On a side note, the Bang & Olufsen filler driver concept is intriguing, getting perfect transient response from a 3 way with higher order bandpass slopes. Ive done some research on it, but haven't found practical info which I could implement. Not sure if the D7608 can cope with the bandwidth requirements, but it looks like a sound concept with the right capable drivers.
This one? http://www.audax.com/archives/HM210C0 - Catalogue 1994.pdf
How did you load the woofers (sealed, ported)?
I heard the M74Bs and really liked them. They sounded very clean, sort of like a smoother Alu/Mg dome but with better character. The huge peak needs to be filtered out for best sound, even with steeper slopes, but rhats realt the only drawback (aside from the cost).
I have T34Bs and wanted to buy a pair of M74Bs, but the Alu dome M74A is honestly just as good if you consider how close they are, not to mention only about half the price. The Alu version also sounds a little more lively and colored in a good way to my ears.
I have T34Bs and wanted to buy a pair of M74Bs, but the Alu dome M74A is honestly just as good if you consider how close they are, not to mention only about half the price. The Alu version also sounds a little more lively and colored in a good way to my ears.
Hi Thanh,
I tested that midrange "as is", with no any chambers
And so I must forget to reach FC=6kHz with the D7608 to transit to the tweeter, then ?
T
Found this thread just right now, I was speaking about my very good experience with Audax HM serie in some of my recent 10-20 posts, I have almost like the same experience as you, its my reference speaker for almost ten years, hm170g8+raal64-10 in transmision line, keeping it for years and simply not want to sell them, invincible : ) Even before Raal I had TW025A16 and it was nice too. Its open public for those who is interested, posted it in one of my recent post.
edit: https://www.diyaudio.com/community/threads/best-8-inch-woofer.278192/post-7562003
I'm realy thinking to put 3 more hm170g8 in serie/paralel for an 8ohm bass configuration in the same box so that sensitivity increase so that Raal get no atenuation, instead an simple 4.7uF, might become realy interesting! : ) Those Raal64-10 have their internal inductance 0.62mH and reguire first order crossower (it form electricaly seccond order at 3k). If attenuation is need more than 1ohm than an L must be put as like what I do per manufacturer recomndation, confirmed by simulation.
Last edited:
Yes - the works of Mr. Bakegaard. That's this idea which inspired me for my 3-way S-// Xover : midrange filling speaker.
I'll add another RLC couple for the High-Mid dome speaker - for the moment, the favori is the SB SATORI MD60N - here's the look of that 4-Ways project :
T
You have to utilize part of the natural rolloff slope to add with electrical response ie LR2 with -6dB points between 3 - 4k. I honestly wouldn't go steeper than LR2 or derivative thereof to preserve smooth phase response. The baffle will dictate much of the response in that area, so you need to factor in diffraction and baffle step losses.
Yes indeed, I would stay in a simple crossover 6-12dB S-// configuration - the one that I experimented with success on my 3-Ways 375L Bass-Reflex (shown some post above) - by adding just the supplementary RLC cell for the high-mid dome midrange (quick principle drawing below). I think it is certainly doable, as don't want to shift into complex crossover with 12dB/Oct. or higher.
Here is the FR response of the SBAcoustics SATORI MD60N, as provided by the manufacturer's data sheet - if it is taken on a plane CEI baffle, then chances that it is realistic... The interest is the 94-95dB/w/m efficiency and the good linearity up to 6-7kHz 😎 :
That said yes, nonetheless a baffle step correction may be required : it has to be evaluated by tests, on a prototype of my project.
Here's the FR of the D7608, according to the official SS specs, showing a lower efficiency and a faster response drop 😕 :
At least on the specs comparison 1-10kHz below - assuming all things equal - the FR response of the MD60N seems more favourable for my project than the FR of the D7608, if I plan to reach 6-7kHz, notably for a near field listening, at incidence 15-20° horizontal :
So I am tempted to go for the MD60N, but having no real experience about dome midranges : am I missing something, @profiguy, that would finally turn to the advantage of the D7608 vs. the MD60N ? 🤔
T
Last edited:
Be aware the specs on the MD60N have changed at least once. There is a small issue with the original design around 1k due to the air space behind the VC gap causing a secondary peak.
If you look at the impedance plot from SB, it shows only a slight asymmetry at the upper end of the peak. The Hificompass test shows some significant asymmetry on both ends of the impedance peak. That corresponds with raised 3rd HD around 900 to 1k. I observed this myself with the first set of MD60N I tried and didn't keep. They sounded like a rubbing VC when in fact it was simply elevated HD around resonance.
They changed the chamber dampening on the 2nd version which I haven't tried. The FR also changed slightly up top and it has the typical mass rolloff kink on the upper end that other large mid domes suffer from.
The sensitivity is pretty high on this dome. The xmax looks low and I'd tend to agree based on the elevated sensitivity. My reasoning for favoring the D7608 over the MD60N is simply that its a large diameter dome and it has excellent breakup behavior in thr upper rolloff despite a lower cutoff. The price is also very good and I prefer to use a larger chamber with extra flow restriction to lower Qts. The MD60N is also closer to $200 whereas the D7608 is around $115.
So maybe the bugs are worked out on the 2nd version of MD60N. Id like to see 3rs party measurements to see if things are better. I'm perfectly happy with the D7608. It has better top end than the measurements show but off axis it shows its a larger dome. I'd cross it to a smaller HF dome around 3.5k. If you need the extra extension, I'd look at the Morel domes, as they have 2mm p-p xmax and about 6mm smaller diameter. That should be close to the MD60N for low end capability, but the sensitivity isn't as high. Pick your poison.
If you look at the impedance plot from SB, it shows only a slight asymmetry at the upper end of the peak. The Hificompass test shows some significant asymmetry on both ends of the impedance peak. That corresponds with raised 3rd HD around 900 to 1k. I observed this myself with the first set of MD60N I tried and didn't keep. They sounded like a rubbing VC when in fact it was simply elevated HD around resonance.
They changed the chamber dampening on the 2nd version which I haven't tried. The FR also changed slightly up top and it has the typical mass rolloff kink on the upper end that other large mid domes suffer from.
The sensitivity is pretty high on this dome. The xmax looks low and I'd tend to agree based on the elevated sensitivity. My reasoning for favoring the D7608 over the MD60N is simply that its a large diameter dome and it has excellent breakup behavior in thr upper rolloff despite a lower cutoff. The price is also very good and I prefer to use a larger chamber with extra flow restriction to lower Qts. The MD60N is also closer to $200 whereas the D7608 is around $115.
So maybe the bugs are worked out on the 2nd version of MD60N. Id like to see 3rs party measurements to see if things are better. I'm perfectly happy with the D7608. It has better top end than the measurements show but off axis it shows its a larger dome. I'd cross it to a smaller HF dome around 3.5k. If you need the extra extension, I'd look at the Morel domes, as they have 2mm p-p xmax and about 6mm smaller diameter. That should be close to the MD60N for low end capability, but the sensitivity isn't as high. Pick your poison.
Suggest perhaps you also do that comparison with the outgoing D3806 (a bit more expensive, discontinued by Scan recently but still available to buy, very nice build quality with SD1 motor, you won't need to futz around with a DIY chamber). External measurements many places.
Thanks for your insight and knowledge @profiguy and @motokok ! I note all that and will study it... 😎 😎 😎
The bandwidth in which I would use this High-Mid midrange dome is circa from 2-3kHz to 6-8kHz.
The Low-Mid midrange will be my favourite Beyma 8M60N, used since 300-400Hz, and able to reach 5kHz, so no issue to stay at 2-3kHz.
The Tweeter choice that I have on hand is a pair of :
---------brand-new Beyma T-2030---------------- vintage but perfect condition Philips AD1631T8---------excellent Beyma CP21/F...
To be honest, in all my crossover attempts in a 2-Way or 3-ways, I personally never liked too much the tone of a tweeter for a FC lower than 4500-5000Hz (i.e. 2500-3500Hz), even with tweeters having a fs at FS=1100Hz (i.e. Monacor DT-280 )... 😕
T
The bandwidth in which I would use this High-Mid midrange dome is circa from 2-3kHz to 6-8kHz.
The Low-Mid midrange will be my favourite Beyma 8M60N, used since 300-400Hz, and able to reach 5kHz, so no issue to stay at 2-3kHz.
The Tweeter choice that I have on hand is a pair of :
---------brand-new Beyma T-2030---------------- vintage but perfect condition Philips AD1631T8---------excellent Beyma CP21/F...
To be honest, in all my crossover attempts in a 2-Way or 3-ways, I personally never liked too much the tone of a tweeter for a FC lower than 4500-5000Hz (i.e. 2500-3500Hz), even with tweeters having a fs at FS=1100Hz (i.e. Monacor DT-280 )... 😕
T
Yes, there are not many ideal use-cases / optimal scenarios for the D3806 nowadays compared to other mid-domes or standard cones, but your design is actually one of them.
https://www.dibirama.altervista.org...-d3806-820000-tweeter-1-5-6-ohm-200-wmax.html
https://hificompass.com/en/speakers/measurements/scan-speak/scanspeak-d3806820000
https://www.dibirama.altervista.org...-d3806-820000-tweeter-1-5-6-ohm-200-wmax.html
https://hificompass.com/en/speakers/measurements/scan-speak/scanspeak-d3806820000
What you're describing here is something similar I always deal with myself. The ear is super sensitive in the 2 to 5k range, mainly because our pina resonates around these frequencies, which amplifies sensitivity. The exact amount of boost and bandwidth in this range is different for every person and depends on overall playback volume, so its important to tailor the FR to your specific ears if you listen critically ie with studio monitors or headphones.
I strategically avoid a sub 3k xover in 2 way designs if the driver selection allows it. Usually I'll opt for 3 to 4k xover and spread the center frequency by lowering the LF LP and raising the HF HP to create a little bit of a dip, not more than 2 - 3dB total @ 3.5k. I also try to make the off axis response fall off a little more around that frequency range, so (1) I put less 3.5k energy into the room and (2) I can tune the dip by toeing the speakers in or out.
This method of tuning gives me more acoustic control over the listening space if its highly reflective and reduces listening fatigue over time. It works very well for me and sounds more natural to my ears playing back a flat EQed recording, so my ears are aware of the slight discrepancy in FR. I get alot of backlash from "experts" claiming I'm screwing up the actual FR linearity, but due to the fact I'm aware of the difference, I can compensate for it in a mix to get the correct desired EQ results and know the natural balance when I hear it.
I also drop the 2 to 5k range when playing a studio recording back very loud or running a live FOH mix. Not doing so will make your ears bleed fast and its honestly abusive to most people's hearing. This specific loudness curve allows you to play louder for longer periods of time and somewhat protects your hearing compared to listening to flat FR at the same level.
So that said, yes I design all my own speakers with this 3.5k dip to some extent (give or take) depending on how they'll be used ie for monitoring or casual listening, the typical intended listening volume and room acoustics. Some call this the BBC dip, others call it the Fletcher Munson curve, but it can be a bit more sophisticated than that.
Bottom line is do what you prefer for your tastes and not for others, unless you need to factor in pleasing specific other people as well. Flat doesn't always sound good and one person's perceived flat is another person's acoustic nightmare lol.
Thanks @motokok - the problem is that this tweeter has an efficiency limited at 89dB/W/m, while all my other speakers are at least at 93-94dB/W/m... Even with a baffle effect - which will also occur for the Low-Mid and Tweeter - I think that the D3806 used as a Hi-Mid speaker will lack output level.
I aready made that kind of test - below on a small 3-ways at 90dB/W/m (now dismantled) :
I used a little full-range speaker as midrange, nice tone by itself, but with an efficiency at 84-85dB/W/m...Iit could never compete with the two other speakers in terms of loudness. So this project turned out to return to a 2-Ways speaker, re-using the same woofer, and with a FC at 4500Hz, despite a 1100Hz FS tweeter and a planned FC at 3kHz initially :
T
Touching on the Morel MDM55s, I plan on using some in this design for small format high output 3 way NF monitors for my son's home studio. He needs something small which can play pretty loud (with good power handling) assisted by subs and not struggle in the lower mids. This is why i chose the MDM55 with 2 x CAW538 midbass and CAT378 tweeter, all in a 20 ltr sealed box. While these Morel woofers aren't the lowest THD out there, they can take ALOT of power from 100hz upwards. Crossover points are planned to be around 1k and 4k 2nd order.
Attachments
I see @profiguy : I agree completely with what you wrote, and we possibly share the same perception of tone, reading you - thanks for your insight again...
Incidentally, when I rebuilt my Magnepan SMGb crossovers in 2021, I came to that similar "dip around 3.5kHz" trick you mention in your post's 2nd paragraph. I finally found that - by ear - the SMGb required a gentle "dip" between 1.5kHz and 3.5kHz grosso-modo, that is to say centered at circa 2.6kHz, (below, hand graph at the right) :
And yes : a flat FR doesn't imply necessarily a good FR...
Then it drives me to the assumption that, if my Hi-Mid dome speaker offers an excess of sensivity (dB/W/m), I can simply balance it with a L-Pad, or add a frequency-focused attenuation, if finally required, to match the desired tone balance.
Conversely, if the sensivity of that speaker is too low (i.e. like 89dB/W/m on the SS D3806 proposed by @motokok), the correction latitude in level is significantly reduced, even insoluble to balance, I'm afraid... 😕
Again, Do I miss something about the behavior of dome midrange setup strategy ? 🤔
T
Ah, shucks I missed your high sensitivity requirement, so I see what drew you to the SB. My table link once more for other ideas