SpeakerTalk

[qguy]

Now that I got an extra amp, I am planning to buy amp my kef 105.2. On x-over diagram their is shared capacitors that is used by the entire circuit. On some other sites, it is advised that this part should be removed any thoughts on this ?

[peejay1959]

Bump, I really want to do this using 2 quad 405's

[audiolabtower]

Do you mean the 900uF (2 x 450 in parallel)?

This is used in the bass equalisation and has a handy benefit of protecting from DC in the event of an amp fault if the SStop fails. If you want the design intent it should be left in. If you want to experiment and bypass it the sensitivity and low bass performance will not be to design but may work in your room.

A further complication is whether the alcaps have drifted over this length of time, but without measuring them you can't tell.

ps I would use a Quad for the B300, and the other channel for the top unit in each channel.

[iso]

[audiolabtower]

[iso]

[audiolabtower]

I agree. When I only had 2 amps I also found driving the B300 was better. Usually bi-amping works well at 2-3 kHz since the ears most sensitive range. However the 105 4th order crossovers work so well at integrating the T52 (if crossover still in spec) I think their performance here is better than most other designs. Given the extended bass response available in the 105, and the tendency of smaller rooms to muddle the high bass and low midrange that extra bit of control either side of the 400Hz crossover is more worthwhile.

[speakerguru]

There is no easy way to bi-amp LF and MF/HF while retaining the the LF alignment, d.c. protection and S-STOP protection for all units afforded by the series capacitors. You would have duplicate some sections of circuitry.

[audiolabtower]

Obviously I kept the large bass series cap for the B300. The midrange has 30uF in series and the tweeter 3.3 uF in series. Providing these are rated at 100 or 150V they should provide reasonable DC protection?

I would have thought the large series bass cap has little effect on the B110 and T52 responses? The loss is the SStop, so to have the same protection as before you would need 3 of these - presumably each specially curved for the voltage and voice coil temperature of the specific units. However I've used my Mk1s for over 30 years with varying amps in single, bi and tri arrangements without the fuses so think those Kef drivers were a well made and reliable bunch.

[peejay1959]

[audiolabtower]

My Mk1s were easy to tri-wire since the crossover is 3 independent parallel circuits to the input terminals, and the mid and treble does not depend on the large series bass capacitor. They have their own third order circuit blocking capacitors, which proved they worked when an old amp went 45V dc many years ago with no damage to my speakers. The overall response is 4th order with the drive unit characteristics.

The Mk2 is more complicated because the mid and treble sections are wired behind the large series bass cap and relay. The bass series cap is required for the design bass alignment, I have read of people removing it, but looking at the Kef design notes it was there for a good reason and is an expensive component(s), so Kef could easily have left it out if they had wanted to - they did not. When I replaced mine with 360uF of parallel polyesters I did some listening tests clipped either side and decided to leave it in, it was subtle, but a cleaner upper bass I thought.

I would not have thought the bass series cap impedance would have affected the mid and treble sections too much at their frequencies but guru is the expert here. I see no reason why the s-stop cannot just be left open circuit. I assume the 680 ohm is in circuit only when the protection operates so would be circumvented? If you feel there is an imbalance in mid and treble it is easily rectified by tweaking the gain of the 3 amps?

[speakerguru]

[audiolabtower]

Just looking at the input caps, the mid in the later crossovers was 450uF in series with 30uF, simple result around 28.2 uF, a significant 6% down? The treble input is 3.3 uF, simple result 3.27 uF, around 1% down.