Silencio 3 way speaker

Silencio 2021.jpg


Untitled-1.jpg



I have been building DIY speakers for many years and wanted to combine the benefits from some good French designs, like Elipson and Jean-Marie Reynaud with improved housing, materials and speaker units. The idea was to have as less material possible beside the drivers to avoid baffle diffraction and off set the drivers for time alignment which has a profound influence to get a coherent wave front. I opted for a closed enclosure because of the huge impact this has on the low midrange performance and a more tight, time correct bass.

The listing position meant that the bass enclosure could not be more then 80 cm high and the total height not more then 110 cm. Other important point was as less filter components as possible so the unit's needed to be able to handle 6 dB filtering, except for the tweeter which has a 12 dB slope for protection and low distortion. The Dynaudio D54 54 mm dome is designed to handle 6 dB filtering (cross over 1.5 KHz - 6 KHz) and the Vifa 8 inch woofer has a flat curve and easy reproduces 2 Khz and forms a great match with the dome midrange. The dome mid range has a natural fall off at about 6 KH but drops under 2 KHz, so it is limited in the lower mid range (600 - 1000 Hz). Because of the huge magnet and motor structure, it is very fast and dynamic. Down side is it's limited low end so you need a woofer who play's neutral and fast enough in the mid band up to 1,5 KHz.

There is no coil for the high cut off because of this natural fall off. The magnetostat tweeter on the other hand has a 12 dB / 6 KHz filter because these small diaphragms do not like big amplitudes or lower frequencies.

The driver lay out is more like a 2+ way because the woofer and dome midrange cover 30 - 6000 Hz together. This means in practice that the tweeter does only the upper high frequency's and the woofer and dome midrange acts almost like a tweeter. The dome mid range is the core driver of this design because it's wide dispersion in the room gives the sound a natural character, even if you are outside the room you hear a coherent voice, not like out of a funnel. This specific character of dome mid range, like the famous ATC SM75-150 and the Dynaudio D54 you hear very well when voices, guitar and saxophone is reproduced. It has this natural dispersion of sound in a room in the same manner as a voice or instrument does. This unique feature you won't find in any cone shaped driver.

Other objective was to avoid resonance and standing waves in the enclosure and the use of a "dead" material without much ringing. The 30 mm MDF baffle is covered in the inside with 6 mm ceramic tiles (glued) to avoid sound leaving the enclosure. It is forming a acoustic barrier because of the two very different resonance frequency's of MDF and ceramic. The inside of the PVC tube is covered with 1 cm felt and the enclosure is filled with synthetic fiber.

The bass unit needed to be able to play up to 2 KHz without ringing (resonance) and be able to reproduce the low-mid frequency's (voices / saxophone ect.) with clarity and low distortion. This means a light (23 gram) non resonant poly cone with good damping like this filled poly cone from the Vifa 21WP250, which by the way play's very low in this 40 Liter enclosure. A usable 30 Hz is reproduced with medium volume. This speaker is ment for rooms up to 30 m2, not for high volume playing which would mean to much excursion for the bass driver. There is a huge difference between electric and mechanic power handling, the bass voice coil can handle 80 watt RMS but in reality it already long before that, reaches it's cone excursion limits. That where my thoughts designing this speaker 33 years ago and still love to upgrade details like capacitors, filter slopes ect. speaker DIY is a great hobby, never ends.

images.jpg


DSC04843.JPG
 
Last edited by a moderator:
Yes, these are the RHT12S, i believe they where made in Japan by Matsushita and sold by Visaton. They have a samarium cobalt (front) and ceramic magnets (back side). Like them.. They do not play low, i crossed them at 5 KHz and recently 4 KHz. It's more of a super tweeter. I was thinking to replace them but not sure about that.. maybe a AMT or so.
 
Last edited:
Hi Dave, that's a good idear, i have thought about that.. i don not know if it will make a difference because the dome mid and tweeter have a big over lap because i do not filter off the highs from the dome, it falls off above 5 KHz. Do you think it will improve?


D54.jpg
 
Yes, these are the RHT12S, i believe they where made in Japan by Matsushita and sold by Visaton. They have a samarium cobalt (front) and ceramic magnets (back side). Like them.. They do not play low, i crossed them at 5 KHz and recently 4 KHz. It's more of a super tweeter. I was thinking to replace them but not sure about that.. maybe a AMT or so.
They don't use any ferrite magnets, only rare earth type magnets (not much neo used back then) in push pull arrangement front and back. I had mine apart before. They definitely look to be designed by Matsushita, as the diaphragm construction and terminals are a giveaway.
 

Attachments

  • 20230422_040820.jpg
    20230422_040820.jpg
    592.5 KB · Views: 75
  • 20230422_040921.jpg
    20230422_040921.jpg
    471.9 KB · Views: 93
No, those are definitely ferrite. My RHT12S have rare earth magnets front and back, so I'm not sure when Visaton changed their design over the years. I bought mine back in 89-90, so depending on how old those were you show, there was some kind of production change. I'd be curious to compare measurements, which may be tricky between 2 different mic calibrations.
 
The D54 did never gave signs of stress so far and the natural fall off is quit a lot, starts at 2 KHz dropping. I do not see the benefit of a LP because the drop off above 5 KHz is significant. 😉 I had before a 22 uF and later 33 uF. I do not know what a elliptical HP is? Never considered more components as strictly necessary and i do not run them with high volume, also because the bass will be the first one to hit it's mechanical limit. D54 is a monster about power (250 watt RMS / 1000 peak). What do you suggest?
 
Last edited:
Do you think it will improve?

Ideally the 2 drivers being XOed have a centre-to-centre distance within a quarter wavelength at the XO. Given this us near impossibe with typical tweeters as close as possible is the goal. You should be able to take at least a couple cm out… if you can trim bezels or use other tricks to get them closer…

dave
 
The D54 did never gave signs of stress so far and the natural fall off is quit a lot, starts at 2 KHz dropping. I do not see the benefit of a LP because the drop off above 5 KHz is significant. 😉 I had before a 22 uF and later 33 uF. I do not know what a elliptical HP is? Never considered more components as strictly necessary and i do not run them with high volume, also because the bass will be the first one to hit it's mechanical limit. D54 is a monster about power (250 watt RMS / 1000 peak). What do you suggest?
If the level restriction isn't an issue for you, just leave it the way it is. The big issue thought is the amount of
LF that gets through a 1st order filter that low. Its more than you think and even though it isn't audibly strained, it still pushes the mid very hard and the woofer running in the overlapping band will mask most of the distortion. No issue if you dont play it loud, but if you want proof of what I'm saying, disconnect the woofer and listen to some piano or male vocals.
 
  • Like
Reactions: Henk Haring
If the level restriction isn't an issue for you, just leave it the way it is. The big issue thought is the amount of
LF that gets through a 1st order filter that low. Its more than you think and even though it isn't audibly strained, it still pushes the mid very hard and the woofer running in the overlapping band will mask most of the distortion. No issue if you dont play it loud, but if you want proof of what I'm saying, disconnect the woofer and listen to some piano or male vocals.
After long testing i gave it a try change the 33 uf to 16 uf, 6 dB at around 1200 Hz and indeed, the sound improved a lot, less distorsion i think.. 33 uf was to low. Untill low / med. volume it was o.k. but above not so much. Even the imaging improved, more detail ect. I put two 33 uf in series because i had some old MKP Solen chatteua rough capacitors that i combine with Hifidiy Live Zilver Mkp to make the 16 uF. Is one 16 uF better?