Should you buy a commercial speaker with a ferrofluid tweeter or not? I read lots of cases where the ferrofluid corrupts with age and I don't fancy taking on the problem of cleaning and replacing ferrofluid in a tweeter.
I have a pair of early Wharfedale speakers Program 30D-6 and I have no idea if the tweeter has ferrofluid in it or not - can't find any info. I was thinking of buying a later pair, like the Diamond 9.1, but the tweeters do have ferrofluid in them and I'm concerned that this will degrade with age.
Given that ferrofluid deteriorates over time, should one simply stay well away from tweeters containing it?
I have a pair of early Wharfedale speakers Program 30D-6 and I have no idea if the tweeter has ferrofluid in it or not - can't find any info. I was thinking of buying a later pair, like the Diamond 9.1, but the tweeters do have ferrofluid in them and I'm concerned that this will degrade with age.
Given that ferrofluid deteriorates over time, should one simply stay well away from tweeters containing it?
Early ferrofluid lasted 15-20 years, the stuff we have had the last 45 years lasts at least 30 years(note: I do not live sans a/c in a hot stinking desert).
I would say watch out for non-maintainable components, so a tweeter should be screwed or clipped together, not glued.
If surrounds, caps, and fluid require a little love every thirty years or so, that's better than any vehicle or other electronic device you will ever own.
Some dimple-dome tweeters perform very well for the money, and giving up servicability could be an informed option.
I would say watch out for non-maintainable components, so a tweeter should be screwed or clipped together, not glued.
If surrounds, caps, and fluid require a little love every thirty years or so, that's better than any vehicle or other electronic device you will ever own.
Some dimple-dome tweeters perform very well for the money, and giving up servicability could be an informed option.
Most of the points in this speech are true, except for the generalized statements about Ferrofluid linearity: Ferrofluid is a newton liquid, which has a viscous counter force proportional to speed. This is by definition linear behavior. The nonlinear coefficients that ferrofluid contributes to driver non-linearities are far lower than those from other mechanical components and magnetic design.
Beside that, nowaday's ferrofluids are not the same as those from 1970 or 1980. They improved a lot. The premium series from Ferrotec has >10× the lifetime of older series. In a lot of speakers the end of life is not determied by ferrofluid, but from other components. On the bottom line, Ferrofluid is like any other compnent in a speaker an extension in degrees of freedom of trading parameters against each others. For example: Ferrofluid reduces power compression, but there are also other measures to reduce power compression and face you with other drawbacks.
Beside that, nowaday's ferrofluids are not the same as those from 1970 or 1980. They improved a lot. The premium series from Ferrotec has >10× the lifetime of older series. In a lot of speakers the end of life is not determied by ferrofluid, but from other components. On the bottom line, Ferrofluid is like any other compnent in a speaker an extension in degrees of freedom of trading parameters against each others. For example: Ferrofluid reduces power compression, but there are also other measures to reduce power compression and face you with other drawbacks.
"generalized" perhaps - but the strain on the fluid in the gap (relating to obstructions and turbulence - it's not a laminar flow) alters deformation. Notably has he's mentioned in the video it can even move out of the gap.
Of course the most notable change (subjectively) isn't viscous damping of linear motion, rather it's the damping of the VC itself not it's linear excursion/pumping.
Of course the most notable change (subjectively) isn't viscous damping of linear motion, rather it's the damping of the VC itself not it's linear excursion/pumping.
Last edited:
I've got a pair of B&W P series speakers from the 90s with FF tweeters and had convinced myself (paranoia) I needed to check on the ff. I had the service diagrams and with some careful disassembly I was able to check. They were fine. I got the tweeters back together too
So deterioration is not a given.
The flow in the ferrofluid is certainly never turbulent. I can only refer to Reynolds' stability criterion. Read it, make a calculation and you will quickly find that the flow in a VC gap is very distant from exceeding stability criterion for a liquid. You may dig a little deeper into flow modelling science, and you will find that the magnetic properties of the fluid cause the limit to shift up by an order of magnitude. It is flow stabilizing. So in reality it is absolutely impossible to exceed Reynold's stability criterion and turn the flow in a VC away from laminar. If fluid moves permanently out of the gap this is just outside the designated operating area, that means either overload condition or badly engineered driver. In both cases origin of nonlinearities cannot be blamed to the ferrofluid.
For used drivers its a concern I have 7x M200 compression drivers with ferrofluid and trying to work out if differences between them are due to ferrofluid or in production design changes is annoying. The best I can do is buy more drivers than I need and select on performance characteristics: https://www.diyaudio.com/community/...per-scripts-and-m200-fun.385493/#post-7003214
I don't think it would be a concern if buying new drivers due to the long lifespan.
I don't think it would be a concern if buying new drivers due to the long lifespan.
My search involves less time:
Look up "ferrofluid turbulence".
(..and while most examples concern stable magnetic fields, note that in the case of a driver the field is NOT stable).
Yes, less time maybe, but i don't know which of the 70000 results that google spits out i could refer to? Ans i am not sure what you mean by "stable" magnetic field in a driver. It does not reverse or flip, the direction and the magnitude do not change, there is no rotation or something like that. So by definition it is stable.
..and perhaps none of the 70000 results that google can spit out that you could refer to have any relevance with your statement.
The magnetic field in the gap, particularly as it's leaving the gap isn't uniform (force isn't linear): notably the field itself "expands and contracts" due to stress from "back" emf (and most particularly as you extend further from the gap's center assuming a fairly typical "bell" shaped force curve).
Last edited:
Ok, got it. Of course ferrofluid can have turbulent flow. But my statement was particularly about fluid in a VC gap. And as i said: take the shear velocity, the material viscosity, the gap width and calculate the Reynolds number, and you will see stable laminar conditions! Turbulent flow in a VC is just one of these rumors, which are recited endlessly. I have never seen a calculation or measurement that proves this claim...and perhaps none of the 70000 results that google can spit out that you could refer to have any relevance with your statement.
The magnetic field in the gap, particularly as it's leaving the gap isn't uniform (force isn't linear): notably the field itself "expands and contracts" due to stress from "back" emf (and most particularly as you extend further from the gap's center assuming a fairly typical "bell" shaped force curve).
The point I was trying to make is that I don't think it's that simple: calculating the Reynolds number (under its standard equation) isn't going to work IMO - it's to "general".
Just reading over Reynolds - it doesn't seem to really apply: it's looking at a condition from a laminar flow to a turbulent flow, sort of where a constant "stream" of flow becomes turbulent. Here we have a VC (tweeter) that's moving back and forth 20,000+ a second, it's not just moving forward (or only rearward) continuously.
Moreover, even if all the other attributes of ferrofluid in a non-stable magnetic field were eliminated and you had an effective monolithic "plug" (high viscosity) with zero friction moving in and out within the gap - at some point because it's a liquid the "ends" of the "plug" are "sloshing" back and forth - and at those "ends" your not-real monolithic plug is generating turbulence (which I'm reasonably certain is effecting at least some *distance further in toward the center of that "plug").
*and considering the average length of a tweeter's VC, that's not much distance at all.
Just reading over Reynolds - it doesn't seem to really apply: it's looking at a condition from a laminar flow to a turbulent flow, sort of where a constant "stream" of flow becomes turbulent. Here we have a VC (tweeter) that's moving back and forth 20,000+ a second, it's not just moving forward (or only rearward) continuously.
Moreover, even if all the other attributes of ferrofluid in a non-stable magnetic field were eliminated and you had an effective monolithic "plug" (high viscosity) with zero friction moving in and out within the gap - at some point because it's a liquid the "ends" of the "plug" are "sloshing" back and forth - and at those "ends" your not-real monolithic plug is generating turbulence (which I'm reasonably certain is effecting at least some *distance further in toward the center of that "plug").
*and considering the average length of a tweeter's VC, that's not much distance at all.
Opinions are not argumentative. The Reynolds model is not only applicable to round pipes: It has a factor Dh, hydraulic diamater, which is nothing but the fluid cross section divided by the contact surface, in our case the surface of the pole plate. By this, the model is applicable to (almost) any geometry. And exactly this Dh is very small in a speaker: The cross section is small over the surface, which results in a small Reynolds number. Even if you were right and the math model was not applicable or not complete enough: Where is a plausible argument proof for turbulence like a measurement, cam observation…. If you claim it you got to proove it.
And you mix up unrelated things: in any driver the diaphragm speed, as well as the excursion decrease with frequency. The velocity has always a maximum at the system resonance. In your example at 20kHz the excursion of a tweeter gets very, very small. It's in the scale of micrometers. So, with small excursion the argument of whatever "leaving the gap" is obviously not applicable. Even if ferrofluid moves partly out of the gap at long stroke systems with large excursion, this has absolutely nothing to do with turbulences. Unlikely to see turbulences here, because the velocity maximum is at zero crossing of the VC position, not at it's reversal points! If you look to the analysis of Richardson/Kolmogorov, you see that built up of turbulences also takes a certain travel path and time, which are cleary given in a simple continous pipe flow, but hardly in a speaker vc. Consequently, the condition for turbulence flow of a ferrofluid in a speaker is never met.
And you mix up unrelated things: in any driver the diaphragm speed, as well as the excursion decrease with frequency. The velocity has always a maximum at the system resonance. In your example at 20kHz the excursion of a tweeter gets very, very small. It's in the scale of micrometers. So, with small excursion the argument of whatever "leaving the gap" is obviously not applicable. Even if ferrofluid moves partly out of the gap at long stroke systems with large excursion, this has absolutely nothing to do with turbulences. Unlikely to see turbulences here, because the velocity maximum is at zero crossing of the VC position, not at it's reversal points! If you look to the analysis of Richardson/Kolmogorov, you see that built up of turbulences also takes a certain travel path and time, which are cleary given in a simple continous pipe flow, but hardly in a speaker vc. Consequently, the condition for turbulence flow of a ferrofluid in a speaker is never met.
Opinions are of course often argumentative; perhaps you intended "facts"?
I'm not saying that Reynolds isn't truly applicable - I'm saying in this instance it appears woefully incomplete.
I've given you reasonably logical reasons why I don't think so - as for proof: I'm not the one trying to make the *initial argument - you are, and you haven't provided any of the proof that you "require". Besides this isn't the forum for that.
*that Reynolds is directly (and fully) applicable in this instance.
^ this doesn't make much sense.
Obviously 20kHz is signal dependent (though with harmonics of course) and most of the tweeter's excursion is derived from its lower freq. use (which is most of the time). So excursion under normal operation is relative to much lower freq.s., not operation at 20kHz.
Yes assuming it's under normal conditions/proper design/use, perhaps as much as .25mm (.5mm total): so maybe as much as 250 micrometers.
Why would we assume that? How much fluid was placed in? What's the fluid's viscosity? How much force is there to keep the fluid in the gap and at any length within the gap? etc.
There are many variables here. Simplicity is fine if what's being discussed is actually simple - this isn't.
Even if the fluid stays perfectly in place (as designed) it's *vibrating at its periphery/"ends" (as previously mentioned) generating turbulence: a chaotic change in flow of the ferrofluid (and it's not just flow either, it's also pressure because that part is exposed to air and likely because that part doesn't have quite as much force in the gap as it does at center and this is with the false assumption that the force in the gap is actually uniform/stable - it is NOT).
*what I referred to before as "sloshing".
Anyway, I'm done - keep thinking it's as simple as you want it to be, I realize I'm not going to change your opinion of this and I'm OK with that.
Deterioration of the fluid in tweeters was primarily due to being driven too hard and overheating. (cooking the oil in the FF)
If you're worried about FF then just remove it, there are some tweeters that are the same except one has FF and the other has not. Perfomance changes slightly but not so much that the crossover needs modification. Some deliberately remove the FF as it gives a slightly more lively sound and also improves performance at low sound levels.
If you're worried about FF then just remove it, there are some tweeters that are the same except one has FF and the other has not. Perfomance changes slightly but not so much that the crossover needs modification. Some deliberately remove the FF as it gives a slightly more lively sound and also improves performance at low sound levels.