I did notice that a lot of manufacturers used staples along with glue on bitumen pads. The staples must give a degree of shock resistance. Me, I just stick to rubbery stuff and carpet adhesive.
Sorry to hear you've abandoned this for now, DrBoar.
It's all about BW3 with these designs, IMO. Very hard to model, and you have to play with the time alignment a lot to line it up. I haven't quite got the hang of it, but it does some neat things on power delivery. I seem to make a worse sounding speaker when I get nearer LR4, even though it looks flatter on frequency response. 🙂
Hello Dr. Boar
How are you have been a good many years since. 🙂
I am recovering a project parked aside since 2013 period. I have done up the KEF104aB cabinet but ran into a bottle neck because the KEF SP1039 I have are now too old to be worthy to be placed into a good cabinet.
Appreciate your curves shared as these confirm my electrical simulation and some measurement. The T27 is very uneven from 15khz upwards. A similar curve was also reported in the 1979 Audio report on the 104aB.
SPICE simulation show that the high pass 3.3u/0.3/10uf with 0.6uf T filter actually starts at 4khz point. The KEF 2.7mH/5uf/0.6mH low pass start to attenuate from 1000hz, and the B200 curve meets the T27 downward curve at just below 3khz point. The spice modelling of B200 SP1039 shows the roll off was rapid from 1.5khz (the acoustic measured sum of T27+B200 at 3khz was even). I have not seen the raw KEF B200 amplitude curve sweep, so I assume it has a big hump near the 3khz range, and it breaks up at 3500hz as mentioned by KEF. Both the B200 and T27 were operating at the limits at 3khz, and there maybe not much margin for both.
As the SP1039 are too old (deformed curtains, coils rubbing against the magnet etc) I am still desperately trying to revive them. The wide curtain rubber is hard to find. Unlike 2011, SP1039 are getting more scarce to source. I would expect most of them are considered End of Life, sadly. I have however found the Monacor SPM205/8 (Qt=0.7) and the Monacor SPM205/4 (Qt=0.5) as potential replacement drivers.
Simulation shows a much better bottom end if I use a 680uf -> 1.5mH ->8uf ->0.3mH low pass with the Monacor (with Zobel). The curve is more even and cross with the T27 at 3khz. Without actual measurement, this simulation is still a guesstimate.
The KEF 104aB cabinet caters for the 3khz 3rd order cross over to form the designed Polar and lobe distribution. I will still experiment with the new Monacor 205 driver and the next step will be active cross over. The active can take the form of Butterworth 3rd order or LR4 (in phase).
How are you have been a good many years since. 🙂
I am recovering a project parked aside since 2013 period. I have done up the KEF104aB cabinet but ran into a bottle neck because the KEF SP1039 I have are now too old to be worthy to be placed into a good cabinet.
Appreciate your curves shared as these confirm my electrical simulation and some measurement. The T27 is very uneven from 15khz upwards. A similar curve was also reported in the 1979 Audio report on the 104aB.
SPICE simulation show that the high pass 3.3u/0.3/10uf with 0.6uf T filter actually starts at 4khz point. The KEF 2.7mH/5uf/0.6mH low pass start to attenuate from 1000hz, and the B200 curve meets the T27 downward curve at just below 3khz point. The spice modelling of B200 SP1039 shows the roll off was rapid from 1.5khz (the acoustic measured sum of T27+B200 at 3khz was even). I have not seen the raw KEF B200 amplitude curve sweep, so I assume it has a big hump near the 3khz range, and it breaks up at 3500hz as mentioned by KEF. Both the B200 and T27 were operating at the limits at 3khz, and there maybe not much margin for both.
As the SP1039 are too old (deformed curtains, coils rubbing against the magnet etc) I am still desperately trying to revive them. The wide curtain rubber is hard to find. Unlike 2011, SP1039 are getting more scarce to source. I would expect most of them are considered End of Life, sadly. I have however found the Monacor SPM205/8 (Qt=0.7) and the Monacor SPM205/4 (Qt=0.5) as potential replacement drivers.
Simulation shows a much better bottom end if I use a 680uf -> 1.5mH ->8uf ->0.3mH low pass with the Monacor (with Zobel). The curve is more even and cross with the T27 at 3khz. Without actual measurement, this simulation is still a guesstimate.
The KEF 104aB cabinet caters for the 3khz 3rd order cross over to form the designed Polar and lobe distribution. I will still experiment with the new Monacor 205 driver and the next step will be active cross over. The active can take the form of Butterworth 3rd order or LR4 (in phase).
I tried reducing the 2.6mH coil to 1.2 mH to try to fill out the 2-3 kHz gap
Black curve. It worked but I also got a nasty peak at 2 kHz.
I tried a 1.8 mH coil and it really did not get better, I am also less than impressed with the T27. This is compounded by the fact that a 20mm dome will strain to reach down to a 200mm driver.
So the new game plan is to use an other vintage dome the Seas H087. This is 1.5" dome to the tweeter will have larger diameter voice coil than the woofer!
This will be crossed over all the way down at 2 kHz using a 24 dB/octave LR filter. I expect to top off the high range with small super tweeter using a 12 dB high pass filter in the 8-10 kHz range. I have a ribbon, some 1/2" domes and even some cone tweeter that might do the trick
Last edited:
Thanks for the thoughts🙂
Try to either resurrect the KEFs or go for a totally different design. Have you tried to get the rubber surrounds softer my silicone oil?
I will try to sell off my KEF B200, I have other projects coming up and need the space!.
I assume that the 680 microF cap is in series with the bass driver to in practice lower the Q of the system resonance and give a more gradual bass slope.
Try to either resurrect the KEFs or go for a totally different design. Have you tried to get the rubber surrounds softer my silicone oil?
I will try to sell off my KEF B200, I have other projects coming up and need the space!.
I assume that the 680 microF cap is in series with the bass driver to in practice lower the Q of the system resonance and give a more gradual bass slope.
My original intent was to resurrect the KEF 104aB. I will still do so as I am sorting out the B200 SP1039. I tried Armour All leather/rubber restore. Hope the rubber gets better.
The 400-680uF are to lower the Q. Will open a new thread when done.
I still have another restored KEF 104aB (year 2011) playing in the living room. Still a lovely pair of speakers over the years for relaxation.
The 400-680uF are to lower the Q. Will open a new thread when done.
I still have another restored KEF 104aB (year 2011) playing in the living room. Still a lovely pair of speakers over the years for relaxation.
- Status
- Not open for further replies.