Member
Joined 2009
Paid Member
Until the Alpair6 (paper cone) shows up I'm thinking of using the Fountek FR88Ex. Compared to the now obsolete FF85k the Fountek has the same Qts (although half the Qms), similar Vas and a lower Fs. It may not be of the same ilk but perhaps the Fountek could be used to make some kind of micro-Onken (aka planet10 uFonken) ?
Going on the premise that the Onken is an EBS alignment (enhanced bass shelf) the best information I could find was here [http://www.diyaudio.com/forums/multi-way/17460-calculating-ebs-alignment.html] where GM gives some guidelines that I've followed using the Fountek FR88Ex data:
Where Qt is the total Q of the system = roughly 0.6
Vb = 7.95*Vas*Qt^2.21 = 3.3L
Fb = 0.471*Vas*Qt^-0.677 [Fs instead of Vas?] = 63Hz
F3 = 0.33*Fs*Qt^-1.01 = 52.5Hz
Using the box simulation here []Loudspeaker enclosure calculating with Thiele Small parameter] I explored this alignment, at least very simplified. Using Vb 3.3L, port diameter of 3.5cm and port length of 20cm gives me something that looks like an EBS. It seems to indicate a lower group delay around Fs at the expense of worse delay below Fs. I Used a low QL in this simulation (if I understand properly what this means) to represent the lossy nature of the resistive ports.
Feels like I'm just scratching the surface here, the simulation doesn't know it's an Onken, so what else do I need to know ?
Going on the premise that the Onken is an EBS alignment (enhanced bass shelf) the best information I could find was here [http://www.diyaudio.com/forums/multi-way/17460-calculating-ebs-alignment.html] where GM gives some guidelines that I've followed using the Fountek FR88Ex data:
Where Qt is the total Q of the system = roughly 0.6
Vb = 7.95*Vas*Qt^2.21 = 3.3L
Fb = 0.471*Vas*Qt^-0.677 [Fs instead of Vas?] = 63Hz
F3 = 0.33*Fs*Qt^-1.01 = 52.5Hz
Using the box simulation here []Loudspeaker enclosure calculating with Thiele Small parameter] I explored this alignment, at least very simplified. Using Vb 3.3L, port diameter of 3.5cm and port length of 20cm gives me something that looks like an EBS. It seems to indicate a lower group delay around Fs at the expense of worse delay below Fs. I Used a low QL in this simulation (if I understand properly what this means) to represent the lossy nature of the resistive ports.
Feels like I'm just scratching the surface here, the simulation doesn't know it's an Onken, so what else do I need to know ?
Attachments
That driver models really nice for a uFonken. And since Fountek starts with an F we can call it a uFonken... to avoid confusion i dub it uFonkenEX. More bass extension (at the expense of efficiency).
Take one uFonken and adjust the slot height & width to meet the specs in the attachment. Adjust cutout & holey brace.
Damn... just had a shipment leave from Madisound today. A pair of those will need to wait till next time, but well worth a try i'd say.
dave
Take one uFonken and adjust the slot height & width to meet the specs in the attachment. Adjust cutout & holey brace.
Damn... just had a shipment leave from Madisound today. A pair of those will need to wait till next time, but well worth a try i'd say.
dave
Attachments
Qa = box losses due to absorption (stuffing, etc.)
Ql = box losses dues to leakage (poor construction, etc.)
Qp = port losses due to friction (too small, too high aspect ratio, damping in the port, etc.)
So a combination of Qa, Qp will show the response change trend, though no easy way to tell how much of what matches the sim without measurements, so usually whatever amounts sounds OK as a balance between driver protection and sounding over-damped.
GM
Ql = box losses dues to leakage (poor construction, etc.)
Qp = port losses due to friction (too small, too high aspect ratio, damping in the port, etc.)
So a combination of Qa, Qp will show the response change trend, though no easy way to tell how much of what matches the sim without measurements, so usually whatever amounts sounds OK as a balance between driver protection and sounding over-damped.
GM
Member
Joined 2009
Paid Member
I ended up getting my drivers from Solen, by the time I cover shipping from the US the price difference was negligible but I guess the volume of stuff you normally buy makes it worthwhile to get the better US prices.
Your suggested port dimensions are not what I understood to be an Onken alignment, it has no extended bass, if anything the extension is reduced, falling somewhere between a sealed box and optimally flat vented enclosure.
Why do you choose this approach instead of the EBS approach ?
Attachments
My miniOnkens only share the high aspect ratio ports with an "official' Onken.
I'm not a big fan of EBS alignments. I use my approach because it has proven in practise to work REALLY well. At least 10 successful basic designs (+ lots of variants)
dave
I'm not a big fan of EBS alignments. I use my approach because it has proven in practise to work REALLY well. At least 10 successful basic designs (+ lots of variants)
dave
Member
Joined 2009
Paid Member
Thanks Dave. I'll go with your wisdom on this.
I may already have some wood for the uFonkenEX, just need to watch Postie for the drivers.
If we did not already have a 2L box design, would there be any advantage in a bigger box for the Fountek driver - perhaps bringing the driver closer to ear height ?
I may already have some wood for the uFonkenEX, just need to watch Postie for the drivers.
If we did not already have a 2L box design, would there be any advantage in a bigger box for the Fountek driver - perhaps bringing the driver closer to ear height ?
It worked so well in 2 litres i didn't bother playing with anything else. Put a tilt on the baffle (or prop up the front) if you need to fire it at your head -- or as i will eventually try, i'm going to hang a set of milli or micro size off of some surplus legs off flat screen iMacs.
dave
I just got in those big buck ScanSpeak 3" (nicely made but nothing in the build that justifies them costing as much as FE126En). I'll be angling towards some sort of miniOnken for these too.
dave
dave
Member
Joined 2009
Paid Member
Glad to hear that the (F)Onken series will have a life beyond Fostex and you will widen the net to consider some other drivers. I think a lot of newbies like me look to you and Chris and Scott for advice and we tend to follow your lead towards the kinds of drivers we consider. For awhile it seems that Fostex was king and now it's Mark Audio. But a bit more variety would be welcome - although I recognize that it's a bit more work for you guys.
Member
Joined 2009
Paid Member
Hi Bigun,
since I can measure TS parameters myself, I found that product brochures are too often dictated by marketing dudes rather than R&D engineers.🙁
My values for the FR88EX are: Fres 106Hz, Qts 0,70, Vas 1,05 l.
Apart from that, very nice sound. I use them in 2l CB, supported by a bandpass subwoofer as PC speakers. Notch filter 0,33mH/8,2yF/4,7Ohm ( Spectrumaudio rec. for the Needle ).
Joe
since I can measure TS parameters myself, I found that product brochures are too often dictated by marketing dudes rather than R&D engineers.🙁
My values for the FR88EX are: Fres 106Hz, Qts 0,70, Vas 1,05 l.
Apart from that, very nice sound. I use them in 2l CB, supported by a bandpass subwoofer as PC speakers. Notch filter 0,33mH/8,2yF/4,7Ohm ( Spectrumaudio rec. for the Needle ).
Joe
DO keep in mind that the measurements you make & the (even hugely) different manufacturers T/S data can be correct.
The measure conditions of the factories software (LMS & LEAP) are usually quite different than us guys measuring with our computers.
This was rammed home when Mark Fenlon measured T/S of a set of FE127eN i sent him. My numbers were way off factory, his -- measured with factory kit -- were almost bang on the factory numbers. (i have posted those numbers in a number of posts, i'm in the field right now, so don't have access to them)
One must realize that T/S are not scalars but curves. How close our numbers come vrs factory numbers is usually more an indication of how flat the T/S curves are.
dave
The measure conditions of the factories software (LMS & LEAP) are usually quite different than us guys measuring with our computers.
This was rammed home when Mark Fenlon measured T/S of a set of FE127eN i sent him. My numbers were way off factory, his -- measured with factory kit -- were almost bang on the factory numbers. (i have posted those numbers in a number of posts, i'm in the field right now, so don't have access to them)
One must realize that T/S are not scalars but curves. How close our numbers come vrs factory numbers is usually more an indication of how flat the T/S curves are.
dave
Member
Joined 2009
Paid Member
Hi Joe,
What is a "CB" - is this closed box ?
why the notch filter, the published curves don't seem to suggest the need ?
Dave, thanks for updating your thoughts on the T/S, I remember you were concerned about my Moon Onken speakers being based on factory 127E spec - it's good to know that this is a non-issue.
Member
Joined 2009
Paid Member
I think you had measured a bunch of 127's and found the Fs was higher than the spec. sheet. I'm not sure what else was different. Still, it's all now irrelevant with the demise of the 127.
Yes, closed box. Well, the published curves are measured with the speaker in an IEC baffle ( I guess).
Spectrumaudio, a german expert for FR speakers, suggests this notch filter for the "Needle", baffle width 12,5cm (5"). (Value of the resistor should be smaller when placed directly in front of a wall...)
PS: with small 3" FRs, I always get higher Fres. Maybe because I measured them "free-air". With W5-876 woofers I got exactly the factory values, so my measuring equipment ( Bruel&Kjaer 😎) seems to be OK...
Joe
Member
Joined 2009
Paid Member
Well I'm too lazy to try and measure any of the parameters, I'm going to trust the spec. for this first build. I believe the final result will not be too sensitive to the driver parameters because it's not a butterworth vented box alignment rather something closer to the CB.
Dave - your data shows quite a spread, suggests that box designs should be tolerant to this kind of thing.
Interestingly, the box they came in includes a printed statement that this is a matched pair, which I wonder out loud as it doesn't seem very economical for the manufacturer to do this unless the matching is simply within their 'factory spec' for all drivers.
Dave - your data shows quite a spread, suggests that box designs should be tolerant to this kind of thing.
Interestingly, the box they came in includes a printed statement that this is a matched pair, which I wonder out loud as it doesn't seem very economical for the manufacturer to do this unless the matching is simply within their 'factory spec' for all drivers.
Last edited:
- Status
- Not open for further replies.