SE amps to drive ES panels.

Design and use of the various types.
Post Reply
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

SE amps to drive ES panels.

Post by dave slagle »

SO i met SLK, JJ, and Frank at Izaz's place in DC for 24 hours of fun. I think i learned a few things, but i'll start slow with this one.

Ijaz was using a 300B PP amp to drive his stax ES panels and the sound left me wanting for some dynamics and some emotion. Historically the lack of dynamics makes me like horns over panels, and the lack of life makes me prefer SE over PP. Sure that is a pair of overly simple statements, but it was my experience up until the visit, and it still holds true after. It could very well be that i like a distorted coloured sound but a boy can have an opinion can't he :-)

When JJ's 304TL amp made it in front of the panels, the dynamics stayed the same, but the emotion factor went up considerably. Trust me, i really want to like the concept of PP, but have yet to hear it do what is claimed so the only conclusion i can draw is I have different tastes. I will still work to get that se sound from pp, but for now SE still wins.

Now onto the panels. I first heard ESL's with the SS driven martin logans. Back then i was all about stereophile, imaging and all of the associated parlour tricks. The same Krell amps driving some ESL-57's were much more musical to me. Then i stumbled over to the dark side of high eff. and SET. While I was waffling with my tastes, i heard Stewarts 8B's driving some ESL-63's and heard a lot of the magic, but still was not convinced. Stewart insisted that a 212 amp driving his 63's would do the deed but i needed to choose and i went the lowther path.

This trip to DC was to be the slam dunk that i made the right decision. While the 304TL amp sounded way better at moderate levels, when pushed the sound quickly became confusing and un-involving. When everyone was eating i had the couch to myself and i played anal audiophile. Luckily those pigs ate quick and I only had about 15 minutes so my ears stayed clear. Here is what i noticed.

I kept turning up the volume to get dynamics, but when i reached a point it all fell apart. I left the next morning with a lot on my mind but why the SE struggled when it was pushed was on the top of my list. From what i understand the 304TL amp was putting out 20W which should be in the same ballpark as a PP 300B amp. Then i was told that the 300B amp had some feedback and things started to make sense. The one thing the 300B amp did was keep its composure when pushed, a little feedback goes a long way to reducing the Zout making the reactive load of the panels less of an issue. This made me think that the problems i was hearing with the SE amps had more to do with damping factor than with power output.

If you think about it, everything you do with a nfb SET to get more power tends to reduce damping factor. The 304TL amp was going through a "5K" OT and needed some more B+ to get the full 50W output expected from the beast. This makes me wonder what would happen if the B+ were increased and at the same time the turns ratio reduced to keep the 20W output and net some gains in DF to deal with the panels reactive nature. Could less power and more control of the panels be the path so SET-ESL nirvana?

of course the other option is to add some feedback or use some 6C33's :-)

dave
shinebox
Posts: 88
Joined: Fri Jan 28, 2005 9:44 am

Post by shinebox »

Yo Dave,
A subject close to my heart this... and uncanny how it matches my experience:

A long time ago, Pieter T made me some 1 to (0.75+0.75) SE stepups to drive some ML CLS panels directly from SE. The ratio was determined on the basis of maximising voltage swing driving up to 3kHz (where the panels are beaming and some ribbons would be a better idea) and not running out of slewing current. Ie not asking the tube for more than 60 pct or so of the standing current and keeping the capacitive impedance nice and high...much like what you're saying.

I had a VV52 lash up running at 500V or so - sounded nice but very quiet. The plan was always the 212...which is what the txs were spec'd for.

Paul Barker then paid a visit and we hooked up his 212 breadboard. I was disappointed... sounded dynamically lacking and uninvolving, but I hoped that a big part of this was the lack of EQ, and *still* limited volume, even with the 212s at 1500V or. Confusion on anything remotely complex at decent levels.

Basically, I had *exactly* the same response to the sonics as you did, after such high hopes. I really thought I was an ESL man after an old experience of the 57s.

On a couple of tracks, things were good. Paul at the time reckoned it was the best reproduction of a cello he'd heard but I know his overall fave speakers are the Edgarhorns, which I aint heard but can well imagine.

That's when I decided that for this to have a chance of working with the big complex music I like, I'd have to try a 450TL running at 3500V or so, upping the lower x/o point so the transformers would cope with the higher voltage swing (Pieter reckoned the insulation was good). Oh, and doubled up panels (picture an 8ft high ML CLS with adjacent ribbon tweeter array).

I reckoned that the panels should cover 200 or 250Hz to 2 or 3k. The idea was to roll the top off with resistors in series with the stators, which would also really tame the load seen by the SET. And get a half decent radiation pattern from them, better matching the directivity of the tweeters at crossover. Dipole woofers on the bottom... don't ask about that.

I amassed a few 450TLs, an array of ribbon tweeters and woofers but never had the cash or time to finish the project so it is on long term hold. And bloody unlikely in the near future...

So on to the multiway horn system (originally intended for listening in Summer!) ... then there was the move to a little apartment in Queens which put that on hold.

Well, I'm finally settled in NYC now. Treating myself to some Fertins for a nice, simple, open baffle system that actually has a chance of being completed. Then sort out the amp in a format that won't whack the girlfriend.

Anyways, should hopefully have an afternoon off sometime next month and catch up - long overdue.

Best
cv
IslandPink
Posts: 56
Joined: Thu May 12, 2005 8:01 am
Location: Denbigh, North Wales

ESLs various

Post by IslandPink »

Hi Dave, Chris ,

Can concur with your experiences ( again ) from encounters with Quad 989s and various DIY ESL speakers at our recent Yorkshire meetings . On the DIY ESLs , we had difficulty driving one of the large pairs with anything other than a 300B PP amp with NFB . All other options, small SE , large SE and Aurora VV32 PP were showing distortion at some level . Subsequent models of these speakers have shown better, and in one case quite pleasing , but it took a GM70 SE amp with 1800v on the anode , glowing red, to get them to sound half dynamic . Tone was great , but edge definition was still lacking .
On the Quad 989’s , owned by a local ( North Wales ) guy , he started with SS amps, then ran them with a rather hard-sounding 6550 PP amp which had enough guts ( and NFB ) but wasn’t a great listening experience . He then moved on to a 211 Ongaku-copy which was a lot nicer but more mellow . I took the Auroras ( PP triode, no feedback ) round at that stage , running VV32B pairs , and these amps were the best match I heard on those speakers . Female vocal and certain instruments came out as well as anything I’ve heard that music played on – but we were still down on dynamics and edge definition , beyond the low-level stuff (which was beautifully rendered) . Interestingly the midrange and tone on those 211 SE amps was otherwise ( on other speakers ) slightly better than the Auroras , but not so on the ESLs .

I suppose they are a tough capacitive load, and the bigger the panels the greater the challenge .

I don’t know what the answer is with speakers really . I suspect a well-crafted horn system ( yikes ! ) but how do you guys, say with something like a TAD or JBL horn set-up, integrate any bass from 500-600 Hz downwards ? Seems a bit of a mystery to me .

Mark
slowmotion
Posts: 20
Joined: Wed Mar 09, 2005 7:04 pm
Location: Norge

Post by slowmotion »

Thanks for a good read, guys...

I don't have the tecnical knowledge of you guys, so wont
comment on what it would take to get valve amps to work on panels.
IMHO horns and active crossovers are the way to go,
but it's all so subjective...

I've had good experience with increasing the Zout with frontloaded horns......


cheers
Ed Sawyer

stax panels, etc

Post by Ed Sawyer »

Dave, what stax panels were they? ELS-F81? F83? 8XBB ? ;-) All the stax panel speakers were HUGELY inefficient, on the order of 79 to 80db 1/wm. That may have some bearing on what you heard perhaps? were they running directly off the tube without output trans? or via the normal step down/up method that most electrostatics use?

FWIW, I use Martin Logan CLS I's on the Eluv 845 amps and it's a great match.

thanks
-Ed
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

hey guys,

Mark, It is interesting that the PP amps bettered the SE amp since both were non-feedback designs at about the same power. The reactive load is what made me think that the PP amps typically fare better since they often used feedback to achieve a lower Z-out. Maybe the typically wider bandwidth of the PP amp had some bearing on the results.

Ed,

I'm not sure what stax they were but I was told they were in the 78dB range. The amps both had the traditional 8 ohm outputs. I would have expected the PP 300B amps ot have less power than the 304TL, yet they played louder.

The ESL panels had an overload light on them and with the PP 300B the light would illuminate frequently with no perceived change in sound. With the 304TL amp, if the light would even blinlk occasionally the music would sound "strained" yet a slight reduction in the volume would make a huge difference.

dave
Ed Sawyer

thanks for the update

Post by Ed Sawyer »

Makes sense re: 78db/1wm - that sounds like the smaller ones (F81 or F83 probably). the 4xbb and 8xbb were huge and rare.

I hadn't heard about the overload lights on those before. Must be some sort of voltage sensor? Interesting the 300B played louder and with more authority on those than the 304TL. Not what I would have expected either. Could be some sort of interaction maybe with the electronics of the speaker itself (and it's transformer, etc.)

Stax made a big-ass pure Class A 150wpc solid-state stereo amp that was designed expressly to power their speakers. it was from way back, 1978-1984 era or so I think, called the DA-300. They had later smaller versions that were similar (though class AB I think) called DA-80. They also had some enormous monoblocks (DMA-X2). I have a DA-300, it works well, puts out a fair bit of heat to say the least, and is kinda retro in a sense. it has switchable AC and DC coupling, which is interesting.

I forget the exact details of how/why it was the best amp to use on their speakers, but they had reasons for the way they did it. Literature is hard to come by on those old stax products.

anyway, just a bit of stax/electrostatic trivia.

-Ed
shinebox
Posts: 88
Joined: Fri Jan 28, 2005 9:44 am

Post by shinebox »

Allo
Is it not just the ellptical loadline presented by the ESLs? PP better able to deal with that?

Look at an ESL's capacitive loadline on SE and you quickly see you want to drive a higher impedance than usual to minimise current variation (the "width" of the ellipse)...

But then, that's for direct drive - dunno how the ESL load would look after the speakers internal electronics (x/o or eq)...

Just some thoughts...
DowdyLama
Posts: 104
Joined: Thu Jan 27, 2005 5:42 pm
Location: Blue Ridge Mtns, NC
Contact:

Post by DowdyLama »

This is totally unrelated to tubes or transformers...it's purely nostalgic.

I lived in Kansas City from 1978 through 1981 - during that time I recall a listening session at someone's house who had Stax ES speakers driven by a large pair of Stax SS amps [my young ears thought they sounded pretty good].

But, the thing I remember the most was that the amps were water-cooled - you could actually hear the water gurgling during quiet passages - does anybody recall these amps?
slowmotion
Posts: 20
Joined: Wed Mar 09, 2005 7:04 pm
Location: Norge

Post by slowmotion »

In the mid '70s I was lusting after those Stax amps,
but they were way too expensive for me....
Still I almost got me that preamp they had back then.

:wink:
JeffreyJ
Posts: 320
Joined: Fri Jan 28, 2005 2:53 am
Location: stuck in the USA for now...
Contact:

Se on stats...

Post by JeffreyJ »

hello, all...

I just wanted to pop in and add my thoughts.. basically, they echo dave's... you could hear how the PP300B amp held its composure longer than the SE amp... but it didn't matter... the presentation of the SE amp was so much more my taste... but, it was even warmer there than at my place as this amp just coasts on my speakers... mabe we could have gotten away with the 15E driver on those speakers?

and Dave, the "Sid Special" 300b amp *DID* have feedback... I got to peek at the schematic.. there was an entire folder of Sid's PP thoughts and circuits... the pag ethat Ijaz specificaly pulled out for me did say something about modifying for the same gain without feedback, but it was drawn with feedback... we really should clarify with Ijaz.... to me, that would explain a lot....

we thought we were going to be playing on Quads.. which 20 watts should have been fine on.... but the Stax panels were 76dB.. maybe 78... can't recall exactly.. and which model? I, too, have forgotten.. but they were about seven feet tall and a good three feet wide... not small by any margin... definitely larger than stacked quads...

and, as dave and I discussed more than a year ago, if you are going to build an amp for stats, you definitely should go for the biggest amp you can stand to power/heat/carry and then throw away as much power as you can afford to lose in the step down... this will minimize the variance of the SE amp into the crazy impedance of the stats... still, I didn't hear a dramatic change in the lower registers with the SE amp...

so.. I would call the trip/experiment a success... we know how to build this for a stat customer.... besides, Ijaz's record collection made it worth it for me... what a great guy and the fact that our music tastes overlap some 75% was one hell of a bonus... and then there was the Sarod cincert in the living room... and the demos of the Garrards.. and the Serugi/EMT/Olympos.... man... great time... thanks, Ijaz...

Peace,
Me
Ed Sawyer

stax history

Post by Ed Sawyer »

Jim I am not aware of any liquid-cool stax amps, maybe they had them... the DA-300 I have is dual-fan-cooled. Their biggest amps were the DMA-X1 I think:

http://www.stax.co.jp/OLD/Photo/Dmax1.jpg

There's a history of stax stuff here:

http://www.stax.co.jp/Export/History.html

which shows some of their speakers and some of the amps and other stuff. THere's a few other stax history links out on the web somewhere, they are not that easy to turn up info on.

Jeff those big ones sound bigger than the ones I had envisioned. the F-81 and F-83 were the common ones. it sounds like the ones you were listening to were the larger ELS-8 or ELS-4 models. those were quite rare. THey had a battery-biased version too (BB). Have never heard any of those but would love to someday.

FWIW
-Ed
JeffreyJ
Posts: 320
Joined: Fri Jan 28, 2005 2:53 am
Location: stuck in the USA for now...
Contact:

It was these...

Post by JeffreyJ »

[/url]http://www.stax.co.jp/OLD/Photo/8XBB.JPG[url]


identified by sight... agreed, Dave?
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Image

Yes, that appears to be them.

what is the battery bias all about???

dave
Ed Sawyer

wow

Post by Ed Sawyer »

If you guys heard some ELS-8X's that's quite a feat. those are exceedingly rare. I am jealous. ;-)

the battery bias - not sure. Presumably something to do with the power supply for the panel. Either to bias the stators or panel? Not sure, Stax info is hard to come by. will see what I can dig up.

They had BB and non-BB versions, the BB's were the later ones. (ELS-8XBB vs ELS-8X I think)

-Ed
Ed Sawyer

More info

Post by Ed Sawyer »

http://www.audiocircuit.com/A-HTML/AA-S ... itcode=941

apparently the batteries biased the panel to 4kv!

Stax did some direct-drive transformless amps too apparently, for driving these.

interesting reading...

-Ed
lisajackusde
Posts: 1
Joined: Sat Dec 08, 2012 6:51 am

Post by lisajackusde »

FWIW, I use Martin Logan CLS I's on the Eluv 845 amps and it's a great match.
liuyelian
Posts: 1
Joined: Mon Jul 01, 2013 8:05 am

Post by liuyelian »

That may have some bearing on what you heard perhaps? were they running directly off the tube without output trans?
suqun1
Posts: 4
Joined: Thu Dec 26, 2013 8:48 am

Post by suqun1 »

I was disappointed... sounded dynamically lacking and uninvolving, but I hoped that a big part of this was the lack of EQ, and *still* limited volume, even with the 212s at 1500V or. Confusion on anything remotely complex at decent levels.
Gnnett
Posts: 1
Joined: Mon Jan 30, 2006 12:40 am

Re: SE amps to drive ES panels.

Post by Gnnett »

Prodding phoenix, so to speak, or more like stoking the fire (risks).
I know that shows in the not too distant past have featured some beautiful stacked ELS57's with dedicated 300B amplifiers. I had always assumed these were direct drive amplifiers, however from this thread, is it possible they are actually using low ratio step up transformers, or autoformers?
I would really enjoy hearing a little more background detail on how these wonderful statement beasts actually operate, including how the different step up ratios between the bass and mid/hi transformers are accommodated and the associated equalisation for the 3 way panel system is achieved.

Obviously not interested in any commercially sensitive, or PI details, just conceptual stuff for inquisitive minds.

Kind Regards

Grant
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Re: SE amps to drive ES panels.

Post by dave slagle »

You are indeed correct that the panels are driven by step-up transformers from the plates of the tubes. It always made little sense to me to step down the voltage 25X for a cable run to only step it up 270X to drive the panels. Asking two transformers to do this (amp output / quad input) and maintain 20-20k bandwidth is a daunting challenge. Splitting the signal at line level and biamping allows for a single transformer to be designed for the frequency ranges in question.

dave
Get Your Fix
www.hifiheroin.com
Post Reply