I need to build some boxes already!
I wondered about the basket design myself. It is going to be very interesting comparing these two drivers as point sources...
As I said in the post above, I really wish I had more time to explore... We are getting there... Last night of Mid-night shift, so that should help!
Wesayso beat me with the SD's, at least I got the maths right
I wondered about the basket design myself. It is going to be very interesting comparing these two drivers as point sources...
As I said in the post above, I really wish I had more time to explore... We are getting there... Last night of Mid-night shift, so that should help!
Wesayso beat me with the SD's, at least I got the maths right
1 mm Xmax, but there are 75 drivers. The question is what is the efficiency at 300 Hz. It is rated at 83 db/1W/1m. IIRC, efficiency increases by 6 db per doubling of cone area. Conservatively, if we start at 80 db/1W/1m assuming that it won't be 83 db/1W/1m at 300 Hz, we get 116 db/1W/1M (going from 1 driver to 2 drivers, then to 4, then to 8 and so on). If you get 12 db/octave roll off below 300 Hz, you'll be at 92 db/1W/1m at 75 Hz. Not bad. But it won't go much louder than 95 db maybe at 75 Hz. Better to stay above 150 Hz.
This is why I ultimately decided it was not worth the punt. And TC9 would be a safer bet. But the array would be very easy to construct. It would have almost no comb-filtering, i.e., a more idealized version of the array.
25 years ago, there were not many drivers such as these. With a bit of convincing, I'm sure we can get Tymphany to make these to our specs. Anyone for a group buy?
That shows promise...
ra7, what where your listening impressions of this driver?
I really like the off-axis Frequency Response.The cost for 75 with the parts-express discount is $568/array, so double the TC9. That I do not think is unreasonable, if they indeed perform "better" for high frequencies. I wish I could find a CSD for these...
My concern is getting down to 80 Hz, as 75 of these only has an overall SD of 367.5 cubic cm, compared to the 907.5 cubic cm of 25 TC9s. And this driver has a "fixed" enclosure volume.
I am interested in how these would perform with high frequencies in an array. My gut is telling me, these would lend themselves better to a two-way design crossed between 200 and 300 Hz. If I was to do this myself, I would consider a line of small bass drivers. Thankfully, at 200 - 300 Hz, one could easily get the two lines close enough...
Yes, it would be a good idea to pair these with maybe some 8" woofers crossed at 200 Hz. That would work. Could make a great array.
I bought four of these to try. I put them on a cardboard sheet, hung them in the corner and made some measurements. I cannot really say how they sounded. They sound fine. It's not fair to judge how four of these unequalized would sound compared to 75 of these equalized to perfection.
Attachments
I found this catalogue while looking for an easier and possible cheaper way to build your beasts wesayso....
Alluminium boat masts!
These particular profiles would need some machining to create a viable driver baffle but the potential is there. Strong, long and would be sexy when finished properly.
See Page 41.
http://www.capral.com.au/Aluminium-Extrusion-Catalogue-PDF-View
This is just the first link I found, there must be thousands of different profiles built over the years...
Now looking through the secondhand market for a semi damaged unit, possible for free... ;P
Any input gents?
Alluminium boat masts!
These particular profiles would need some machining to create a viable driver baffle but the potential is there. Strong, long and would be sexy when finished properly.
See Page 41.
http://www.capral.com.au/Aluminium-Extrusion-Catalogue-PDF-View
This is just the first link I found, there must be thousands of different profiles built over the years...
Now looking through the secondhand market for a semi damaged unit, possible for free... ;P
Any input gents?
Attachments
Last edited:
Four won't be telling you much.
I built a 9-driver array prior to my 16-driver arrays.
Even with 9 drivers, the sweet spot was tiny. The moment I turned my head, it was gone. And the presence was nothing like I had with 16 drivers.
There's just nothing like a long row of drivers. You HAVE to be on-site and experience it, no amount of reading about comb filtering and correction will tell you anything of the feeling you get when you physically listen to a nice pair of arrays.
I see they even have an open baffle version
.The sizes seem a bit small to me, I build quite a solid mast in comparison
.The biggest one still seemed small to me. I did see a fun oval shape in there, only 50 deep x 300 wide (page 73). A bit like a mini Genesis Dragon
:Attachments
Last edited:
1 mm Xmax, but there are 75 drivers. The question is what is the efficiency at 300 Hz. It is rated at 83 db/1W/1m. IIRC, efficiency increases by 6 db per doubling of cone area. Conservatively, if we start at 80 db/1W/1m assuming that it won't be 83 db/1W/1m at 300 Hz, we get 116 db/1W/1M (going from 1 driver to 2 drivers, then to 4, then to 8 and so on). If you get 12 db/octave roll off below 300 Hz, you'll be at 92 db/1W/1m at 75 Hz. Not bad. But it won't go much louder than 95 db maybe at 75 Hz. Better to stay above 150 Hz.
This is why I ultimately decided it was not worth the punt. And TC9 would be a safer bet. But the array would be very easy to construct. It would have almost no comb-filtering, i.e., a more idealized version of the array.
25 years ago, there were not many drivers such as these. With a bit of convincing, I'm sure we can get Tymphany to make these to our specs. Anyone for a group buy?
Yes, it would be a good idea to pair these with maybe some 8" woofers crossed at 200 Hz. That would work. Could make a great array.
I bought four of these to try. I put them on a cardboard sheet, hung them in the corner and made some measurements. I cannot really say how they sounded. They sound fine. It's not fair to judge how four of these unequalized would sound compared to 75 of these equalized to perfection.
Made me think of the JBL CBT line arrays...
Attachments
As an experiment I ran the APL_TDA software on a loop back test with my DAC in the chain. Just to see it's output. Here's what the graph looks like:
I was busy optimising settings in JRiver and figured I might as well try the loop back test in the TDA software as well. This loop back lets me see the processing chain which can be useful to get a grip on it all. JRiver's WDM driver can be finicky though and I had to change a couple of settings to get a clear picture. Sometimes the driver would simply be late in responding until I removed the setting in REW to play silence at startup for hardware synchronisation.
I re-ran my DRC correction and plotted that in APL_TDA too:
With the accompanying DFR (Delay Frequency Response) plot:
4 ms at ~25 Hz, not bad... is that what they call fast bass?
These last two graphs are from a stereo pair of arrays at the listening position.
Ambient channels are turned off. I shifted the plots to -23 ms to be able to see the front side, the start of the peak. Just to see that side and check for visible pré-ringing etc.
I'll have to experiment some more, I notice that even in a loopback measurement if I set my mic to 44.100 it suddenly drops off at ~18 to 19 KHz, if I set the mic to 48000 KHz it records all the way up. Even when I record in 44.100, basically creating a mismatch. Some more experimenting to be done. I might even try to record the impulse with the Audiolense measurement suite to see if/how that differs from REW results.
I was busy optimising settings in JRiver and figured I might as well try the loop back test in the TDA software as well. This loop back lets me see the processing chain which can be useful to get a grip on it all. JRiver's WDM driver can be finicky though and I had to change a couple of settings to get a clear picture. Sometimes the driver would simply be late in responding until I removed the setting in REW to play silence at startup for hardware synchronisation.
I re-ran my DRC correction and plotted that in APL_TDA too:
With the accompanying DFR (Delay Frequency Response) plot:
4 ms at ~25 Hz, not bad... is that what they call fast bass?
These last two graphs are from a stereo pair of arrays at the listening position.
Ambient channels are turned off. I shifted the plots to -23 ms to be able to see the front side, the start of the peak. Just to see that side and check for visible pré-ringing etc.
I'll have to experiment some more, I notice that even in a loopback measurement if I set my mic to 44.100 it suddenly drops off at ~18 to 19 KHz, if I set the mic to 48000 KHz it records all the way up. Even when I record in 44.100, basically creating a mismatch. Some more experimenting to be done. I might even try to record the impulse with the Audiolense measurement suite to see if/how that differs from REW results.
I'm glad you're still enjoying finicking with your arrays!
On an other note, I though of you tonight.
I watched the Scorsese/Jagger first episode of Vinyl, a new TV show.
Robert Plant and Led Zep were in there ... well, the actors portraying Led Zep!
Pretty good TV show, but some parts are not for watching with the kids!
On an other note, I though of you tonight.
I watched the Scorsese/Jagger first episode of Vinyl, a new TV show.
Robert Plant and Led Zep were in there ... well, the actors portraying Led Zep!
Pretty good TV show, but some parts are not for watching with the kids!
Sensitivity, Sd and dynamics....
Please note this is incorrect, you only gain 3 dB of sensitivity (efficiency remains unchanged regardless of how many drivers you use) per doubling of surface area.
You can gain an additional 3dB sensitivity if you half the impedance, but as the driver starts at a low 3.5 Ohms you dont have much wiggle room!
Also please note the difference between the 83dB sensitivity and the true 1 watt 79dB sensitivity .....The 83dB is achieved at 2.83 volts....Not 1 watt.
The 79.5 dB figure is the true 1 watt input figure.
Re Sd, I have found that in average rooms with typical music SPL's most folks need a minimum of 550 square cm ( equal to a single 12 inch driver) per side plus separate subs is what it takes to deliver lifelike dynamics and even room coverage. YMMV....!
IMO, this little driver really will need to be combined with an array of bass / mids crossed over at approx 600Hz. Then you may still need to crossover the bass/ mids to subs at 60Hz to 80Hz.
Hope this helps and all the best
Derek.
Please note this is incorrect, you only gain 3 dB of sensitivity (efficiency remains unchanged regardless of how many drivers you use) per doubling of surface area.
You can gain an additional 3dB sensitivity if you half the impedance, but as the driver starts at a low 3.5 Ohms you dont have much wiggle room!
Also please note the difference between the 83dB sensitivity and the true 1 watt 79dB sensitivity .....The 83dB is achieved at 2.83 volts....Not 1 watt.
The 79.5 dB figure is the true 1 watt input figure.
Re Sd, I have found that in average rooms with typical music SPL's most folks need a minimum of 550 square cm ( equal to a single 12 inch driver) per side plus separate subs is what it takes to deliver lifelike dynamics and even room coverage. YMMV....!
IMO, this little driver really will need to be combined with an array of bass / mids crossed over at approx 600Hz. Then you may still need to crossover the bass/ mids to subs at 60Hz to 80Hz.
Hope this helps and all the best
Derek.
I'm glad you're still enjoying finicking with your arrays!
On an other note, I though of you tonight.
I watched the Scorsese/Jagger first episode of Vinyl, a new TV show.
Robert Plant and Led Zep were in there ... well, the actors portraying Led Zep!
Pretty good TV show, but some parts are not for watching with the kids!
Have to check that out!
I saw a promo of it the other day, seemed like a fun show!
Derek,
If it is 2.83V @ 8 ohms then it is 1 watt. So that can be a bit deceiving when a driver has a different impedance. It is interesting that the Beyma driver has such a long voicecoil compared to the gap length and the Xmax is not greater, probably the surround is the limitation otherwise there is an awful lot of that coil length that would never see the gap. Not saying anything about how it sounds just that the coil height is long for that Xmax.
If it is 2.83V @ 8 ohms then it is 1 watt. So that can be a bit deceiving when a driver has a different impedance. It is interesting that the Beyma driver has such a long voicecoil compared to the gap length and the Xmax is not greater, probably the surround is the limitation otherwise there is an awful lot of that coil length that would never see the gap. Not saying anything about how it sounds just that the coil height is long for that Xmax.
Wrong, but only by 110%!
Kindhorn,
The Tymphany driver is not 8 Ohms it is 3.8 Ohms....It will be drawing over 2 watts @ 2.83 volts ....Not, as you state 1 watt!
It is not "a bit deceiving" its basic Ohms law stuff, but it is fundamentally important to understand the basics.
Also worth noting, as the power handling (calculated in watts not voltage) remains the same, your max SPL calculations will be out by over 100%...
There are lots of articles on this, here is a simple user guide with useful explanations:
Loudspeaker Sensitivity & Impedance Explained | Audioholics
The most relevant sections are :
" If the amplifier is set to 2.83V, the 4 ohm speaker will draw 2 watts of power from the amplifier while an 8 ohm speaker draws 1 watt. (Power = Voltage2/resistance. So, 2.832 = 8. 8/8ohms = 1 watt.) This potentially gives the 4 ohm speaker a +3dB SPL advantage over the 8 ohm speaker for a given input level. This seems to be unfair. Therefore, some engineers and reviewers adjust the input voltage so that it equates to 1 watt at the impedance of the input frequency (or band of frequencies) of the speaker under test.
There is an international standard that defines sensitivity, and it specifies that sensitivity be referred to SPL @ 1m for 2.83V input. The actual measurement can be made at any distance and input level, but must be calculated back to 1m and 2.83V equivalent. A 4-ohm speaker will under these conditions appear to have an advantage, and many manufacturers take unfair advantage of this. This is why it is also required to quote nominal impedance, so that one can see that if the nominal impedance is low then the speaker will in fact be drawing more power from the amplifier and hence the amplifier will have to be able to deliver this. Not all manufacturers follow the industry standards in every instance of published specifications, so the consumer needs to pay close attention and read the fine print. "
I hope this helps and all the best
Derek.
Kindhorn,
The Tymphany driver is not 8 Ohms it is 3.8 Ohms....It will be drawing over 2 watts @ 2.83 volts ....Not, as you state 1 watt!
It is not "a bit deceiving" its basic Ohms law stuff, but it is fundamentally important to understand the basics.
Also worth noting, as the power handling (calculated in watts not voltage) remains the same, your max SPL calculations will be out by over 100%...
There are lots of articles on this, here is a simple user guide with useful explanations:
Loudspeaker Sensitivity & Impedance Explained | Audioholics
The most relevant sections are :
" If the amplifier is set to 2.83V, the 4 ohm speaker will draw 2 watts of power from the amplifier while an 8 ohm speaker draws 1 watt. (Power = Voltage2/resistance. So, 2.832 = 8. 8/8ohms = 1 watt.) This potentially gives the 4 ohm speaker a +3dB SPL advantage over the 8 ohm speaker for a given input level. This seems to be unfair. Therefore, some engineers and reviewers adjust the input voltage so that it equates to 1 watt at the impedance of the input frequency (or band of frequencies) of the speaker under test.
There is an international standard that defines sensitivity, and it specifies that sensitivity be referred to SPL @ 1m for 2.83V input. The actual measurement can be made at any distance and input level, but must be calculated back to 1m and 2.83V equivalent. A 4-ohm speaker will under these conditions appear to have an advantage, and many manufacturers take unfair advantage of this. This is why it is also required to quote nominal impedance, so that one can see that if the nominal impedance is low then the speaker will in fact be drawing more power from the amplifier and hence the amplifier will have to be able to deliver this. Not all manufacturers follow the industry standards in every instance of published specifications, so the consumer needs to pay close attention and read the fine print. "
I hope this helps and all the best
Derek.
Last edited:
Beyma design principals
Kindhornguy,
It is not my place to question or answer for the Beyma driver designers, but I can point you in the direction of thinking and if you email Pablo at Beyma he may (under NDA) divulge a bit more .....
Beyma are very conservative with their Xmax, but also they have very high Dmax (driver travel before damage) and this ensures they withstand the abuse and over driving that Live sound dishes out....Well over enthusiastic bands and DJ's dish out!
Hope this helps and all the best
Derek.
Kindhornguy,
It is not my place to question or answer for the Beyma driver designers, but I can point you in the direction of thinking and if you email Pablo at Beyma he may (under NDA) divulge a bit more .....
Beyma are very conservative with their Xmax, but also they have very high Dmax (driver travel before damage) and this ensures they withstand the abuse and over driving that Live sound dishes out....Well over enthusiastic bands and DJ's dish out!
Hope this helps and all the best
Derek.
Comparing apples to apples even when building oranges!
I think this is why it is so hard to pick out drivers for arrays, and why many struggle with that leap of faith to build one. But, Wesayso and I agree: An array of drivers is one driver on steroids, so like the sound of the single driver first...
The same could be said, if one wants to try a two-way.
In the case with the PMTs, I believe the fairest way to compare and test them, would be to make a good two-way: one PMT with one 6"-8" mid-bass. I would try 300 Hz, even 600 Hz, see how well the integration works. Get it all dialed in with EQ and DSP. Likewise, I would do the same with One TC9 crossed at 80 Hz to a sub (In ra7's case). These systems would for sure be limited in volume and will not have the "presence" of an array, but I think it would give us an idea of what the drivers sound like, and if they have some "character" flaws that can not be solved with EQ or DSP.
ra7, do You have a pair of mid-basses to at least try with the PMTs? For laughs and giggles, You could use one TC9 (if you have an extra pair laying around), then You would really get an idea of what the PMT does compared to a TC9 full-range, and if the challenges of integrating a two-way system are worth the effort.
I will be exploring some of these options myself, so I will keep You all posted.
I think this is why it is so hard to pick out drivers for arrays, and why many struggle with that leap of faith to build one. But, Wesayso and I agree: An array of drivers is one driver on steroids, so like the sound of the single driver first...
The same could be said, if one wants to try a two-way.
In the case with the PMTs, I believe the fairest way to compare and test them, would be to make a good two-way: one PMT with one 6"-8" mid-bass. I would try 300 Hz, even 600 Hz, see how well the integration works. Get it all dialed in with EQ and DSP. Likewise, I would do the same with One TC9 crossed at 80 Hz to a sub (In ra7's case). These systems would for sure be limited in volume and will not have the "presence" of an array, but I think it would give us an idea of what the drivers sound like, and if they have some "character" flaws that can not be solved with EQ or DSP.
ra7, do You have a pair of mid-basses to at least try with the PMTs? For laughs and giggles, You could use one TC9 (if you have an extra pair laying around), then You would really get an idea of what the PMT does compared to a TC9 full-range, and if the challenges of integrating a two-way system are worth the effort.
I will be exploring some of these options myself, so I will keep You all posted.
Derek,
I wasn't saying that was a one watt for the Beyma driver, I was only stating that the standard way that speakers have been rated was really a poor way to be stated, the one watt at one meter at 2.83V is based on an 8 ohm speaker. I'm not saying that this is what many speakers truly are but that is where that number always comes form. What I am saying is that it is many times deceiving and I actually agree with your math. it is just so commonly stated that way for marketing purposes. Buyer beware.
I don't need anyone to tell me about why they chose an overhung voicecoil, I understand that completely. I just personally don't like that construction method, not my cup of tea.
I wasn't saying that was a one watt for the Beyma driver, I was only stating that the standard way that speakers have been rated was really a poor way to be stated, the one watt at one meter at 2.83V is based on an 8 ohm speaker. I'm not saying that this is what many speakers truly are but that is where that number always comes form. What I am saying is that it is many times deceiving and I actually agree with your math. it is just so commonly stated that way for marketing purposes. Buyer beware.
I don't need anyone to tell me about why they chose an overhung voicecoil, I understand that completely. I just personally don't like that construction method, not my cup of tea.
Hi Derek,
I had to go back and look at the D'Appolito book to verify the formula. Here is the formula for SPL:
SPL db/2.83V/1m = 20 x log10 (Bl x Sd/Re/Mms) + 88.5 db.
Now, if you assume the term (Bl x Sd/Re/Mms) to be 1, you get 88.5 db. If you assume it is two, you get 94.5 db, a 6 db increase. So, double the area, get 6 db more sensitivity. I don't see how Re comes into the picture. We are not changing Re, Bl or Mms in the above term. Besides, in an array, you can get back to the same impedance by connecting groups of drivers in series and parallel.
Agree with you on the difference between sensitivity and efficiency, should've chosen my words more carefully there.
Doubling of Sd gives a 3dB increase if Ohm is maintained
Hi ra7
This question is asked and explained here...
Speaker sensitivity for multiple drivers - AVS Forum | Home Theater Discussions And Reviews
Using the log calculation you must allow for the Ohm change.
Ie Regardless of whether your single driver is 4Ohm, 6Ohm, 8Ohm or 16Ohm you only gain 3 dB each time you double the Sd.
Now each time you half the impedance you gain an additional 3dB sensitivity.
So if you start with "X " Ohms and finish with " X " Ohms you will only gain 3dB each time the Sd doubles.
Of course if you start with an 8 Ohm driver and finish with a 4Ohm array you will gain a "one off" extra 3dB over and above your doubling of Sd gains.
Hope this makes sense?
I am happy to help work out actual examples if you want to look at a few options of driver numbers / Ohms / Sd / power handling etc.
All the best
Derek.
Hi ra7
This question is asked and explained here...
Speaker sensitivity for multiple drivers - AVS Forum | Home Theater Discussions And Reviews
Using the log calculation you must allow for the Ohm change.
Ie Regardless of whether your single driver is 4Ohm, 6Ohm, 8Ohm or 16Ohm you only gain 3 dB each time you double the Sd.
Now each time you half the impedance you gain an additional 3dB sensitivity.
So if you start with "X " Ohms and finish with " X " Ohms you will only gain 3dB each time the Sd doubles.
Of course if you start with an 8 Ohm driver and finish with a 4Ohm array you will gain a "one off" extra 3dB over and above your doubling of Sd gains.
Hope this makes sense?
I am happy to help work out actual examples if you want to look at a few options of driver numbers / Ohms / Sd / power handling etc.
All the best
Derek.
Derek is correct on this as far as I understand the question. Typically you will see the comparison of a single 8 ohm speaker and then two of the same 8ohm speakers in parallel so you are really dropping the impedance in half and doubling the Sd. Now take a single 4 ohm speaker and compared that to two 8om speakers in parallel with the identical Sd of both examples and you will get that simple 3db increase in output if both sets of speakers started out with the same sensitivity.
I want to believe you guys, but can you explain how the math works? If I have 4 drivers, 4 ohms each, and I connect two each in parallel and then the two pairs in series, I would get back the 4 ohm impedance. But I would've quadrupled the area. I can see how a 12 db increase in SPL would be phenomenal in this case, and I would doubt it myself now that I think about it, but I am not seeing how the math works out given the formula for SPL.