Speaker cone 'sucking in' at power-on

[nathanscribe]

I've been using a pair of KRK Rokit 5 for a few years – powered speakers, each has its own IEC lead, hooked up to my interface/mixer via balanced cables. I've been noticing a momentary loud hum in the left speaker at power-up (usually lasts a second or so, starts loud and fades over that time) – and today I noticed that when it does that, the main cone recedes into the body a substantial amount, then returns to the normal resting position. It does this without the audio connection in place too, so it's not something coming from the interface.

Clearly this isn't great. My knowledge of speaker electronics is comparatively flimsy, but would I be on the right track suspecting some kind of DC offset that's present at first power application? Would this indicate something like decoupling caps needing attention?

[James Perrett]

Assuming that they use +/- power rails, it probably means that the power rails are coming up at different speeds. It could be that one of the main reservoir capacitors has reduced in value but it could also be that something is taking more current from one rail than the other.

Do they use speaker protection relays? If so then it may be worth checking the delay circuit on those to make sure that they are activating on startup.

[Hugh Robjohns]

Yes, the cone deflection is due to one side of the power supply coming up quicker than the other side. That is not that unusual, but it should be transitory, not sustained. The hum is due to the regulation taking time to sort itself out, and taken together I would agree that it implies the caps in the power supply maybe losing their efficacy.

[Mike Shand]

One of mine started doing that and after a while it died completely. Just hummed loudly all the time. It was, as noted above, a power supply issue.

I got it repaired but a few years later the other one suddenly died.

I gave up and bought kh80s. The difference was amazing.

[nathanscribe]

Thanks for the quick replies – these are probably 12 years old so probably due a checkup anyway. Will get them on the bench this week. I'm seeing reports there's also potting compound that goes bad over time, so that's another thing to look forward to!

[ef37a]

In my experience that is not a supply problem but the output transistors being 'unbalanced' as the circuit stabilizes. Their DC conditions are all over the shop until the DC negative feedback pulls the output offset voltage back to a few mV.

Still probably a capacitor but in the amp. Get it done or one day you will have a burned out amplifier and an open circuit speaker.

Note, I keep losing connection and it seems to only happen posting to SoS?

Just thought, the PA in that speaker is likely an integrated module? Makes repair easier IF you can get hold of one!

Dave.

[nathanscribe]

Got the thing on the bench and it turns out the black goop of doom has been quietly corroding sevearal components across two boards. Oh dear.

[ef37a]

Got the thing on the bench and it turns out the black goop of doom has been quietly corroding sevearal components across two boards. Oh dear.

Ah, there is no diagnosing that! Used to destroy Hitachi tellies.

Dave.

[Philbo King]

This guide might help with repairs

https://gearspace.com/board/geekzone/11 ... akers.html

[ef37a]

Interesting and I am sure very helpful Philbo.

Just a couple of points? Yes, fit Metal Film resistors wherever you can, low noise and practically indestructible.

No point though in fitting electro-caps of a higher voltage rating than the originals, assuming they were adequately rated, in fact even if a cap is a few volts under a supply voltage it will 'reform' very quickly. Not a recommended procedure but worth a do in extremis!

Dave.

[ajay_m]

that's kind of odd though, the article implies that the glue holding components down is degrading in a way that causes it to become conductive, rather than that the caps have leaked electrolyte over the PCB.
Normally that brittle glue is just a nuisance to clean away when it's glueing down something you need to replace, I haven't seen this before. Perhaps some kind of material degradation like the glue on old Roland (I think) keyboards that simply degrades into goop.

But these power amps are usually DC coupled (as was the case with my tweeter-eating S5) and will happily put the DC rail into a speaker if something driving them has a DC offset, so yes, caps associated with the amp itself would be suspect for sure.

[ef37a]

that's kind of odd though, the article implies that the glue holding components down is degrading in a way that causes it to become conductive, rather than that the caps have leaked electrolyte over the PCB.
Normally that brittle glue is just a nuisance to clean away when it's glueing down something you need to replace, I haven't seen this before. Perhaps some kind of material degradation like the glue on old Roland (I think) keyboards that simply degrades into goop.

But these power amps are usually DC coupled (as was the case with my tweeter-eating S5) and will happily put the DC rail into a speaker if something driving them has a DC offset, so yes, caps associated with the amp itself would be suspect for sure.

When we were getting complaints about 15625 Hz emitting from CTVs the only material the mnfctr, Hitachi IIRC, said was proven safe was PVA glue to encapsulate the offending component. This was a "transductor" in the convergence circuit. Had to be removed, glooped up and left to dry overnight.

Fixed it mind!

Dave.

[nathanscribe]

After some time scraping the black goop off things, I can confirm some findings:

- the power rail electrolytics measure good with my Peak ESR meter, though I'll be replacing them with new ones anyway
- the black goop is what you might call 'liberally applied', and entirely smothers things that seem hardly necessary to smother – eg tiny resistors, jumpers, signal diodes
- the black goop shows resistance in the order of mega-ohms across gaps of several mm
- the black goop has absolutely corroded several components – one particular resistor and one ceramic disc cap simply broke away when I was working. Legs just fell off. There's fuzz on the jumpers as well, so those are coming out.

As it stands, there's no sign at all of leaking electrolytics, all the visible faults appear to have been caused by the goop of doom, and I've got to swap out about half a dozen bits because of it (excluding the caution of those 4 power rail caps).

So yeah.

If anyone can recommend an anti-vibration goop that would make a suitable replacement, or whether it's worth it, please do say.

[ef37a]

"If anyone can recommend an anti-vibration goop that would make a suitable replacement, or whether it's worth it, please do say."

IMHO don't bother. Think of the millions of guitar amps that were built 'pre gloop'? Never had a vibration problem. Even when the industry went over to printed circuits 'we' did not gloop things and still had no trouble. I would aver the vibrations inside a 50W combo are an order greater than a dinky 'domestic' monitor? A bass combo? Geez!

I always thought the glue was there simply to hold components in place while being wave soldered? Seems an easy and cheap solution but not so good in the long run. I think it only started with Japanese electronics?

If you have particularly heavy or 'tall' components, big caps come to mind, then a better way to secure them is some semi-rigid foam and a cable tie. I doubt however that it is ever really necessary? Maybe in a tank or a fighter?

Dave.

[ef37a]

On the matter of holding components in place on a PCB? This can be done using pre formed 'fixtures' of silicone rubber material on the component side.

These do of course cost more and take more time to design than a random squirting of ***t over the board!

Dave.

[nathanscribe]

On the matter of holding components in place on a PCB? This can be done using pre formed 'fixtures' of silicone rubber material on the component side.

These do of course cost more and take more time to design than a random squirting of ***t over the board!

Dave.

Thanks Dave, that's a neat idea. As you suggest though, I think for the time being I'll just not bother – am starting to think of this as an interim fix while I ponder an upgrade.

[ef37a]

On the matter of holding components in place on a PCB? This can be done using pre formed 'fixtures' of silicone rubber material on the component side.

These do of course cost more and take more time to design than a random squirting of ***t over the board!

Dave.

Thanks Dave, that's a neat idea. As you suggest though, I think for the time being I'll just not bother – am starting to think of this as an interim fix while I ponder an upgrade.

No no Nat! I am not suggesting YOU do it. Just that it is a better way to secure components whilst they are wave soldered.
You can just bend the pins, solder then crop.

Dave.

[nathanscribe]

Just a little update on the monitors – I managed to fix one of them, the goop had eaten away all but one strand of wire on one of the hookups to the tweeter, jumpers were mostly not terrible once cleaned up, and I had to replace a few resistors and caps, some of which just fell apart while handling. The Goop of Doom is indeed a nightmare.

The other monitor is, pragmatically, a goner. The goop has caused more wreckage in there than I can be bothered to undo, and looks to have been applied by someone wearing a blindfold, using a bucket, across a room.

Anyway, the saved one will have a new life on my test bench, and I'm mixing on headphones till I can afford a decent pair of new monitors.

Thanks folks!

[ef37a]

Just a little update on the monitors – I managed to fix one of them, the goop had eaten away all but one strand of wire on one of the hookups to the tweeter, jumpers were mostly not terrible once cleaned up, and I had to replace a few resistors and caps, some of which just fell apart while handling. The Goop of Doom is indeed a nightmare.

The other monitor is, pragmatically, a goner. The goop has caused more wreckage in there than I can be bothered to undo, and looks to have been applied by someone wearing a blindfold, using a bucket, across a room.

Anyway, the saved one will have a new life on my test bench, and I'm mixing on headphones till I can afford a decent pair of new monitors.

Thanks folks!

Well done to rescue something. Despite certain adverse comments about then in this forum, the Presonus Eris 3.5s are pretty good. "Not monitors" Hugh would say but about the best you will get for a ton. My son has been messing with music for a big chunk of his 52 years and is pretty bloody fussy about his classical guitar sound and he rates his pair.

Dave.