Newtonian adjustanator
If you hold the drivers in you hands while playing sub-200Hz tones and press the magnet covers back to back you'll easily hear that coupling them mechanically, clearly reduces harmonic resonances of the baskets themselves. In free air, it sounds like awful distortion. While playing with finding the pressure needed to accomplish this, I've found that its somewhere between 1/2 and 2 lbs of pressure, and definately needs to be dialed in.
In other words, I don't think you can reliably just cut a dowel close to size and glue it in with a removable caulk.
I have some circular pieces left over from three inch driver cutouts about the same size as the magnets, these make a nice block to screw to a dowel or short 2"x2". The other end of the dowel gets predrilled and fitted with a T-nut fastener. You can then use a threaded rod and lock nuts to adjust the pressure in your test fit.
If you don't want to deal with cutting and filing a threaded rod, you can use anchor fasteners (sometimes called Threaded dowels) these are bolts with no heads, lag screw thread on one end (directly into the dowel) and machine thread on the other.(Into the T-nut)
To make this assembly semi-permanent/removable I'm thinking silicone to one driver and ductseal to the other. Any thoughts?
And has anyone worked out a final measured drawing of their preferred "reference" for our common group build?
If you hold the drivers in you hands while playing sub-200Hz tones and press the magnet covers back to back you'll easily hear that coupling them mechanically, clearly reduces harmonic resonances of the baskets themselves. In free air, it sounds like awful distortion. While playing with finding the pressure needed to accomplish this, I've found that its somewhere between 1/2 and 2 lbs of pressure, and definately needs to be dialed in.
In other words, I don't think you can reliably just cut a dowel close to size and glue it in with a removable caulk.
I have some circular pieces left over from three inch driver cutouts about the same size as the magnets, these make a nice block to screw to a dowel or short 2"x2". The other end of the dowel gets predrilled and fitted with a T-nut fastener. You can then use a threaded rod and lock nuts to adjust the pressure in your test fit.
If you don't want to deal with cutting and filing a threaded rod, you can use anchor fasteners (sometimes called Threaded dowels) these are bolts with no heads, lag screw thread on one end (directly into the dowel) and machine thread on the other.(Into the T-nut)
To make this assembly semi-permanent/removable I'm thinking silicone to one driver and ductseal to the other. Any thoughts?
And has anyone worked out a final measured drawing of their preferred "reference" for our common group build?
Mono-pole vs Bi-pole
I finally got some time to play with v0.2 of the design and the MJK MathCAD software some more.
The mono-pole, bi-pole parallel and bi-pole series have essentially the same frequency response, with the exception that the parallel is about +6db in sensitivity.
I was supprised to see that the mono-pole and bi-pole series have the same response. I would have expected more bass out of the bi-pole. Is there any reason why this is the case?
For tube amps, I like the idea of an extra 6dB that you get with the parallel connection. However, I don't know if most tubes amps can handle the very low impedance (<3.5 ohm!)
Cheers,
Gio.
I finally got some time to play with v0.2 of the design and the MJK MathCAD software some more.
The mono-pole, bi-pole parallel and bi-pole series have essentially the same frequency response, with the exception that the parallel is about +6db in sensitivity.
I was supprised to see that the mono-pole and bi-pole series have the same response. I would have expected more bass out of the bi-pole. Is there any reason why this is the case?
For tube amps, I like the idea of an extra 6dB that you get with the parallel connection. However, I don't know if most tubes amps can handle the very low impedance (<3.5 ohm!)
Cheers,
Gio.
1. Regardless the number of drivers the Fs is still the same.
2. The worksheets do not account for baffle step.
3. Most amps, tube or SS will perform better with a higher load. Tube amps tend to be more sensitive to load then SS. I found that having two FE103E in series sounded cleaner than parallel even when driven by the little S5 amp which is rated for 4 or 8 ohm loads by the designer.
2. The worksheets do not account for baffle step.
3. Most amps, tube or SS will perform better with a higher load. Tube amps tend to be more sensitive to load then SS. I found that having two FE103E in series sounded cleaner than parallel even when driven by the little S5 amp which is rated for 4 or 8 ohm loads by the designer.
Re: Mono-pole vs Bi-pole
GM -- you need to shake the brain-washing...
For a tube amp with proper taps you will get the 3 dB of extra sensitivity whether you connect in series or parallel... 6 dB is an artifact of MJK assumming everyone is using a SS amp with high damping (1/2 of the 6 dB gain from parallel is from assumming that the SS amp doubles its power into 4 ohms and -3 dB in the case of the series connection comes from the SS amp putting out half as much power into 16 ohms. (ie the 6 dB difference between parallel & series is purely that into 4 ohms the amp is providing 2 W & into 16 ohms a 1/2 W -- this is a completely false assumption if you have a tube amp)
dave
gmilitano said:
GM -- you need to shake the brain-washing...
For a tube amp with proper taps you will get the 3 dB of extra sensitivity whether you connect in series or parallel... 6 dB is an artifact of MJK assumming everyone is using a SS amp with high damping (1/2 of the 6 dB gain from parallel is from assumming that the SS amp doubles its power into 4 ohms and -3 dB in the case of the series connection comes from the SS amp putting out half as much power into 16 ohms. (ie the 6 dB difference between parallel & series is purely that into 4 ohms the amp is providing 2 W & into 16 ohms a 1/2 W -- this is a completely false assumption if you have a tube amp)
dave
Timn8ter said:
I went through the design some more and read over the MJK two driver doccument. The actual reason may be that each driver in a bi-pole system has 1/2 the displacement of the driver in the mono-pole version.
In going through the mono-pole and bi-pole design, I discovered that the volume of port on the bi-pole is not double that of the mono-pole design.
The bi-pole port radius should be SQRT(2 * 0.75) = 1.225". I have attached a MJK mathCAD simmulation of the larger port.
I don't know if a 1.25" port can be attained, therefore, one may have to use two 0.75" ports. The two ports would work nicely out the bottom. A while back, Dave had a nice plan for the 40-1197.
http://www.diyaudio.com/forums/showthread.php?postid=390693#post390693
Perhaps we can incorporate this stand in the design to get the desired driver height and the twin ports out the bottom.
Regards,
Gio.
Of course you mean a 2.5" diameter port or two 1.25" diameter ports, don't you? Actually you should be increasing the volume of the port rather than just the diameter. In the dims I gave to SCD I suggested a 2" diameter port.
If a driver is capable of a theoretical LF output limit of 50Hz that is all it will produce regardless whether there is one, or a dozen. With an increase in the number of drivers you will experience a corresponding increase in SPL but not lower frequency response.
Timn8ter said:
Yes and no. The mono-pole uses a port with radius of 0.75", diameter of 1.5". For the bi-pole, the port diameter should be the SQRT (0.75 * 2), which is 1.225". The bi-pole should use either a single port with radius of 1.25", diameter of 2.5" or use two of the ports used in the mono-pole design (two 1.5" diameter by 2" length).
I am thinking that the bi-pole should be double the mono-pole. I don't see why the volume of the enclosure would be exactly double, but not the volume of the port.
Gio.
Re: Re: Mono-pole vs Bi-pole
I hope I am not being brain washed! If so, here is your turn!
Ok, I think I got most of your comments.
What do you mean by a "tube amp with proper taps"?
To summarize what you have said, if one uses a tube amp to drive this bi-pole, the sensitivity will be about 93dB regardless of weather it is wired series or parallel.
Cheers,
Gio.
planet10 said:
GM -- you need to shake the brain-washing...
For a tube amp with proper taps you will get the 3 dB of extra sensitivity whether you connect in series or parallel... 6 dB is an artifact of MJK assumming everyone is using a SS amp with high damping (1/2 of the 6 dB gain from parallel is from assumming that the SS amp doubles its power into 4 ohms and -3 dB in the case of the series connection comes from the SS amp putting out half as much power into 16 ohms. (ie the 6 dB difference between parallel & series is purely that into 4 ohms the amp is providing 2 W & into 16 ohms a 1/2 W -- this is a completely false assumption if you have a tube amp)
dave
I hope I am not being brain washed! If so, here is your turn!
Ok, I think I got most of your comments.
What do you mean by a "tube amp with proper taps"?
To summarize what you have said, if one uses a tube amp to drive this bi-pole, the sensitivity will be about 93dB regardless of weather it is wired series or parallel.
Cheers,
Gio.
When Tim specified the bi-pole he gave a 2" port... for the monopole i divided the area in half, it came out to 1.4x something diameter so i just rounded it to 1.5" since that is a standrad size in NA... it may need to be a tad longer because of that approximation.
Until someone actually builds them, the port size is really just a close approximation.
dave
Until someone actually builds them, the port size is really just a close approximation.
dave
Re: Re: Re: Mono-pole vs Bi-pole
Tube amps have taps on there output transformers, typically at least 4 & 8 ohm, or 8 & 16 ohm (or 5 & 10 ohm on some older German OPTs). The idea is to connect the speaker to the one that is closest to your speaker. Now if your speaker doesn't correspond to the right tap (ie a 16 ohm speaker on an 8 ohm tap) you will get a little less power (in the case of the example shown usually a little less distortion also), but in the end the power output differences between taps is usually not all that great, and the best way to determine which one to use is the one that sounds best.
Yes... real sensitivity of 2 speakers in parallel or series will be 3 dB greater than a single one. And tube amps (and most SS amps for that matter) are happier with higher impedance (it is just that the SS amp typically puts out less maximum power -- ie it has a relatively more robust power supply just by changing speaker impedance)
dave
gmilitano said:
Tube amps have taps on there output transformers, typically at least 4 & 8 ohm, or 8 & 16 ohm (or 5 & 10 ohm on some older German OPTs). The idea is to connect the speaker to the one that is closest to your speaker. Now if your speaker doesn't correspond to the right tap (ie a 16 ohm speaker on an 8 ohm tap) you will get a little less power (in the case of the example shown usually a little less distortion also), but in the end the power output differences between taps is usually not all that great, and the best way to determine which one to use is the one that sounds best.
Yes... real sensitivity of 2 speakers in parallel or series will be 3 dB greater than a single one. And tube amps (and most SS amps for that matter) are happier with higher impedance (it is just that the SS amp typically puts out less maximum power -- ie it has a relatively more robust power supply just by changing speaker impedance)
dave
I have two questions:
First, is the purpose of this reference speaker simply for doing A/B test with other speakers. Or to provide a flat full range response?
Personally i would prefer a reference speaker with a flat response to compare others to, rather than a speaker modelled to sound 'good'..
Secondly, does anyone know where to get Fostex drivers in New Zealand?
Thanks guys, keep up the good work
Matt
First, is the purpose of this reference speaker simply for doing A/B test with other speakers. Or to provide a flat full range response?
Personally i would prefer a reference speaker with a flat response to compare others to, rather than a speaker modelled to sound 'good'..
Secondly, does anyone know where to get Fostex drivers in New Zealand?
Thanks guys, keep up the good work
Matt
planet10 said:
I think i may give these a go next weekend...got some time on my hand...now this is basically my first DIY home audio project. Sence it looks "cute" according the the boss....i got the approval..so with that...i had one question...would anyone have any recommendations on stuffing...as in how much and where exactly.
thank you
Greetings full rangers,
I would like to follow up on Matt's (optical) first question. How are you using the word reference?
By the dictionary definition of the word "reference," you don't have to have flat response. Still, it ought to be a known quantity or quality.
This is, however, different from the way "reference" has been used in ad copy to refer to certain loudspeakers. There the reference to reference has implied superior performance (flatness or whatever).
This brings me to the driver selected by the vote of I believe seven people. One of the voters made the comment about a preference for shielded if the shielded and unshielded versions had the same performance. I have not tested the FE126E or the FE127E. I can only comment on what is listed in the spec sheets. From the spec sheets, however, the drivers are not the same. The motors (magnets and BL) are significantly different. The 126 uses a 440 gram ferrite magnet, the 127 uses a 100 gram ferrite magnet. The 126 has a specified BL of 5.92, the 127 is only 4.14. The 126 is rated at 93 db and the 127 is rated at 91 db.
The published frequency responses of the two drivers are also different. The 126 appears to be up as much as 3 or 4 db between 500 and 7 kHz above the 2 db lesser sensitivity of the 127.
In addition, even Fostex's rather smoothed response graphs show two drivers that do not meet rated response bandwidth even with a plus or minus five db range. For example, both drivers' output is falling fast above 15 to 17 kHz. The two drivers are also significantly different in output between 500 and 7 kHz.
As I already stated, I do not have access to either the 126 or the 127. I do have access to the FE103E. I do know how an engineering quality frequency response test of the FE103E compares with the Fostex published response of the FE103E. While I will not know for certain unless I test, I have little reason to expect that Fostex changed their test set up just for the 126 and 127 drivers. (As an aside, I love the way Fostex changes the vertical sizes of their published graphs. Makes it so much easier to compare frequency response differences from driver to driver.)
Now, I have a one-time proposition for the thread and the thread contributor who owns a company that sells the FE127E. If you will donate a driver, I will test the driver and publish the response in this thread. In addition, I will build the monopole TL box and test the box and the driver.
I do have some experience testing TL designs. I published a couple of reviews of Fried's TL loudspeakers during the 1980s. And I still have the data in a digital format that I can access to use as a TL reference. I also have a reference four-inch driver that I can compare to the 127 in the exact same test set up.
Good designing and good building,
Mark
I would like to follow up on Matt's (optical) first question. How are you using the word reference?
By the dictionary definition of the word "reference," you don't have to have flat response. Still, it ought to be a known quantity or quality.
This is, however, different from the way "reference" has been used in ad copy to refer to certain loudspeakers. There the reference to reference has implied superior performance (flatness or whatever).
This brings me to the driver selected by the vote of I believe seven people. One of the voters made the comment about a preference for shielded if the shielded and unshielded versions had the same performance. I have not tested the FE126E or the FE127E. I can only comment on what is listed in the spec sheets. From the spec sheets, however, the drivers are not the same. The motors (magnets and BL) are significantly different. The 126 uses a 440 gram ferrite magnet, the 127 uses a 100 gram ferrite magnet. The 126 has a specified BL of 5.92, the 127 is only 4.14. The 126 is rated at 93 db and the 127 is rated at 91 db.
The published frequency responses of the two drivers are also different. The 126 appears to be up as much as 3 or 4 db between 500 and 7 kHz above the 2 db lesser sensitivity of the 127.
In addition, even Fostex's rather smoothed response graphs show two drivers that do not meet rated response bandwidth even with a plus or minus five db range. For example, both drivers' output is falling fast above 15 to 17 kHz. The two drivers are also significantly different in output between 500 and 7 kHz.
As I already stated, I do not have access to either the 126 or the 127. I do have access to the FE103E. I do know how an engineering quality frequency response test of the FE103E compares with the Fostex published response of the FE103E. While I will not know for certain unless I test, I have little reason to expect that Fostex changed their test set up just for the 126 and 127 drivers. (As an aside, I love the way Fostex changes the vertical sizes of their published graphs. Makes it so much easier to compare frequency response differences from driver to driver.)
Now, I have a one-time proposition for the thread and the thread contributor who owns a company that sells the FE127E. If you will donate a driver, I will test the driver and publish the response in this thread. In addition, I will build the monopole TL box and test the box and the driver.
I do have some experience testing TL designs. I published a couple of reviews of Fried's TL loudspeakers during the 1980s. And I still have the data in a digital format that I can access to use as a TL reference. I also have a reference four-inch driver that I can compare to the 127 in the exact same test set up.
Good designing and good building,
Mark
