advertisement


Kef KM1 in need of repair.

Success.

I've put 20Hz tone through it, at +/- about 6mm excursion and nothing's hitting anything.

Looking down from the top, into the voice-coil / pole-piece gap, it looks central.

I think I've done what I set out to do.


So, now the stress level of this project is greatly, greatly reduced.


I still need to:

Add ferrofluid to this driver

Fit 4 x midrange dust-caps and then subsequently brush them over with PVA to make them impermeable.

Flip all of the bass drivers 180' (one has slightly sagged and is just rubbing the coil off the magnet, but rotating it will fix that, I'm sure)

Fit 2 x bass driver dust-caps.

Get all of the bits (bass enclosures and chassis) over from the storage place to work, assemble and make a lot of noise doing proper testing.


None of the remaining tasks are difficult or high-risk. There's a bit of logistics and carrying heavy things around, but nothing that's actually technically demanding. Whilst there's still cosmetic work to be done, and of course high-level testing might reveal more problems, I think I'm rapidly approaching the 'fixed' stage.

I'm relieved greatly by this. This pair of speakers have cost me, so far, more than the car I bought to drive to work in a couple of weeks ago. But there's more to it than that... I've been after a pair of these for a very long time, and I don't imagine the opportunity to own a pair of KM1s will come my way again.

I'll update in due course with my progress.
 
This has been a magnificent effort on your part.

Legendary speakers, and I cannot imagine the sheer scale of sound they produce.

I still can't understand why this thread is languishing in the eBay section! It should be moved into Classic, surely?
 
Ferrofluid time for that midrange that I had apart last week.

https://imgur.com/a/OBrLfLr

The bottle is half-full (from the supplier). That's what £270 looks like. Ouch.

I've installed it in the midrange. I used a graduated 1cc syringe with a long needle. I'd estimate that I used somewhere between half and three quarters of a millilitre. The gap had been thoroughly cleaned before, when I had the cone out, so I'd call it 'filling from empty'.

This is what it looks like now:

https://imgur.com/9SFkjMf
 
So, and this is not entirely unexpected, I think I've got to take the other mids apart and re-do their ferrofluid.

The one I've just done has very 'free' movement (it feels as I think it should), much like a normal, modern midrange driver.

The other mids, when you manipulate them, take a noticeably longer time to spring back to their centre position. I've poked a piece of card into the gap on one of them, and my 'sample' of ferrofluid I've extracted feels sticky and viscous.

This isn't entirely unexpected. I'm aware of the problem of ferrofluid drying out and these drivers are somewhere between thirty-five and forty years old.

I'd initially planned to send all the mids and both tweeters off to Wembley Loudspeakers for disassembly and new ferrofluid.

I've bought enough ferrofluid to do multiple pairs of KM1s, so I'm set up to do it.

I need to pluck up the courage to push my luck and pull the other three mids out and take them apart. The process isn't particularly tricky, having done it once.

Time is something I'm a bit short of at the moment, working five days a week at my main job and then weekends on other (very not broadcast or hifi related) stuff for other people.

Perhaps I can do one next Friday evening (disassemble, clean and glue) and then add ferrofluid the next Friday evening, once the glue is 100% dry and I'm confident everything's aligned. Assuming it is, I can pull another mid apart and then repeat the cycle, so working through effectively one driver per week.

I haven't looked at the tweeters yet. Getting them out without chewing up the paintwork on the front baffle is not a job I'm going to rush. That can wait until the mids are done.

Progress is happening...
 
It’s great.

What do you know about selection of ferrofluid? Anywhere you can gain advice on the type for different tweeters?
 
Eoin said:
It’s great.

What do you know about selection of ferrofluid? Anywhere you can gain advice on the type for different tweeters?
Click to expand...
Only a little bit.

Ferrofluid, in loudspeaker applications, has two functions (arguably three but we'll get to that).

Principally, it's a coolant, used to conduct heat energy away from the voice coil windings and into the magnet structure where it then dissipates to the air inside the loudspeaker enclosure.

Secondarily, it's mechanical damping. It's there to slow the movement of the cone / dome to reduce it overshooting and oscillating.

Thirdly, and this is a 'maybe' it also serves to alter the magnetic field density in the gap - think of it as bringing the magnet closer (in contact) to the voice coil.


The main functional difference between different types is the viscosity. In my case, KEF used it just as a cooling measure and the stuff specified for those drivers (and all KEF ferrofluid drivers of that era) is very thin and runny - when you shake the bottle that I've got it seems like water in its viscosity (although the measured viscosity is higher).

Drivers where it's mechanical damping as well as cooling would have had a thicker, more viscous ferrofluid.

There are also specifications for its magnetic performance, but I'm hazy as to how much that's relevant for loudspeaker applications.

In my case, I have been incredibly lucky to have tracked down one of the two development engineers at KEF who were responsible for the design of the KM1 drivers. He has a very good memory for the technical details (and has been very helpful and tolerant of my unceasing barrage of questions) and has told me the (now-obsolete) ferrofluid type that KEF bought from Ferrotec during that period. Ferrotec are quite helpful about which type is the current equivalent in terms of its mechanical properties so I've pretty much replicated what KEF had.

As far as 'what type of ferrofluid do I buy for XYZ model of loudspeakers?' - no idea. There are some vendors who sell a 'universal, for all loudspeakers' ferrofluid, but it clearly isn't suitable for all different types.

If it's for a pair of KEFs I can ask the KEF engineer I've been talking with and see if he can tell me. If it's for another brand, then you're probably dependent on what you can get them to tell you.
 
Last edited:
Time for this week's update.

I've done another midrange driver (well, partly).

So, Friday evening at work - everything's done, it's after hours, everyone else has gone home. Time to re-ferrofluid another midrange.

I examined its movement prior to disassembly, and noticed a clicking sound when manipulating the cone.
Close inspection revealed that, just like the first one, the glued join between the voice-coil former and the cone was starting to fail. A little more poking and prodding and the two separated.

https://imgur.com/a/0pwKfJS

OK. Having closely watched when the first driver got fixed, I now know how to go about it, and happened to have the correct solvents and adhesives to hand. On the one hand, I'd have preferred it if the thing hadn't fallen apart. On the other, if it's going to break, now is by far the best time for it to do so. Fine... I'll fix it.

So, I made up some plastic shims to hold the coil former central in the magnet gap. I cleaned the mating surfaces of the cone's neck and the coil former with isopropanol and applied the primer (heptane) for the adhesive. There's a 5 to 8 minute wait for the primer to do its thing, and then you assemble the two parts and apply a little cyanoacrylate adhesive by means of a syringe and needle. Simples...

And this is where things went wrong. I'm adding micrograms of adhesive at a time, watching it wick into the joint. That's nice.
And then I spotted that not all of it was going where I wanted. The edge of one of the coil's shims was in contact with the inside of the top of the voice-coil former, and I watched in horror as a drop of cyanoacrylate ran down the shim and into the space where the voice-coil former entered the magnet's gap. Disaster - I've just superglued the coil former to the magnet. 'Oh dear, I seem to have messed this up a bit' I said to myself (well, not those words exactly, but you get the picture).

Hmmm... I have very little time do something about this (seconds). OK, I've got another syringe and a bottle of methyl ethyl ketone (death-strength solvent) to hand. If I chuck a millilitre of it into the gluey mess I've created, perhaps it'll stop it from bonding all of the moving parts together. No better idea came to mind, so I did. It appears this was my master-stroke of luck.

In record time, I freed the front surround and rear suspension from the chassis of the loudspeaker and wiggled the cone and coil assembly free, with no obvious damage. A moment of great relief, one might say.

Of course, now I need to clean out the drying cyanoacrylate (and old congealed ferrofluid) off everything.

Although I was unaware of this, it seems that applying methyl ethyl ketone to superglue (whilst using language that would kill a priest) turns it from rock-hard indestructible nastiness into a gooey rubber-like substance that peels cleanly off. It was at this moment that my train of thought went from 'I've just wrecked this' to 'I can fix this', and my pulse rate and blood pressure started to return to normal.

Twenty minutes of q-tip and MEK action cleaned the voice coil and former to perfection, and some MEK and strips of cardboard seem to have cleared the magnet gap completely based on inspection with a microscope.

https://imgur.com/gw0vhK7
https://imgur.com/mRyzIWh

Things I learned in the process:

Cyanoacrylate dissolves in MEK.
Razor blades will cut through the adhesive bond of the front surround just as well as they will cut a chunk out of your thumb.
Methyl ethyl ketone in a cut hurts a lot more than isopropanol does.
You can clear the magnet gap of solvent and old ferrofluid using compressed air remarkably quickly.
If you clear the magnet gap with compressed air, everything in a ten foot radius gets sprayed with ferrofluid.
Ferrofluid is very hard to remove from walls, floor coverings, computer keyboards, toolboxes and skin.

The driver's cone and coil are attached. They're not back in the chassis of the speaker yet, but I see no reason why they won't reassemble without difficulty. The alignment and bonding of the voice-coil former and cone looks good.

Of course, now we're two-for-two on the voice-coil formers detaching from the cone neck, I expect the other two mids to go the same way. If they don't immediately fall apart, I'll gently torture them until they do and then re-glue, but hopefully more straightforwardly than this one...

Progress....
 
Last edited:
Oh my god it sounds like a rollercoaster ride. I don't think I could have coped with all these things going wrong. In the past when something like this has happened to me I've lost patience with myself and bodged something, soon to my regret.

I hope it all goes to plan from here on and you get the speakers you deserve, and the audiophile community has gained a pair of classics from the dead.
 
lawrence001 said:
Oh my god it sounds like a rollercoaster ride. I don't think I could have coped with all these things going wrong. In the past when something like this has happened to me I've lost patience with myself and bodged something, soon to my regret.

I hope it all goes to plan from here on and you get the speakers you deserve, and the audiophile community has gained a pair of classics from the dead.
Click to expand...

I'm making steady progress, although perhaps not as quick as I would have liked. Nonetheless, I'm still on target for getting these working.

There have been ups and downs. There is still considerable risk in what remains for me to do. One real mistake and it's all over. It's been an interesting learning experience.

It'd be no fun if it was easy...
 
Tweeter time.

They're both out, and I've made little crash helmets for the domes to protect them temporarily:

https://imgur.com/B1Ota3w

https://imgur.com/hsGf1uJ

Big enough magnets... the ruler in the first picture is a 30cm one.

Tomorrow they're going to get a ferrofluid top-up. I've been strongly advised (by the man who designed them) not to attempt disassembly. The ferrofluid can apparently be topped up with a long syringe needle through the cable exit holes on the back of the magnet.

I've also pulled the other two midrange drivers out of their enclosure, but they're a job for another day.

Interestingly, the first two mids had been re-coned in 1986. The second two had been done in 1991 and 1995 according to their stickers.
 
Sorry for the tangent, but just out of curiosity - is there an application for ferrofluid outside the field of loudspeakers?

(no pun intended)
 
Yank said:
Sorry for the tangent, but just out of curiosity - is there an application for ferrofluid outside the field of loudspeakers?

(no pun intended)
Click to expand...
Yes, more than one.

I'm led to believe that NASA had some involvement in its development to pump fuel around in zero gravity conditions.

Also, you can use it in place of an O-ring seal on a rotating shaft. There's a maximum upper speed that rubber seals will survive due to frictional heating and wear on the contact surface. As an alternative, a slug of ferrofluid, held in place with magnets will survive much higher speeds. Although you can only get a few PSI of pressure resistance from a single ferrofluid seal, it's possible to stack seals one after another to get a higher total pressure rating.

Hard drives apparently use ferrofluid as sealing around the motor spindle to keep debris out of the enclosure that holds the platters.
 
robbyd said:
Think a lot of people will want to visit you and hear when these are done!
Click to expand...
I wouldn't necessarily be resistant to that. I need to sell my current house first and move, as I've got nowhere with enough space for the KM1s in the present place. It's been on the market for a few months now, and I'm (really, really) hoping I get a buyer for it soon. Leaving aside the loudspeaker issue, the 100 miles a day commute for work in a 2005 Ford Focus is getting a bit tiresome.
 
Currently commuting 120/day and no fun in a car that does 23/gal! Well, actually it is fun!!

Heard these at Heathrow I'd say 1980 because everyone was playing 'Mountain Dance' Dave Grusin...

Would be interesting to hear how they stack up against what we're used to now...

Best regards
 
You must log in or register to reply here.


advertisement


Back
Top