Restoring Monitor Audio R300 bookshelf speakers.

[AllenB]

It's good to hear you are making this kind of progress.

Just to add some spice I'd like to mention Head Related Transfer Function (HRTF). It's controversial and I'm just throwing it out for what it's worth, but attempts to correct for it seem to add a good quality to some peoples speakers.

Another suggestion, although the results are likely to be subtle at best, is to trial removing the capacitor from within the inductor. Since capacitors have a little inductence themselves due to their being wound, there is the possibility of interaction.

Just like to add that I once used Vifa aluminium domes (some time ago), and they clearly sounded like a piece of foil. Apart from that, they had very good qualities and I liked them.

[system7]

Originally Posted by AllenB

It's good to hear you are making this kind of progress.

Just to add some spice I'd like to mention Head Related Transfer Function (HRTF). It's controversial and I'm just throwing it out for what it's worth, but attempts to correct for it seem to add a good quality to some peoples speakers.

Another suggestion, although the results are likely to be subtle at best, is to trial removing the capacitor from within the inductor. Since capacitors have a little inductence themselves due to their being wound, there is the possibility of interaction.

Just like to add that I once used Vifa aluminium domes (some time ago), and they clearly sounded like a piece of foil. Apart from that, they had very good qualities and I liked them.

Flippin' eck, AllenB, you kept me awake worrying about that capacitor within the coil...

Fixed it now. Maybe sounds better too. Dunno really.



Don't think Monitor Audio have done much better in some respects...

[system7]

Currently investigating relative polarity of tweeter and bass. The white (tweeter) and red (bass) wires from the crossover turn out to be hot or positive. The speakers are marked with positive terminals, which mark the forward direction of movement for a 1.5V battery when +ve end connected to hot terminals of the speakers.

Original Filter:



For this crossover I reckon the tweeter has to be wired out of phase with the bass, especially if you reckon the bass has a natural !st. order rolloff, leading to an effective second order butterworth (?) rolloff on both drivers:



In fact they were wired in phase, for whatever reason, which has gotta be wrong and create a 180 degree hole in the frequency response around 3.5 KHz. Anyway, corrected it to my satisfaction.

[AllenB]

There are enough variables that the issue of preferred polarity loses its clarity. I'm reluctant to guess even if I see the frequency response and the schematic. Glad you found your preference. Did you use the pink noise method?

[system7]

Just an update on this "restore blog"...

Thanks to help from Planet 10, aka Dave, I have reinforced the cabinet front and back panels. Just 4X20cm pine cross battens per speaker on front and back to strengthen around holes. The battens don't even go to the edges, since I decided not to interfere with natural damping too much. Access was limited, so no chance of a proper brace really.

Huge number of crossover bits accumulating, but I'm learning as I go along, so we are not near the finished article yet. A Vifa or Morel tweeter with a 94mm metal front plate is beckoning along with some new coils, and I need to rolloff the bass a bit, I reckon.

[strawberry]

I see lots of problems. The crossover point for the tweeter is too high. The woofer needs a Zobel. The inductors shouldn't be parallel like that or the magnetic fields will feed into each other and make a sickening sound. Depending where your speakers are in relation to your ears, it can be motivated to switch polarities to "wrong", plus plus. If they are in front of you or below you, wire the tweeter "wrong". The dip in mid response is because of the flawed crossover design. You have a lot of work to do there, man.

For the tweeter, if you have no bigger inductor, add a 3.3 uf or 4 uf in parallel to the existing 3.3 uf. Put a 3-5 ohm resistor in series with the 0.27 mH inductor. A Zobel for the woofer could be 10 ohm and 10 uf. If you have a 0.5-1.0 mH inductor for the woofer, use that instead. If you have no such inductor, put a 10 ohm resistor in parallel with the 1.6 mH inductor. Move one of the inductors away with extension wiring. Tell me how it sounds.

[system7]

Originally Posted by strawberry

I see lots of problems. The crossover point for the tweeter is too high. The woofer needs a Zobel. The inductors shouldn't be parallel like that or the magnetic fields will feed into each other and make a sickening sound. Depending where your speakers are in relation to your ears, it can be motivated to switch polarities to "wrong", plus plus. If they are in front of you or below you, wire the tweeter "wrong". The dip in mid response is because of the flawed crossover design. You have a lot of work to do there, man.

For the tweeter, if you have no bigger inductor, add a 3.3 uf or 4 uf in parallel to the existing 3.3 uf. Put a 3-5 ohm resistor in series with the 0.27 mH inductor. A Zobel for the woofer could be 10 ohm and 10 uf. If you have a 0.5-1.0 mH inductor for the woofer, use that instead. If you have no such inductor, put a 10 ohm resistor in parallel with the 1.6 mH inductor. Move one of the inductors away with extension wiring. Tell me how it sounds.

Thankyou for that, strawberry. Yes, you are right! The two coils are probably way too close together and badly aligned for mutual effects...I now have taken it apart for a rebuild. The hot-glue softened in hot water. It may be the root of the top-end harshness.

I am looking into baffle-step correction. Which is sort of what your 10R adjustment amounts to. 1.6mH is kinda awkward except with 4.5 ohms on the 10" baffle. Lots to do, as you say. It's a bit borked really. No idea on an exact impedance correction for the woofer, but I'm guessing it's about 0.25mH Le. FWIW, the current tweeter is 0.05mH and Re of 6.2 ohms. Woofer Re is 6.0R.

This is all tremendously interesting.

[WaVeInFoRm]

mate, unless your the biggest flippin genius since Leonardo da Vinci, i dont see you pulling this off... acoustic modeling is painfully lacking

[WaVeInFoRm]

infact your not far away from where i live, im in kent. maybe when ive finished my speakers you can pop over and have a listening session with a few beers, then you can let me know if they where as unpulloffable as you first supposed... be interesting to see if you hold to your originnal viewpoint

[sreten]

Hi,

Edit : misunderstood the pictures .....

1.6mH on the bass/mid = some included baffle step, and as you've noted in
another thread zobelling the bass / mid will completely screw up its roll-off.
The bass unit looks like a paper coated Audax to me or possibly Elac.

Zobelling an attenuated tweeter is an accepted method of making
it sweeter, and of course you don't have to stick to zobel values.

Not saying the x/o can't be improved, it probably can, but not by conjecture.

Assumptions about driver roll-offs are exactly that and it is wise to assume
they are as such that the original driver phasing is correct, not incorrect.

Other than the correct point about inductor alignments, Strawberry's
musings are the usual nonsense that only illustrate just how little he
actually knows about practical x/o design. I'd suggest a thorough
perusal of the x/o designs at : Zaph|Audio, anyone open to and
willing to learn can learn a lot there, and elsewhere.

rgds, sreten.