Hi BarG01,
It is based on one Tangband W8-2145 8" full range driver:
W8-2145 - 8" Paper Full Range - TB SPEAKER CO., LTD.
And two Eminence Alpha 15A:
https://www.eminence.com/pdf/Alpha_15A.pdf
Crossover filter:
W8-2145 (no filtering).
2 x Eminence Alpha 15A in parallel, 4.7mH choke.
This crossover filter was designed for use with the PowerDAC.
Impedance 2.67 ... 8 Ohm.
Crossover frequency (bass) 135 Hz.
Sensitivity: 90dB.
Height: 122 cm
Width: 58cm
depth: 46cm
It is constructed from 12mm and 18mm plywood.
I attached pictures of the speaker construction. The boxes that hold 6th order Linkwitz Riley low / hi pass filters are no longer required.
We plan to offer these open baffle speakers for use with our PowerDACs. These speakers will be introduced together with the PowerDAC, introduction is planned for late 2021.
It is based on one Tangband W8-2145 8" full range driver:
W8-2145 - 8" Paper Full Range - TB SPEAKER CO., LTD.
And two Eminence Alpha 15A:
https://www.eminence.com/pdf/Alpha_15A.pdf
Crossover filter:
W8-2145 (no filtering).
2 x Eminence Alpha 15A in parallel, 4.7mH choke.
This crossover filter was designed for use with the PowerDAC.
Impedance 2.67 ... 8 Ohm.
Crossover frequency (bass) 135 Hz.
Sensitivity: 90dB.
Height: 122 cm
Width: 58cm
depth: 46cm
It is constructed from 12mm and 18mm plywood.
I attached pictures of the speaker construction. The boxes that hold 6th order Linkwitz Riley low / hi pass filters are no longer required.
We plan to offer these open baffle speakers for use with our PowerDACs. These speakers will be introduced together with the PowerDAC, introduction is planned for late 2021.
Attachments
Do you know how well the new fractal DAC architecture fares in its measured performance (THD+N, IMD) ?
There was some measurements of the older MOS DACs on SBAF and the measurements were a little disappointing but im hoping this was improved upon with fractal.
Hi John,
I don't know anything about your fractal DAC so I cannot argue, but the H bridge configuration suffers the same issues of a complementary pair in a R2R ladder: all the mosfets will never switch at the same time, that means huge current spikes (several amperes) on the power rail, and this inevitably generates unwanted noise.
I have thought to drive each gate separately nulling the delay with an FPGA, but this complicates a bit the design and mostly it injects jitter in the switches. And the H bridge configuration is more complex to cure than a mosfet pair.
How did you solve this issue (obviously if you want share your solution)?
It's very similar to the speaker system I'm building, if ever I will find the time to finish it.
A couple of 15" driver for the bass section, the AE LO15. Then a fullrange driver, the Supravox 165 GMF. And finally the Great Heil to refine the high section.
The sensitivity is around 97 dB.
I use a minimal crossover although very different from yours. The AE drivers operates up to around 100 Hz, but I cut them with acoustical filter rather than electric. Then a single series capacitor for the Supravox to avoid very low frequencies that only generate distortion. Again there is an acoustical filter at about 7 kHz for the fullrange. Finally a single capacitor for the AMT driver that could be damaged by high escursion.
After years of listening I believe that the Open Baffle is the only way to get realistic sound.
Attachments
Supravox better than the chineese L CAO alnico or the EMS-speakers ?
I surmise the open bafle is very widthy to have a so low cut-off and the overlapp between the bass and the little 165 from Supravox (which has no a big x-max by design) : many good testimonials about the Supravox but many finally given up it for compression driver + mid-bass horn. But the EMS-speaker driversare very special (very clear). No look at PHL Audio with higher X-Max ?
I surmise the open bafle is very widthy to have a so low cut-off and the overlapp between the bass and the little 165 from Supravox (which has no a big x-max by design) : many good testimonials about the Supravox but many finally given up it for compression driver + mid-bass horn. But the EMS-speaker driversare very special (very clear). No look at PHL Audio with higher X-Max ?
Last edited:
The Supravox 165 GMF has very nice mid, one of the best I ever heard. It's more linear than the L CAO, while I have never listen to the EMS drivers.
The baffle will be enough wide but the excursion of the Supravox will be limited by a high pass filter with a single capacitor. I think to cross it at around 100 Hz, but the best crossover frequency will be experimentally derived.
Yes Andrea, I agree this particular Supravox (and the 16") has something quite nice. Papers cone from PHL seems better though but mayve not as FR than the 165 GMF. I believe the new Fertins with EMS-Speaker to be a little above old SUPRAVOX -mms very low that causes break-ups in FR which they are made for -
think if you HP with a single cap you should have to cut-off high and it will suffer more with the natutal acoustic short of OB in these frequencies. SO certainly the bass driver must overlap and be a little Linkwitz-transform helped.
ok. Sorry John for the little disgress.
think if you HP with a single cap you should have to cut-off high and it will suffer more with the natutal acoustic short of OB in these frequencies. SO certainly the bass driver must overlap and be a little Linkwitz-transform helped.
ok. Sorry John for the little disgress.
Last edited:
Hi Andrea,
Interesting project. Are you sharing the progress of this OB design by any chance on another thread? Thanks
So this is the reason you limit the DAC's to 96khz .
Have you not found a way to circumvent the DIR9001 as it seems this bottlenecks and limits all performance of a DAC .
Hi ecdesigns,
what do you think about to get completely rid of S/PDIF and develop a device similar to the MSB PRO USB:
Pro USB Adapter - MSB Technology
This device is also USB bus powered.
Thank you
Matt
Last edited:
Just the way I have suggested in post #7686
Building the ultimate NOS DAC using TDA1541A
except for the synchronous mode.
Using a FIFO all the sources including USB can work asynchronously.
The goal is slaving the FIFO to the clock of the DAC, namely to a SOTA oscillator.
Building the ultimate NOS DAC using TDA1541A
except for the synchronous mode.
Using a FIFO all the sources including USB can work asynchronously.
The goal is slaving the FIFO to the clock of the DAC, namely to a SOTA oscillator.
Having heard the new EC designs U192 and Fractal DAC with ElectroTos there is no sense of any limitation or bottleneck. The way this combo renders music has to be heard.
It's not true, please read accurately my post #7686 and then tell me how the jitter of the source can affect the DAC.
The LRCK is fully isolated from the source, it comes directly from the dividers via copper, so the only jitter that affects the DAC comes from the oscillators/dividers.
I have already published the phase noise plot of the best oscillator ever for digital audio, that doesn't add any jitter to the DAC.
The high jitter of BCK and DATA can be neglected because they does not affects the DAC timing.
And no EMI/RFI can reach the DAC from BCK and DATA because the optic fiber is a real brick wall.
The only issue is that such this way is a little expensive.
Hi matthiasw,
I tried that before, many times, absolutely convinced that direct I2S interface is the best way to go. I tried every practical isolator and every practical way of combining master clocks & FIFO re-clocker.
The end result of these exhaustive tests is the ElectroTos interface and the ElectroTos low jitter protocol.
I will list its advantages again:
- Perfect galvanic insulation (LED -> Toslink optical receiver, very low coupling capacitance).
- Low jitter (coaxial interlink, unique low jitter protocol).
- Extra jitter reduction inside the DAC with modified PLL.
- Fewest practical signal changes to transmit a given amount of data (minimum data bandwidth required, lowest switching noise energy).
- Band-limited output (20MHz, greatly reduced noise injection, greatly reduced issues with parasitics inside the DAC circuits).
- Jitter / noise is tackled -before- the signal is routed to the DAC, this way it can't spread.
The problem with S/PDIF coaxial interfaces:
- Ground loop noise injection (insulation transformers have high input-output coupling capacitance).
- Unlimited data bandwidth (maximum noise injection).
- Protocol promotes high jitter.
The Problem with Toslink interfaces:
- High jitter
- Issues with reflections (optical).
- Protocol promotes high jitter.
The problem with USB interfaces:
- Unlimited data bandwidth (maximum noise injection).
- (RF) ground loop noise.
- Not designed for digital audio, highly problematic.
The problem with I2S interfaces:
- Unlimited data bandwidth (maximum noise injection).
- (RF) ground loop noise.
- Not designed for external DACs, only chip to chip communication. I2S stands for Inter-IC Sound.
I tried that before, many times, absolutely convinced that direct I2S interface is the best way to go. I tried every practical isolator and every practical way of combining master clocks & FIFO re-clocker.
The end result of these exhaustive tests is the ElectroTos interface and the ElectroTos low jitter protocol.
I will list its advantages again:
- Perfect galvanic insulation (LED -> Toslink optical receiver, very low coupling capacitance).
- Low jitter (coaxial interlink, unique low jitter protocol).
- Extra jitter reduction inside the DAC with modified PLL.
- Fewest practical signal changes to transmit a given amount of data (minimum data bandwidth required, lowest switching noise energy).
- Band-limited output (20MHz, greatly reduced noise injection, greatly reduced issues with parasitics inside the DAC circuits).
- Jitter / noise is tackled -before- the signal is routed to the DAC, this way it can't spread.
The problem with S/PDIF coaxial interfaces:
- Ground loop noise injection (insulation transformers have high input-output coupling capacitance).
- Unlimited data bandwidth (maximum noise injection).
- Protocol promotes high jitter.
The Problem with Toslink interfaces:
- High jitter
- Issues with reflections (optical).
- Protocol promotes high jitter.
The problem with USB interfaces:
- Unlimited data bandwidth (maximum noise injection).
- (RF) ground loop noise.
- Not designed for digital audio, highly problematic.
The problem with I2S interfaces:
- Unlimited data bandwidth (maximum noise injection).
- (RF) ground loop noise.
- Not designed for external DACs, only chip to chip communication. I2S stands for Inter-IC Sound.
Looks like an hybrid between the Bufallo teleporter device and what we can see everywhere with ime filtering and isolator chips... nothing special imho reading the link but one has to test in to really know of course.