A Story of OTLs and Ribbon Tweeters

OTL Asylum

OTL, Output Transformerless Amplifier User Group.

C.B.

Here's the most recent development concerning my scratch-built OTLs and Newform Research (quasi) ribbon speakers that I thought might be of interest to inmates here.
I've been using my (4) Bruce Rozenblit-designed T8s in a bi-amp set-up for about a year and a half now. The lower end on each speaker consists of (2) ScanSpeak 5" mid-bass drivers, crossing over at 900 Hz via a Marchand electronic xover to the 45" ribbons. There is also a Velodyne ULD-18 II subwoofer, operating below 50 Hz. You can see and read about the set-up on my webpage:
http://mywebpages.comcast.net/cbrodersen/home.html

Things were basically OK until a few months ago, when I noticed a certain resonance or "ringing" in the ribbons. On material with a strong treble content at just the right frequencies, say trumpet or a soprano voice, there was noticeable distortion. I immediately suspected the (factory-mandated) series capacitor used with the ribbons to shape the response--basically a kind of high-pass filter with a corner frequency of about 2.7 kHz--perhaps the caps were creating problems or interacting with the T8s. I never got around to upgrading the stock caps (12uF Solen metalized polyester) that came with the ribbons, as many people have done, since "audiophile" film-and-foil types would have set me back around 100 bucks.
I wanted to try an idea that I've been kicking around for a long time--replace the large value caps at the input to the ribbons with smaller value caps at the input to the amps. Doing the math, I came up with the value of 570pF for the series cap, which, together with the 100k input resistor of the T8, would theoretically yield the same response as the RC network at the speaker.
Sure 'nuff, it worked! The T8s are now driving the ribbons directily, and are obviously MUCH happier with the pure, non-reactive load of 4.8 ohms (claimed by the manufacturer) than before. The Marchand xover, with its super-low output impedance, seems to have no problem driving the relatively small value RC network at the input to the amps.
Moral of the story? If you need to do any kind of response-shaping with your speakers (and I guess that includes crossovers), it's much better to do it at the line-level stage than at the output of the power amp.

This post is made possible by the generous support of people like you and our sponsors:

VH Audio

Topic - A Story of OTLs and Ribbon Tweeters - C.B. 13:22:21 01/18/06 (1)

Similar solution for Newform 645's... - mpathus 14:42:46 01/22/06 (0)

You can not post to an archived thread.