Interesting! I agree that precisely centering a diaphragm is annoying, and I only know of the tedious adjust-and-listen-to-sinewaves method.
So what does the impedance curve say about diaphragm (or rather, voice-coil) alignment? If it's off-center, how does it look different? I'd like to know!
I can see running the current measurement through a distortion analyzer, and looking for residual asymmetries in the waveform, but I'd really like to know what the frequency vs impedance curve reveals.
Thanks in advance!
Unfortunately, I have not saved the trial and error process. But the reference impedance curve is a really good help.
Start point was a sinus about two times the frequeny at impedance maximum. Voltage was really low. ClIO pocket has a high output impedance but we do not want to damage the driver or our ears
.If I remember well it was about 0.1 to 0.2 Volts. Then a clear sinus could be heard but the following impedance measurement showed a different curve with another maximum. I searched alot in the web and found some people who reported the same. If I really remember well all other positions than "perfect" one showed a higher center frequency of the maximum resonance. In our example if the maximum is around 630 Hz not perfectly aligned showed 660 Hz. But do not treat these absolute numbers as totally confident.
But about what I am very confident is that at the perfect position the impedance curve shows the highest peak at all. So if you change the position and the peak becomes higher this position is better centered than that before.
It would be nice if you could give a feedback next time you align a diaphragm following this procedure if it worked for you.
Hi Lynn!
I have found the source for this procedure very well documented. Unfortunately, in german:
Kalottenhochtoner reparieren - Austausch der Spule?
But the pictures should speak for themselves. All the credits go to this guy! And I can absolutely confirm what he wrote.
So, it is quite easy to recognize the correct alignment but not how to find it mechanically
I have found the source for this procedure very well documented. Unfortunately, in german:
Kalottenhochtoner reparieren - Austausch der Spule?
But the pictures should speak for themselves. All the credits go to this guy! And I can absolutely confirm what he wrote.
So, it is quite easy to recognize the correct alignment but not how to find it mechanically
What I always block the driver acoustically by placing it face down. This nulls any acoustic effects except at the very highest frequencies. It is the most stable way to compare driver only effects through the impedance curve.
One can extract all the relevant info from a compression driver with a few tests like this. Another is a back off test face down.
Do I understand correctly that you measure the driver terminated with it's opening against the infinite impedance of the table live a closed box? PWT Sounds more logical to know more Abort the driver in nearly perfect environment.
Good suggestion; simple and effective. What's your impression of using the magnitude of the Zmax to assess centering of the voice coil?
R2R DAC - first impressions
Just by way of a quick follow-up to my previous thread(s) about the RATOC multi-bit R2R dac:
it arrived from Japan (I found one brand new in box for a great price), and I installed it in my chain, feeding it via SPDIF from my network player, while retaining connection of the latter to the preamplifier, too, so as to be able to quickly A/B the multi-bit DAC to the built-in Sigma-delta DAC of the network player itself.
In all honesty, I must say first and foremost that the difference is SUBTLE.
MUCH subtler, for instance, than the one I experienced when I switched from my previous class AB (mostfet) integrated to my current pure class A (mosfet) power amp. So, all those who are convinced that "all amps sound the same", you can stop reading right here.
And of course, MUCH, MUCH, MUCH subtler than any readily-apparent change brought about by tweaking the passive crossover of the loudspeakers.
OK, so with that out of the way... what are these tiny differences between the two dacs?
For lack of a better way of expressing it, the RATOC multi-bit dac sounds slightly more "organic" and "fluid" and less "showy" and with less "digital glare".
On some tracks, though, it also seems to sound a smidgen less dynamic (which may be a psychoacoustic effect due to its "smoother" sound, I'm not sure).
In the end, I am satisfied with my purchase... BUT, would I recommend it as an upgrade to the more common run-of-the-mill sigma-delta chips found in most players and standalone dacs today?
Not sure. Not so quick. In fact, the differences are so subtle that I could happily live with either and never look back.
In other words, in all honesty, I would roughly give the following % as to what contributes to obtaining a "great" sounding chain (excluding the room, which is VERY important but deserves a separate thread):
- Loudspeakers (incl. choice of drivers, crossover, etc.): 90%
- Amplifier (incl. valve vs. solid state, class A vs. B vs. D, etc.): 7-8%
- DAC (incl. multi-bit vs. sigma-delta, and assuming in all cases a good jitter-free source/transport): 2%
- Cables (incl. line-level and speaker-level): 1% or less.
I think this apportionment (as crude as it is) should go a long way towards setting one's priorities on where to focus one's efforts.
Marco
Just by way of a quick follow-up to my previous thread(s) about the RATOC multi-bit R2R dac:
it arrived from Japan (I found one brand new in box for a great price), and I installed it in my chain, feeding it via SPDIF from my network player, while retaining connection of the latter to the preamplifier, too, so as to be able to quickly A/B the multi-bit DAC to the built-in Sigma-delta DAC of the network player itself.
In all honesty, I must say first and foremost that the difference is SUBTLE.
MUCH subtler, for instance, than the one I experienced when I switched from my previous class AB (mostfet) integrated to my current pure class A (mosfet) power amp. So, all those who are convinced that "all amps sound the same", you can stop reading right here.
And of course, MUCH, MUCH, MUCH subtler than any readily-apparent change brought about by tweaking the passive crossover of the loudspeakers.
OK, so with that out of the way... what are these tiny differences between the two dacs?
For lack of a better way of expressing it, the RATOC multi-bit dac sounds slightly more "organic" and "fluid" and less "showy" and with less "digital glare".
On some tracks, though, it also seems to sound a smidgen less dynamic (which may be a psychoacoustic effect due to its "smoother" sound, I'm not sure).
In the end, I am satisfied with my purchase... BUT, would I recommend it as an upgrade to the more common run-of-the-mill sigma-delta chips found in most players and standalone dacs today?
Not sure. Not so quick. In fact, the differences are so subtle that I could happily live with either and never look back.
In other words, in all honesty, I would roughly give the following % as to what contributes to obtaining a "great" sounding chain (excluding the room, which is VERY important but deserves a separate thread):
- Loudspeakers (incl. choice of drivers, crossover, etc.): 90%
- Amplifier (incl. valve vs. solid state, class A vs. B vs. D, etc.): 7-8%
- DAC (incl. multi-bit vs. sigma-delta, and assuming in all cases a good jitter-free source/transport): 2%
- Cables (incl. line-level and speaker-level): 1% or less.
I think this apportionment (as crude as it is) should go a long way towards setting one's priorities on where to focus one's efforts.
Marco
Last edited:
I think your assessment is "in the ballpark."
Although I would add that the speaker and amp matching is an important quotient. AND lets not forget the quality of the recording which may trump everything.
Me thinks the quote "it's the speakers stupid" could be injected into 95% of all discussions regarding audio.
The problem with speakers is that they are so damn complicated and amps and turntables are cooler.
Glad you agree.
True. Especially in regard to how the frequency response is affected by the output impedance of the amplifier. With valve (tube) amps having non-negligible output impedance (often 1 ~ 3 Ohm), any deviation from a flat resistive load is going to be "mirrored" in a similar deviation of the frequency response. And THIS will be audible, no doubt.
100% agreed.
Although I'd argue that "cool" is in the eye of the beholder.
+1 for audio money investment proportion. For low budget friends wanting a better sound, I always recommend to start with speakers. That said, most people like amplifiers due to that Ogie said: is easier...
I spend more time with amplifiers than speakers and else, but not money: most money goes to speakers, at least .
marco_gea (about DAC),
For me also the DAC are things with not huge sound differences, but in long run and in extended time listenings, the subtle difference contributes for less listening fatigue.
I spend more time with amplifiers than speakers and else, but not money: most money goes to speakers, at least
.marco_gea (about DAC),
For me also the DAC are things with not huge sound differences, but in long run and in extended time listenings, the subtle difference contributes for less listening fatigue.
Marco, have you tried comparing the frequency responses of the two DACs? The "digital glare" versus "smooth/subdued" subjective description fits what a slight roll off in the upper frequency range might sound like. Like you said, it was subtle. Might be worth measuring.
I'm convinced that almost everything we hear can be captured by the humble frequency response. Or least, a lot of what we hear can be captured in the FR.
Last edited:
Didn't realize that "trump" is also a verb
Use it all the time. Came from watching my father play Sheepshead, a card game. They might have stolen that from Bridge, which likely came first. Generally I apply it when overriding someone's stance with one having better support or a stronger position. English is weird, as I suppose all languages are.
Didn't realize that "trump" is also a verb
Since 1520 or thereabouts, as used above; since 1230 with its other meaning as a verb. Dictionary.com
Last edited:
Multi-bit vs. delta-sigma? Or is it simply the Frequency response??
Now this was interesting.
Intrigued, I went and measured the frequency response of my system when reproducing pink noise (24bit/96kHz WAV, FR flat out to 48kHz), first
- (A) using the RATOC (R2R multi-bit) dac, and then again
- (B) using my network's built-in (delta-sigma) dac as a point of reference.
For both measurements I used my calibrated Beyerdynamic MM1 mic and the TrueRTA software package. My mic preamp/ADC only works up to 48kHz sampling rate, so that is what I used (the mic itself isn't calibrated beyond 22kHz anyway).
I then computed the DIFFERENCE between measurements (A) and (B), and this is what came out (see attachment).
Basically, the RATOC dac seems to employ a "smooth roll-off" low-pass filter rather than the more common/standard "brickwall" filter such as the one that is used in my "reference" dac, and this results in a frequency response with a gentle slope past ~12kHz which leads to a subtle attenuation of approx. -1dB at 20kHz.
Might this small frequency response difference be the main (only?) reason for the subjectively slightly "smoother" sound of the RATOC dac? (Rather than the whole multi-bit vs. delta-sigma thing?)
Interesting!
Marco
Now this was interesting.
Intrigued, I went and measured the frequency response of my system when reproducing pink noise (24bit/96kHz WAV, FR flat out to 48kHz), first
- (A) using the RATOC (R2R multi-bit) dac, and then again
- (B) using my network's built-in (delta-sigma) dac as a point of reference.
For both measurements I used my calibrated Beyerdynamic MM1 mic and the TrueRTA software package. My mic preamp/ADC only works up to 48kHz sampling rate, so that is what I used (the mic itself isn't calibrated beyond 22kHz anyway).
I then computed the DIFFERENCE between measurements (A) and (B), and this is what came out (see attachment).
Basically, the RATOC dac seems to employ a "smooth roll-off" low-pass filter rather than the more common/standard "brickwall" filter such as the one that is used in my "reference" dac, and this results in a frequency response with a gentle slope past ~12kHz which leads to a subtle attenuation of approx. -1dB at 20kHz.
Might this small frequency response difference be the main (only?) reason for the subjectively slightly "smoother" sound of the RATOC dac? (Rather than the whole multi-bit vs. delta-sigma thing?)
Interesting!
Marco
Attachments
Last edited: