Before people get their nickers in a bunch, I think you both are talking about different aspects of ESL driving.
The Quads are not very efficient, but can't handle much power either. So 15 watts of the best power you have is enough. That Hood amp has 20 volt rails. perfect for the 57's but not great for some ESL's. (it would be fine for my ESL's though)
There's a mix of direct drive amp discussion and this topic on this thread. To clear it up for everybody, ESL's are capacitors, crappy capacitors. If they are direct driven, they act almost like perfect capacitors. When driven through a matching transformer, there is an inductive component as well.
So, no matter how you drive the ESL, you have to drive a nearly perfectly reactive load. So whether it's a direct drive amp or an amp driving the ESL through a step-up transformer, your amp will have to be stable into those loads. This means that marginal amp design won't cut the mustard.
For me, I decided early that direct drive isn't the way to go. It looks great on paper, but in real life it's a nasty proposition. There are more compromises in a DD amp than in a matching transformer in my opinion.
That still leaves what amp would do that job of driving an ESL the best. For my 20" x 48" panels, I waffle between my Pass Aleph 3 and a variety of tube amplifiers. Currently I am liking the Heath W5M which is a williamson design with a nice peerless transformer.
Sheldon
The Quads are not very efficient, but can't handle much power either. So 15 watts of the best power you have is enough. That Hood amp has 20 volt rails. perfect for the 57's but not great for some ESL's. (it would be fine for my ESL's though)
There's a mix of direct drive amp discussion and this topic on this thread. To clear it up for everybody, ESL's are capacitors, crappy capacitors. If they are direct driven, they act almost like perfect capacitors. When driven through a matching transformer, there is an inductive component as well.
So, no matter how you drive the ESL, you have to drive a nearly perfectly reactive load. So whether it's a direct drive amp or an amp driving the ESL through a step-up transformer, your amp will have to be stable into those loads. This means that marginal amp design won't cut the mustard.
For me, I decided early that direct drive isn't the way to go. It looks great on paper, but in real life it's a nasty proposition. There are more compromises in a DD amp than in a matching transformer in my opinion.
That still leaves what amp would do that job of driving an ESL the best. For my 20" x 48" panels, I waffle between my Pass Aleph 3 and a variety of tube amplifiers. Currently I am liking the Heath W5M which is a williamson design with a nice peerless transformer.
Sheldon
Thanks Sheldon for bringing everything back on track a bit of sanity.... I'm in the process of untangling my knickers right now
lest they strangle me altogether. you make a lot of sense
trouble with a forum like this is everyone knows better than the other person and determined to have their say; all I wanted to do is point the enquirer in a direction which I thought; wrongly it seems would help him hence the reference to the quad ESL and the associated little jlh circuit modified to do the job which it appears it does very well the only proviso is that like all things not all ESL's were created equally
cheers and thanks for the cold shower
TJB
lest they strangle me altogether. you make a lot of sense trouble with a forum like this is everyone knows better than the other person and determined to have their say; all I wanted to do is point the enquirer in a direction which I thought; wrongly it seems would help him hence the reference to the quad ESL and the associated little jlh circuit modified to do the job which it appears it does very well the only proviso is that like all things not all ESL's were created equally cheers and thanks for the cold shower TJB
One thing to keep in mind...........Most modern DIY ESL's will require a "balanced" amplifer if you want to direct drive it. both the front and rear stators need to be driven out of phase with each other. One is pushing while the other one is pulling. The JLH amp was single nded, and would only push one of the stators. In order for it to push both, you would have to build two such amps and run them balanced.
The JLH circuit did look very nice. I would be temped to build one if I wasn't half way in to my new Zen V4 pair.
One last thing to consider when direct driving ESL's is safety. In order to drive that big speaker/capacitor, you need pretty high volages between the stators. If you, or someone in your family (small children), were to touch both the fron stator and back stator at the same time, it could be deadly.
Even after all of these problems, I would still love to built a DD amp for my ESL's. What we need is a high voltage (several thousand volts) output and wide frequency response. Most ESL's don't go very low, so bass response wouldn't be as important. Any ideas? How about a massively paralell MOSFET design?
If someone can do that, I'll move my Zen to the nursery for my son to listen to. It's never too early to appreciate good sound.
Cheers,
Zach
The JLH circuit did look very nice. I would be temped to build one if I wasn't half way in to my new Zen V4 pair.
One last thing to consider when direct driving ESL's is safety. In order to drive that big speaker/capacitor, you need pretty high volages between the stators. If you, or someone in your family (small children), were to touch both the fron stator and back stator at the same time, it could be deadly.
Even after all of these problems, I would still love to built a DD amp for my ESL's. What we need is a high voltage (several thousand volts) output and wide frequency response. Most ESL's don't go very low, so bass response wouldn't be as important. Any ideas? How about a massively paralell MOSFET design?
If someone can do that, I'll move my Zen to the nursery for my son to listen to. It's never too early to appreciate good sound. Cheers,
Zach
Direct drive
Solid state is going to be difficult and complex at these voltages. Tubes are the best technological solution here, I believe. And there are widely-available, relatively inexpensive tubes which will let you swing 5 or 6 kilovolts or more. Zeus willing and the river don't rise, my prototype will be done mid-summer.
Solid state is going to be difficult and complex at these voltages. Tubes are the best technological solution here, I believe. And there are widely-available, relatively inexpensive tubes which will let you swing 5 or 6 kilovolts or more. Zeus willing and the river don't rise, my prototype will be done mid-summer.
Re: Direct drive
That's true, in the past tubes were the only way to go. But I'm hearing of FM transmitters now being made with squallid state devices, so there must be some high voltage devices out there which are pretty high voltage.
That said, I'll stand by my assertion that there is more compromises in a 5KV swinging amp than in a good step-up transformer when it comes to driving an ESL.
that said, I'm really anxious to see what you come up with. I wimped out and went the iron core route, but you shouldn't.
Sheldon
SY said:
That's true, in the past tubes were the only way to go. But I'm hearing of FM transmitters now being made with squallid state devices, so there must be some high voltage devices out there which are pretty high voltage.
That said, I'll stand by my assertion that there is more compromises in a 5KV swinging amp than in a good step-up transformer when it comes to driving an ESL.
that said, I'm really anxious to see what you come up with. I wimped out and went the iron core route, but you shouldn't.
Sheldon
Just yesterday I collected a pair of custom wound transformers at AE to be used on a "semi DD" amp to drive my full range ESL's.
What I will do is build a KT88pp amp with these custom OPTs to go directly to the panels. This way you have one transformer instead of two and it is only needs a 1:4 step-up instead of 1:100 which should make it a lot easier to build right.
Also, I once had the chance to listen to this solid state DD amp. It sounded absolutely amazing. The idea was to offer it as a kit. I don't know what the status is at the moment though
What I will do is build a KT88pp amp with these custom OPTs to go directly to the panels. This way you have one transformer instead of two and it is only needs a 1:4 step-up instead of 1:100 which should make it a lot easier to build right.
Also, I once had the chance to listen to this solid state DD amp. It sounded absolutely amazing. The idea was to offer it as a kit. I don't know what the status is at the moment though
thanks Peterr,
the DD-amp of your link is the one I heard, and it sounded very good. We will wait until the designer is ready to output it's product to the esl-club.
I have also experimented with ultra-lineair output transformers as in the way you described:
I connected the output valves anodes to the screen tabs of the xformer and the anode tabs to the stators of my ESL.
The output xformer is now a step up, with symmetrical drive.
This works for 3k or less primery impedance xformers, otherwise the high-frequency response suffers.
Dick.
the DD-amp of your link is the one I heard, and it sounded very good. We will wait until the designer is ready to output it's product to the esl-club.
I have also experimented with ultra-lineair output transformers as in the way you described:
I connected the output valves anodes to the screen tabs of the xformer and the anode tabs to the stators of my ESL.
The output xformer is now a step up, with symmetrical drive.
This works for 3k or less primery impedance xformers, otherwise the high-frequency response suffers.
Dick.
That reminds me, I was going to add something to that affect. Wind a triple-tapped choke of the same nature, so the tubes connect to the taps either side of the CT, and only have to swing maybe 500 volts tops; the outer ends of the choke connect to the ESL stators. This would also allow efficient classes of operation.
You still have the problem of L and C, of course.
Tim
You still have the problem of L and C, of course.
Tim
The Alphonse Wuyts Way.
Hi,
Another way to go semi DD is to build an OTL tube amp and drive the ESLs, it usually is a match made in heaven.
Cheers,
P.S. For all of you wondering who Alphonse Wuyts was: Fons, as he was known, died at the aged of 28 some ten years ago leaving behind his beloved wife and more importantly to him, I think, an incredibly good, no, amazingly good sounding ELS design of his own design (well over 2M high) pair of ELSs and EL509 based OTLS driving them.
Some of our Dutch members may have known him too, especially the active ELS crowd in Holland.
Sorry to go sentimental and off topic on you all, this young man was DIY incarnated, I miss him.
Hi,
Another way to go semi DD is to build an OTL tube amp and drive the ESLs, it usually is a match made in heaven.
Cheers,
P.S. For all of you wondering who Alphonse Wuyts was: Fons, as he was known, died at the aged of 28 some ten years ago leaving behind his beloved wife and more importantly to him, I think, an incredibly good, no, amazingly good sounding ELS design of his own design (well over 2M high) pair of ELSs and EL509 based OTLS driving them.
Some of our Dutch members may have known him too, especially the active ELS crowd in Holland.
Sorry to go sentimental and off topic on you all, this young man was DIY incarnated, I miss him.
.
I wouldn't exactly say "secret," but I hate talking about something that I don't know is going to work until I have it working. That goes double for this project, since I've never done a HV design nor have I ever tried to stabilize an amp driving a (nearly) pure capacitance. I'll say this much- it will be based on TV sweep tubes, it will NOT be class A, and it won't be made from Fashion parts (anyone have some 10KV Black Gates?). And if I can get it to work, I will absolutely publish all details.
The limitation of my design idea is that it will only work for 'stats that have insulated stators.
The limitation of my design idea is that it will only work for 'stats that have insulated stators.
Fair enough. I was mostly wondering what tubes you were going to be running; the ESLs will likely require a fair bit of current to avoid slewing distortion and all the usual suspect TV tubes happy working at 5kV are quite wimpy looking; at least, the ones I've come across are.
Meanwhile, I got some crazy ideas about 845 diff pairs with resistive loads at +- 1.8kV... but gotta finish an SE 4212 first with phase splitting output transformer...
Best,
cv
Meanwhile, I got some crazy ideas about 845 diff pairs with resistive loads at +- 1.8kV... but gotta finish an SE 4212 first with phase splitting output transformer...
Best,
cv
According to an article published by Peter Walker in the Journal of the Audio Engineering Society in 1980, flat panel ESL's should actually be driven with a frequency-independent current to obtain a flat far-field response, at least when the membrane mass is negligible (thickness 4um or less) and the suspension doesn't mess things up.
According to a colleague of mine (Frank Verwaal) who is working on DIY electrostatic loudspeakers, for practical listening distances, you also have to limit the area transmitting high audio frequencies to a small value, 50cm^2 or so, or you get problems with cancellation effects limiting the high-frequency response.
Anyway, when you design your ESL's such that they require frequency-independent current drive, the problem with slewing disappears. You then need as much or as little current at 20kHz as you do at 100Hz.
At the moment, I am working on a mixed transistor-valve (PL519) class A amplifier for directly driving Frank's loudspeakers. If I survive and if it becomes a success, I'll give you more details.
According to a colleague of mine (Frank Verwaal) who is working on DIY electrostatic loudspeakers, for practical listening distances, you also have to limit the area transmitting high audio frequencies to a small value, 50cm^2 or so, or you get problems with cancellation effects limiting the high-frequency response.
Anyway, when you design your ESL's such that they require frequency-independent current drive, the problem with slewing disappears. You then need as much or as little current at 20kHz as you do at 100Hz.
At the moment, I am working on a mixed transistor-valve (PL519) class A amplifier for directly driving Frank's loudspeakers. If I survive and if it becomes a success, I'll give you more details.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.