has anyone driven k-horns or la scalas with a transconductance [constant-current] amp?

[m8o]

As the title reads, I'm wondering if anyone has driven k-horns or la scalas with a transconductance [constant-current] amp before? Did you try it tri-amp'ed? I plan to do so driving my la Scalas using a modified power-follower circuit driven by my Little Dot III+ amp. Anyone try it before?

I heard transconductance drive works well with horns, so thought I'd give it a try. Ideally, transconductance amps should be connected right to a full range drivers without typical parallel type crossovers designed to be driven by a voltage source amp inbetween the amp and driver. My finish line is using one custom built tube pre + ss constant current power follower per bass, squaker & tweeter, electronically crossed over of course. But trying this route with the three drivers + parallel x-over network driven by one constant-current amp for now.

[DrWho]

I have no experience with these amps, but if they are truly ideal current sources, then the crossovers aren't going to function as they should....so apart from bypassing all of the carefully built-in EQ, you're going to blow up them squakers and tweeters.

[PrestonTom]

I have no experience with these amps, but if they are truly ideal current sources, then the crossovers aren't going to function as they should....so apart from bypassing all of the carefully built-in EQ, you're going to blow up them squakers and tweeters.

Mike,

You hit the nail on the head regarding the issue of crossovers. If the application is tri-amped with the filtering done prior to the amp (the way it really should be done), then you will get around the headache.

-Tom

[m8o]

hmmm... I know overal acoustical output is affected "un-naturally" (vs if the driver was driven direct) because @ constant current the input voltage level 'follows' the impedance curve, which is altered by the network components in addition to the driver's impedance .... but that doesn't 'stop' a capacitor's impedance from rising as frequency goes down, or inductor's impedance from rising as frequency goes up. I do know the equalization you mention is more trouble then good to the frequency response / acoustical output when driven this way. I was sure x-over still functions though.

I will be approaching this with 'measured' steps, both figuratively and literally. Meaning applying power cautiously, as well as actually measuring the response to the input terminals of the drivers measured with one of these that I have: http://www.terrasonde.com/products/color.php.

Using the Little Dot III+ headamp/preamp/poweramp already has transconductance-like drive characteristics when driving speakers; output impedance is much higher then a typical amp. That's what 1st whet my whistle to this. That, and the fact that, - transconductance drive of the squawker driver provides a natural dip in power to the mid portion of its operating region where impedance is low - transconductance drive of the hors, bass horn especially, provides a natural increase in power delivered at frequency where the impedance is high.

Thanx for the warning. I'll be sure to perform as FFT/RTA scan of the frequencies each individual driver is seeing when driven by a constant-current source.

[DrWho]

You might try throwing on some dummy load resistors in place of the drivers when doing your measurements...and then you don't have to worry about blowing stuff up.

[m8o]

I have no experience with these amps, but if they are truly ideal current sources, then the crossovers aren't going to function as they should....so apart from bypassing all of the carefully built-in EQ, you're going to blow up them squakers and tweeters.

Ah, now I understand the error of my ways!!! I didn't understand the whole issue. Based on this post here, I now understand:

http://www.audioasylum.com/forums/tubediy/messages/60883.html The operative statement being: I have a three-way with xover at 300 and 3500Hz. The output impedance of the amp is in series with the impedance of the speaker drivers, so the crossover must be modified for approx 55ohms, and not 8ohms.

It's so simple makes me ask "why didn't I think of that". Looks like before trying anything I'll be pulling out my good old Pioneer SF-850 selectable electronic crossover and will need to build three, not one, transconductance power follower. Argh! ...though, I have built plenty of x-overs in the past. We'll see.

[DrWho]

If you're going to build 3 of the amps, then you might as well just move to an active crossover and get rid of a lot of the guesswork with trying to make a new passive crossover...not to mention it's not going to cost too much more and you'll have the option of playing with time-alignment too.

[m8o]

By my final thought, I meant maybe I'd do just one amp and re-do the x-overs at the higher combined impedance of the source & load.

Either way, this got more expensive; there's no "cheap" way to enter into this experiment ... Either I have to go fully active with three cc amps, or one cc amp and new x-overs. I thought I could just 'try it out' but see now it's a bit more of a comittment.