There's an interesting thread on HF right now pertaining to the K1000 - it's interesting because one of the engineers involved in it's design is actively participating.  As I don't go to HF too much these days and figured some others don't either, I'm just going to cull some of the info from the thread into a Q&A format.  If I'm an idiot and re-phrase something incorrectly (especially the technical stuff) feel free to edit this.


So who is this guy? 


From my extensive five minutes of googling, he is Heinz Renner, R&D engineer from 1972-1989 with AKG, then director of R&D at Phillips/NXP/Knowles.  At Phillips he also worked on headphones for a few years that were OEM's for Sony, Nokia, Apple, RIM, etc.  He is currently retired and on pension, and is freelancing and is a specialist in "renewable energy and the environment"


Here's his LinkedIn: [size=78%]http://at.linkedin.com/pub/heinz-renner/18/973/275[/size]


In the K1000 he was mainly responsible for the membrane, filter, and magnet system.  He was tapped from developing dynamic microphones, which were his specialty at the time because of his know-how for very light mass membranes.  He designed the K1000 together with two other engineers, Helmut and Ewald.


Q: What led up to the design of the K1000?


The K1000 was developed in the wake of the K240DF.  DF stands for diffuse field, which, simply put, is designed so that the frequency response that hits the ear is the same as from far field listening - synthesizing listening to ideal speakers in a listening room.


There was a large following for the product, especially in Germany.  We still thought there was more to be done as engineers - sound performance varied depending on the user's ear [note- seems like HRTF definitely comes into play with this type of headphone] and the closed design was hot and uncomfortable. 


So in 1985, we decided to design a headphone with the 240DF's pluses without it's minuses - thus was the K1000 born.  It was clear that due to missing sound pressure chamber in front of the ear, that the speaker driver has to deliver a very low resonance frequency and relative huge displacement.  Also, the effective membrane surface has to be very large, stiff, and lightweight.


The developement of the membrane took approximately 3 years with the dearth of simulation programs available at the time.


We debated alternatives like orthodynamics and electrostatics for the magnet/coil system, but ultimately chose to go with the dynamic route.  This is because the special design of the radial magnet system which manages to be very effective while still very tiny.  This enabled the acoustic inductance of the complete K1000 to be lower than any other driver system we benchmarked.  In a nutshell: When measuring at the ear entrance wearing the K1000 with the sound coming from the free field, is almost the same as without the headphone.



Q: Is a DF adjusted headphone ideal for all types of music?


The clean sound of DF adjusted headphones are not great for all genres.  We recognized this at AKG which is why we tuned different headphones differently at the time.  I'm partial to DF but maybe that's my love of Pink Floyd


Q: Was the K1000 a no-holds-barred project or were there cost limitations? 


There was no absolute cost limitation.  This was a first for AKG.  That's because the three engineers (myself, Helmut, and Ewald) developed it under the table on our own time.  After initia l prototype listening tests, an official development order came through.


Q: Can it be improved upon with today's materials?


Here is what could be done better:
1) The membrane we used was not brittle enough.  We tried to compensate for this by putting shellack around the coil at the membrane surface to stiffen the area
2) The damping within the membrane material (a 5 layer type) is not as high as would be ideal.  Today's very thin materials can offer this easily.
3) The magnet force at the time was approximately half what it is today.  This could be spent into higher efficiency and/or into damping
4) Mechanic precision in manufacturing was not on today's level.  This would add efficiency and/or lower resonance frequency which translates to more bottom end.
5) The way the headphone interacts with the head was rushed through developement.  We could do much more today to increase comfort, such as gel filled cushions.
6) Perhaps the amplifier to convert standard headphone output towards K1000 suitability could be integrated into the headphone.


Q: Max input power is 1000mW equivalent to approximately 100 dB - this means they can't sound louder without degrading the sound, with higher THD  breaking down - why impose this limit on the headphone?  Also, what's up with it's sensitivity?


I'm not sure about those specs.  It is very difficult to compare the performance of speakers to headphones with specs.  This is because the signals are so different and the sensor type and position (microphone or artificial ears), not to mention commercial interests.


The K1000 is a simple dynamic system with a very low Q factor.  This means that the fundamental resonance frequency is offering the low frequency level only.  The K1000 was originally designed to hit 25Hz and below that would be the slope of 12dB/octave.  [Note - to my knowledge, this means the crossover slope drops by 12dB per octave/per halving.  So it would drop the first 12 at 12.5Hz, etc] The problem was, to manufacture those drivers we would have needed higher manufacturing precision than was available.  Otherwise the fall off rate would be high.  Therefore, after I left AKG, it was decided to move the f0 to 35 and later 40 Hz.  This means approximately 10 Db less fundamental bass level.  Therefore it is important to inform yourself what the production date is on your K1000 [note - this seems to explain bass heavy/bass light with regards to the K1000]


Any dynamic speaker slopes down with 12dB per octave below the fundamental resonance.  The K1000 is designed in a way that only the movement of the coil is the limiting factor.  The thermal limits of the coil as well as the mechanical limits of the membrane far exceed the limits of the movement of the coil.  Simply put: If you listen to the headphone with music which offers high bass levels and drive with the amplifier until the point of clipping, its fine - the THD is always very low, and when driven to extremes it has "only" second order harmonic distortion, which is not easily audible [note - or necessarily unpleasant].  I'm sure you can measure 115dB without any significant THD. 


As to sensitivity, it is poor.  This was the design drawback.  Yes, it could be theoretically improved today, by about 6dB, but would still be poor compared to traditional headphones.  The K1000 is relatively high impedance-  for speakers.  Any reasonable amp can drive it.


Q: Bass-light/Bass heavy?


This would d epend on:
1) Tightness to the ear channel and
2) The acoustic adjustment.


Amplifiers wouldn't affect this much, as long as it offers enough voltage without distortion and a flat frequency response.


Q: With DF equalized headphones, any two heads with different shaped pinnea/acoustic impedance will yield different DF eq curve. How do you design a DF headphone for everyone?  Is it because the recording head doesn't matter as long as you're using the right calibration curve to assess performance, or is it because you get more variation between FF and DF than you get between any two measuring devices?


The DF design of the K1000 is not dependent on the user's ear.  This is what I like so much about it.  It is acoustically invisible, which is why it works so well.  It's just open like ideal speakers in far field inside a standard listening room.  This means that the design is not dependent on the ear or the artificial ear.  (This conversation reminds me of a long discussion with my previous boss... he did not understand what I am talking about.)
In other words, you get different results with different ears, but the correct measurement always compares the far field result with the headset result, and they are the same.


Q: Honestly, do you think the K1000 has compromises, specifically: 1) potentially large amounts of distortion at low frequency because the driver has to be excessively pushed to compensate for front/back cancellation. 2) Inevitable presence of breakup modes in a large diaphragm at mid/high frequencies, and 3) Diffraction from the edge of the baffle which probably can yield some "speaker in front of the ears feeling" if the mind is tricked by this


Here are the compromises of the K1000:
More than +-3.5mm membrane displacment will occur recognizable distortion to the human ear.  First K2 and K3 later when driving even more.  So if you drive the bass too much, there will be some distortion.  But to be honest, most human ears distort before this point.  Due to the natural sound of the K1000, most humans underestimate the real sound level on the ear. 


Breakup mode is not a problem with the K1000 membrane design - it took us more than 3 years to realize that.


A two-way driver system can only help if the "one driver" system its compared to has problems with transient oscillation.  This behavior is dependent on the real moving mass.  This is not only the mechanical moving mass, but the acoustic and the mechanic together.  Acoustic mass is underestimated.  Imagine how much acoustic mass is inside the front and rear holes of a capacitive driver... it is much more than the total moving mass value of the K1000 system.


Mechanical suspension of the K1000 is linear until almost the maximum movement.  This is because we didn't compromise any of the 4 edges of the square, we just cut the edges out and left them filled close enough to reach 25Hz on the bathtub level


Q: Would it be reasonable to expect a re-release or update to the K1000 these days?


It would be great to re-start or update production.  But, according to my estimates, it would cost $400,000 in developement and $650,000 in manufacturing costs for 5000 units per year.