I own a pair of Fostex FE 206 E drivers in BR boxes (2.5 mk II from David Dicks). Since there is almost no bass I plan to built a BSC filter proposed by Martin J. King and put it between the amp and the drivers. If I understand it right this circuit attenuates mids and highs for a few db allowing the bass to become alive.
On Decware site I read that removing the dust cap and instaling a steel phase plug improves and extends the hi frequences of the driver. I would like to give it a try but dont know how to do it. Is there anybody out there who actually did it (instaled a new phase plugs) ? to explain the easiest way to do it.
levon
On Decware site I read that removing the dust cap and instaling a steel phase plug improves and extends the hi frequences of the driver. I would like to give it a try but dont know how to do it. Is there anybody out there who actually did it (instaled a new phase plugs) ? to explain the easiest way to do it.
levon
I installed "phase plugs" in my FE206E. I also used a sharp hobby knife and removed the dust cap.
After the dust cap had been removed I carefully remove a little of the black glue to leave stright surface on the inside.
I used a socket wrench top as described at decware, but just placing the top made me realise that I proberly never gonna get them off again. The only reason for removing them would be to test with other types of socktes.
The top and upper midrange did improve don't know about bass or if the linear xmax has change since I use them at 250hz and up.
After the dust cap had been removed I carefully remove a little of the black glue to leave stright surface on the inside.
I used a socket wrench top as described at decware, but just placing the top made me realise that I proberly never gonna get them off again. The only reason for removing them would be to test with other types of socktes.
The top and upper midrange did improve don't know about bass or if the linear xmax has change since I use them at 250hz and up.
sreten beat me to the punch, his suggestion is a good one and does extend the high end response. You might want to try using a 1, 2, or 3 uF cap in parallel with the BSC resistor. I have done this with my Lowther ML TLs and it does provide more high frequency extension. You should try all three cap values, I settled on the 2 uF value for my drivers. I would probably try that treak before altering the drivers permanently.
For the Fostex FE-206E driver in my ML TL enclosures people have been using the same Zobel as I used and for a BSC circuit a 3 mH coil and a 10 ohm resistor. For your wider enclosures the value of the inductor will be different as calculated using my write up. The resistor value will probably be about 10 ohms, part of this value is for BSC and part is to raise the Qts of the driver, if you are using a solid state amp. If you are using a tube amp, with a low damping factor, the resistor value will probably be in the 6 to 8 ohm range.
Please let me know how you make out either in a private e-mail or in this thread. I am very interested in your results and how the circuit effects the sound balance of David Dicks enclosure design.
Hope that helps,
For the Fostex FE-206E driver in my ML TL enclosures people have been using the same Zobel as I used and for a BSC circuit a 3 mH coil and a 10 ohm resistor. For your wider enclosures the value of the inductor will be different as calculated using my write up. The resistor value will probably be about 10 ohms, part of this value is for BSC and part is to raise the Qts of the driver, if you are using a solid state amp. If you are using a tube amp, with a low damping factor, the resistor value will probably be in the 6 to 8 ohm range.
Please let me know how you make out either in a private e-mail or in this thread. I am very interested in your results and how the circuit effects the sound balance of David Dicks enclosure design.
Hope that helps,
I would start with either a 2.5 or 3.0 mH iron core inductor and a 10 ohm resistor (assuming a SS amp). My guess is that a 2 uF cap in parallel with the BSC circuit would help with the highs. You need a Zobel circuit for this to be predictable. People have been using a 3 uF cap in series with a 15 ohm resistor for the FE-206E drivers in my ML TL design. That would be my starting point.
The circuit can be adjusted to raise or lower the mids and high end. I have tried to summarize the steps I take in my BSC write-up on my site under the General Speaker Related Articles heading.
Hope that helps,
The circuit can be adjusted to raise or lower the mids and high end. I have tried to summarize the steps I take in my BSC write-up on my site under the General Speaker Related Articles heading.
Hope that helps,
Phase plugs work in two different ways.
By acting as a big lump in the middle of the cone they lessen high frequency distortion by absorbing and redirecting MF and HF that would otherwise bounce around inside the cone. This is often seen as peaks in measured response that can falsely be attributed to cone breakup. This type are usually soft material, like rubber
Secondly, they can act to partially horn load the HF to improve directivity and response. This type is usually hard, such as metal or wood.
By acting as a big lump in the middle of the cone they lessen high frequency distortion by absorbing and redirecting MF and HF that would otherwise bounce around inside the cone. This is often seen as peaks in measured response that can falsely be attributed to cone breakup. This type are usually soft material, like rubber
Secondly, they can act to partially horn load the HF to improve directivity and response. This type is usually hard, such as metal or wood.
MJK said:
MJK said:
Hi Martin,
I've read your BSC article and I'm afraid to say
you are giving confusing and conflicting advice.
Regarding BSC :
The standard circuit is R||L.
This creates a step down of midrange and treble. Due to
the drive units inductance there will be slow recovery of
some sensitivity towards to the top end of the driver.
Adding C, to give R||L||C will accentuate this recovery
back to the units extreme treble sensitivity.
Adding a Zobel to the driver for the R||L circuit will prevent
the general recovery. Consequently adding C to give R||L||C
+ Zobel will give a more controllable recovery consequently
its more tunable to recover extreme treble only.
However none of what is outlined above will affect Qts of
the driver and selecting R the way you describe does not
make sense.
The only selection that will affect Qts is the DCR of the inductor.
To affect Qts you need a Rs--(Rp||Lp) circuit, for the R||L above,
Rs is the DCR of the inductor, which can be added to by making
the total circuit :
(Rq--L)||R||C or Rq--(L||R||C) + with/without unit Zobel.
You need to clarify the basic BSC circuit, with the various options
of Zobel unit compensation and added C and their effects.
And then show how to incorporate Qts correction into the above,
and describe the consequent loss of overall sensitivity it causes.
http://www.mhsoft.nl/spk_calc.asp#newqts
Rq needs to take into account the inductor DCR.
sreten.Hi sreten,
"I've read your BSC article and I'm afraid to say
you are giving confusing and conflicting advice."
I am sorry you feel that way, but I am looking at things outside of the classical rules for BSC and Zobel circuits. If you are dealing with a standard bass reflex or closed box using a driver that fits in the standard T/S alignment tables then what I wrote in my BSC Sizing article should work very well. But in the case of the Fostex FE-206E, or the Lowther drivers I am currently using, the classic BSC circuit will provide a marginal improvement in the speaker's performance. This is due to the very low Qts and very high efficency of these types of drivers.
I have used a series resistance and a BCS circuit with these types of drivers and it worked OK. But the more I thought about it, the more I felt that the series resistance was not very effectiv and could be accomplished in an easier way. I will outline my thought process below.
1. By adding the Zobel I have turned the speaker into a pure resistance above about 100 Hz. Below 100 Hz the impedance peak(s) due to the closed or ported box dominate the impedance curve. So I mentally divid the impedance curve into two region with about 100 Hz as the boundary.
2. Any series resistance (typically a couple of ohms) that is added will have a minimal effect below 100 Hz because the the impedance of the speaker/box combination is so large that voltage division will apply essentially the entire signal across the speaker terminals. The series resistance will only be effective above 100 Hz.
3. Adding a BSC circuit to a very low Qts driver will help with the baffle step effect but will do nothing for the rolled off low end that typically starts in the several hundred Hz region. If you put a low Qts driver in a bass reflex enclosure and add a BSC circuit the bass will still be light and not pleasing.
4. If you combine the BSC resistor with the series resistor, that you want to apply to raise the effective Qts of the driver, then the response will improve. Remember that the series resistance has very little effect below 100 Hz and the BSC circuit is probably providing correction centered at 300 to 500 Hz for baffle widths associated with eight inch full range low Qts drivers in a closed, bass reflex, or ML TL enclosure. In reality, what this resistor is doing is reducing the effiency of the driver over the mid range and high end. This resistor is not really changing the Qts of the driver at the resonant frequency one bit.
So by increasing or decreasing the parallel resistor in the BSc circuit you are accomplishing all of the results that you describe as the classic method. My way is accomplished with fewer parts to buy and adjust. By changing the value of the inductor or the resistor of the BSC circuit you have all of the adjustability required using just two components. Then by adding a by-pass cap, you can also have adjustability in the high end extension and efficiency.
What I described above is my method of dealing with a low Qts/high efficiency driver. The penalty is a loss of some efficiency. For my Lowthers this means the drives stock 97 to 98 dB efficiency becomes closer to 93 to 94 dB. This is still very efficient and if driven by a solid state amp is very effective. I have used this method on several of these types of drivers and been very happy with the results. Others have reported similar results. No question that what I recommend is not the classical approach, you have to look at the real function of each component and think outside the box. Don't restrict yourself by formulas found in a book. My appologies if it is confusing. I know that most Lowther and Fostex purist will object to what I am doing but please don't knock it until you have tried it. A few dollars in parts can make a substantial imporvement to the balance of the system. It really works.
Hope that helps,
"I've read your BSC article and I'm afraid to say
you are giving confusing and conflicting advice."
I am sorry you feel that way, but I am looking at things outside of the classical rules for BSC and Zobel circuits. If you are dealing with a standard bass reflex or closed box using a driver that fits in the standard T/S alignment tables then what I wrote in my BSC Sizing article should work very well. But in the case of the Fostex FE-206E, or the Lowther drivers I am currently using, the classic BSC circuit will provide a marginal improvement in the speaker's performance. This is due to the very low Qts and very high efficency of these types of drivers.
I have used a series resistance and a BCS circuit with these types of drivers and it worked OK. But the more I thought about it, the more I felt that the series resistance was not very effectiv and could be accomplished in an easier way. I will outline my thought process below.
1. By adding the Zobel I have turned the speaker into a pure resistance above about 100 Hz. Below 100 Hz the impedance peak(s) due to the closed or ported box dominate the impedance curve. So I mentally divid the impedance curve into two region with about 100 Hz as the boundary.
2. Any series resistance (typically a couple of ohms) that is added will have a minimal effect below 100 Hz because the the impedance of the speaker/box combination is so large that voltage division will apply essentially the entire signal across the speaker terminals. The series resistance will only be effective above 100 Hz.
3. Adding a BSC circuit to a very low Qts driver will help with the baffle step effect but will do nothing for the rolled off low end that typically starts in the several hundred Hz region. If you put a low Qts driver in a bass reflex enclosure and add a BSC circuit the bass will still be light and not pleasing.
4. If you combine the BSC resistor with the series resistor, that you want to apply to raise the effective Qts of the driver, then the response will improve. Remember that the series resistance has very little effect below 100 Hz and the BSC circuit is probably providing correction centered at 300 to 500 Hz for baffle widths associated with eight inch full range low Qts drivers in a closed, bass reflex, or ML TL enclosure. In reality, what this resistor is doing is reducing the effiency of the driver over the mid range and high end. This resistor is not really changing the Qts of the driver at the resonant frequency one bit.
So by increasing or decreasing the parallel resistor in the BSc circuit you are accomplishing all of the results that you describe as the classic method. My way is accomplished with fewer parts to buy and adjust. By changing the value of the inductor or the resistor of the BSC circuit you have all of the adjustability required using just two components. Then by adding a by-pass cap, you can also have adjustability in the high end extension and efficiency.
What I described above is my method of dealing with a low Qts/high efficiency driver. The penalty is a loss of some efficiency. For my Lowthers this means the drives stock 97 to 98 dB efficiency becomes closer to 93 to 94 dB. This is still very efficient and if driven by a solid state amp is very effective. I have used this method on several of these types of drivers and been very happy with the results. Others have reported similar results. No question that what I recommend is not the classical approach, you have to look at the real function of each component and think outside the box. Don't restrict yourself by formulas found in a book. My appologies if it is confusing. I know that most Lowther and Fostex purist will object to what I am doing but please don't knock it until you have tried it. A few dollars in parts can make a substantial imporvement to the balance of the system. It really works.
Hope that helps,
Variac said:
Unless you are using an active equalizer circuit to increase the gain of the lower frequencies, this shouldn't be a problem as a passive BSC circuit will only attenuate the upper frequencies. Now, obviously, this means that he'll have to use more amplifier power to get the same SPL, but unless he's going for super high SPL's accross the board, I wouldn't anticipate any problem. If you want to be on the safe side, you could install a series high pass filter centered a little bit below where the driver starts to roll off to keep the excursions from notes the driver can't actually reproduce to a minimum. However, it may have an adverse effect on sound quality since you're adding more components to the signal path. I would only do it myself if I noticed some >or= xmax excursions.
As for the phase plug, I can't help you there, but the advice Timn8tor and Flemming J P gave sounds good to me, at least if you have a nice steady hand.
Hi MJK,
Your previous approach was to use series resistance to modify Q.
Reading your reply I can see this has been superceded.
So we are talking serious step compensation that also includes some
of the bass rolloff due to the drivers low Qts - seems fair enough.
but you did say :
Which is not the case, hence my post.
I can now see you meant :
"and part is due to the bass roll caused by the drivers low Qts"
I can now see that with the 10R resistor value baffle width is no
longer directly the issue, the inductor has to be tuned to allow
compensation of some bass roll-off whilst still operating correctly
over the baffle step region in the higher part of the transition.
The 6-8R would need near the calculated inductor value,
but the 10R value would need to be higher than calculated.
Now I understand your approach I find it faultless, given the options.
sreten.
Your previous approach was to use series resistance to modify Q.
Reading your reply I can see this has been superceded.
So we are talking serious step compensation that also includes some
of the bass rolloff due to the drivers low Qts - seems fair enough.
but you did say :
Which is not the case, hence my post.
I can now see you meant :
"and part is due to the bass roll caused by the drivers low Qts"
I can now see that with the 10R resistor value baffle width is no
longer directly the issue, the inductor has to be tuned to allow
compensation of some bass roll-off whilst still operating correctly
over the baffle step region in the higher part of the transition.
The 6-8R would need near the calculated inductor value,
but the 10R value would need to be higher than calculated.
Now I understand your approach I find it faultless, given the options.
sreten.I have decided already to put a cap paralel to BSC to bypass the influence of attenuation on the high frequences. It seems to me everybody agrees on that. If I go further and use the BSC with Zobel and add a new steel (pipe) phase plug? Do you think this combination will give even better results. And last but not least what about impregnating the speakers cone with suitable material (Steve Deckert did it with his HDT speakers) ?
Levon
Levon
TBH I'd take SD's musings with a pinch of salt.
"Special black ink" indeed, well it would be wouldn't it.
Though a phase plug should work.
A steel one only distortion measurements would tell the real
tale on this, but one thing is certain a steel phase plug would
increase Qts and lower sensitivity somewhat.
sreten.
"Special black ink" indeed, well it would be wouldn't it.
Though a phase plug should work.
A steel one only distortion measurements would tell the real
tale on this, but one thing is certain a steel phase plug would
increase Qts and lower sensitivity somewhat.
sreten.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.