Help Modeling TQWL on Eton 5-880 driver

[jmillerdoc]

Hello,
I was hoping somebody here to crunch some numbers for me. I have looked at the different articles that are written by the supposed experts in TL theory and have decided I like the idea but I don't really want to try to figure out for myself how to actually do the calculations.....There are things I love to learn but..... after a little while of trying to figure out the computations myself I have chosen the easier path.....ask someone else to do it for you! I don't love it that much.

I really am interested to see how a properly designed TL would sound with high quality components. After reading about the great advantages of TL designs I am suprised there have been so few manufacturers that have implemented the technology to make off the shelf TL's.

Anyway, I would appreciate the help very much if somebody might be willing to help me model the proper TQW TL for the Eton 5-880. This is a fantastic little driver I have used in other projects and would like to see if it will work out in this application.
This driver has the following specs (cut and paste from Madisound.com)

Znom 8 ohm
Re 6.6 ohm
Le@1kHz 0.48 mH
fs 52 Hz
Qms 2.29
Qes 0.46
Qts 0.38
Mms 9.6 g
Cms - mm/N
Sd 82 cm2
BL 4.5 N/A
Vas 8.7 ltrs
Xmax 3 mm peak
VC Ø 25 mm
Sensitivity
1W / 1m 87 dB
Nom. Power DIN 50 W

Again, thanks to anybody willing to assist! I am not committed to any particular variation of TL's so if someone out there has any suggestions or alternate ideas, but still TL, I would be willing to entertain them.

Thanks again.
Sincerely,
Jeff Miller
lawton, OK

[GM]

These specs are way off in that BL is way too low for Qes or vice versa and Fs is a bit low based on Mms, Vas, so not sure which way to go. Did you use just the basic T/S specs (Fs, Vas, Qts) to get your other good performing speakers?

GM

[jmillerdoc]

Yep, used the basics....I am not familiar with "BL", maybe I am familiar with another term. Can you elaborate?

[GM]

Technically, it's B*L where 'B' is the VC gap magnetic flux and 'L' the length of wire in the gap, i.e. the motor's force factor that Qes is derived from, so if the two don't jive, then one has to change to get an accurate sim if both are used to create it. Since BL is measured (not Qes/Qts per se), it's considered the correct one to use.

In this driver's case it increases Qes/Qts significantly to 1.057/0.723 and what I would normally use, but depending on what your other designs are using a much lower Qts (Fs is only off a couple of Hz based on published Vas/Mms, so overall not a major consideration), what you and I consider high SQ may be just as far apart, so just trying to get a 'feel' for what sounds good to you.

IOW if I design a TL based on the basic specs and the higher Qts is more accurate, then it will be severely over-damped, sounding 'choked' with little/no mid-bass, LF without significant BSC whereas if I use the higher Qts, then it will be larger with a flatter response (less/no BSC required), but will be severely under-damped if the lower Qts is the most accurate. This at least yields more gain BW tuning flexibility, so short of measuring your driver's specs would be the way I would go.

GM

[jmillerdoc]

I put a query into Eton's engineering dept in Germany. I asked them to clarify this problem. I will post again once they have responded.
Many thanks,
Jeff

[jmillerdoc]

I received an email with the attached "correct" specs. Appearently they have mixed the specs up on several drivers in the .pdf file Spec sheets they release to their distributors. Purchasers of Eton products beware!!! You may need to check with them if you have their drivers. they assure me they are correcting this problem with their distributors.

The "correct spec" are as follows:

Technical Data Symbol Wert/Value
Nominal impedance Zn 8
DC resistance Re 6.6
Resonance frequency1 W fr 44
Resonance frequencyTS-parameter-meas. fr 58
Suspension compliance CMS 0.83
Mechanical Q QMS 2.44
Electrical Q QES 0.48
Total Q QTS 0.40
Mechanical resistance RMS 1.35
Total moving mass (including air mass) MMD 8.87
Effective piston area SD 80
Voice coil diameter 25
voice coil former Aluminium
Voice Coil length 12
Voice Coil layers 2
Voice Coil inductance Le 0.53
Flux density B 1.15
Force factor BL x l 6.64
Height of the gap 6
Diameter of magnet 90
Height of magnet 17
Mass of speaker
Characteristic sound pressure level 1 W1m 88
Rated power 50
Equivalent air volume of suspension VAS 7.49

[GM]

Originally Posted by jmillerdoc

Appearently they have mixed the specs up on several drivers in the .pdf file Spec sheets they release to their distributors.

Total moving mass (including air mass) MMD 8.87

Greets!

No comment.

Mmd is without air load, so this is the Mms spec.

No Xmax listed, but derived from VC length, magnet gap height it's nominally 3 mm.

Anyway, whether they are close enough to yours for ~accurate design remains to be seen, but at least they jive close enough from a mathematical POV now, so at a glance a TL tuned to Fs will need BSC to be tonally flat and a maximally flat alignment TL that will need little/no BSC will be tuned to around 143 Hz, so there's its practical tuning range.

What's it to be?

GM

[jmillerdoc]

I would like it tuned for the lowest possible bottom end if that is feasible. I know it is a small driver and such, probably not a very good one for a TL. Let me know what you can do. Thanks again for your help.
Sincerely,
Jeff

[GM]

You're welcome!

Well, as asked, the only correct answer is a TL equivalent of at least a 10x Vas sealed cab (IB) since it will load to Dc which means the TL will in theory need to be infinitely long. It would make for a life-long DIY speaker project though, keeping the Devil from having yet another pair of idle hands to generate mayhem........

Converted to a more realistic size TL with minimal performance trade-off, tuned to Fs has been the traditional solution with a high stuffing density used to smooth it out to mimic the IB's response down to Fs with a low Q roll off below it in a cab of 'only' ~3x Vas, so both will have a Qtc (Qp) ~ = driver Qts = over-damped = high F3 = no mid-bass/bass of note with a relatively high Fs, medium Qts driver and especially with a small, low Xmax one. Nowadays it would be tuned to whatever T/S specs dictate though, or ~57.56 Hz for this driver's published ones.

Anyway, here's three alignments based on various design trade-offs to choose from since from my POV the high tuned TL best meets the usable bottom end, but you may mean a near IB one or even somewhere in-between. Not that it matters much size wise since its acoustic net bulk is ~the same regardless of the one chosen with only its baffling/stuffing density changing.

All dims approximate, L = axial pipe length, CSA = cross sectional area of pipe, zdriver = location from the closed end and simmed stuffing density is based on using loose polyfil which according to MJK isn't accurate since it exceeds 1.0 lb/ft^3, so strictly a starting point for experimentation:

57.56 Hz Fp:

L = 56.437"
CSA = 31.937"^2
zdriver = 19.687"
stuffing density = 1.5 lbs/ft^3

91.51 Hz Fp:

L = 34.312"
CSA = 50.812"^2
zdriver = 11.937"
stuffing density = 1.75 lbs/ft^3

143.06 Hz Fp:

L = 20.500"
CSA = 79.437"^2
zdriver = 7.187"
stuffing density = 2.0 lbs/ft^3

Looking forward to a review plus final design details!

GM