Kit to re-magnetise speaker drivers

[YNWaN]

As I understand it, the capacity/strength of the magnet in a given driver is determined by the chemical composition, size and design of the magnet assembly. Any re-magnetising equipment normally vastly over 'magnetises' (swamps) the material which can only magnetise up to its original specification. You cant overdo it.

Ah, yes, that makes sense.

[Edward]

What are the benefits of remag'ing a speaker?

§

[Barry]

What you need to know is the stated gap flux of the speaker magnet you wish to re-magnetise. From that, you can calculate the total stored energy of the magnet, and this will give you some idea of the energy requirement of the re-magnetiser.

The device shown is a pulsed energy device having a maximum pulse energy of 10kJ (the manufacturer quotes this as 10,000Ws). This ought to be sufficient for small magnets, but I'm not sure if it is enough for large speaker magnets - but I'm no expert.

Unfortunately I can't open up and read the link cited in post #9, so cannot comment on the stability/longevity of Alnico magnets. Whilst Alnico magnets are more prone to demagnetisation than the newer rare-earth types, from my limited reading of the subject, AlNiCo(5) magnets only 'loose' < 3% of their field strength after 100,000 hours (approx. 11.5 years), dependant on operating conditions. So the condition of a 70 year-old Alnico speaker magnet is debatable.

[Edward]

Unfortunately I can't open up and read the link cited in post #9, so cannot comment on the stability/longevity of Alnico magnets.

I've downloaded the relevant page (it is on the internet archive) to a pdf and placed it here.

[Barry]

Thanks Edward

Unfortunately there is a screen navigation pop-up at the top of each page, covering an important part of the text which discusses both the contribution to the demagnetisation by an audio signal, and the re-magnetising capability of the pulse magnetiser. But what I can read supports what I understand of the problem (though I hadn't realised that older speakers use a much larger diameter voice coil than modern speaker designs). The typical gap flux of a modern speaker is ~ 12,000 Oe, so the quoted magnetising field of 20,000 Oe ought to be sufficient to provide saturated re-magnetisation.

[Edward]

Thanks Edward

Unfortunately there is a screen navigation pop-up at the top of each page, covering an important part of the text which discusses both the contribution to the demagnetisation by an audio signal, and the re-magnetising capability of the pulse magnetiser. But what I can read supports what I understand of the problem (though I hadn't realised that older speakers use a much larger diameter voice coil than modern speaker designs). The typical gap flux of a modern speaker is ~ 12,000 Oe, so the quoted magnetising field of 20,000 Oe ought to be sufficient to provide saturated re-magnetisation.

Oh I see what you mean. Pasted below is the text. The relevant website no longer exists and the business closed down so I'm claiming 'fair usage rights'.

-----------

There are five primary mechanisms working in concert to demagnetize Alnico magnets - fortunately only a couple of them are likely to affect loudspeakers...


1. Age:


The flux in all magnets diminishes slowly over time, due to thermal fluctuation of the magnetic moment. All grades (3, 5, 5G, 7, & 8) of Alnico suffers from this problem more than ceramic ferrite magnets, and what we call "permanent magnets" are not quite permanent after all. The lower the permanence coefficient of the magnetic circuit, the faster the magnetic domains return to their random equilibrium. In typical Alnico 5 or 5DG magnets, the flux drops very slowly, at 2~3% per decade, for broadband sensitivity loss of only 1dB over 40-years, although losses at high frequencies may be more severe. Sumitomo Special Metals Co, of Osaka, Japan, have documented this phenomenon in considerable technical detail, but the research is not available on-line.


2. Thermal losses:


There are two distinct forms of temperature related flux loss: reversible, where the flux recovers to its original level as the temperature returns to normal; and irreversible, where the flux never recovers. Good thermal design in the magnetic structure minimizes reversible losses, and (mercifully) the temperatures that endanger alnico alloys, the Curie point, are far higher than those observed in loudspeakers, at 520 degrees Celcius. In loudspeakers, thermal losses in magnetic energy can be neglected.


3. Pulse loss:


Alnico speakers are intrinsically susceptible to modulation of the flux by the AC field from the voice coil. When operating normally, this momentary (AC) flux is superimposed over the permanent (DC) magnetic field. During normal listening, there the magnetic flux is dynamically modulated, causing third harmonic distortion. This non-linearity is a major contributor to the 'Alnico Sound', and is often welcomed by musicians and Alnico speaker lovers. Unfortunately, as the input power rises, the strength of the modulating field may exceed the intrinsic coercivity of the Alnico, partially degaussing the magnet. A single high power 'pop' can reduce sensititivity by 2dB, and repeated pulses, such as loud cracks from a faulty connection will cause additional (smaller) losses. Alnico drivers become noticeably quieter, especially at the top of their range. A speakers' sensitivity to discharge is proportional to the aspect ratio of the Alnico slug, the length of voice coil winding, and amount of available amplifier power. Virtually all Alnico woofers that have been enjoyed loudly show some flux loss, and those with short slugs and long coils suffer most. Most such designs date from the 1950s and 60s, when 35-watts was considered high-power, and amplifiers lacked the ability to endanger magnets. The problem effects virtually all Alnico drivers that have been used with modern, high power amplifiers.


For example: Low frequency drivers with 4-inch voice coils (JBL D120, D130, D140, LE15, 2205, 2215, 2231, Gauss, or TAD drivers) can lose 2~3dB, becoming 'dull' or 'slow' sounding, unless they have been grossly abused, in which case they more and sound dreadful. Drivers with short Alnico slugs and 3-inch voice coils, (JBL 2213, 123A, etc) are effected more severely, so typically show 3dB sensitivity loss, while small alnico woofers (JBL 116A, 125A, 127A etc) are commonly 4dB down after quite modest use. Conversely, speakers with higher coercivity ( taller Alnico slugs) and short voice coils, such as Tannoy dual concentrics, Altec and EV woofers, rarely suffer unless they have been badly abused. Users report that Altec 416, 515, and 604 drivers become 'mushy' after heavy use, but most of their loss of definition is due to fatigued cones. Remagnetization can restore the magnet, but cannot address cone fatigue.


Most Alnico compression drivers (JBL LE85, LE175, 375, 2410, 2420, 2440, 2441, Altec 288, 290, 802, 804, 806, 808, Emilar EA-175, TAD 2001, 4001, Goto Unit, etc) have tall magnets, short voice coils, and modest power inputs, so do not suffer degaussing. Those compression drivers with low aspect ratio magnets, especially if used <500Hz (Vitavox S2, University ID series, Klipsch K-55-V, Electrovoice 1823M etc) can suffer flux loss, and will benefit from remagnetization.


4. Shock:


Knocking or hammering on an Alnico magnet causes a sudden, dramatic reduction in magnetic energy as the shockwave knocks the closely aligned magnetic domains from their parallel pattern into a more random pattern. Shocks during shipping (from accidental drops) can cause a reduction in sensitivity. Mercifully, beating expensive loudspeaker magnets with hammers is not a common pastime. The solution is remagnetization, and, providing nothing else was broken, original performance is restored.


5. Disassembly:


If an Alnico magnet structure is opened, such as to re-centre a shifted pole-piece, repair a broken throat, or remove a bug-screen, the magnetic field collapses instantly. While this type of damage is not common, ill-advised repairs create very unwelcome problems. If a compression driver is dropped, aside from the possible 'shock' loss (see above) the magnet may shift inside the structure, breaking the throat molding. In these circumstances, the output may drop 20dB, as the magnet is seriously discharged. If there are other issues such as shifted pole, broken throat, corrosion, or loose internal parts, we can discharge the magnet, repair the problem, then remagnetize the driver, back to factory spec.


THE SOLUTION


Recharging an Alnico magnet can be performed without driver disassembly, or leaving any marks. Vancouver Audio Speaker Clinic use an industrial capacitor-discharge magnetizer that can deliver 440 volts at 13,000 amps into the water cooled charging-fixture that completely surrounds the magnet. The 10ms pulse generates over 20,000 Oersteds, completely saturating any magnet structure, returning it to its absolute maximum flux. In refining our recharging process (using our own drivers) we measured (before and after) flux density, sensitivity and frequency response.






We charge a flat rate fee for recharging Alnico drivers with magnet structures up to 190mm (7.5") diameter. If there is no cone or diaphragm in the driver (allowing access to the gap) we measure and record the flux density before and after recharging. Our probe is 0.85mm (0.034") thick, so fits all but the narrowest gaps. Narow gap drivers incl; JBL 'bullet' variants (075, 076, 077, 2402, 2403, 2404 and 2405, and Beyma tweeters. We can remagnetize these drivers, but measurement is not practical. We recharge woofers, midranges extended-ranges and full-range units, incl: Lowther, Fostex, and Celestion guitar speakers. There are some drivers that rarely need recharging, except if they are accidentally damaged, incl: JBL LE85, LE175, 2410, 2420, 2421, 2425, 2460, 2461, 2470, 375, 376, 2440 & 2441, Altec 288, 290 & 291, 802, 804, 806 & 808 as well as Emilar, Renkus Heinz, Coral, Goto Unit, B&C;, RCF and Yamaha models.


The good news for terminally careless audiophile is that there is no limit to how many times an Alnico magnet can be recharged. There is an additional (modest) labour charge if the magnet also needs to be recentred and/or reglued.