I haven't tried current drive on compression drivers but have used it on direct radiators. There are some subtle differences between constant current and constant voltage.
The primary change is that the voltage will track the impedance curve. Notionally, imagine the normal constant voltage frequency response of the compression driver and then think of adding an EQ curve that exactly follows the impedance curve. That is what you will get for frequency response.
A second aspect is that there will be minimized change with temperature. As the compression drive heats up under high drive its impedance will rise. With constant voltage drive level will drop as resisitance rises. With constant current it will stay constant.
Distortion effects will be a mixed bag. With conventional drivers the only distortion difference I saw were down to the current distortion causes by magnet material hysteresis effects. None of the excursion related (BL vs. X or stiffness vs. X) effects were touched by current drive, once the changed frequency response was taken into account.
Since the largest distortion factor in a compression driver is 2nd harmonic from air nonlinearity, I wouldn't expect to see a difference in distortion due to current drive.
Knowing how messy the impedance curves of horn/compression driver combos are, I can't imagine that current drive would be an improvment. I know for a fact that the frequency response of some horns that I am familiar with would be distinctly worse. That is, where the impedance bumps (and level will rise under current drive) tends to correspond with frequencies where there are already acoustical peaks.
Regards,
David S
The primary change is that the voltage will track the impedance curve. Notionally, imagine the normal constant voltage frequency response of the compression driver and then think of adding an EQ curve that exactly follows the impedance curve. That is what you will get for frequency response.
A second aspect is that there will be minimized change with temperature. As the compression drive heats up under high drive its impedance will rise. With constant voltage drive level will drop as resisitance rises. With constant current it will stay constant.
Distortion effects will be a mixed bag. With conventional drivers the only distortion difference I saw were down to the current distortion causes by magnet material hysteresis effects. None of the excursion related (BL vs. X or stiffness vs. X) effects were touched by current drive, once the changed frequency response was taken into account.
Since the largest distortion factor in a compression driver is 2nd harmonic from air nonlinearity, I wouldn't expect to see a difference in distortion due to current drive.
Knowing how messy the impedance curves of horn/compression driver combos are, I can't imagine that current drive would be an improvment. I know for a fact that the frequency response of some horns that I am familiar with would be distinctly worse. That is, where the impedance bumps (and level will rise under current drive) tends to correspond with frequencies where there are already acoustical peaks.
Regards,
David S
Hi Dave. Our Messr Le Cléac'h did some testing with the TD-2001 that showed considerably less distortion and somewhat flatter frequency response than with voltage drive:
http://www.diyaudio.com/forums/mult...ver-qts-you-cant-tuna-fish-2.html#post2745626
The results are intriguing, to say the least. As you said, I would have expected huge bumps in the frequency response, following the impedance curve.
I observed a similar phenomenon when adding series resistance to my TD-2002s; the frequency response remained more or less unchanged. I was quite surprised, but have no explanation.
http://www.diyaudio.com/forums/mult...ver-qts-you-cant-tuna-fish-2.html#post2745626
The results are intriguing, to say the least. As you said, I would have expected huge bumps in the frequency response, following the impedance curve.
I observed a similar phenomenon when adding series resistance to my TD-2002s; the frequency response remained more or less unchanged. I was quite surprised, but have no explanation.
Last edited:
Those are interesting curves. It looks like about 10dB reduction in 2nd and some in 3rd as well. Frequency response is a wash between the two. I assume the zero level is the same for both?
I'm surprised the 2nd harmonic changed. For current drive to matter it means that the actual current flowing through the driver is being distorted (not just some mechanical aspects of the driver model) and so linearizing current reduces distortion. As I said, I only found that in conventional drivers for hysteresis effects in magnet material.
Regards,
David
I'm surprised the 2nd harmonic changed. For current drive to matter it means that the actual current flowing through the driver is being distorted (not just some mechanical aspects of the driver model) and so linearizing current reduces distortion. As I said, I only found that in conventional drivers for hysteresis effects in magnet material.
Regards,
David
I've tried pure current drive with the CD of my Beyma coaxials (basically a CP755Ti) and got mixed results. Down low voltage drive fared better (and increase of damping to lower than it's own Re did even better), probably the suspension is very nonlinear/progressive and current drive eliminates any corrective feedback in the driver.
Ended up with what I also like to use for cone drivers, a -k*Re + j*w*L drive impedance profile, k ~ 0.5. Not usable for high power applications, for exaggerated power compression and truly nasty distortion when overloaded (too high excursion).
I don't get when people say a non-flat / ragged reponse is a disadvantage, since you gotta EQ anyway, so what.
Haven't tried true ribbons, but folded ribbons (AMTs) did significantly improve with pure current drive.
Ended up with what I also like to use for cone drivers, a -k*Re + j*w*L drive impedance profile, k ~ 0.5. Not usable for high power applications, for exaggerated power compression and truly nasty distortion when overloaded (too high excursion).
I don't get when people say a non-flat / ragged reponse is a disadvantage, since you gotta EQ anyway, so what.
Haven't tried true ribbons, but folded ribbons (AMTs) did significantly improve with pure current drive.
Well, that's part of the mystery here. If you look at Le Cléac'h's results, there's no evidence of that. The response marches right across, unperturbed by the impedance peaks.
Maybe it was distortion coming from the amp itself. Adding lpad that would give the same level reduction than the single resistor would then cause the same result.
I have done some testing with inline resistors to simulate current drive with voltage amps.
Midwoofers did benefit somewhat with a reduction in 3dr order distortion (and more so with an "old" motor, without shorting ring). Compression driver did not benefit at all, expect for the aforementioned reduction of the distortion comming from the amp, but no more than what a lpad or an autoformer could do.
Could be, although generally the impedance of a compression driver still has substantial peaks, even on a good horn.
Imagine a setup where you put a small resistance in series with the driver (say 0.1 ohm). Now lets drive the unit with constant voltage and sine waves and measure distortion of the voltage we see across the 0.1 ohm. This would be distortion in the current flowing through the driver.
If the distortion you measure in the current is significant then you have the potential of reducing driver distortion with current drive.
If it isn't you don't.
Beyond that, every difference you see will be due to altered frequency response. For example most curves I've seen by others current drive a woofer and so the Q goes way up at resonance. That means excursion is up and hence distortion is up. They proclaim "a big difference due to current drive".
Read through the thread referenced above (the whole thread is good, especially my contributions
) and you will see curves where I took measurements with a woofer under the control of a compressor (feedback) loop to hold level constant between voltage drive and current drive. This is essential to making a real comparison. Otherwise the comparison is meaningless.Now if you really want to reduce distortion in a compression driver, then you should experiment with cancelling the 2nd harmonic with an out of phase but same level amplifier 2nd harmonic (say a single ended triode amplifier). But you better hurry up before Earl jumps in and reminds you that compression drive nonlinearity is immaterial!
Regards,
David
I cannot see how current drive on a compression driver would be any advantage. And I agree with Dave, we should see some rather pronounced changes in the frequency response between voltage driver and current drive, unless there is a substantial crossover in the way, which would tend to null out any response changes. But the kinds of nonlinearities in a compression driver should not have much effect from current driver. So I have doubts about any test that shows that it does. Since everyone here who has done some measurements seems to get different results, I suspect that there isn't really much of an effect and that we are simply seeing differences in the measurements not differences in the results.
I seemed to have missed the tests that showed that nonlinearity was audible.
If any nonlinearity is going to be inaudible then second order is certainly the most likely.
Just spinning a thought,
could it be that current drive does that intrinsically because one mainly injects force on the air, instead of establishing mainly a diaphragm position (actually a velocity) that in turn injects a force?
I don't think I was. I would like to use current driver for my midbass (TAD TM1201H, with an old motor design), but I cannot find any readily available amplifier on the market that could do that, and adding a resistor is just really a waste of power.
I am using autoformer on my compression drivers, with active filtering, in order to reduce distortion (and mainly background noise in fact) from my amplifiers.
I can (re)post measurement of my midbass with or without inline resistors if needed, as well as the same compression driver with inline resistor, lpad, and autoformer
Hello,
Mills and Hawksford used to write a good paper on the distortion reduction phenomenon in current driven loudspeakers.
AES E-Library Distortion Reduction in Moving-Coil Loudspeaker Systems Using Current-Drive Technology
Additionally I have to say that with good compression drivers as the TAD TD2001 mounted on a horn providing an essentially resistive acoustic load to the driver and thus giving very few peaks on the impedance curve, perturbations like the resonance inside the cavity behind the voice coil is dramatically reduced by current drive. With the TAD TD2001 this reflects in a filling of the famous hole around 1600-1800Hz which is most often seen on the frequency response of the TAD TD2001.
See a short study I did on that :
https://docs.google.com/file/d/0BwTS40IaQa16bEZTNEx1YUk3ODg/edit?usp=sharing&pli=1
I would advocate to current drive your compression driver but only if the impedance curve of the driver mounted on the horn shows very few peaks. (With truncated horns you have a lot of impedance peaks due to the multireflection of the wave from the mouth to the throat).
For conventional high efficiency current driven drivers, you must also read Nelson Pass :
http://www.firstwatt.com/pdf/art_cs_amps.pdf
Best regards from Paris, France
Jean-Michel Le Cléac'h
Mills and Hawksford used to write a good paper on the distortion reduction phenomenon in current driven loudspeakers.
AES E-Library Distortion Reduction in Moving-Coil Loudspeaker Systems Using Current-Drive Technology
Additionally I have to say that with good compression drivers as the TAD TD2001 mounted on a horn providing an essentially resistive acoustic load to the driver and thus giving very few peaks on the impedance curve, perturbations like the resonance inside the cavity behind the voice coil is dramatically reduced by current drive. With the TAD TD2001 this reflects in a filling of the famous hole around 1600-1800Hz which is most often seen on the frequency response of the TAD TD2001.
See a short study I did on that :
https://docs.google.com/file/d/0BwTS40IaQa16bEZTNEx1YUk3ODg/edit?usp=sharing&pli=1
I would advocate to current drive your compression driver but only if the impedance curve of the driver mounted on the horn shows very few peaks. (With truncated horns you have a lot of impedance peaks due to the multireflection of the wave from the mouth to the throat).
For conventional high efficiency current driven drivers, you must also read Nelson Pass :
http://www.firstwatt.com/pdf/art_cs_amps.pdf
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
Thank you Jean-Michel. As I described earlier, the frequency response of my TD-2002 on a 511 horn was also unaffected by adding series resistance. I admit I haven't measured the impedance on the horn; perhaps it is, as you say, far flatter because of the acoustic loading. I should go measure it
Thank you Jean-Michel. As I described earlier, the frequency response of my TD-2002 on a 511 horn was also unaffected by adding series resistance. I admit I haven't measured the impedance on the horn; perhaps it is, as you say, far flatter because of the acoustic loading. I should go measure it
Hello,
Additionally to the links in my last message you can llok at my results ;
http://www.melaudia.net/zfoto/rueil1010/TD2001J321Rs-934x541.gif
http://www.melaudia.net/zfoto/rueil1010/TD2001_Melaudia_J321-694x377.gif
see also the amelioration on the impulse response itself I obtained:
http://www.melaudia.net/zfoto/rueil1010/Impulsions_TD2001J321-667x360.gif
Additional links to current drive (and related topics):
Distortion example: Vifa | Current-Drive - The Natural Way of Loudspeaker Operation
http://www.eighteensound.com/staticContent/technologies/products/aic.pdf
High Performance Loudspeakers - Martin Colloms - Google Livres
http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/Power_Distortion.pdf
Loudspeaker operation: The superiority of current drive over voltage drive | EDN
Current driving
http://doc.utwente.nl/58981/1/Schurer94modeling.pdf
http://brightnoise.nl/Loudspeaker-driving.pdf
As it seems thatMills and Hawskfir paer is no ore on line at the Sussex Univ., here are the scans :
Index of /jahonen/Audio/Papers/Loudspeaker Distortion Reduction in Moving-Coil Loudspeaker Systems Using Current-Drive Technology
or as a pdf :
J12 Distortion Reduction MC Current Drive
Good reading!
Best regards from Paris, France
Jean-Michel Le Cléac'h
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.