The LA25-NP uses a closed loop Hall effect feedback circuit, which sends a current through a coil to cancel the magnetic field induced by current in the primary.
It has a bandwidth from DC to 150kHz, and worked well in my application.
I saw them widely used in servomotor drives.
Boden;
I discussed the system on an old thread on this forum.
The LEM modification simply replaced the resistor that I used to sense current, and it had a gain adjustment so the signal level was the same.
I changed the power supply to an SMPSU at the same time, and another one with +/-15v output for the LEM and other circuits.
Unfortunately, the system broke when a metal stand pierced one of the cones on an obsolete speaker, an Audax Aerogel, HM170, I think
Before that, an amplifier blew the equally obsolete Focal tweeter, which I replaced with a current model from another manufacturer.
Lots of fun was had, but it was time to move on.
I discussed the system on an old thread on this forum.
The LEM modification simply replaced the resistor that I used to sense current, and it had a gain adjustment so the signal level was the same.
I changed the power supply to an SMPSU at the same time, and another one with +/-15v output for the LEM and other circuits.
Unfortunately, the system broke when a metal stand pierced one of the cones on an obsolete speaker, an Audax Aerogel, HM170, I think
Before that, an amplifier blew the equally obsolete Focal tweeter, which I replaced with a current model from another manufacturer.
Lots of fun was had, but it was time to move on.
Yes they are popular in idustrial applications. Where higher currents are involved. Worked with boards with LEM devices where they used inside feedback loop to make current drive for power LED’s for lasers. For audio they are quite noisy, non linear limited bandwitch. Can be used for subs, but small Rsense is also an option
Can you explain a bit more? Are you saying current drive gives more distortion?
Cool what you did back then. Respect👍 if you see IR diference with 0.3 and 14sec it’s a slow varying effect. I am worried about modulation in audible frequencies. This is what Angelo is talking about in the paper you refering.
Ah, I see, thanks. I didn't check the datasheet and thought it were a simple AC current measurement step-up transformer.
Actually, it works like my Tektronix AM503+A6302 current probe in the LF/DC region.
I had to find out the hard way that sweep direction made a difference even at levels that I had thought to be safe wrt thermal compression when measuring small AMT tweeters. This was with the typical ~2 second sweep, two runs (from the CLIO system).
I tried that in some experiments, using a notch at the resonance frequency in the current-derived feedback path filled with the complementary band-pass from the voltage path. Wasn't worth the effort.
I also tried steeper 2nd slopes for a high-pass for the current drive path which, like in the case above, must sum to unity (zero phase, no allpass response) with the voltage path. So you need a linear-phase crossover. Simple subtractive filter only gives 1st-order slope for the derived path no matter how steep the primary filter is. Higher order on both slopes gives large peaks left and right of the split frequency for the individual paths, the phase offset is 120 degrees or more and hence the whole filter response is very sensitive to the slightest tolerances and parameter shifts.
For me, simple inductive slope above resonance is good enough, transitioning to resistive and finally capacitive slopes. That keeps the feedback circuit small and manageable. Also, I have some sort of gut feeling (no hard evidence, though) that higher that 1st-order slopes may give rise to hidden instabilities and side-effects.
If possible, use drivers with "too low" Qes to start with as then the base resistance, before transitioning into inductive, can be quite a few Ohms.
The appendix of the Fryer paper includes listening tests on 'modulation' type distortions. These are 'non-linear' and hence audible. Our listening tests suggests they are often not objectionable. But non-TI stuff are among the things that make speakers sound 'small' ... and is one of the audible benefits of current drive.
Us speaker designers often find it convenient to pretend speakers are LTI
... even thought they aren'tyes, current drive will often increase distortion down around the bass resonance and below (compliance and damping controlled range). Reduces distortion above (mass controlled frequency range )Can you explain a bit more? Are you saying current drive gives more distortion?
At the frequencies of the downspikes visible in the D3 Current under current control, the D3 current under voltage control actually does reduce the acceleration error?
I'm completely lost here, in translation as well as technically, so I would like to ask for your indulgence. I know a question mark doesn't make a question.
Last edited:
Impedance peaks - plural - so you talk about current driven bass reflex Speakers?
@ >KSTR< : I read/see/feel in your postings the effort to develop an adapted current drive amplification for bass reflex?
@>All< : Please let us differentiate between current drive for closed box speakers, and current drive for bass reflex speakers.
Last edited:
The system for driving electromagnetic vector displays is of some side interest, when the E and I waveforms in the deflection coils are examined on a scope. The amplifier's feedback is developed across a low value resistor in series with the deflection coil. To make a linear deflection response, which is proportional to the coil current in these displays, the voltage waveform must become very distorted and often have quite high peak voltages. There is little mass to move, just an electron beam to shove around, but the current feedback principle is clear as far as controlling the load current is concerned and there are many vector display schematics and manuals on the web.
The sense resistor vs speaker resistance gives a certain ratio, say 0.5 ohm and 8 ohm nominal, which controls the amount of feedback and gain. So, if the speaker impedance spikes up to, say, 20-50 ohm in the bass, the amplifier gets several times less feedback, which is a mechanism for higher HD (at least inside the amplifier itself).Can you explain a bit more? Are you saying current drive gives more distortion?
An active notch filter adds at least 1 buffer stage, which has to be immune to wild capacitive loads.
If the amplifier starts off with 0.0001% THD then that may be moot, but I also want it to be 'simple', given that my project is already made bigger by including active filters. Voltage control in the bass at least ensures that the peak (or peaks @Hörnli , for BR or other box designs) will be precisely matched in frequency, even accounting for seasonal variation.
Conjecture / edumated guesses:
In the bass resonance area, VC heating could be one source of distortion, especially if it's a non-conductive VC that's driven hard with bass boost. Current drive would allow hot spots to be ignored (while potentially making them worse.) OTOH, @ bass resonance, the middle of the coil in the magnetic gap would hog nearly the entire voltage regardless of how it's driven.
I've seen graphs where the motor strength seems to vary 2:1 across the swept range of voice coil displacement. So current control would not fix that. But I see no obvious reason why voltage control should be better either, unless there's some attempt to subtract distortion by modulating the other parameters (like the restoring force of the suspension) and the manufacturer has a preferred output resistance that the speaker was designed for.