Use a low noise DSP/amp system. That’s the only way. Doesn’t need to be low wattage. The Purifi 1ET400A modules are 450w into 4 ohms. Good for any active build for residential usage.
And MiniDSP finally released a low noise DSP setup. Although limited to a 2 way. A friend of mine ordered one, so I can share feedback after it arrives. I’ll measure it as well to see if it meets the manufacturers specs.
https://www.minidsp.com/products/minidsp-in-a-box/flex
This is an example of why the whole system gain staging has to be considered.
Reduce the voltage gain. You need almost no voltage to make a highly efficient compression driver and horn to be way too loud.
Most amplifiers are not set up for that situation. The buffer or input stages have 26 to 30dB of voltage gain which is not a good match for your situation. Hypex and Purifi amps can have the buffer gain changed through a resistor, but because of the them being fairly small surface mount components doing it yourself needs either the right equipment or some skill and practice to avoid doing damage.
It does but then the input stage is bypassed and the input impedance of the amp stage is low, so the preceding stage needs to be capable of driving it and some voltage gain in the input might still be needed. This is why the whole system has be considered to end up with the best result.
The MiniDSP Flex has a 200ohm output impedance from the balanced outputs. So would be able to drive the 4.4K input impedance of the UN-buffered 1ET400A module just fine. Especially with the light load of the horn to drive. And the 4v output would drive a 130dB efficient horn to very high levels. More than his Ncore amp driven to full output anyways.
As plasnu hasn't said he is using either of those devices I'm not sure how it helps him, but certainly what you outline there should work well.
My point was that where and how the signal is amplified makes a bigger difference with very sensitive drivers, even if the amp and DSP is very quiet that might not be enough with some compression drivers and horns if the amp is set up with a lot of voltage gain for low efficiency drivers.
Yes and thankfully a low cost, and low noise DIY solution can now be had with the Purifi Eval kit and MiniDSP Flex. Likely an 8 channel version coming out next.
Use low noise amp and skip the DSP. My TPA3255 easily makes 100w has about 1uV rms noise floor in audio band. The tweeter has no hiss. It’s silent and cannot tell it is powered on except for LEDs glowing.
Flat thin film power resistors from EBG like EBG UXP-300 with a fan cooled heatsink work very well.
There’s far more advantages to DSP than the noise. And with low cost front ends like the MiniDSP flex on the market, high noise is no longer an excuse to not going with DSP. You also need to subtract the passive crossover parts cost out of the build too. If you go with high end parts, the passive crossover costs more than the MiniDSP flex.
https://purifi-audio.com/vare/spk5-crossover-kit/
It’s limited by the THD+N of the audio interface I use: Focusrite 2i4. When I used regular wire wound resistors, the minimum measured THD was in the 0.01% range with the EBG, I was able to go to limit of measurement interface at around 0.0007%. Much of it described here in the Howto thread for REW. You should check out REW’s measurement capabilities. Not sure what software you use but your plots are not as clear as what REW produces. We are often scratching our heads trying to see what you are showing
https://www.diyaudio.com/community/threads/howto-distortion-measurements-with-rew.338511/
Here is a typical FFT plot from REW:
https://www.diyaudio.com/community/threads/howto-distortion-measurements-with-rew.338511/
Here is a typical FFT plot from REW:
I’m skeptical that it has noise and distortion that could keep up with my analyzer. But it would be cool if it did. I haven’t been able to find anyone on this forum that has an analyzer that can measure down to -116 THD+N, and 100V. So all the measurement threads on here haven’t been helpful.
I‘m using the dScope software. I don’t have any issues reading these THD+N plots.
Last edited:
We are off topic again. But one and see if it helps. I have not been able to find a better or cleaner dummy load for high wattages. Bolt to a CPU cooler with fan.
https://www.ebay.com/itm/ONE-EBG-UX...2349624.m46890.l49286&mkrid=711-127632-2357-0
I am not sure any of this matters as speaker distortion levels are typically much higher.
Even PTT6.5 is higher than distortion in typical amps. It is rather the shape of the harmonic profile in frequency space that is more audible.
https://www.ebay.com/itm/ONE-EBG-UX...2349624.m46890.l49286&mkrid=711-127632-2357-0
I am not sure any of this matters as speaker distortion levels are typically much higher.
Even PTT6.5 is higher than distortion in typical amps. It is rather the shape of the harmonic profile in frequency space that is more audible.
Is there any reason that the 800W version wouldn't be as good? Specs look the same
https://www.ebg-resistors.com/filea...oduktdownloads/01_EBG_Series_UXP-350_2019.pdf
https://www.ebg-resistors.com/filea...oduktdownloads/03_EBG_Series_UXP-800_2019.pdf
Vishay makes an equivalent in 800w/4.7ohm
https://www.digikey.com/en/products/detail/vishay-sfernice/LPSA800H4R70JB/11657781
https://www.ebg-resistors.com/filea...oduktdownloads/01_EBG_Series_UXP-350_2019.pdf
https://www.ebg-resistors.com/filea...oduktdownloads/03_EBG_Series_UXP-800_2019.pdf
Vishay makes an equivalent in 800w/4.7ohm
https://www.digikey.com/en/products/detail/vishay-sfernice/LPSA800H4R70JB/11657781
Last edited:
Yeah the IMD is much more important for sound. But you need a clean dummy load for that as well. I use the THD+N measurement for QC mainly. Because when the design is finalized, using the THD+N test during the QC process tells you if everything is working to spec or not.
I’ll tell you guys a little trick to making a 3 way fully active Purifi based speaker using that 4 channel MiniDSP flex unit. So let’s say you want to make a 3 way using one of the Purifi 6.5” midwoofers, and 1+ Purifi 6.5’s, or the upcoming 8’s as a woofer. What you do is put the mid in a sealed cabinet of 6-12L volume, and let it roll off naturally to the woofers. Then connect one of these speaker level to balanced XLR output converters up to the outputs of the amp driving the mids:
https://www.bhphotovideo.com/c/prod...s_sp_2sx_speaker_to_line_level_converter.html
Feed the outputs of that box into a pair of Hypex Fusion mono plate amps in whatever power level you want. Then use the DSP in the plate amp to tune the woofers in. And use one of the sealed box alignments I shared the simulations of earlier in the thread for best sound. You’ll be able to boost all you want without boosting the mid.
https://www.bhphotovideo.com/c/prod...s_sp_2sx_speaker_to_line_level_converter.html
Feed the outputs of that box into a pair of Hypex Fusion mono plate amps in whatever power level you want. Then use the DSP in the plate amp to tune the woofers in. And use one of the sealed box alignments I shared the simulations of earlier in the thread for best sound. You’ll be able to boost all you want without boosting the mid.
This thread is off topic. Threads are free if you want to start your own.
That's just an indication of extremely poor system design. Why would you use an amp that would produce 150dB at one meter in a domestic environment?? A low gain 100mW amp would be far more suitable. All your test proved is that you are no engineer.
Well if the horn is the tweeter of a passive multiway system, I can see the 125w not being too much power. Horns tend to be sensitive, but 130dB will pull hiss out of the most quiet amps. You could probably hear 5uV rms noise with it! It would need -40dB attenuation to level match most dynamic mid range and woofers.