DIY ESL's

[Jazzman53]

Below are a link to my DIY ESL website, a video link, and a couple of speaker photos-- enjoy!

Jazzman's DIY Electrostatic Loudspeaker Page
Video from Carverfest 2018

 

[ack]

Oh freaking WOW! I just read your blog on these speakers. Goodness, I have to listen to these speakers one day. Of particular interest to me - a modder - was the following that you wrote:

The transformer's winding capacitance adds to the load capacitance and its leakage inductance combines with the load capacitance to generate an ultrasonic resonance peak in the frequency response and rapid rolloff above it. Coincident with this response peak is an impedance minimum which can be a difficult load for the amplifier. When series resistance is added on the primary side it dampens this resonance peak. However, as previously noted, too much resistance over-damps the resonance, rolling off the audible highs.

It's interesting to me exactly because I have been playing with series resistance to my panels lately, in my custom crossover, and the results are fascinating. Indeed, too much damping rolls off the highs and affects speed as well. The right series resistance also improves timbre all around the panel, as the treble is adjusted. And at the same time, soundstage expands!

I have been reading for years that electrostatics sound so good partly because the crossover is simple or non-existence. Nonsense, my crossover is probably on the very complex side, and unlike yours, I had to employ a couple of capacitors and inductors.

The general guidance is to omit the section one resistors on the secondary side, add a 1? series resistor on the primary side and give it a whirl.

Funny, I was experimenting with 1? until I settled for 1.8?

From there the only tuning, if any, is adjusting the series resistance on the primary and/or the first two stator sections to dial in the treble response. Less resistance increases treble and visa versa.

BINGO! I have not read such an incredible blog as yours! And Bob Carver visited? I don't think anyone here has read your blogs, and they are missing out. Also wow about http://jazzman-esl-page.blogspot.com/2008/03/a-segmented-wire-stator-esl-with.html And you mean to say you built the stators like this http://jazzman-esl-page.blogspot.com/2010/01/building-stat-panels.html ??? OMG, I need to spend more time on DIYaudio

Heartfelt congratulations!!! I would strongly encourage you to post all these pictures on this forum.

 

[Steve Williams]

I would strongly encourage you to post all these pictures on this forum.

I totally agree

 

[Al M.]

I totally agree

Yes, it's impressive!

 

[jeff1225]

Agreed....these speakers are very impressive. I bet the wood work is as impressive as the sound. Can you please post the rest of the system? And welcome to WBF.

 

[Jazzman53]

Oh freaking WOW! I just read your blog on these speakers. Goodness, I have to listen to these speakers one day. Of particular interest to me - a modder - was the following that you wrote:



It's interesting to me exactly because I have been playing with series resistance to my panels lately, in my custom crossover, and the results are fascinating. Indeed, too much damping rolls off the highs and affects speed as well. The right series resistance also improves timbre all around the panel, as the treble is adjusted. And at the same time, soundstage expands!

I have been reading for years that electrostatics sound so good partly because the crossover is simple or non-existence. Nonsense, my crossover is probably on the very complex side, and unlike yours, I had to employ a couple of capacitors and inductors.



Funny, I was experimenting with 1? until I settled for 1.8?



BINGO! I have not read such an incredible blog as yours! And Bob Carver visited? I don't think anyone here has read your blogs, and they are missing out. Also wow about http://jazzman-esl-page.blogspot.com/2008/03/a-segmented-wire-stator-esl-with.html And you mean to say you built the stators like this http://jazzman-esl-page.blogspot.com/2010/01/building-stat-panels.html ??? OMG, I need to spend more time on DIYaudio

Heartfelt congratulations!!! I would strongly encourage you to post all these pictures on this forum.

 

[Jazzman53]

Oh geez guys, thanks so much for the compliments on my speakers and website. I seldom build speakers these days (even DIY ones aren't cheap) but I get almost as much satisfaction from my website. I've sent years attempting to make it a one-stop resource to convince other DIY'ers to take the plunge--- because I know what the effort and result would do for them.

You may have noticed that my speakers don't have passive crossovers like most Martin Logan and other commercial ESLs. Two reasons for that: First; I'm not smart enough to build a passive crossover for an ESL. The standard crossover math and calculators assume (and require) that the drivers have a relatively stable impedance. And an ESL's impedance is all over the map (falls with rising frequency). ML figured out how to do it by linearizing the panels' impedance somehow and probably spent a ton of time and money perfecting it for each model-- something most DIY'ers can't really do-- especially math-challenged folks like me. I just do it the easy way by bi-amping the woofer and panel separately and inserting an active DSP crossover upstream of the amps (not cheap but way easier than building a passive crossover that actually works well). Reason 2: Active bi-amping sounds better and the bass really tightens up without a passive crossover inductor in the way.

I'm a pretty fair woodworker but I struggle with the electronics and math part of building ESLs. The real brains behind my latest projects are the two guys I credited on my website; Rod White and Steve Bolser. Rod is a PHD physicist who wrote the white paper that my latest stators are based on. And his collaborator, Steve, is an aerospace engineer/mathematician with a lot of patience and a real gift for making the complex understandable. I can't tell you how much those guys helped me with my builds, and with proofing my website.

BTW; Bob Carver didn't visit my home to hear my speakers. We met and became friends at the annual Carverfest retreat in NC and we hang out there for 10 days each year. In 2015 I brought my beam splitter speakers to Carverfest and in 2018 I brought the Audi speakers. Basically; it's a big party with really loud music and lots of food and vintage Carver audio gear.

As time permits I will post some more pics of my speaker builds here.

Thanks, again guys for the great welcome!

 

[Jazzman53]

....Also wow about http://jazzman-esl-page.blogspot.com/2008/03/a-segmented-wire-stator-esl-with.html And you mean to say you built the stators like this http://jazzman-esl-page.blogspot.com/2010/01/building-stat-panels.html ??? OMG, I need to spend more time on DIYaudio

Heartfelt congratulations!!! I would strongly encourage you to post all these pictures on this forum.

Yes; I've built several different types of panels and the welding rod panels were an interesting project. Before I learned about electrical segmentation I had no idea it was possible to widen a flat panel's dispersion and curve its wave front electrically.

Generally; unsegmented flat panels have narrow dispersion which gives the best imaging and slam, and curved or segmented panels have wider dispersion at the expense of somewhat less precise imaging. But it occurred to me when building the welding rod panels that it would be easy to make a segmented panel's dispersion selectable (wide or narrow) by merely inserting a switch to jump over the segmentation resistors.

The photo below shows a segmented welding rod panel, its resistor network, and a box containing a rotary ON/OFF switch to jump (defeat) the resistor network. The switch was a Soviet military multi-pole rotary type with silver contacts I found for $4 each on Ebay from Belarus.

I had fun with it but the novelty wore off after a while because the switch was only rated for 300 volts, so I had to power down to make the switch and then re-EQ the panel each time. I thought about submerging the switch in linseed oil to allow switching modes with the panel playing without frying the switch, but I would still have to re-EQ the panel each time so it didn't make sense to take it that far.

Absent any EQ'ing, a flat panel (or even a curved panel to lesser extent) has a naturally rising frequency response due to its beaming the highs and its low frequency dipole roll off. A typical unsegmented ESL's crossover network will include a compensating shelving filter which pulls down the highs to flatten response. ML on the other hand, curves the panel to spread the highs, and they use the "distributed resonance" approach to partially offset the panel's low frequency dipole roll off-- that is; they tension the diaphragm quite high (to increase its drum-head resonance frequency) and use unequal spacing between support spars to distribute this resonance over a wider bandwidth. In this way, the diaphragm's drum-head resonance is used to boost the upper bass, as opposed to pulling down the highs with a shelving circuit, which would lower the efficiency.

On the other hand; a properly segmented wire panel has a flat frequency response from the top octave all the way down to the dipole roll off (3db down point frequency occurs where the bass wavelength reaches 4X panel width).

The segmented welding rod panel and mode switch box: