No. There were more complications than benefits. I also didn't find an improvement in sound. My conclusion was that the current drive might make sense only in an active speaker without passive crossover.
Just a little idea I had. With your schematic, if any break occurs in the second coil, the amp will have no more feed back, and will suddenly work in open loop, full gain: Not good for both the speaker and itself, right ?
Why not use traditional feedback, then compare the sensor signal (reduced to the level of the input signal) to the input one and add in opposite phase the resultant "error signal" to the input one ?
I can see some advantages: this will fix the problem, could work like a "protection" for both speakers and amplifier if you add a power breaking system like in this thread:
An ultimate amp protection circuit ?
and could be added to any existing amplifier. Adding the ease to adjust the % of the correction signal.
It's a good idea. We actually discussed it a billion posts ago. Adding a feedback resistor from the output back to the pickup tap. High enough resistance that it doesn't impede the feedback, but if the pickup coil opens, the amp would simply revert to normal voltage feedback. the resistor can be sized to have volume lower at break for a noticeable difference, or retain the same volume level for a pro application where volume is everything.
I have mentioned running a current drive with the tap lead as a second measure of error.
jn
Maybe we can even find a way to partly use the second coil for driving also, which would be really nice in a real application. And a way to factor out the thermals without a complicated servo, plus adding some degeneration around resonance which is again needed for any practical solution with most drivers/alignment...
Horn loading might be a nice try as it keeps excursion low and if we could get rid of BL(i) and Le(i) that should give a nice improvment. No/little damping needed.
Horn loading might be a nice try as it keeps excursion low and if we could get rid of BL(i) and Le(i) that should give a nice improvment. No/little damping needed.
Last edited:
I apologize.
When I was a small child, I was already bored after the first round of the carousel (Imagine after a billion). Reason why I often read this forum diagonally.
4 years ago I started with a passive crossover box, realized that it was incorrect and then tested individual tweeters. Few days ago I continued with a new tweeter and found an improvement of almost 20dB in reduction of odd harmonics under current drive. This results I did not achieve 4 years ago, but it seems that the old driver tested might have been defective.
No need to apologize. This thread move so fast, that you have to keep up, otherwise a thousand posts will pass you up in no time. I'm guilty of the exact same diagonal read as well..
ps. I really wish this site had the gallery working, that way I could start locking a lot of this stuff in time and location.
Interesting. But now, isn't a tweeter kinda not moving too much? I've a hard time thinking about excursions which would hit the Le(x) curve.
What is the current thinking on what is giving you the third, as well as the difference (I recall a 1k harmonic).
I can't help but think there may be some eddy/pole/face magnetic interaction going on. (course, as I said before, if all ya got is a hammer...😀 )
Are tweeter faceplates and pole pieces ferrite or solid iron? I would hope ferrite or at least front plate laminated stack outta .1mm flat stock.. (I like ferrite much better..)
jn
I do not think that the excursion differs when we are enough above resonance. Around resonance, driving current must be reduced by means of pre-filter.
The difference tone (1k) was almost unaffected and is same for voltage and current drive. This means, asymmetric source of distortion was not cured.
"Skirts", odd harmonics, were reduced of almost 20dB. So, symmetric source of distortion was highly suppressed. Now we can speculate 🙂
I have updated my page
Proudove buzeni repro a zkresleni repro
4800+5000Hz plots show reduction of both harmonic and intermodulation odd distortion components under the current drive.
See attachment in Ian Hegglum’s post .
Ltspice .asc file with driver and box sim parameters, no gyrator 🙂
mix current drive with voltage drive at LF?
George
Of course, you may use transformer instead or nothing, but then you need to make mechanic - electric transformation, which means series becomes parallel and you calculate through Bl. In all cases, you need to understand electro-mechanics transformation. Gyrator is most logical and is used in serious publications. Sorry, this is irrelevant to the subject.
Thanks, J.C. Touched.
Happy Christmas to you... and all of you here 🙂
Hoping that the new year that will come soon will bring more understanding and kindness between the members of the big family who lives on our little planet.
Thank you for taking your time to elaborate for me. I hope all primary distortion mechanism has been accounted for in your estimate.
- Status
- Not open for further replies.