Back in the mid 90's I built a pair of speakers along the lines of the Pro-ac Supertower. I purchased all Scan Speak and Dynaudio tweeters drivers from Simply Speakers in the US. The Scan Speak units were all made in Denmark.
The 18W speakers were refoamed after about 7 years and after another 6 years or so they were showing signs of needing a repeat operation. I was fairly stretched financially at the time so I didn't do the refoam - I simply left the speakers sit idle in a spare room and used another pair of speakers.
This week I decided I should do something about getting the18W/8542's back in operation so I stripped them from the cabinets in preparation. I was surprised to find that the first speaker I started work on had absolutely no voice coil travel whatsoever - it was firmly stuck. Inspection of the magnets and rear pole piece plate strongly indicated that the magnet and pole piece had shifted considerably and thus trapped one side of the voice coil in the air gap. The second speaker also exhibited scraping when the cone was depressed and the same indication of magnet/pole piece movement was evident. Inspection of the other two speakers indicated that one of them was also exhibiting signs of magnet/pole piece shifting.
The speaker with the trapped voice coil was the obvious candidate to dismantle first so I started out on that. Whenever I have tried to remove ceramic magnets from modern drivers with burnt out voice coils I have found it extremely difficult if not impossible to separate the magnet and pole piece but on this Scan Speak driver I found it very easy to get the blade of my pocket knife between the magnet and pole piece plate using nothing more than a gentle tap on the back of the blade with a plastic screwdriver handle. By gently levering up the rear plate and inserting plastic spacers all round and progressively increasing the size of the spacers I was able to reduce the magnetic attraction to a point where I could lift the pole piece fully out. I then found that it was just as easy to separate the magnet from the front plate so that I could eventually release the voice coil. Sadly the voice coil and former had been damaged to some extent such that I would not be confident the driver would perform optimally even if I could get everything back in place.So I decided that this speaker was a write-off.
Now why would the adhesive used to fix the magnet to the pole piece and front plate just fail? A close inspection of the adhgesive on the areas on the magnet and metal plates showed it was extremely thin and not an epoxy or similar strong adhesive. Firstly the areas covered by the adhesive were no more than 30% of the total area on both faces of the magnet. When the adhesive was scraped with the blade of my knife it came of off easily and was fairly hard and brittle and left an essentially clean surface when removed. The adhesive was clear and had a similar appearance to dried cyanoacrylate (superglue) which I thought was very odd. Whatever the adhesive was it was NOT suitable for holding a strong ceramic magnet and iron pole pieces together in speaker drivers .
I then attacked the second speaker and found exactly the same problem. Fortunately, I was able to rescue the voice coil undamaged on this unit so I am hopeful that I can restore it to full working order using a more suitable adhesive - possibly Araldite or a polyurethane type. I suspect that I will also have to perform a similar operation on the remaining drivers since they were all from the same batch.
I would never have suspected that Scan Speak would use anything but proven techniques and top grade materials to construct their speakers since they are relatively expensive. I think the current 18W/8542-10 sells for something like AUD319 each plus freight.
Has anybody experienced a similar failure with Scan Speak or other supposedly reputable speakers?
The 18W speakers were refoamed after about 7 years and after another 6 years or so they were showing signs of needing a repeat operation. I was fairly stretched financially at the time so I didn't do the refoam - I simply left the speakers sit idle in a spare room and used another pair of speakers.
This week I decided I should do something about getting the18W/8542's back in operation so I stripped them from the cabinets in preparation. I was surprised to find that the first speaker I started work on had absolutely no voice coil travel whatsoever - it was firmly stuck. Inspection of the magnets and rear pole piece plate strongly indicated that the magnet and pole piece had shifted considerably and thus trapped one side of the voice coil in the air gap. The second speaker also exhibited scraping when the cone was depressed and the same indication of magnet/pole piece movement was evident. Inspection of the other two speakers indicated that one of them was also exhibiting signs of magnet/pole piece shifting.
The speaker with the trapped voice coil was the obvious candidate to dismantle first so I started out on that. Whenever I have tried to remove ceramic magnets from modern drivers with burnt out voice coils I have found it extremely difficult if not impossible to separate the magnet and pole piece but on this Scan Speak driver I found it very easy to get the blade of my pocket knife between the magnet and pole piece plate using nothing more than a gentle tap on the back of the blade with a plastic screwdriver handle. By gently levering up the rear plate and inserting plastic spacers all round and progressively increasing the size of the spacers I was able to reduce the magnetic attraction to a point where I could lift the pole piece fully out. I then found that it was just as easy to separate the magnet from the front plate so that I could eventually release the voice coil. Sadly the voice coil and former had been damaged to some extent such that I would not be confident the driver would perform optimally even if I could get everything back in place.So I decided that this speaker was a write-off.
Now why would the adhesive used to fix the magnet to the pole piece and front plate just fail? A close inspection of the adhgesive on the areas on the magnet and metal plates showed it was extremely thin and not an epoxy or similar strong adhesive. Firstly the areas covered by the adhesive were no more than 30% of the total area on both faces of the magnet. When the adhesive was scraped with the blade of my knife it came of off easily and was fairly hard and brittle and left an essentially clean surface when removed. The adhesive was clear and had a similar appearance to dried cyanoacrylate (superglue) which I thought was very odd. Whatever the adhesive was it was NOT suitable for holding a strong ceramic magnet and iron pole pieces together in speaker drivers .
I then attacked the second speaker and found exactly the same problem. Fortunately, I was able to rescue the voice coil undamaged on this unit so I am hopeful that I can restore it to full working order using a more suitable adhesive - possibly Araldite or a polyurethane type. I suspect that I will also have to perform a similar operation on the remaining drivers since they were all from the same batch.
I would never have suspected that Scan Speak would use anything but proven techniques and top grade materials to construct their speakers since they are relatively expensive. I think the current 18W/8542-10 sells for something like AUD319 each plus freight.
Has anybody experienced a similar failure with Scan Speak or other supposedly reputable speakers?
