Hi Sreten,
thank you.
Now a bit more.
I have been using Barry Porter, Bullock and just recently diysubwoofers.org to analyse reflex cabinets/alignments.
I have a pair of 18inch Celestions.
Qms =1.87425167
Qes =0.299014516
Qts =0.257873821
Fs=44Hz
Vas ~ 200L
Sensitivity quoted as 101db/2.83V for the pair in the cabinet.
probably 96db/W for a single 8ohm driver.
In standard form mounted in a 300L cabinet tuned to about 31Hz (twin 187diam ports 380long) rather than the Classic 48Hz. Massive power handling and SPL but no deep bass. Looks like the chevy you just mentioned.
I have added mass (+120gms) to one driver and the new T/S are
Qms = 2.302433759
Qes = 0.362596985
Qts = 0.313263005
Fs=31.6Hz
Fiddling about with one driver in that big cabinet or two drivers, I cannot reach a conclusion on which to use.
Choices are modify both drivers to an Fs near 31Hz,
mount one or both in the cabinet,
block off one of the twin ports,
make the remaining port smaller but of similar length and tune to somewhere in the 20Hz to 26Hz region.
Lots of different graphs. No real answers. I want to retain the Tannoy cabinet (B950) at least meantime.
Any guidance?
thank you.
Now a bit more.
I have been using Barry Porter, Bullock and just recently diysubwoofers.org to analyse reflex cabinets/alignments.
I have a pair of 18inch Celestions.
Qms =1.87425167
Qes =0.299014516
Qts =0.257873821
Fs=44Hz
Vas ~ 200L
Sensitivity quoted as 101db/2.83V for the pair in the cabinet.
probably 96db/W for a single 8ohm driver.
In standard form mounted in a 300L cabinet tuned to about 31Hz (twin 187diam ports 380long) rather than the Classic 48Hz. Massive power handling and SPL but no deep bass. Looks like the chevy you just mentioned.
I have added mass (+120gms) to one driver and the new T/S are
Qms = 2.302433759
Qes = 0.362596985
Qts = 0.313263005
Fs=31.6Hz
Fiddling about with one driver in that big cabinet or two drivers, I cannot reach a conclusion on which to use.
Choices are modify both drivers to an Fs near 31Hz,
mount one or both in the cabinet,
block off one of the twin ports,
make the remaining port smaller but of similar length and tune to somewhere in the 20Hz to 26Hz region.
Lots of different graphs. No real answers. I want to retain the Tannoy cabinet (B950) at least meantime.
Any guidance?
Hi AT,
Just some musings (sounds more like a Bessel than a Chebyshev) :
With a 340L cabinet and each driver Vas ~ 200L, not much point not using both.
How have you added the 120g ? Seems a good idea to do to both.
Closing off one port will tune to ~ 22Hz. Alternatively line the ports with 3cm foam.
What I would do : modify the feedback loop of the driving amplifier with an RC circuit
to boost low bass, say 12dB centred on say 36Hz. This will intereact with the low
bass roll-off of the amplifier to give a gently peaking active bass boost.
(Or simply use a passive line level circuit prior to the amplifier).
20Hz a few dB down (for room gain) should be no problem. (graph - Fs=32Hz, Qts=0.31)
/sreten.
Just some musings (sounds more like a Bessel than a Chebyshev) :
With a 340L cabinet and each driver Vas ~ 200L, not much point not using both.
How have you added the 120g ? Seems a good idea to do to both.
Closing off one port will tune to ~ 22Hz. Alternatively line the ports with 3cm foam.
What I would do : modify the feedback loop of the driving amplifier with an RC circuit
to boost low bass, say 12dB centred on say 36Hz. This will intereact with the low
bass roll-off of the amplifier to give a gently peaking active bass boost.
(Or simply use a passive line level circuit prior to the amplifier).
20Hz a few dB down (for room gain) should be no problem. (graph - Fs=32Hz, Qts=0.31)
/sreten.Attachments
Hi,
yes. It was when I downloaded diysubwoofers I began to realise that twin drivers could be made to work in the small box and that restores 6db of sensitivity.
The change in doubling the cone mass from about 110gms + air load to 230gms + air load does not calculate out accurately with the formulae I was given. But I added 40gms at a time and the T/S test results were fairly consistent at each stage.
I beat out a 16g thick strip of lead (20gms) to about 20mm wide by 100mm long. It is pretty thin and weighs about 1gm/sqcm.
I was warned that the sub cone could accelerate at upto about 10g, but I reckon the double sided tape could cope with 200g at that weight per area ratio.
If they start to fall off or vibrate/rattle as the tape starts to fail I can beat them out a bit thinner next time.
I was a bit unsure about attaching the tape to the paper cone but bit the bullet and went for it.
Just now the single 18inch driver has more volume than the active main speakers and both are driven by about the same power (50W). After I change to twin modified drivers I will have to turn the gain down yet again.
I guess that the basic 95db/W is now about 89db/W (is -6db right for doubling the cone mass?) with twin drivers I will need just 25W per driver to sound far too strong for the main speakers.
I have argued this, for months, with others on this Forum that want to fit kW amps to sub-bass drivers and few if any want to believe. They have been told they need mega power.
yes. It was when I downloaded diysubwoofers I began to realise that twin drivers could be made to work in the small box
and that restores 6db of sensitivity. The change in doubling the cone mass from about 110gms + air load to 230gms + air load does not calculate out accurately with the formulae I was given. But I added 40gms at a time and the T/S test results were fairly consistent at each stage.
I beat out a 16g thick strip of lead (20gms) to about 20mm wide by 100mm long. It is pretty thin and weighs about 1gm/sqcm.
I was warned that the sub cone could accelerate at upto about 10g, but I reckon the double sided tape could cope with 200g at that weight per area ratio.
If they start to fall off or vibrate/rattle as the tape starts to fail I can beat them out a bit thinner next time.
I was a bit unsure about attaching the tape to the paper cone but bit the bullet and went for it.
Just now the single 18inch driver has more volume than the active main speakers and both are driven by about the same power (50W). After I change to twin modified drivers I will have to turn the gain down yet again.
I guess that the basic 95db/W is now about 89db/W (is -6db right for doubling the cone mass?) with twin drivers I will need just 25W per driver to sound far too strong for the main speakers.
I have argued this, for months, with others on this Forum that want to fit kW amps to sub-bass drivers and few if any want to believe. They have been told they need mega power.
Well if you want heavy and massive bass (real life explosions and stuff) from a far too small open baffle with a truckload of eq. on it...
Sometimes power is just awesome
, but for me 300W/ch. would be absolutely fine...
Agree about the R/C filters, before the amp so to say...filters on the output are just too expensive, large and get pretty stressed from the LOUD signal...
Sometimes power is just awesome
, but for me 300W/ch. would be absolutely fine... Agree about the R/C filters, before the amp so to say...filters on the output are just too expensive, large and get pretty stressed from the LOUD signal...
writing a book/article for beginners
There are many "experts" here who know more about designing speakers than novice person like myself should seriously looking at writing a book/article on designing speakers for beginners rather discrediting(language used is over the top) books like LDC.
LDC(I don't own one) must have some merits otherwise it won't be in its 6/7th edition. If it has any deficiencies/challenges to his ideas the new author(s) can highlight them in the new book/article rather than doing it here.
The book/article would be a new approach towards speakers design an alternative to LDC.
Something to think about......seriously....At the end of the day most new DIYers here will benefit..
There are many "experts" here who know more about designing speakers than novice person like myself should seriously looking at writing a book/article on designing speakers for beginners rather discrediting(language used is over the top) books like LDC.
LDC(I don't own one) must have some merits otherwise it won't be in its 6/7th edition. If it has any deficiencies/challenges to his ideas the new author(s) can highlight them in the new book/article rather than doing it here.
The book/article would be a new approach towards speakers design an alternative to LDC.
Something to think about......seriously....At the end of the day most new DIYers here will benefit..
Hi Richie,
I am not at all sure in the internal pressure differences between a transformed sealed enclosure and a relex loaded enclosure.
I suspect I can seal off one port effectively and make it sufficiently strong and airtight to survive as a ported box but the existing design limits what I can do ( and I want it reversible so that I have the possibility of selling at a later date).
Besides, the biggest amps I have that I can rely on are 190W. I have a stereo 350W+350W I bought as a design comparison exercise but I do not trust the SOAR of 4pair 2sa1943/c5200 running on 92Vdc supply rails.
If Tannoy recommended 600W to 1200W when in ported mode then does the same apply if converted to sealed mode?
Finally, the Celestion drivers appear to have very limited Xmax. They are designed as music reinforcement, not sub-bass. Xmax~4 to 5mm.
If this were usable then a new box may be the solution.
I am not at all sure in the internal pressure differences between a transformed sealed enclosure and a relex loaded enclosure.
I suspect I can seal off one port effectively and make it sufficiently strong and airtight to survive as a ported box but the existing design limits what I can do ( and I want it reversible so that I have the possibility of selling at a later date).
Besides, the biggest amps I have that I can rely on are 190W. I have a stereo 350W+350W I bought as a design comparison exercise but I do not trust the SOAR of 4pair 2sa1943/c5200 running on 92Vdc supply rails.
If Tannoy recommended 600W to 1200W when in ported mode then does the same apply if converted to sealed mode?
Finally, the Celestion drivers appear to have very limited Xmax. They are designed as music reinforcement, not sub-bass. Xmax~4 to 5mm.
If this were usable then a new box may be the solution.
If you are going to have problems sealing off the vent then that will be a problem.
190 w, is that per channel or in total? Per channel feeding large drivers with limited xmax it might be ok.
If the speakers are 8 ohm load I think your 350 w amp would be fine.
The power handling depends on the design. There is the voicecoil thermal handling (which is commonly the quoted power rating) and the excursion limited handling which is affected by the box design. You would need to model the excursion, Bullock provides reliable equations for this that I would choose to use over WinISD.
190 w, is that per channel or in total? Per channel feeding large drivers with limited xmax it might be ok.
If the speakers are 8 ohm load I think your 350 w amp would be fine.
The power handling depends on the design. There is the voicecoil thermal handling (which is commonly the quoted power rating) and the excursion limited handling which is affected by the box design. You would need to model the excursion, Bullock provides reliable equations for this that I would choose to use over WinISD.
Andrew,
Can you confirm if the cabinet already has the two drivers in?
It's not a poor solution if he doesn't want to mess with the cabinet or the amplifier.
And how can you say it's not an LT, surely if he builds an LT and designs it properly, what else can it be?
Can you confirm if the cabinet already has the two drivers in?
sreten said:Model it. Its a very poor solution compared to what I suggested.
And it would not be a "linkwitz transform", simply bass boost.
It's not a poor solution if he doesn't want to mess with the cabinet or the amplifier.
And how can you say it's not an LT, surely if he builds an LT and designs it properly, what else can it be?
Hi,
a few suggestions flying about.
Yes, a two 18inch driver 300L cabinet with twin ports. Each port a triangular cut half a square with sides 235mm and about 380mm long area equivalent to 265mm diameter.
Can I remind you what I think applies here.
A 5th or 6th order alignment based on the existing box will use up more than the Xmax available. Any alignment to extend the low bass below the Fb uses enormous Xmax at fairly low SPL.
A transform of a sealed box will use up Xmax at relatively low SPL and require big power.
I think those two options are out if I have understood the requirements correctly.
Can anybody confirm that the original 101db/2.83V (two 8ohm drivers in parallel) becomes a single driver 95db/W?
Similarly that 95db/W is reduced by 6db when the moving mass is doubled?
If this is the case then I will be starting with a pair of 89db/W drivers, if the added mass route is the one to follow.
My first experiment with just one modified driver with the wrong port tuning is massively better than the original (it goes much lower down and seems on tiny power to go louder than the active mains.
I know and expect it can become better yet.
a few suggestions flying about.
Yes, a two 18inch driver 300L cabinet with twin ports. Each port a triangular cut half a square with sides 235mm and about 380mm long area equivalent to 265mm diameter.
Can I remind you what I think applies here.
A 5th or 6th order alignment based on the existing box will use up more than the Xmax available. Any alignment to extend the low bass below the Fb uses enormous Xmax at fairly low SPL.
A transform of a sealed box will use up Xmax at relatively low SPL and require big power.
I think those two options are out if I have understood the requirements correctly.
Can anybody confirm that the original 101db/2.83V (two 8ohm drivers in parallel) becomes a single driver 95db/W?
Similarly that 95db/W is reduced by 6db when the moving mass is doubled?
If this is the case then I will be starting with a pair of 89db/W drivers, if the added mass route is the one to follow.
My first experiment with just one modified driver with the wrong port tuning is massively better than the original (it goes much lower down and seems on tiny power to go louder than the active mains.
I know and expect it can become better yet.
Hi Sreten,
I agree that SPL and Xmax are tied together.
It's my understanding that in a sealed box the bass drops off at 12db/octave once below the resonance of the air volume cone mass frequency.
One can boost the low end at upto 12db/octave to extend the bass lower down and this uses power and Xmax.
In a reflex box the Xmax is rising as frequency drops from above Fb but drops as the frequency approaches and reaches a minima at about Fb, then rises again below Fb. Once the frequency is well below Fb the Xmax approaches the unloaded cone values.
Again my understanding is that by lowering the box tuning one uses up more of the Xmax above the new Fb but at around the Fb the minima still exists and one can achieve higher SPL at the chosen lower frequency than is available from that driver than by any other non horn method.
If one were to select Fb and driver parameters appropriately then the highish Xmax in the frequency range just above Fb can give acceptable SPL and the high efficiency minima around Fb avoids Xmax problems in this narrow band of frequencies.
One would still have the inherent problem of low power handling, due to Xmax limitations, below Fb and low SPL must be the result of reflex loading at very low frequencies.
If all the preceeding is true then there is little point in using these low Xmax drivers with any kind of power boosting to bring up the low frequency response and with it low frequency SPL. They will put simply "run out of Xmax".
Back to the guidance I need.
If I were to manipulate the T/S parameters by adjusting Fs and the Qs, I suspect, nay am convinced, that improved low frequency response (and low frequency SPL) is available from this box and driver set up than the optimised set up that the standard model has achieved for music reinforcement.
I agree that SPL and Xmax are tied together.
It's my understanding that in a sealed box the bass drops off at 12db/octave once below the resonance of the air volume cone mass frequency.
One can boost the low end at upto 12db/octave to extend the bass lower down and this uses power and Xmax.
In a reflex box the Xmax is rising as frequency drops from above Fb but drops as the frequency approaches and reaches a minima at about Fb, then rises again below Fb. Once the frequency is well below Fb the Xmax approaches the unloaded cone values.
Again my understanding is that by lowering the box tuning one uses up more of the Xmax above the new Fb but at around the Fb the minima still exists and one can achieve higher SPL at the chosen lower frequency than is available from that driver than by any other non horn method.
If one were to select Fb and driver parameters appropriately then the highish Xmax in the frequency range just above Fb can give acceptable SPL and the high efficiency minima around Fb avoids Xmax problems in this narrow band of frequencies.
One would still have the inherent problem of low power handling, due to Xmax limitations, below Fb and low SPL must be the result of reflex loading at very low frequencies.
If all the preceeding is true then there is little point in using these low Xmax drivers with any kind of power boosting to bring up the low frequency response and with it low frequency SPL. They will put simply "run out of Xmax".
Back to the guidance I need.
If I were to manipulate the T/S parameters by adjusting Fs and the Qs, I suspect, nay am convinced, that improved low frequency response (and low frequency SPL) is available from this box and driver set up than the optimised set up that the standard model has achieved for music reinforcement.
Andrew, most of what you say above is correct, except the bit about boosting.
If you do manage to lower Fs with addes mass etc, and thus can tune the vent lower to supplant output to achieve more deep bass SPL, there is a danger that somewhere above tuning you might run out of Xmax.
At the end of the day, whether you add mass to manipulate Fs, or apply EQ, Xmax is unchanged. And for a given Xmax, Vas, box volume and tuning frequency the danger of over excursion is the same - the boosted case simply uses more input power.
Your only saving grace with the added mass method is that Vas will be increased, which could mean there is sufficient air load to prevent Xmax being reached above tuning.
Some extensive simulation is needed.
If you do manage to lower Fs with addes mass etc, and thus can tune the vent lower to supplant output to achieve more deep bass SPL, there is a danger that somewhere above tuning you might run out of Xmax.
At the end of the day, whether you add mass to manipulate Fs, or apply EQ, Xmax is unchanged. And for a given Xmax, Vas, box volume and tuning frequency the danger of over excursion is the same - the boosted case simply uses more input power.
Your only saving grace with the added mass method is that Vas will be increased, which could mean there is sufficient air load to prevent Xmax being reached above tuning.
Some extensive simulation is needed.
richie00boy said:
Hi,
Agree with A + C , B not at all, it stays the same.
What really counts is Vd - volume displacement
- churlish to complain about low Xmax for twin 18" drivers.
Running out of X max at low frequencies ? not if you match the
ampifier power correctly and they are loud enough for you needs.
Absolutely.
Starting off with a lower FS is always a good idea if you are after low bass.
What are the main speakers and intended c/o point ?
Retuning to 22Hz will increase max SPL at low frequencies.
The other changes really just increase efficiency low down.
/sreten.
sreten said:
I agree.
sreten said:
I also agree, despite the slightly off wording.
However, you are thinking under static conditions. Under dynamic conditions which happen to be those under which T/S parameters are measured and the driver used, once you add mass you alter the damping of the system (hence Qts raises - it's less damped now). I wrongly implied that Vas will change, what I should have said is that the required box is now bigger due to the increased Qts.
richie00boy said:
Hi,
True. The box is too big for the "standard" drivers in terms of
alignments, i.e. driver Vas and Qts do not suit the box. One
reason why two drivers is better than one. By adding mass
and increasing Qts the match to the "too big" box improves.
At the time Fbox (sealed) drops - always good for transient response.
The added mass route (and both drivers) is the way to go.
/sreten.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.