Challenge: Can an Aleph J-ESL be built?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thank you very much for commenting, Mr. Pass.

So using these FETs (maybe even a bridged config) would make it for sure.

My over-interpretation of your comment is of course that you fully support the Aleph DirectDrive-ESL idea, and a suggested design will show up soon. :p

Joking aside, if you ever listened to ESL57s, you probably understand the interest of tweaking these wonders --- and using Linkwitz LX521 dipole subwoofers to offload and fill the the missing lowest octave (20-40 Hz) makes them even more impressing.
 
Last edited:
Joking aside, if you ever listened to ESL57s, you probably understand the interest of tweaking these wonders --- and using Linkwitz LX521 dipole subwoofers to offload and fill the the missing lowest octave (20-40 Hz) makes them even more impressing.


I've been listening to mine since the early 90s. Since they sound awfully good, those transformers can't be so bad, methinks.

I bought the B5 for use with the Quads. The bi-amped setup really makes them more dynamic and less congested. A great place to start if you're looking to enhance them.
 
Agree that 57's need offloading to not sound congested, so I'm using the LX Mini Crossover (tweaked to a -24 dB/octave high-pass @ 95 Hz now)


Awesome! I never settled on a crossover frequency. The Quads go pretty low, but a lot of people like to cross them over above 150 Hz. You can get rid of the bump around 125 Hz in the process, which is less welcome when the whole system goes lower.


What are you using for woofers? So far, I've been using a single Eminence Beta 15" per side, but I would like to build ripoles when time allows. The setup I have now serves as a stand for the Quads, and you can also attach a baffle to the top. My Quads need a repair at the moment, and I'm listening to the SAL 08Cs in the meantime - a very nice speaker in its own right.
 
Straying a bit OT to answer Luxwinter...

There seem to be many views and no consensus on best ESL57 xover frequency
Some say 40-50 Hz due the high-quality ESL bass and mid-bass
Some say 80-100 Hz as a good compromise (the famous stacked ESL57s + Decca ribbon tweeter + 24" woofer system built by Mark Levinson crossed over at 92.5 Hz)
Some say 150 Hz to offload the 57s (and transformers :eek:) as much as possible.
The well-known Gradient SW57s cross at 150 Hz.

The only consensus today seem to be to avoid omnipolar subs and use dipole radiation to keep the same radiation pattern.

Link to the subwoofers i built (20-120 Hz, changed to 95 Hz) is here ---->
www.linkwitzlab.com/LX521/Description.htm

It is part of the Siegfried Linkwitz top-of-the line dipole speaker
and uses 2x 10" long-throw woofers per channel - SEAS L26RO4Y, D1004-04 with Alu cone that are Push-pull mounted in V-frame baffle of 24"H x 13"W x 15"D.

The xover and equalisation needed is somewhat complex -- but can also be built in the digital domain using a miniDSP 4x10HD version.... or you may even build it with the PASS LX mini Crossover (I have the bass part done using op amps and the QUAD HP part built with the PASS LX Mini Crossover.
I did also build "the bridge" above the bass dipole box to isolate the QUADs from a lot of LF vibrations...bass is a straight line down to 18 Hz measured in-room using REW. Send private mail if you need more info :)

so...connecting back to this thread...my *wishful thinking* set-up to reach new levels of sound (maybe ;)

1.Multichannel amp 4x250 W to drive the 4 woofers mentioned above (heavy EQ)
2.Aleph J for QUAD 57 bass panels via stock xformer (8-40 ohm load)
3.Aleph ESL for QUAD 57 midrange/treble direct drive , approx 400 pF load :eek:

1. and 2. done and playing beautifully
3. ----> "Nothing is impossible until proven so" :wrench:
 
These SEAS drivers are expensive (for a 10") , but weigh 22 pounds each, have a force factor of 18 N/A and make a kick drum sound like a kick drum....:p

Yes, I did try laying the boxes on the side and rejected it due to the bloated bass -- the added floor support was overwhelming my rather small listening room (15'x14'x10').

Also I know the recommendations are to raise the ESL 57s some 14-18" from the floor (running full range), but since they only need to reproduce above 95 Hz for me the additional height (at 26") works well.
(not including WAF here, good thing with a dedicated room :))

Not to say it can't work laying them on the side but that would likely need adjusted EQ and/or a bigger listening room. If you want I can do some REW measurements showing the FR differences between standing up and lying on the side.
 
Back to ESL direct drive, some more focused Internet searching reveals that this area is not completely new territory (HV MOSFETs were already used back in 2012)

High-voltage class-D direct-drive audio amplifier for electrostatic loudspeakers
John Kota, Daniel Holman, +3 authors Malik E. Elbuluk Published in IEEE 55th International Midwest Symposium on… 2012
DOI:10.1109/MWSCAS.2012.6292085


Electrostatic loudspeakers (ESLs) are capable of low distortion and have desirably flat frequency response within the audio frequency range. In this paper, a novel approach is proposed for audio amplification using a high-voltage direct-digital solid-state class-D amplifier as a method that replaces impedance matching ESL audio transformers. Optical coupling is used to achieve isolation from the high-voltage side. The proposed stacked MOSFET output stage provides a solution to the voltage swing limitations of the available switching power MOSFETs.

(Research done at Univ of Akron, OH)

MOSFETs used were STFW3N150 Vds=1500 V @ 2.5A

But Class A would be the real thing!!! :p:p

ESL DD.png
 
Last edited:
Attached is an output amplifier stage that runs off of + /- 1200V rails.
It is not an audio product. You could call it an industrial product. It is probably overkill for audio.

I have built ESLs including everything, panels and power supply. I used UTC transformers to step up the audio voltage.

One technical hurdle is how much current your speakers will require. This current at high voltage will likely be lethal.

From an amplifier standpoint, the heat problem may not be huge. If the load is largely capacitive, the current will be out of phase with the voltage.

To the extent that the speakers actually consume real power, which they must to make acoustid SPLs there will a resistive component to the speaker and so some current will be in phase with some voltage.

The good news is that most likely the actual acoustic watts will be a few watts and the speaker will probably only consume a few watts. Just a guess. Measurements would tell the actual story.

If I really wanted to know, I would place a resistor in series with a normal ESL system and have a look at phase of the current vs voltage at various frequencies and SPLs.

The Acoustat Answer Man on the Exotics forum here may know how much real power is consumed by the venerable Acoustat Monitor X direct drive system as well as how much current is needed at the high voltage.
 

Attachments

  • OutputStage.jpg
    OutputStage.jpg
    299.7 KB · Views: 137
AX tech editor
Joined 2002
Paid Member
The ESL itself is almost purely capacitive, 1nF or more ballpark. The dissipation will be in the amp due to the in-phase I-V product in the output stage. Plus something in the transformer when used, and some resistive components in the filter.

The need for a relatively powerful amp is due to the large current resulting from that 1nF transformed up with that 1:150 or thereabouts transformer.

Jan
 
The ESL itself is almost purely capacitive, 1nF or more ballpark. The dissipation will be in the amp due to the in-phase I-V product in the output stage. Plus something in the transformer when used, and some resistive components in the filter.

The need for a relatively powerful amp is due to the large current resulting from that 1nF transformed up with that 1:150 or thereabouts transformer.

Jan



Just curious how much is relatively powerful? I've always liked stats but heard they need a little as 50 watts to needing 100s. I guess that depends on if it full range a hybrid though...
 
AX tech editor
Joined 2002
Paid Member
Just curious how much is relatively powerful? I've always liked stats but heard they need a little as 50 watts to needing 100s. I guess that depends on if it full range a hybrid though...

My personal experience is with a Martin Logan Sequell II, which is a hybrid. I had a commercial 50W amp, don't remember the brand, but nothing special, and it didn't sound well.
When I hooked it up to my powerhouse Sony TAN-7B everything sounded much better and cleaner. That TAN-7B produces 200W max and gobs of current.

But there are many differences between speakers and amps, especially current capacity, which may or may not follow from power rating. So you need to trial and error, or go with overkill just to be sure.

Jan
 
My personal experience is with a Martin Logan Sequell II, which is a hybrid. I had a commercial 50W amp, don't remember the brand, but nothing special, and it didn't sound well.
When I hooked it up to my powerhouse Sony TAN-7B everything sounded much better and cleaner. That TAN-7B produces 200W max and gobs of current.

But there are many differences between speakers and amps, especially current capacity, which may or may not follow from power rating. So you need to trial and error, or go with overkill just to be sure.

Jan


Interesting. Looks like i'll be adding this to my laundry list of journals I need to read. Thanks for the insight
 
I used Perreaux 150w/ch MOSFET amps circa 1980s for my Martin Logan CLS and subwoofer array. Rattled the pictures on my walls. I upgraded from DIY panels to ML.


To Jan's point, capacitors want to see high current at low voltage and high voltage at low current. That means a big bad beefy amp that, essentially, can produce high voltage at high current.
 
AX tech editor
Joined 2002
Paid Member
To Jan's point, capacitors want to see high current at low voltage and high voltage at low current. That means a big bad beefy amp that, essentially, can produce high voltage at high current.

Emmm, that's not really my point. If you want to charge a cap to a higher voltage, you either have to charge it longer, of faster meaning higher current. It all depends, also on the cap value.

The point I was making is that if you have a 1:150 transformer, a 1Megohm on the secondary looks like 1Meg/(150^2) on the secondary side, which is about 4.5k.
For the ESL, 1nF at the secondary side looks like about 22.5uF at the primary side, what the amp sees. That kind of cap requires a LOT of current when you want to drive it with several kHz of say 20V signal.

Jan
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.