Most compression drivers will have a much higher sensitivity than the mid range driver you want to use below it. For example I was just looking at the specs of a CD that is 109dB. Suppose for example you have had a win on the horses 😉 and you want to use it with the Purifi 10” which has a sensitivity of 88dB. That is a one hell of a difference to compensate for. If you go active then using an amp on the CD that has a lower gain would get you part of the way there. If you use DSP then you could reduce the gain digitally, but I’ve read that has a negative sonic impact if you go too far with it. Having used an autoformer for volume control I suspect that’s the best way of dealing with any remaining imbalance, but good autoformers are expensive. Consequently you may well end up with an Lpad on the CD.
So this begs the question, do those resistors in the signal path noticeably degrade the sound quality? Are there other undesirable effects? For example, part of the reason horn systems sound so dynamic may be due to their high sensitivity so are you throwing out the baby with the bath water? I’d be very interested hear the thoughts of serial speaker builders. And please don’t just tell me to choose a different midrange driver 🤨, this is all just moot. Besides, I don’t gamble . . . .
So this begs the question, do those resistors in the signal path noticeably degrade the sound quality? Are there other undesirable effects? For example, part of the reason horn systems sound so dynamic may be due to their high sensitivity so are you throwing out the baby with the bath water? I’d be very interested hear the thoughts of serial speaker builders. And please don’t just tell me to choose a different midrange driver 🤨, this is all just moot. Besides, I don’t gamble . . . .
You're overthinking the gain control in an active setup. Many power amps have a gain or volume control, and if yours doesn't it's easy to add one. Even a simple resistor divider, on the input, can be used to level match the CD amp with the Mid amp. There's zero need to use DSP for this function.
Why would the resistors cause a problem? If you've read something somewhere, post a link so we can set things straight.
For example resistors could cause changes to the response or conditions the amp is operating under, but these are not a problem with the resistance per se.
For example resistors could cause changes to the response or conditions the amp is operating under, but these are not a problem with the resistance per se.
Your post reads as though you are stating a settled fact whereas it is in fact a controversial statement given the number of people who contend that every component in the signal path has an impact on sound quality to some degree. I’m sure that’s not news to you. Personally, I don’t have enough information to form a view, hence my opening post.
There could be differences, my intention was to challenge what the cause of them may be.
The effects of components are well known and measurable, and many of them happen at frequencies higher than the audio band.. but there are also issues of implementation where minor variations in values are not taken into account when making an assessment. Another along those lines is where capacitors with differing parasitic resistance are used. Then of course, the amplifier might be in a range where distortions are different, or where current needs are challenging the supply.
If you can find some text which states that resistors are a problem, it would be interesting to get to the bottom of why that is..
The effects of components are well known and measurable, and many of them happen at frequencies higher than the audio band.. but there are also issues of implementation where minor variations in values are not taken into account when making an assessment. Another along those lines is where capacitors with differing parasitic resistance are used. Then of course, the amplifier might be in a range where distortions are different, or where current needs are challenging the supply.
If you can find some text which states that resistors are a problem, it would be interesting to get to the bottom of why that is..
Last edited:
IMHO padding down the driver passively has some advantages in comparison to direct/active drive: Most significant is that noisefloor of amp (often an issue with high sensitivity CDs) is reduced also the same. Minor effect - especially by padding with series resistor - is introduction of local current feedback that reduces system power compression / thermal modulation (not very significant for domestic use, can be an advantage when the driver gets really pushed to the power limit) and current distortion effects (also often not very significant as modern Neodymium CDs have saturated iron circuit indicated by B in the range 2-2.2 Tesla, so current distortion is minimized by design). Take a good resistor, most technically ideal is iron free wire resistor with bifilar windings, or metal foil resistor.
Which is why you wouldn't begin to entertain that idea.
Every part of a speaker is designed with the others in mind.
Never mind the effect of resistors, just get 16 of those Purifi woofers per side and be done with it.
Are you going to put monster truck tires on your Corolla?
Match, don't fight.
Compression driver impedance is non-flat in the passband, so a series resistor or an L-pad will change the frequency response. See here for examples:
https://audioxpress.com/article/tes...dymium-compression-driver-from-eighteen-sound
https://audioxpress.com/article/tes...ompact-1-compression-driver-from-b-c-speakers
If you're matching one of those to a direct radiator, you could minimise the damage by paralleling the CD with a 1 or 2 ohm ohm resistor before adding the series R. Or you could go with an active crossover.
https://audioxpress.com/article/tes...dymium-compression-driver-from-eighteen-sound
https://audioxpress.com/article/tes...ompact-1-compression-driver-from-b-c-speakers
If you're matching one of those to a direct radiator, you could minimise the damage by paralleling the CD with a 1 or 2 ohm ohm resistor before adding the series R. Or you could go with an active crossover.
@Studley,
So we have a 20db difference with the drivers only, without even talking about voltage gain into amp ( along 25db range).
Then i would go multiamp with a power buffer for your CD ( eg: Passlab F4). From there the voltage gain will be defined by your dac drive/output voltage.
Put your pad between dac output and your power buffer so no impedance issues (compromised FR) as it's bridging drive at line level.
I run something like that for more than 10 years now without issues (but not with CD drivers).
The only downside* would be in Jonhson noise which will be a bit higher than without analog pad, but you'll run your dac without compromising on digital bits ( dynamic range) as they can drive their full upward dynamic range ( to 0dbfs)... it's a compromise on noise front, either analog ( a tiny bit of hiss) or digital ( quantification noise (and dithering) and loss of dynamic resolution at digital stage).
I made my choice. 😉
*if we except the multiamp architecture which can be an issue by itself regarding whole system's ergonomic (volume control if you want a one knob analog solution, having to turn on/off multiple amps + loudspeaker management unit (dsp or computer+soundcard) in incremental/decremental order,...).
So we have a 20db difference with the drivers only, without even talking about voltage gain into amp ( along 25db range).
Then i would go multiamp with a power buffer for your CD ( eg: Passlab F4). From there the voltage gain will be defined by your dac drive/output voltage.
Put your pad between dac output and your power buffer so no impedance issues (compromised FR) as it's bridging drive at line level.
I run something like that for more than 10 years now without issues (but not with CD drivers).
The only downside* would be in Jonhson noise which will be a bit higher than without analog pad, but you'll run your dac without compromising on digital bits ( dynamic range) as they can drive their full upward dynamic range ( to 0dbfs)... it's a compromise on noise front, either analog ( a tiny bit of hiss) or digital ( quantification noise (and dithering) and loss of dynamic resolution at digital stage).
I made my choice. 😉
*if we except the multiamp architecture which can be an issue by itself regarding whole system's ergonomic (volume control if you want a one knob analog solution, having to turn on/off multiple amps + loudspeaker management unit (dsp or computer+soundcard) in incremental/decremental order,...).
Last edited:
Yup, it could a rude awakening to fire up that awesome CD/horn to discover the system's true noise floor, ground loops, EMI/RFI, tube noise, etc. I've used things like TKD carbon pot to fancy ladder resistor attenuator at amp input with no subjective sonic degradation. In fact, due to decreased background noise, it sounds better!
Dynamics aren't lost with level matching.
Horn does what it does.
In live audio at blasting SPL the only dynamic change can be distortion.
88/89 dB woofer with baffle step and compensation
will loose even more.
Crossover is low for directivity.
Large horn and suitable driver, doubt 109 dB
more like 102 to 104 dB
Better to ignore everything in signal path is horrible crowd.
Since a audio chain will have a lot regardless.
passive, resistors are fine. Active, none needed nor transformer.
People with 1 to 3 watt listening levels aren't plagued with unbearable noise = nonsense.
Horn does what it does.
In live audio at blasting SPL the only dynamic change can be distortion.
88/89 dB woofer with baffle step and compensation
will loose even more.
Crossover is low for directivity.
Large horn and suitable driver, doubt 109 dB
more like 102 to 104 dB
Better to ignore everything in signal path is horrible crowd.
Since a audio chain will have a lot regardless.
passive, resistors are fine. Active, none needed nor transformer.
People with 1 to 3 watt listening levels aren't plagued with unbearable noise = nonsense.
In a word no. Pot's can get noisy so you hit them we deox or just used fixed attenuation to avoid the issue. When I bi-amp I always have a pad on the compression driver makes it easier to gain match and reduces hiss.
Rob 🙂
Not a deal breaker. I have a system that pads down the CD to 89db. No issues at all.
Rob 🙂
Last edited:
88dB sensitivity is the figure for the 4 ohm version of the Purifi 10", at 2.83V. The 8 ohm version is 84.2dB at 2.83V. If you wanted baffle step compensation, take 4dB (maybe) off those figures.
I'd say that the Purifis don't really follow that philosophy.
IME, resistors are colouring the sound (every type its own way) but that even can be heard as an improvement subjectively.
Resistors doesn't change the own efficiency of a driver, they just change the net efficiency/sensitivity, but since resistors does reduce the noise that coming from the amp (whereas active level matching without resistors usually doesn't), the S/N ratio are improved with them, compared to active level matching.
Last edited:
If these things are issues padding down the output and pretending they don't exist is probably not the best solution IMHO. I have 107db horns with no attenuators and absolutely zero detectable noise hum or hiss with my ear inside the mouth of the HF horns. This wasn't easy to accomplish with tube amplification and a tv/PC in the loop but it's possible.