pretty big so if you have trouble reading it download and zoom in to witness the analysis madness

[phoenixdigital]

OK I think I solved it!!!!

After replacing all of the components in the opamp power supply part of the circuit. The resistor R157 kept overheating and releasing magic smoke.

Poking around with my multimeter I was measuring -6V on the negative rail instead of -12V so this would explain why it was burning up. I was kinda stumped because everything looked right.

So I built a simulated circuit to verify what I was expecting to see and of course it should be generating +12V and -12V.

https://www.multisim.com/content/MLSGsZ2h5g6DvTXqThewnM/subwoofer-opamp-power/

So I took off all the components and started measuring around looking for shorts. Sure enough I found some conductivity between the -V rail and GND measuring about 140ohms. A lot of dremelling later and it finally measured OL on the multimeter.

I'll be waiting for the epoxy resin to dry then put all the components back in but I'm feeling moderately confident it's solved. I'd say over time the board got hotter and hotter and the carbon from charring started conducting till BOOM!!!!

Will report back.

[Ground_Loop]

I recommend get that black soot off the board. It's usually conductive. A stiff wire brush and alcohol should make quick work of it.

[phoenixdigital]

Thanks for the tip. I cleaned up as much as I could.

Soldered everything back in and tested it out and was getting +13V and -13V. A little higher than I expected but closer to where I wanted to be.

The major hum has gone but there is a subtle hum now. It doesn't sound like 50Hz (Aust mains freq). But kinda sounds close to 100Hz with some minor popping and crackling. Would a recitfied 50Hz sine wave sound like 100Hz?

Possibly still some ground issues.

I don't think there is much more I can do until the oscilliscope arrives so I can pinpoint what part of the circuit the noise is being introduced.

[phoenixdigital]

I had a thought maybe the inputs weren't grounded properly as I hadn't put the screws in place for the inputs which likely ground it to the backplate.

Sure enough I've got rid of the 100Hz hum and random popping. Still haven't plugged it in for a signal test which I'll do some other time.

There is still a very very subtle hum but it appears to be coming from the components and not the woofer speaker itself. Possible the active relay hum. It's really not obvious unless you are up close.

HUGE PROGRESS!


Edit: Looking at some of the many many pics I took of the internals. The input jack screw sockets are plastic so those screws wouldn't have made a difference. However the two switches on the right
 left screw sockets do appear to be metal so the screws there likely assisted with the grounding. My money's on the phase switch. If I can be bothered I'll remove it and see if the 100Hz hum comes back.

[Ground_Loop]

I recommend get that black soot off the board. It's usually conductive. A stiff wire brush and alcohol should make quick work of it.

I didn't see all the other pics with notes.  Apologies for being captain obvious.

[phoenixdigital]

I didn't see all the other pics with notes.  Apologies for being captain obvious.

No need for apologies at all. I was glad to be getting tips from someone apart from me posting into the void

I probably wouldn't have been as diligent if you hadn't suggested that.

The good news is I've been using it for the last 2 days and it's been rock solid without issue. I haven't put all the screws back in yet as I plan to pull it apart one more time to put some silicon sealant next to the caps I replaced as well as other parts I removed the goo from to keep things from "flapping around in the breeze". I was going to hot glue gun it but saw mixed responses online saying it was risky. The general consensus was RTV compound the best choice.

https://www.bcf.com.au/p/selleys-marine-sealant-75g/166322.html

While I can see what caused this to fail what I don't understand is how it started to fail in the first place. What caused the conductivity to start between the -12V and GND rail? Without a conductive short this part of the board shouldn't really get hot so the charring should never have really started. I can see once a short starts though it would just get worse and worse with time ending up in the state I got it in. Understanding the root cause could help me put some prevention in place to stop it happening again. There are a few spots on the underside of the circuit board where the +12 and -12V rails are literally 1 mm from the GND rail. Those are the points of concern much the same as the point which was the root cause here.

The underside of the circuit board is facing upwards when placed in the cabinet so I suppose it's possible enough dust built up for it to start to become conductive which then cascaded from there. Maybe I'll just put some electrical tape over this area or something non conductive and non flammable ontop of the board.

I'd considered putting the RTV sealant above on the underside of the circuit board near where it shorted to minimise the chance of this becoming conductive but I'm not sure if that runs the risk of making it happen again faster as the sealant starts to age.

[erikbrenn]

Hi phoenixdigital,

I for one enjoyed this thread. It's a while since you posted this, buy may I ask where you got the service manual/schematic for your Dali Concept Sub? 
I have a Dali Basis Sub which I'm trying to repair but I have not been able to find any schematic or service manual for it.

Just like you I have spent a lot of time trying to determine what's wrong with mine, as a learning experience. My Sub seems to have all power rails intact, but there is no sound output. I've desoldered and checked all transistors and they're ok. My next move is to desolder all the opamps and test them out of circuit.

cheers.

Erik

[phoenixdigital]

Hi Erik,

Glad you found it helpful I kinda felt like I was just talking to the ether

As for the service manual I just googled "Dali Concept Sub service manual" which brought me to "DALI CONCEPT SUB REV.B SM" on a few sites. It was an older revision to the one I had but the circuit was pretty damn close to mine.

One of the google result links here - https://elektrotanya.com/dali_concept_sub_rev.b_sm.pdf/download.html#dl

Thankfully I fixed mine before my oscilliscope arrived and the fault was pretty obvious. It's been running fine since my fix. I've been meaning to put some silicon back on the areas I removed it from but haven't bothered yet.

I just tried the same search for your model "Dali Basis Sub service manual" and couldn't find anything useful. That definitely would make it a bit harder to troubleshoot. I had a look at the backpanel of a Basis Sub online and it's pretty damn close to mine. I wonder if they didn't really change much of the circuit for that model?

Good luck. Post your progress.

[erikbrenn]

Thanks,  I cannot find a service manual for my Basis 100 either. 

These amps are basically fully analog, so should be possible to fix this. One thing that confuses me is the pcb that sits perpendicular to the poweramp board. It's fed with signals from the preampboard via a flat cable in one end, and at the other end it's soldered to the poweramp board that hosts the TIP35/36 push-pull pair.  The board has many TO-92 transistors as well as some power resistors that step down the voltage from the main transformer...   (The preamp board has it's own little linear psu not shown in the picture)

I have checked all TO-92 transistors and power transistors on the boards, and they test ok with a component tester. I even desoldered all of them. Also, when I scope the signal out of the preamp board it looks fine, and it follows any volum adjustments I do. So the fault must be after the preamp.  I also ditched the crappy flat cable shown in the picture and replaced it with proper JST connector and cables. And also, I measured ESR on all electrolytics and changed quite a few of them.

Will post progress, if I get any

cheers

[Jwillis]

According to schematic ZD1 should be a 12V zener. The  part that you replaced the resistors is a part of the the power filter and regulation. B D2 is a rectifier. R156a and ZD 104 are for +12V regulation. R157a and ZD 105 are for -12V regulation. If the fault is on the negative 12V side then you need to look for a short on that circuit. Many of the opamps are dual rail, meaning they are supplied by both positive and negative voltages relative to ground.This will include amplifiers for Left and right channels and low pass filters.
The main sub woofer amp is regulated at  +15V and -15V with ZD101 and ZD102 and supplied with its own rectifier BD3 and smoothing caps C122 and C123. If there are no problems with the Sub Woofer amp all the trouble with be at the control/audio filter part.
I apologize if I repeated something you already know. Just have a hard time seeing the lettering to be able to read it all. 

[phoenixdigital]

Your circuit is definitely very different to mine. I suspect part of that vertical board is the power amp based on those fat resistors. I do have the same style resistors on my power amp side of the circuit.

I've uploaded all the pics I took when I repaired mine and you can see them there if you want. Comparing the PCB numbers to the circuit diagram I managed to get you can see they are part of the power amp.

All my pics here if that helps

https://imgur.com/a/2HjJzoO

Do you have any big power transistors like the ones in the 4th image at the back?

As Jwillis said there is likely a +ve and -ve power rail for those preamp circuits.

[erikbrenn]

Yes, completely different circuit unfortunately.

Here are closeups of the poweramp board. The two big transistors are TIP35/36  push pull pair, they test ok with my transistor tester.   The transistor to the right is a smaller one, and it also tests ok.

I have testet all voltage rails and they seem ok. Well actually the one that should be +15V is 18V. Which is strange because the max voltage rating of the opamp 4558 SIP ICs is 18V.  But the preamp seems fine.

The TO-92 to the left is a 10k varistor. The second one from the left is a smaller transistor,  will check that next.

[phoenixdigital]

You've definitely got a harder task than I did. At least mine had obvious burnt components and board.

I'll be watching with interest so keep us updated.

Does the audio signal after the preamp enter the poweramp section OK if you can work out where that is?