Hi All,
I am a diy-er from the Netherlands.
I recently acquired my diy dream: a pair of NOS RSQ8P Philips 'ribbons'.
I have always been keen on designing a sealed enclosure (i love the sound of sealed in the morning) in a 2-way config. The RSQ8P makes this possible. I have some very nice 20cm paper woofers which reach 47Hz -3dB in a parallell config (2 woofers). The volume is appr. 100 liters. The bass extension is phenomenal. Now for the lesser part. Ik cross @ 1150 Hz, which is no problem at all for the ribbon. As a matter of fact, it has been used previously in a similar configuration. But the sound of the lowmids is like it comes from a transistorradio. The low end and highs are superb, but the 500 - 2000 region kills the fun. The crossover i now have is @ 1500 Hz and both sections are exactly -6dB @ 1500 Hz. So, the sum curve is flat. As a matter of fact the entire freqcurve is within 1dB (Arta measurement).
But the problem with the mid remains. My friend and i think the relative slow woofers are at fault. Because it has difficulty reaching the same dynamics in the 500 - 1500 range in order to keep up with the ribbon. Is this plausible or just erronous on our behalf? Any advice would be welcome.
Thanks and regards,
Ramon
PS i so do not want to give up the phenomenal low end of these 20cm woofers.
I am a diy-er from the Netherlands.
I recently acquired my diy dream: a pair of NOS RSQ8P Philips 'ribbons'.
I have always been keen on designing a sealed enclosure (i love the sound of sealed in the morning) in a 2-way config. The RSQ8P makes this possible. I have some very nice 20cm paper woofers which reach 47Hz -3dB in a parallell config (2 woofers). The volume is appr. 100 liters. The bass extension is phenomenal. Now for the lesser part. Ik cross @ 1150 Hz, which is no problem at all for the ribbon. As a matter of fact, it has been used previously in a similar configuration. But the sound of the lowmids is like it comes from a transistorradio. The low end and highs are superb, but the 500 - 2000 region kills the fun. The crossover i now have is @ 1500 Hz and both sections are exactly -6dB @ 1500 Hz. So, the sum curve is flat. As a matter of fact the entire freqcurve is within 1dB (Arta measurement).
But the problem with the mid remains. My friend and i think the relative slow woofers are at fault. Because it has difficulty reaching the same dynamics in the 500 - 1500 range in order to keep up with the ribbon. Is this plausible or just erronous on our behalf? Any advice would be welcome.
Thanks and regards,
Ramon
PS i so do not want to give up the phenomenal low end of these 20cm woofers.
Hi,
According to Zaph's ribbon tests http://www.zaphaudio.com/nondomes/
expecting a ribbon to work well with a low c/o frequency is tempting fate.
How have you arranged BSC ? Considered 0.5 way ? what bass drivers ?
According to what criteria ?
/sreten.http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://www.geocities.com/woove99/Spkrbldg/
Hi Sreten,
That graph is from the Philips datasheet and is already with a 2nd order filter incorporated. Klang und Ton magazine have implemented this ribbon with no problems at all @ 1150 Hz. In conjunction with two Davis 16cm kevlar mids.
Used woofers are somewhat older types from the brand Procus. They use kapton and are very ligthweight paper cones.
By 0.5way you mean that only one of the woofers runs uptil crossover with tweet? (aka 2.5 way).
I am now considering making a 3-way, with the woofers running uptil 100 Hz and very fast mids.
Image of testspeaker:
That graph is from the Philips datasheet and is already with a 2nd order filter incorporated. Klang und Ton magazine have implemented this ribbon with no problems at all @ 1150 Hz. In conjunction with two Davis 16cm kevlar mids.
Used woofers are somewhat older types from the brand Procus. They use kapton and are very ligthweight paper cones.
What do you mean? If they are 6dB down @ 1500Hz then they are down 6dB @ 1500 Hz, is it not? What do you the mean by:'according to what criteria'? If you mean what transferfunction: it is a 4th order LR slope. Which is realized by a second order electrical filter for both lowpass and highpass sections.
By 0.5way you mean that only one of the woofers runs uptil crossover with tweet? (aka 2.5 way).
I am now considering making a 3-way, with the woofers running uptil 100 Hz and very fast mids.
Image of testspeaker:
An externally hosted image should be here but it was not working when we last tested it.
I also want to add that you should consider a 2.5 config as in the current d'appolito approach you still come short to the 1/2lamdba tw-wo spacing and 3 order acoustical ideal. Therefore you are facing large lobing issues originating from the two midbass drivers.
In my experience 1150 Hz isn't a problem, however distortion for the Philips rises sharply below 1500 Hz. I'm using a philips in combination with a 20 cm crossed around 1500 Hz and do not encounter any of your mentioned problems.
I'm not convinced that your setup yields +/- 1 dB, just the Philips alone does not manage to hit these performance figures, check your gating/smoothing settings in your measurement program.
In my experience 1150 Hz isn't a problem, however distortion for the Philips rises sharply below 1500 Hz. I'm using a philips in combination with a 20 cm crossed around 1500 Hz and do not encounter any of your mentioned problems.
I'm not convinced that your setup yields +/- 1 dB, just the Philips alone does not manage to hit these performance figures, check your gating/smoothing settings in your measurement program.
Hi LaMa,
What woofer do you use?
I get the Philips within +/- 1dB by using just a series resistor which compensates for the rise in frequency above 10K. Gated impuls response has been setup correctly and i use 1/12 smoothing.
But i do agree on the lobing problem. The center to center distance between the two woofers is just to large (40cm) for it to be a balanced vertical response.
I therefore have given up this config and wil go for two fast midranges and the woofers uptil 100Hz in push pull on the side of cabinet.
There is a concept called Phlogiston which uses the Davis 16GKLV6M in MTM config.
What woofer do you use?
I get the Philips within +/- 1dB by using just a series resistor which compensates for the rise in frequency above 10K. Gated impuls response has been setup correctly and i use 1/12 smoothing.
But i do agree on the lobing problem. The center to center distance between the two woofers is just to large (40cm) for it to be a balanced vertical response.
I therefore have given up this config and wil go for two fast midranges and the woofers uptil 100Hz in push pull on the side of cabinet.
There is a concept called Phlogiston which uses the Davis 16GKLV6M in MTM config.
1/12 smoothing answered my question ... also what is the resolution of your measurement system (imho you need atleast 5 Hz).
At the moment I'm using a 20 cm custom Audiotechnology 23I driver with improved efficiency to match the Philips.
Crossing at 100 Hz isn't sensible in that case the mid still needs to compensate a great deal of the bafflestep (atleast 3-4 dB). Try to blend in the woofers such that the baffle step is handled by the woofers and you can utilize the full SPL of the mid. This config results in a single mid driver, preferably a 20 cm with 92-93 dB/w/m. However side firing woofers are not viable in that kind of configuration.
At the moment I'm using a 20 cm custom Audiotechnology 23I driver with improved efficiency to match the Philips.
Crossing at 100 Hz isn't sensible in that case the mid still needs to compensate a great deal of the bafflestep (atleast 3-4 dB). Try to blend in the woofers such that the baffle step is handled by the woofers and you can utilize the full SPL of the mid. This config results in a single mid driver, preferably a 20 cm with 92-93 dB/w/m. However side firing woofers are not viable in that kind of configuration.
Do you find 1/12 not useful? 1/24 does not give me much more detail.
Why starting from 5 Hz? I do not see what this adds to the info @ crossover (+1kHz). I also would prefer a single mid. Maybe some high efficiency Audax or Davis. So, you would recommend that the woofers are still crossed @ 500 - 800 Hz region for it to handle baffl step?
Thanks.
Why starting from 5 Hz? I do not see what this adds to the info @ crossover (+1kHz). I also would prefer a single mid. Maybe some high efficiency Audax or Davis. So, you would recommend that the woofers are still crossed @ 500 - 800 Hz region for it to handle baffl step?
Thanks.
RSQ8P said:
You have no slow woofers or something. The vertical dispersion of your long enough ribbon will not 'gel' with your MTM arrangement further than +/- 15deg. So you get direct sound without the proper room phantom bounce at 1500Hz and around with somewhat more than ideal 3rd order distortion on top of it.
Add the power response cancellation of the 2 cones about an octave lower plus the LR24 power deficiency at cross, and here are your threadbare quality 500-2000Hz mids. Smack center to where music breaths its power and heft of color.
My suggestion is you go for 1750Hz-2kHz BW3 filter between upper cone and ribbon, so your cross power is better and the lobe still up. You need to go a bit higher than before, for you lose an order and you want to stay ok with distortion. Just use the lower cone to judiciously fill out your BSC up to about 500Hz.
Regards and good luck with your project. It has potential.
D.A.R.R.Y.L. said:
Yes, exactly. But only 18dB per octave acoustic Butterworth between upper woofer and line source. It will fill out a lot of the power response notch that you experience at crossover with LR4 right now. Also the 3rd order AB phase shift is going to complement the upside down WT configuration and the lobe is going to stay up in crossover. Dropping the MTM for a 2.5 way you are going to rectify the power response cancellation centered at about 750Hz right now too.
As for how much BSC exactly and up to what frequency you are going to use the lower woofer for, your new measurements are going to guide you. I would suppose 3-4dB up to 400Hz for a stereo pair isn't a bad guess. But don't guess, measure and listen. Good luck! Report back!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.