Hi,
As the title says, I'm planing to build a pair of cheap fullrangers for really not-so-critical listening (background music, radio, TV, computer etc).
This cheap fullrange driver i'm interested in is a four inch driver made by "Right", and the model "FR100-8". Of course, the expected quality for this driver is propably very low, but so is the price too. I could buy these from my work (electronics repair&service shop) at price of about 5 euros a piece. So getting a couple of these just for testing and tweaking wouldn't really hurt my wallet.
I modelled the driver in MJK's "TL Offset Driver.mcd". As you can see, there are some dips and peaks in the SPL response.
Because I don't have good knowledge of transmission line design, I wonder if it could be improved? Perhaps the Qts&Qes values are too high, or TL parameters I'm using are not good?
The advertised specs:
Qts: 0,75
Qes: 1,00
Qms: 2,966
Fs: 107 Hz
Vas: 2,84 liters
TL parameters:
Line length: 35 inch
Driver position ratio: 9 inch
S0: 6 x Sd
SL: 1,75 x Sd
Density: 0,2 lb/ft
Please post comments, feedback is welcome!
As the title says, I'm planing to build a pair of cheap fullrangers for really not-so-critical listening (background music, radio, TV, computer etc).
This cheap fullrange driver i'm interested in is a four inch driver made by "Right", and the model "FR100-8". Of course, the expected quality for this driver is propably very low, but so is the price too. I could buy these from my work (electronics repair&service shop) at price of about 5 euros a piece. So getting a couple of these just for testing and tweaking wouldn't really hurt my wallet.
I modelled the driver in MJK's "TL Offset Driver.mcd". As you can see, there are some dips and peaks in the SPL response.
Because I don't have good knowledge of transmission line design, I wonder if it could be improved? Perhaps the Qts&Qes values are too high, or TL parameters I'm using are not good?
The advertised specs:
Qts: 0,75
Qes: 1,00
Qms: 2,966
Fs: 107 Hz
Vas: 2,84 liters
TL parameters:
Line length: 35 inch
Driver position ratio: 9 inch
S0: 6 x Sd
SL: 1,75 x Sd
Density: 0,2 lb/ft
Please post comments, feedback is welcome!
Attachments
With an Fs of 107Hz a 35" line is going to be too long. For example, the Tangband W3-871S with an advertised Fs of 110 Hz works best in ~16" line. Tuning a cabinet below the driver Fs is not beneficial unless you plan on adding heavy EQ. Also, there are some numbers missing such as port radius, length and position and it's going to be difficult getting meaningful models without the driver BL and Re measurements. Overall, I don't think this is a good candidate for a TL but as you pointed out, the price was low enough to take a chance.
Timn8ter,
This is a very puzzling subject for me. I have modeled, using the MJK worksheets, the TB 881s to have a flat FR far below its advertised (and measured) Fs of 111Hz. What should I beware of before I construct the prototype?
Or, why should I expect to use heavy EQ when my predictors demonstrate flat FR?
I'm pursuing this as a personal scientific experiment, but it would be nice to add some more data to my simulations, if you have something to offer in that regard.
Thanks,
Dave
This is a very puzzling subject for me. I have modeled, using the MJK worksheets, the TB 881s to have a flat FR far below its advertised (and measured) Fs of 111Hz. What should I beware of before I construct the prototype?
Or, why should I expect to use heavy EQ when my predictors demonstrate flat FR?
I'm pursuing this as a personal scientific experiment, but it would be nice to add some more data to my simulations, if you have something to offer in that regard.
Thanks,
Dave
To be brief I'll just quote Rick Schultz from his 08/03 AudioXpress article.
"TL myth assumes you can lower system response by building a longer pipe; pushing pipe frequency lower must extend bass performance. This myth started 65 years agoin when Voigt set "length just under one quarter of the lowest frequency at which efficient working is desired." He does not relate quarter-wave pipe frequency to driver resonance. Instead, he related it to whatever frequency "is desired."
Speaker builders have assumed just building a longer pipe will extend response. In fact, building a lower frequency pipe just quiets it. Longer TLs perform much worse than shorter TLs. The lower limit for pipe frequency should equal Fs. I recommend going no lower, any longer."
"TL myth assumes you can lower system response by building a longer pipe; pushing pipe frequency lower must extend bass performance. This myth started 65 years agoin when Voigt set "length just under one quarter of the lowest frequency at which efficient working is desired." He does not relate quarter-wave pipe frequency to driver resonance. Instead, he related it to whatever frequency "is desired."
Speaker builders have assumed just building a longer pipe will extend response. In fact, building a lower frequency pipe just quiets it. Longer TLs perform much worse than shorter TLs. The lower limit for pipe frequency should equal Fs. I recommend going no lower, any longer."
I'm still evaluating the article, so the following comments are a little premature.
I'm still unclear what you meant by "adding heavy EQ." Were you referring to an external EQ unit, passive or active, or were you referring to stuffing as an EQ method?
Oh, and I guess when we're talking about "TL," there's a current problem of referent. What I've modeled, and am about to build, is a variation of the MJK ML-TQWT. I consider that a TL in the broader sense. That's my referent.
Thanks,
Dave
I'm still unclear what you meant by "adding heavy EQ." Were you referring to an external EQ unit, passive or active, or were you referring to stuffing as an EQ method?
Oh, and I guess when we're talking about "TL," there's a current problem of referent. What I've modeled, and am about to build, is a variation of the MJK ML-TQWT. I consider that a TL in the broader sense. That's my referent.
Thanks,
Dave
When I said EQ I was thinking Linkwitz Transform.
I think Martin's worksheet allows for lower driver resonance due to the air mass inside the enclosure loading the cone, however, lower Fs by mass loading changes the Q and thereby the output of the driver, so, are you really gaining anything? My personal experience is that keeping Fb equal to or above Fs is going to yield the best results.
In reference to a 35" long pipe for a driver with a Fs of 107Hz.....well...
I think Martin's worksheet allows for lower driver resonance due to the air mass inside the enclosure loading the cone, however, lower Fs by mass loading changes the Q and thereby the output of the driver, so, are you really gaining anything? My personal experience is that keeping Fb equal to or above Fs is going to yield the best results.
In reference to a 35" long pipe for a driver with a Fs of 107Hz.....well...
Greets!
What are all the specs you are using? Do you know Re, Le, Sd? The IB response (dashed line) is totally whacked for the specs you published so your sim is useless.
Sd = 45.6064 cm^2 (assumes 3" effective dia.)
Re = 8 ohms
Le = 0.6 mH
BL = 3.512 N/A
pipe Vb = ~0.71 ft^3 per T/S
I get:
L = 25.25"
eta = 9.42"
SO = 90.48"^2 (~12.8*Sd)
SL = 6.05"^2 (~0.856*Sd)
density = 0.25
WRT gain, pipe Vb must increase with decreasing Fp (length). When it comes to loading high Q drivers and/or loading a driver below Fs, Vb (air mass) is the acoustic lever required to help a 'weak' driver load the terminus. Voigt understood this. If you don't provide it then RS's statement is at least semi-correct. It's all a matter of what the performance goals/acceptable trade-offs are.
GM
What are all the specs you are using? Do you know Re, Le, Sd? The IB response (dashed line) is totally whacked for the specs you published so your sim is useless.
I disagree! Longer ones are inherently better damped, and with less stuffing density, a 'good thing' when it's a FR driver. That said, 35" is way too long for a reverse tapered pipe since this driver really shouldn't be tuned <~59.3Hz according to T/S. Just running the numbers using:
Sd = 45.6064 cm^2 (assumes 3" effective dia.)
Re = 8 ohms
Le = 0.6 mH
BL = 3.512 N/A
pipe Vb = ~0.71 ft^3 per T/S
I get:
L = 25.25"
eta = 9.42"
SO = 90.48"^2 (~12.8*Sd)
SL = 6.05"^2 (~0.856*Sd)
density = 0.25
I think you're right.
WRT gain, pipe Vb must increase with decreasing Fp (length). When it comes to loading high Q drivers and/or loading a driver below Fs, Vb (air mass) is the acoustic lever required to help a 'weak' driver load the terminus. Voigt understood this. If you don't provide it then RS's statement is at least semi-correct. It's all a matter of what the performance goals/acceptable trade-offs are.GM
Hello all,
Unfortunately, I don't have a clue about these parameters (Re, Le, Sd). The store that sells them neither has any more info. They're very cheap drivers. About 4 euros.
So I had to just guess something: I used like 50 cm2 for Sd, 7,0 ohms for Re, 0,2 mH for Le and 3,0 for BL.
Those two drivers arrive to me on next week, so I'll know at least Re and Sd then.
It's really bad if not evil thing to guess parameters, but these drivers cost only four euros a piece and my primary goal here is to make a pair of speakers for home background music and computer applications. So I don't expect maximum hi-fi sound quality of them anyway...
Looks like I still have much to learn when it comes to designing loudspeakers! There was large peak at around ~100 Hz in my design, and I thought I couldn't make it any smoother easily. That's why I just kept increasing and increasing the line length until the response had weird shape... Hopefully I learnt a lesson here?
I certainly agree: compared to my TL attempt, the response of GM's design looks really better in a shorter line length... Thank you for your help! I will post the SPL response picture of your design here.
Unfortunately, I don't have a clue about these parameters (Re, Le, Sd). The store that sells them neither has any more info. They're very cheap drivers. About 4 euros.
So I had to just guess something: I used like 50 cm2 for Sd, 7,0 ohms for Re, 0,2 mH for Le and 3,0 for BL.
Those two drivers arrive to me on next week, so I'll know at least Re and Sd then.
It's really bad if not evil thing to guess parameters, but these drivers cost only four euros a piece and my primary goal here is to make a pair of speakers for home background music and computer applications. So I don't expect maximum hi-fi sound quality of them anyway...
Looks like I still have much to learn when it comes to designing loudspeakers! There was large peak at around ~100 Hz in my design, and I thought I couldn't make it any smoother easily. That's why I just kept increasing and increasing the line length until the response had weird shape... Hopefully I learnt a lesson here?
I certainly agree: compared to my TL attempt, the response of GM's design looks really better in a shorter line length... Thank you for your help!
I will post the SPL response picture of your design here.Attachments
full range
Hello, I have been building tl's for many years and have found that drivers with a qts .39-.44 are about the limit unless you are willing to experiment with line stuffing alot. Also as qts rises above .4 a line taper 1.25sd to .8- .9 seem to filter a lot of the resonance. You will need to use less stuffing in my experience with higher taper ratios therefore also increasing efficency. I have recently built one with tang band 656 and doesn't use much stuffing all. Woofers under 5" also benefit from much less damping.
Corey
Hello, I have been building tl's for many years and have found that drivers with a qts .39-.44 are about the limit unless you are willing to experiment with line stuffing alot. Also as qts rises above .4 a line taper 1.25sd to .8- .9 seem to filter a lot of the resonance. You will need to use less stuffing in my experience with higher taper ratios therefore also increasing efficency. I have recently built one with tang band 656 and doesn't use much stuffing all. Woofers under 5" also benefit from much less damping.
Corey
Greets!
OK, these aren't required to do basic T/S box designs, but certain assumptions can be made for more elaborate programs such as MJK's WSs. 'Re' can be a common one like 7 ohms for a nominal 8 ohm driver, or for that matter you can just input 8 ohm, same for 'Le' or input zero. 'Sd' can be a typical value from another driver of the same frame size. 'BL' is another matter altogether since its a relative number based on several specs and what was ruining your plot since the number you used didn't match up with the Vas, Sd, Re, Qes you inputted, so you either need to use a program that will calc it for you (I believe the WINISD Pro freeware will, among others) or do like me and load this formula into a spreadsheet to calc it:
BL = SQRT((rho*(c*0.0001)^2*Sd^2*Re)/((2*PI*Fs)*(Vas*0.001)*Qes))
where for MJK's WSs:
rho = 1.21
c = 342
While the FR may not be the most accurate, it will be 'close enough' to decide if you want to build it or not.
You're welcome! Hope so!
GM
OK, these aren't required to do basic T/S box designs, but certain assumptions can be made for more elaborate programs such as MJK's WSs. 'Re' can be a common one like 7 ohms for a nominal 8 ohm driver, or for that matter you can just input 8 ohm, same for 'Le' or input zero. 'Sd' can be a typical value from another driver of the same frame size. 'BL' is another matter altogether since its a relative number based on several specs and what was ruining your plot since the number you used didn't match up with the Vas, Sd, Re, Qes you inputted, so you either need to use a program that will calc it for you (I believe the WINISD Pro freeware will, among others) or do like me and load this formula into a spreadsheet to calc it:
BL = SQRT((rho*(c*0.0001)^2*Sd^2*Re)/((2*PI*Fs)*(Vas*0.001)*Qes))
where for MJK's WSs:
rho = 1.21
c = 342
While the FR may not be the most accurate, it will be 'close enough' to decide if you want to build it or not.
You're welcome! Hope so!
GM
Hi,
These two fullrange drivers arrived to me. First thing i did was measure Re with multimeter and Sd with a ruler. Got values of 7,6 ohm and about 55 cm2. These drivers have foam surrounds and about 5 mm of one-way cone travel.
I tried to calculate BL with WinISD, and got a value of 5,6. I changed the new measured/calculated values to "TL Offset.mcd" and it showed different results. Mainly a bit deeper bass extension, but with some lower sensitivity.
I found some ~75cm long cardboard tubes with 12 cm diameter and 2 cm wall thickness at our garage. I decided to use those as testing-T-Lines. Sound is really "honky" because there's is no any amount of stuffing inside...
Afterall, I think that the overall sound of these is quite bad. I'm no expert, but I think the upper midrange/lower treble sounds too loud and sounds like those "1500W PMPO" plastic computer boomboxes. My extremely bad "enclosure" setup might also have something to do with this.
But again considering it's 4 euros... I attached a "david/goliath" pic of it
These two fullrange drivers arrived to me. First thing i did was measure Re with multimeter and Sd with a ruler. Got values of 7,6 ohm and about 55 cm2. These drivers have foam surrounds and about 5 mm of one-way cone travel.
I tried to calculate BL with WinISD, and got a value of 5,6. I changed the new measured/calculated values to "TL Offset.mcd" and it showed different results. Mainly a bit deeper bass extension, but with some lower sensitivity.
I found some ~75cm long cardboard tubes with 12 cm diameter and 2 cm wall thickness at our garage. I decided to use those as testing-T-Lines. Sound is really "honky" because there's is no any amount of stuffing inside...
Afterall, I think that the overall sound of these is quite bad. I'm no expert, but I think the upper midrange/lower treble sounds too loud and sounds like those "1500W PMPO" plastic computer boomboxes. My extremely bad "enclosure" setup might also have something to do with this.
But again considering it's 4 euros... I attached a "david/goliath" pic of it
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.