Another Desktop Speaker Project (FE103E)

Status
Not open for further replies.
I got the FE103s today and I'm trying to measure the Thiele-Small parameters.

I'm not just trying to build a set of speakers, but I'm trying to understand the theory behind the design.

I'm using Audio Tester V2.2 to try to measure the parameters. If I add the 20gm of clay and de-Vas via cabinet I get positive numbers. If I select Vas via cabinet I get a negative number for Vas and 0 for the other parameters.

Some of the results I'm getting appear in the right range (Re, Fs,0, Qms) but others like Vas appear to me to be wrong.

An externally hosted image should be here but it was not working when we last tested it.


Can anyone shed some light on this?

Here is the phase and impedance plot. I ran it on two speakers and got very similar results.

An externally hosted image should be here but it was not working when we last tested it.


Thanks.

Steven
 
planet10 said:
Geddes' research shows a threshold of something like 25%. ie it has to be pretty bad to concern yourself about when attention could be be elsewhere more important.

In his book he has a chapter on distortion. Now after his research he says not to bother reading it...

dave
Click to expand...

Doesn't this depend greatly on which harmonics we are talking about. Odd in particular are more noticeable, and higher order harmonics tend to be more harsh?
 
How i interperet Geddes results is that THD measures are measuring the wrong thing... the nature of a dynamic speaker usually means very little above 3rd and monotonic.

dave
 
TheGimp said:
Doesn't this depend greatly on which harmonics we are talking about. Odd in particular are more noticeable, and higher order harmonics tend to be more harsh?
Click to expand...

I agree. 25% THD of odd harmonics would sound like a guitar fuzz pedal....actually 25% of even harmonics would sound like a fuzz pedal too...

25% THD...so a 3:1 ratio of sine to overtones...that only represents something like -6dB for the sum of the harmonics amplitudes. i dont get how that is possibly correct, otherwise i struggle to see how the ear could distinguish the phase problems of a 2 way speaker over a FR type when listening to harmonically rich instruments like sax. Obviously, a full fourier anaylses would need to be done to determine the full extent of the harmonics produced by any given driver, but being natural in nature, they would be less at 5th and 7th than at 3rd harmonics.

I would have believed 2.5% THD or even .25% THD, my experience with amps, although small, showed the onset of audible THD to be somewhere between 0.5 and 1%THD. I by no means have 'golden ears' but i do not have a sunshine halo syndrome afflicted ones either. 25% of THD may only just be annoying for those who blast an personal stereo at full volume(in the days beforee mp3), or for those that have pretty underpowered valve gear, but I CAN hear it way before that. I know this as i measured the THD and commented on audibilty for my thesis in E/EE. by no means is it absolute, as 2nd harmonic THD was not tested.

I think this is more of a case of how much THD is acceptable to the listener, and what type is acceptable (especially if it artificially extends the HF BW of a driver), and not a case of what is strictly audible. Whilst i respect Toole's work, his latest work HAS recieved some justified criticism, though not from me i might add.
 
I tried 5grams and still got way off numbers. Vas was 669.97.

I emailed Uli to see what he says to do, but was hoping someone here had already measured some 100mm speakers and could shed some light on the process.

Thanks.
 
sorry to jack the thread inadvertedly.

Personally i have never used the added weight method to determine Vas. From all i have learned and read, and heard from others, this method is not nearly as reliable. I use the 'test box' or known box method...I cant remember the real name for it. Im sure Planet10 will know it.

Measuring Loudspeaker Driver Parameters

the page shows how to do either test to measure Vas, although i recommend the 'known box' method.

sorry im not much help here.

Good luck
 
I don't know about added mass vrs known box. I use added mass.

I do know that the T/S generated are f(V, weather). If meaured using the same or similar kit the factory specs are reasonable... and the boxes designed with them seem to do what you expect.

The numbers i get are quite different. Attached is data i measured, and that Mark Fenlon measured on the same drivers.

Both sets are equally valid IME.

Someone might say that my measures must be done incorrectly, but my measures have been the "same" as those made by others (including Zaph & MJK)

dave
 

Attachments

  • Fenlon-FE127-measures.gif
    Fenlon-FE127-measures.gif
    7.2 KB · Views: 408
Compared to the measurements posted by dave, my measurements are ok, except Cms (off by 3 orders of magnitude) and Vas (off by two orders of magnatude).

How do you mount the speaker when taking these measurements? This might be a factor. I have the speaker setting face up on several towels to isolate it from the workbench. Clip leads are used to make the connections.

What altitude are you at? Since sound velocity effects the measurement this would be a factor. I'm at 1000' which I did not expect to have a great impact.
 
Mr. Müller recommended that I decrease the amount of clay until I got the amount which would shift the resonance down by 1/1.41, so I did that. I'm down to 3 grams of clay.

I get better numbers but with one speaker I still get a negative value for Vas.

I'm using a simple buffer amp made from a TL082 with a 2N2222A and a 2N2907 on the output to buffer the soundcard 0utput (Soundblaster Audigy).

The speaker is now suspended on a string about 1.5M from the ceiling.

I may try mounting the speaker on a tripod and setting it up outside where there is nothing nearby to provide an echo to the speaker to see if that will help.
 
[comment]Looks like i never did any drawings, the email to Chris says an mFonken with ports shortened to 103mm (i didn't see the syncronicity of that till just now, honest)
[/comment]
is that the internal port length to replace the 136.5mm from the fe127 version which would give a port opening of 45.5mm (against the 12mm of the fe127 version?)
 
captain pugwash said:
[comment]Looks like i never did any drawings, the email to Chris says an mFonken with ports shortened to 103mm (i didn't see the syncronicity of that till just now, honest)
[/comment]
is that the internal port length to replace the 136.5mm from the fe127 version which would give a port opening of 45.5mm (against the 12mm of the fe127 version?)
Click to expand...
 
sorry folks, still haven't got the hang of this quoting malarky, the above seemingly random question is based on this quote below from earlier;

"Looks like i never did any drawings, the email to Chris says an mFonken with ports shortened to 103mm (i didn't see the syncronicity of that till just now, honest)

dave"
 
captain pugwash said:
sorry folks, still haven't got the hang of this quoting malarky, the above seemingly random question is based on this quote below from earlier;

"Looks like i never did any drawings, the email to Chris says an mFonken with ports shortened to 103mm (i didn't see the syncronicity of that till just now, honest)

dave"
Click to expand...


well since he never got around to a drawing for this one, based on the above comment, it might look something like the attached (Dave - correct me if this is wrong)

notes:
1)drawn for 15mm material thickness - more than adequate for a driver of this horsepower- if anything, go for thinner ie. 12mm (1/2"), and retain the internal dimensions;
2) even at this size, decent plywood or solids like pine, cedar, VG fir, etc will sound much better than MDF;
3) parts drawn in red are optional (but recommended) bracing;
4) port spacer strips extend full depth as part of bracing scheme;
5) lightly damp walls with 1/2" or so natural felt if available, before buttoning up final side - avoid occluding the opening of vent slots
6) no apologies to imperialists for the metric dimensions 😉
 

Attachments

LOL! I too had the same SWIMBO requirements imposed on me. So built some microFonkens:
microFonken-tn.jpg

For a tone-deaf bloke these little puppies SING and bring vocals to life. Well worth & punchy little mites t'boot. Fantastic for the size. TY P10.

Andy

BTW, due to an ever increasing list of 'SWIMBO requirements' I am no longer with my wife and now I can bloody well build what I damn well like! So there.

Love, Me & Sue
An externally hosted image should be here but it was not working when we last tested it.
 
Fe103 Fonken

chrisb said:
well since he never got around to a drawing for this one, based on the above comment, it might look something like the attached (Dave - correct me if this is wrong)

notes:
1)drawn for 15mm material thickness - more than adequate for a driver of this horsepower- if anything, go for thinner ie. 12mm (1/2"), and retain the internal dimensions;
2) even at this size, decent plywood or solids like pine, cedar, VG fir, etc will sound much better than MDF;
3) parts drawn in red are optional (but recommended) bracing;
4) port spacer strips extend full depth as part of bracing scheme;
5) lightly damp walls with 1/2" or so natural felt if available, before buttoning up final side - avoid occluding the opening of vent slots
6) no apologies to imperialists for the metric dimensions 😉
Click to expand...

I only need cut the champfers to finish a pair of these cabinets. I fear I may have the speaker hole in the wrong location as I picked up the dimension from the cabinet bottom to the speaker center line from the original drawing for the FE 127. Any advice?
 
Status
Not open for further replies.
Back
Top