keith_correa
Well-Known Member
Re: Speaker Kit-Insane value
Hari, what's this network compensating for? Baffle step loss?
Hari, what's this network compensating for? Baffle step loss?
Tymphany custom OEM surplus drivers group buy
- Thread starter linuxguru
- Start date
Re: Speaker Kit-Insane value
In my experience using full-range drivers, the mids and highs over power the lows and sounds quite shrill if used without compensation. Hence the harmonics needs to be balanced so that the lows can be audible and dont get masked by the mids and the lows. In absence of this compensation the lows will not have any punch or will sound laid-back.
The compensation network is not purely a BSC as it does not use the standard forumula of the BSC, but its actually derived after simulation. Also it will need the use of a zobel and other componets (additional inductors and capacitors) to extend the band-width. If you see the simulated response of the published driver you will see an example of this. The red curve is the free-air response of the driver and the yellow response is with the compensation network. The speaker system has a extended response of around 4dB in the mid-bass region to compensate for the baffle step loss, but also the mid-range is extended till 9KHz before the driver begins its natural roll-off. Using only the standard BSC will not allow the higher mids to be extended where it rolls-off at 7KHz. The compensator will actually alter the impedance of the speaker system by some fractions of ohms, but if a 18swg wire is used it will not be a problem (this will increase cost though). Also if thick wires are used the system Q will be benefited and so will the SPL level. I will leave it to FMs to decide if they want a thicker wire and a probable increase of cost by atleast Rs.300/-. I have estimated with 20SWG wire.
In my experience using full-range drivers, the mids and highs over power the lows and sounds quite shrill if used without compensation. Hence the harmonics needs to be balanced so that the lows can be audible and dont get masked by the mids and the lows. In absence of this compensation the lows will not have any punch or will sound laid-back.
The compensation network is not purely a BSC as it does not use the standard forumula of the BSC, but its actually derived after simulation. Also it will need the use of a zobel and other componets (additional inductors and capacitors) to extend the band-width. If you see the simulated response of the published driver you will see an example of this. The red curve is the free-air response of the driver and the yellow response is with the compensation network. The speaker system has a extended response of around 4dB in the mid-bass region to compensate for the baffle step loss, but also the mid-range is extended till 9KHz before the driver begins its natural roll-off. Using only the standard BSC will not allow the higher mids to be extended where it rolls-off at 7KHz. The compensator will actually alter the impedance of the speaker system by some fractions of ohms, but if a 18swg wire is used it will not be a problem (this will increase cost though). Also if thick wires are used the system Q will be benefited and so will the SPL level. I will leave it to FMs to decide if they want a thicker wire and a probable increase of cost by atleast Rs.300/-. I have estimated with 20SWG wire.
keith_correa
Well-Known Member
Re: Speaker Kit-Insane value
I'm sorry but I did not understand the bit about the wires. Which wires?
I'm sorry but I did not understand the bit about the wires. Which wires?
Re: Speaker Kit-Insane value
SWG - Standard Wire Guage
AWG - American Wire Guage.
Inductor wire guage - 18 SWG or 20 SWG
SWG - Standard Wire Guage
AWG - American Wire Guage.
Re: Speaker Kit-Insane value
I have 2 parameters in this driver which concerns me for design,
Vifa TG9FD-10-04 3-1/2" Glass Fiber Cone Full Range
The SPL - 84dB/w/m which will require a higher powered amplifier to reach decent sound levels. The compensation network will further lower the sensitivity by 1dB to 2dB. So the max SPL you can expect is around 82dB to 83dB/w/ which as per me is quite low for decent quality speakers. If compensation network is not used it will sound shrill.
RMS power - 5 Watts / Max Power - 10 Watts - This can handle at the most a 20 watts to 25 Watts amplifier after which they may be damaged beyond repair.
To reach the SPL you need higher power amplifier which this driver does not allow. I am still not sure how this driver will be suitable for music applications. If linuxguru has used this driver successfully his comments will be highly valued here.
The Peerless S081E driver does not suffer from these issues because its SPL is 91dB/W/M and its Max Power handling is 100 Watts. I can for sure mention that I have successfully designed with the Peerless and am also using it with my Denon AVR-1604 at high volumes without any strain or issues.
Cheers,
I have 2 parameters in this driver which concerns me for design,
Vifa TG9FD-10-04 3-1/2" Glass Fiber Cone Full Range
The SPL - 84dB/w/m which will require a higher powered amplifier to reach decent sound levels. The compensation network will further lower the sensitivity by 1dB to 2dB. So the max SPL you can expect is around 82dB to 83dB/w/ which as per me is quite low for decent quality speakers. If compensation network is not used it will sound shrill.
RMS power - 5 Watts / Max Power - 10 Watts - This can handle at the most a 20 watts to 25 Watts amplifier after which they may be damaged beyond repair.
To reach the SPL you need higher power amplifier which this driver does not allow. I am still not sure how this driver will be suitable for music applications. If linuxguru has used this driver successfully his comments will be highly valued here.
The Peerless S081E driver does not suffer from these issues because its SPL is 91dB/W/M and its Max Power handling is 100 Watts. I can for sure mention that I have successfully designed with the Peerless and am also using it with my Denon AVR-1604 at high volumes without any strain or issues.
Cheers,
Last edited:
goldyrathore
Active Member
Re: Speaker Kit-Insane value
The low sensitivity makes this driver less suitable for single driver designs.
OTOH, there are a couple of line array designs where these drivers can really perform excellent.
Transmission line array -> http://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Pipeline_copy.pdf
Just ensure that either the line is very long or very small (the listener has to be placed in near field or far field). This applies for just about any straight array.
For the more adventurous, there is Don Keele's excellent CBT array. This can be just about any size in terms of height depending on how low one wants the array to control the pattern. Or one can just clone this http://www.parts-express.com/projec...ct-cbt-36-circular-arc-line-array-loudspeaker
The high performance/price ratio of such good drivers available locally really opens up many project possibilities. I have picked up some very fine OEM driver in the past in the 2.5" form factor. Ive also measured some of the impedance curves and frequency responses. There is very small piece to piece tolerance between these units. :clapping:
A word of caution though, the heavy magnets make shipping very difficult and if proper packing is not done then often the drivers will end up in the garbage. By proper packing I mean double layer packing that can survive the courier guys' drops and throws. I ve burnt my fingers quite a few times.:sad:
The low sensitivity makes this driver less suitable for single driver designs.
OTOH, there are a couple of line array designs where these drivers can really perform excellent.
Transmission line array -> http://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Pipeline_copy.pdf
Just ensure that either the line is very long or very small (the listener has to be placed in near field or far field). This applies for just about any straight array.
For the more adventurous, there is Don Keele's excellent CBT array. This can be just about any size in terms of height depending on how low one wants the array to control the pattern. Or one can just clone this http://www.parts-express.com/projec...ct-cbt-36-circular-arc-line-array-loudspeaker
The high performance/price ratio of such good drivers available locally really opens up many project possibilities. I have picked up some very fine OEM driver in the past in the 2.5" form factor. Ive also measured some of the impedance curves and frequency responses. There is very small piece to piece tolerance between these units. :clapping:
A word of caution though, the heavy magnets make shipping very difficult and if proper packing is not done then often the drivers will end up in the garbage. By proper packing I mean double layer packing that can survive the courier guys' drops and throws. I ve burnt my fingers quite a few times.:sad:
Last edited:
musiklava79
Active Member
Re: Speaker Kit-Insane value
On the contrary, I am currently breaking-in a pair of Fountek FE87 3" FRs in a pair of Mini Cornu Spiral Horn Speakers that are being driven by a Lepai 2020 T-Amp with Audio Cds as the source. These drivers also have similar Sensitivity of 84 db. But I've had no problems with these drivers whatsoever. Incidentally, I've set the Cornus up in my bedroom measuring 16" x 28". Even with the volume set at 10 O'clock or so the speakers sound very loud. I've also played the speakers through the Topping TP20 Mk II Amp, which BTW require speakers with higher sensitivity, but even here, most Cds except those with lower recording levels, can be played at reasonable sound levels, albeit at about 12 to 1 O'clock setting.
I'm burning in the Founteks for mounting in a pair of Tabaq speakers which I'm presently building. The cost of materials so far (12mm plywood and PVA glue) has not exceeded Rs.500.00!
The Tymphany speakers in question will suit both types of speakers, in my opinion.
On the contrary, I am currently breaking-in a pair of Fountek FE87 3" FRs in a pair of Mini Cornu Spiral Horn Speakers that are being driven by a Lepai 2020 T-Amp with Audio Cds as the source. These drivers also have similar Sensitivity of 84 db. But I've had no problems with these drivers whatsoever. Incidentally, I've set the Cornus up in my bedroom measuring 16" x 28". Even with the volume set at 10 O'clock or so the speakers sound very loud. I've also played the speakers through the Topping TP20 Mk II Amp, which BTW require speakers with higher sensitivity, but even here, most Cds except those with lower recording levels, can be played at reasonable sound levels, albeit at about 12 to 1 O'clock setting.
I'm burning in the Founteks for mounting in a pair of Tabaq speakers which I'm presently building. The cost of materials so far (12mm plywood and PVA glue) has not exceeded Rs.500.00!
The Tymphany speakers in question will suit both types of speakers, in my opinion.
Re: Speaker Kit-Insane value
Thanks for chipping in. In fact, Line Arrays are what I and Hydrovac have decided to build and have confirmed our requirement to Linuxguru.
As one of the very few who have experimented with line arrays on HFV, your views would carry a lot of value.
How many drivers per side do you recommend?
Thanks for chipping in. In fact, Line Arrays are what I and Hydrovac have decided to build and have confirmed our requirement to Linuxguru.
As one of the very few who have experimented with line arrays on HFV, your views would carry a lot of value.
How many drivers per side do you recommend?
Re: Speaker Kit-Insane value
hyeah:
Thanks, May be I have lost my listening sensitivity.On the contrary, I am currently breaking-in a pair of Fountek FE87 3" FRs in a pair of Mini Cornu Spiral Horn Speakers that are being driven by a Lepai 2020 T-Amp with Audio Cds as the source. These drivers also have similar Sensitivity of 84 db. But I've had no problems with these drivers whatsoever. Incidentally, I've set the Cornus up in my bedroom measuring 16" x 28". Even with the volume set at 10 O'clock or so the speakers sound very loud. I've also played the speakers through the Topping TP20 Mk II Amp, which BTW require speakers with higher sensitivity, but even here, most Cds except those with lower recording levels, can be played at reasonable sound levels, albeit at about 12 to 1 O'clock setting.
I'm burning in the Founteks for mounting in a pair of Tabaq speakers which I'm presently building. The cost of materials so far (12mm plywood and PVA glue) has not exceeded Rs.500.00!
The Tymphany speakers in question will suit both types of speakers, in my opinion.
hyeah:Re: Speaker Kit-Insane value
To rephrase Rajesh's question, is 4 drivers per side workable for a short line array? That way, one can maintain 6 ohms impedance with series-parallel wiring, and one could (at least in principle) stack 2 of those 4-driver cabinets to get a 8-driver array if needed.
The other update is that the 12-ohm driver seems to be an exception, so those who're designing line arrays are advised to stay with 6 ohms.
In my experience, the SPL with a single 6-ohm driver in a sealed enclosure driven by modest gainclones (MiniRef 1875, 2x 20W) shows that it is more than adequate for normal indoor listening levels (similar to the experience of Musiklava above with small T-amps driving the Founteks).
(I'll reply to the (modest) backlog of PMs and emails in the evening or within a day, please bear with me.)
To rephrase Rajesh's question, is 4 drivers per side workable for a short line array? That way, one can maintain 6 ohms impedance with series-parallel wiring, and one could (at least in principle) stack 2 of those 4-driver cabinets to get a 8-driver array if needed.
The other update is that the 12-ohm driver seems to be an exception, so those who're designing line arrays are advised to stay with 6 ohms.
In my experience, the SPL with a single 6-ohm driver in a sealed enclosure driven by modest gainclones (MiniRef 1875, 2x 20W) shows that it is more than adequate for normal indoor listening levels (similar to the experience of Musiklava above with small T-amps driving the Founteks).
(I'll reply to the (modest) backlog of PMs and emails in the evening or within a day, please bear with me.)
Re: Speaker Kit-Insane value
Do expand on the line array idea. This was the first thought that came to my mind on seeing this thread.
Do expand on the line array idea. This was the first thought that came to my mind on seeing this thread.
Re: Speaker Kit-Insane value
And what would be the impedence if 3 drivers are wired in parallel?
Didn't get you. Are these only three in number?
Yes, looking for various ideas. Do share if you have come across any.
musiklava79
Active Member
Re: Speaker Kit-Insane value
Maybe or maybe not coz I'm no spring chicken either!!
Thanks, May be I have lost my listening sensitivity.
Maybe or maybe not coz I'm no spring chicken either!!
musiklava79
Active Member
I've PM'ed you the link. This is the plan for the 3" FR drivers. You can safely mount a 3.5" driver if you plan to use the Tymphany drivers from the group buy. The speaker hole cut out diameter is suitable for this driver.
Last edited:
Guys, rather than restricting sharing of information via PMs, all the information of nonpersonal nature could be shared on the thread for the benefit of all.
Re: Speaker Kit-Insane value
@Captrajesh,
I could do a quick simulation with published parameters for the 4 ohms drivers for discussion purpose. I understand that after measurements and with 6 ohm driver the situation will be quite different. I have taken the series-parallel connection.
4 drivers in line array below is the simulated response without any compensation network. The roll-off begins at around 2.5KHz as per the simulation. SPL is at 88dB/W/M. Impedance is 3.33 ohms in the linear region. i.e. 2 drivers in 4 ohms in series = 8 ohms and then in parallel = 4 ohms
3 drivers in line arry below is the simulated response without any compensation network. The roll-off again begins at around 3KHz as per the simulation. SPL is at 92dB/W/M. Impedance is 2.33 ohms in the linear region. i.e. 2 drivers in 4 ohms in series = 8 ohms and then in parallel with one 4 ohm driver = 2.66 ohms
Its interesting to note that with 3 drivers the speaker has better sensitivity and extended bandwidth than using 4 drivers. Also the system Q is well controlled in the 3 drivers than the 4 drivers line array. The only catch is the impedance falls by 1 ohm. The extended bandwidth be due to less cancellation of co-incident waves in 3 drivers. I am expecting an additional +3dB lift if the low end from 50Hz to 110Hz region (half octave below resonance of the driver) if the drivers are used in a linear array TL enclosure. It will still be required to be in an advanced TL instead of a classic TL to have realistic dimensions.
Compensation network however will be required for extended bandwidth and also for compensation for the baffle step.
Cheers,
@Captrajesh,
I could do a quick simulation with published parameters for the 4 ohms drivers for discussion purpose. I understand that after measurements and with 6 ohm driver the situation will be quite different. I have taken the series-parallel connection.
4 drivers in line array below is the simulated response without any compensation network. The roll-off begins at around 2.5KHz as per the simulation. SPL is at 88dB/W/M. Impedance is 3.33 ohms in the linear region. i.e. 2 drivers in 4 ohms in series = 8 ohms and then in parallel = 4 ohms
3 drivers in line arry below is the simulated response without any compensation network. The roll-off again begins at around 3KHz as per the simulation. SPL is at 92dB/W/M. Impedance is 2.33 ohms in the linear region. i.e. 2 drivers in 4 ohms in series = 8 ohms and then in parallel with one 4 ohm driver = 2.66 ohms
Its interesting to note that with 3 drivers the speaker has better sensitivity and extended bandwidth than using 4 drivers. Also the system Q is well controlled in the 3 drivers than the 4 drivers line array. The only catch is the impedance falls by 1 ohm. The extended bandwidth be due to less cancellation of co-incident waves in 3 drivers. I am expecting an additional +3dB lift if the low end from 50Hz to 110Hz region (half octave below resonance of the driver) if the drivers are used in a linear array TL enclosure. It will still be required to be in an advanced TL instead of a classic TL to have realistic dimensions.
Compensation network however will be required for extended bandwidth and also for compensation for the baffle step.
Cheers,
Last edited:
@captrajesh
Here you go 3 drivers with my own design compensation network to take care of baffle step and extend the bandwidth of the driver from 2.5KHz to 7KHz. With the compensation the roll-off now is at 7KHz. This will also be a very easy load for the amplifier to drive as the impedance phase shows close to a resistive load to the amplifier. With the network the impedance at the linear region is around 3 ohms.
Here you go 3 drivers with my own design compensation network to take care of baffle step and extend the bandwidth of the driver from 2.5KHz to 7KHz. With the compensation the roll-off now is at 7KHz. This will also be a very easy load for the amplifier to drive as the impedance phase shows close to a resistive load to the amplifier. With the network the impedance at the linear region is around 3 ohms.
Last edited:
Get the Award Winning Diamond 12.3 Floorstanding Speakers on Special Offer