I would like to test drive my Elsinore speakers with current drive...
In order to do that you'll need to remove the Zobel impedance components of the crossover. You'll then need to measure and equalise the response back to flat after the current amplifier changes it.
When building elsinore speakers I was under the impression that these speakers can be driven with voltage or current source as is.
This is what Joe had to say about Zobel. Did I misunderstand this?
"The Elsinore Mk6 are unusual in the sense that it can be driven by a transconductance amplifier, that is, from a power amplifier that outputs, not volts, but current (sometimes also called current steering). That pre-Zobel is needed to EQ the current from the amplifier and it means that it can be driven by any source impedance. So it is semi-optional, but more compulsory than optional in my mind."
I have posted Daniel/Duo’s 3875 chipamp transamp, Joe’s is based on that.
As i understand things, the addition of a sense resistor in the feedback loop.
dave
It appears you've been given only part of the story.
The reason the Zobel components allow you to run from a current source amp is because they defeat the nature of the amp to make it behave like a Voltage source.
More to the point, the amplifier is unable to supply constant current to the speaker where some of the current is being siphoned off through the Zobel components (which additionally can have their own effects). This is a simple premise in electronics terms. For some reason there is a shroud of confusion regarding how this works.
Lars Risbo of Purify has become the latest to state this...
vacuphile said:
lrisbo said:
Now, if you remove the Zobel components and use a current source amplifier, the response will change... So in order to make a proper comparison, this change needs to be measured and corrected with buffered equalisation.
So reading between elsinore thread and this thread and current amp threads, if I understand correctly, elsinore speakers with current crossover with zobel components will run on current amp (or any other type of amp) but it will not be of any benefit because of zobel components?
From that I gather if you remove zobel components from the crossover which allow speakers to run from current amps that this would also damage speakers?
Final conclusion: don't mess with current amplifiers if you don't know what are you doing... 😅
From that I gather if you remove zobel components from the crossover which allow speakers to run from current amps that this would also damage speakers?
Final conclusion: don't mess with current amplifiers if you don't know what are you doing... 😅
Well, without zobel there would be great excursion on the woofers at resonance, which you would need to EQ before amp to tame. I guess this is what is referred to be what could destroy speakers, excursion might reach limits.
With current amp system response would change with impedance and you would compensate for it somehow. Doing it with zobel is fine, it just makes the current flow through the zobel instead of the driver. Also allows some driver distortion current to flow as described and somewhat negates the benefits. Active EQ better for this reason. Although I'm not sure how much of a difference there is, distortion is probably low enough anyway not to disturb in home use, except for some small speakers / fullrange stuff.
Great bonus of the zobel and flattish impedance load for amplifier is that pretty much any amplifier works fine with it, without much of change in system frequency response no matter what the amplifier output impedance is.
I see greatest benefit in all this in active loudspeakers, be it current amplifiers or just the series impedance. If there is need to reduce speaker distortion, its easy to do with series impedance.
With current amp system response would change with impedance and you would compensate for it somehow. Doing it with zobel is fine, it just makes the current flow through the zobel instead of the driver. Also allows some driver distortion current to flow as described and somewhat negates the benefits. Active EQ better for this reason. Although I'm not sure how much of a difference there is, distortion is probably low enough anyway not to disturb in home use, except for some small speakers / fullrange stuff.
Great bonus of the zobel and flattish impedance load for amplifier is that pretty much any amplifier works fine with it, without much of change in system frequency response no matter what the amplifier output impedance is.
I see greatest benefit in all this in active loudspeakers, be it current amplifiers or just the series impedance. If there is need to reduce speaker distortion, its easy to do with series impedance.
I've used Zobel compensation between the amp and crossover for different amps to keep the response flat. Since it addresses the problem directly, I think it make some sense.
It's not the only way to equalise...
No, I don't see it that way. Nelson Pass has written about using a high output impedance to intentionally alter the response of fullrange drivers according to their impedance curves - https://www.passdiy.com/pdf/cs-amps-speakers.pdf
Take a look at the response variations there. In that case the output impedance isn't always infinite but even if it was, the response isn't going to go beyond a certain level.
Running a current amp without a speaker connected at all is going to trouble it the way that a Voltage amp doesn't like to be shorted.
I like current drive in this scenario as well. If you leave it open it will output rail voltage. Just when connecting the load to that rail voltage output you hear small click from loudspeaker.nothing to worry about. If you short current drive amp you get current what the input tells. Nothing to worry about, nothing burns even without protections. Zero to infinity load.
Maybe I'm missing something but that thread from 2013 seems to have lost the Schematics .. at least from the earlier posts. Anyone got the 'final' Schematics?
Even https://www.customanalogue.com/diy/Trans_Amp_Schema_Wiring_Caution.gif which Joe posted, doesn't work. Mr. Rasmussen, could you post it here again please.
Last edited:
This isn't a Current Drive amp. It's just a Voltage amp with wonky frequency response.This is one version:
I presume this is temporary, we discussed web site issues recently - https://www.diyaudio.com/community/...uctance-current-amplifier.239321/post-7126809
Joe has previously stated he hosts the images on his site in order to be able to update them.
@Joe Rasmussen
Who says it´s a current drive amp?
Not me.
I am exactly showing what jjasniew described, period.
which my schematic matches to a T.
Don´t waste your time (and mine) commenting on things I didn´t say.
There are some ways to bias transistors in class-B or -AB while they are connected like that. You can do it with some sort of class-AB bias loop or with a Monticelli output stage. (And of course with local feedback that turns it into two complimentary voltage followers again, like CFP output stages, but that defeats the purpose.) See also https://www.diyaudio.com/community/threads/monticelli-or-other-rail-to-rail-output.325128/ and https://patents.google.com/patent/US4570128A/en
Considering the subject of the thread, it is logical to assume that any amplifier schematics shown in it have a high-impedance current output unless otherwise noted.