Got a good deal on Stax ESL-F81 electrostatic speakers. They are in good working condition.
I am trying to figure out what amps would drive F81 best. I have a Class A 20W custom made ECC88/MOSFET Amp: works great with high sensitivity speakers but fails badly with Stax, McCormack DNA-1 Deluxe: drives Stax fine but is not as fluid as Aleph 3, Pass Aleph 3: works fine but clips at higher SPL and Symphonic Line RG11 that works nicely but does not bring anything special to the table. As you can see these are very demanding speakers.
With impedance all over the place from 2Ω to 500Ω depending on the frequency what would be an ideal amp? Please provide commercial models as well as DIY choices and general ideas.
I am trying to figure out what amps would drive F81 best. I have a Class A 20W custom made ECC88/MOSFET Amp: works great with high sensitivity speakers but fails badly with Stax, McCormack DNA-1 Deluxe: drives Stax fine but is not as fluid as Aleph 3, Pass Aleph 3: works fine but clips at higher SPL and Symphonic Line RG11 that works nicely but does not bring anything special to the table. As you can see these are very demanding speakers.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
With impedance all over the place from 2Ω to 500Ω depending on the frequency what would be an ideal amp? Please provide commercial models as well as DIY choices and general ideas.
Stax F81 amplifier recommendations;
The STAX esl seems to have a synergistic match with OTL amps like the Futterman, Counterpoint SA4 or AtmaSphere, much like the SoundLab esls. A second choice would be a high power PP tube amp of at least 120 watts like the Jadis, VTL, Tube Research Labs or other heavy iron known for a superior midrange performance. You can forget about any flea power SET amps and low power class A amps, start with at least 100 watts output. One of the problems with the STAX F81 and even its big brother the F83 which is just a stacked pair of 81s, is that they are basically a midrange and up speaker, no real bass to them at all. Oh they played games with the tuning of the diaphragm but that just produces a sort of one note bass performance even worse than the original Martin Logan CLS had. The STAX is a polite speaker best suited for small ensemble works and especially vocals at which it possibly excels the vaulted QUAD 57 one of the best midrange speakers ever. The character of the midrange is such that it is not easy to match any dynamic cone subwoofer to the system. What is needed is a midbass driver covering the range of 60hz to 150hz, maybe a pair of Magnepan bass panels perhaps. That will necessitate an electronic crossover and separate bass amp. Marchand offers the XM26 and Bryston and FM acoustics both have high-end and costly crossovers; on the budget side look for a used Audio Research EC2 crossover but update its capacitors and diodes. Some suggest that the bass amp can be different than the high pass amp but that transition is always noticeable on highly transparent speaker systems. I recommend using the same model for both the bass and midrange amplifiers. One could build a pair of DIY OTL amps as the schematics are out there, either the totem pole Futterman circuit or the circlotron based AtmaSphere; although the 60 may not be enough power you really want at least 100 watts or more from the tube amp. The second major caveat of the STAX is its astoundingly low efficiency, the two worst offenders I've run across in this regard were the original 3 way Apogee Full Range ribbon and the STAX esl, both need a high power amp to achieve even normal listening levels while at the same time will overload and compress if driven too hard. There is a narrow volume range at which the speaker will perform at its best.
You might ask over on the AudioAsylum planar forum.
The STAX esl seems to have a synergistic match with OTL amps like the Futterman, Counterpoint SA4 or AtmaSphere, much like the SoundLab esls. A second choice would be a high power PP tube amp of at least 120 watts like the Jadis, VTL, Tube Research Labs or other heavy iron known for a superior midrange performance. You can forget about any flea power SET amps and low power class A amps, start with at least 100 watts output. One of the problems with the STAX F81 and even its big brother the F83 which is just a stacked pair of 81s, is that they are basically a midrange and up speaker, no real bass to them at all. Oh they played games with the tuning of the diaphragm but that just produces a sort of one note bass performance even worse than the original Martin Logan CLS had. The STAX is a polite speaker best suited for small ensemble works and especially vocals at which it possibly excels the vaulted QUAD 57 one of the best midrange speakers ever. The character of the midrange is such that it is not easy to match any dynamic cone subwoofer to the system. What is needed is a midbass driver covering the range of 60hz to 150hz, maybe a pair of Magnepan bass panels perhaps. That will necessitate an electronic crossover and separate bass amp. Marchand offers the XM26 and Bryston and FM acoustics both have high-end and costly crossovers; on the budget side look for a used Audio Research EC2 crossover but update its capacitors and diodes. Some suggest that the bass amp can be different than the high pass amp but that transition is always noticeable on highly transparent speaker systems. I recommend using the same model for both the bass and midrange amplifiers. One could build a pair of DIY OTL amps as the schematics are out there, either the totem pole Futterman circuit or the circlotron based AtmaSphere; although the 60 may not be enough power you really want at least 100 watts or more from the tube amp. The second major caveat of the STAX is its astoundingly low efficiency, the two worst offenders I've run across in this regard were the original 3 way Apogee Full Range ribbon and the STAX esl, both need a high power amp to achieve even normal listening levels while at the same time will overload and compress if driven too hard. There is a narrow volume range at which the speaker will perform at its best.
You might ask over on the AudioAsylum planar forum.
by "synergy" do you mean 3-5 Ohms series output R between amp and ESL-F81's xmfr?
it certainly looks like that could help flatten the on axis frequency response in the top octave: http://www.audioasylum.com/images/staxELS.pdf
but resistors could be added to any SS amp output too
ridiculous conservatism would suggest resistor power rating similar to amp power - 200 W can be had for ~$20-30 ea
but realistic music 3 KHz power bandwidth and peak overload ratings of big reisitors mean you could proably get away with much lower rating - just put them in a flame proof box with enough surface area to get rid of the heat
it certainly looks like that could help flatten the on axis frequency response in the top octave: http://www.audioasylum.com/images/staxELS.pdf
but resistors could be added to any SS amp output too
ridiculous conservatism would suggest resistor power rating similar to amp power - 200 W can be had for ~$20-30 ea
but realistic music 3 KHz power bandwidth and peak overload ratings of big reisitors mean you could proably get away with much lower rating - just put them in a flame proof box with enough surface area to get rid of the heat
Last edited:
bridged & paralleled chip amps could be OK, wrap in a quality op amp overall feedback loop, other ideas from http://www.diyaudio.com/forums/chip-amps/164634-chip-amp-rival-hi-end-design-advice.html#post2144636
extrapolating from Stereophile's Golden Eared reviewer's experience you couldn't expect anyone to do better: http://www.diyaudio.com/forums/solid-state/12752-blind-listening-tests-amplifiers-55.html#post152392
extrapolating from Stereophile's Golden Eared reviewer's experience you couldn't expect anyone to do better: http://www.diyaudio.com/forums/solid-state/12752-blind-listening-tests-amplifiers-55.html#post152392
Last edited:
Stax F81 and OTL amps
No, I wasn't referring to transfer characteristics, I meant the sound quality or character of the two products compliment each other. Furthermore, a resistor between a solid state amp and the step-up transformer of the Stax is not going too improve matters any. As far as I remember from the 80s when the STAX F81 was manufactured I do not recall any solid state amp working very well with it, strictly a vacuum tube amp type of speaker.
In regards to vacuum tube amps they should have an output impedance of less than 5 ohms. Interestingly, Roger A. Modjeski of RAM tubes fame designed the SA 4 for Counterpoint and I recall Counterpoint quoting a damping factor of 4, I could be mistaken on that. Damping factor is not as important for electrostatic speakers as it is for conventional dynamic cone speaker systems. OTL amps will deliver more power into a high impedance load ( by that I mean above 32 ohms and certainly above 100 ohms) than a solid state amp will. Now solid state is great for driving a sub one ohm ribbon speaker but not the best choice for driving a one hundred ohm electrostatic. I sense a reluctance to try out an OTL amp, why not at the very least audition an AtmaSphere amp at a dealer as it seems to be the most commonly available OTL on the market. And, while OTL amps can be problematic with conventional speaker systems they really do work quite well with electrostatic speakers both Beveridge and Acoustat had direct drive tube amps for their electrostatic speakers. Or, if you already have a collection of electronic parts why not build a Futterman OTL, if not to your liking you could always sell the power tubes? 6AS7 and 6LF6 tubes always have a ready buyer..
No, I wasn't referring to transfer characteristics, I meant the sound quality or character of the two products compliment each other. Furthermore, a resistor between a solid state amp and the step-up transformer of the Stax is not going too improve matters any. As far as I remember from the 80s when the STAX F81 was manufactured I do not recall any solid state amp working very well with it, strictly a vacuum tube amp type of speaker.
In regards to vacuum tube amps they should have an output impedance of less than 5 ohms. Interestingly, Roger A. Modjeski of RAM tubes fame designed the SA 4 for Counterpoint and I recall Counterpoint quoting a damping factor of 4, I could be mistaken on that. Damping factor is not as important for electrostatic speakers as it is for conventional dynamic cone speaker systems. OTL amps will deliver more power into a high impedance load ( by that I mean above 32 ohms and certainly above 100 ohms) than a solid state amp will. Now solid state is great for driving a sub one ohm ribbon speaker but not the best choice for driving a one hundred ohm electrostatic. I sense a reluctance to try out an OTL amp, why not at the very least audition an AtmaSphere amp at a dealer as it seems to be the most commonly available OTL on the market. And, while OTL amps can be problematic with conventional speaker systems they really do work quite well with electrostatic speakers both Beveridge and Acoustat had direct drive tube amps for their electrostatic speakers. Or, if you already have a collection of electronic parts why not build a Futterman OTL, if not to your liking you could always sell the power tubes? 6AS7 and 6LF6 tubes always have a ready buyer..
the point of the Carver Challenge is that the "best ears in the industry" Did Lose that bet - when the midprice SS amp was tweaked to match in frequency response, output impedance to a "5 figure SOTA Tube amp"
I agree that you don't need continuous power drive capability at 20 KHz - but you might want the current output capability to be there for transients like snare drum hits - so you do need an amp capable of pushing current into the lowest input impedance of the speaker
That's where a custom chip amp solution can save some bucks - TAD7293/4 handle high ps V but you wouldn't have to size the heatsink, ps for continuous power into the 4 Ohms at 20 KHz unless you just want to watch the 'scope trace (I wouldn't want to be in the room with the speakers!)
if you need the reassurance of audiophile snob appeal - the chips have been used in Kilobuck "audiophile" amps from Linn and others
Last edited:
Hi,
when You want to use chips like the TDAs with ESLs then buy them in Kilos and make sure You can easily replace them.
That such chips are used in kilobuck amps (no audiophile ones for sure) just shows how expensive You can sell cheap stuff when You´ve learned Your marketing lessons.
jauu
Calvin
when You want to use chips like the TDAs with ESLs then buy them in Kilos and make sure You can easily replace them.
That such chips are used in kilobuck amps (no audiophile ones for sure) just shows how expensive You can sell cheap stuff when You´ve learned Your marketing lessons.
jauu
Calvin
just use enough in the 1st place and they won't be over stressed - 4x parallel with the "eye-ball" amp output series R target of 4 Ohms for the Stax would give 8 Ohm resistors in series with each chip amp in bridged-parallel
I certainly hope its pretty hard breaking a chip amp with 8 Ohm real resistance in series with any passive RLC combo on the output and "music" signals
they're certainly cheap enough that you can even double those numbers and together the chips still cost less than tubes in OTL capable of 10% of the output current of just one of the chips
I certainly hope its pretty hard breaking a chip amp with 8 Ohm real resistance in series with any passive RLC combo on the output and "music" signals
they're certainly cheap enough that you can even double those numbers and together the chips still cost less than tubes in OTL capable of 10% of the output current of just one of the chips
If you own a ESL it is a little late to complain about low efficiency
The series R does “waste” considerable % of the power at 20 KHz – but there really shouldn’t be much power in music there – conventionally music spectral content has been approximated as having ~1st order roll off above ~3 KHz
I guesstimate the Stax apparent load at the step up xfmr pri as ~5uF at audio frequencies from the impedance plot
After looking closer I would choose ~3 Ohm series R to help tame the 20 KHz >6 dB on axis SPL peak in the manual’s FR plot
The 3 Ohms series R would waste ~27% of the circulating power at 3 KHz, the expected R loss with music decreasing proportionately to f^2 below the music spectra breakpoint
that roughly equals Class B 30% loss at max power
The series R does “waste” considerable % of the power at 20 KHz – but there really shouldn’t be much power in music there – conventionally music spectral content has been approximated as having ~1st order roll off above ~3 KHz
I guesstimate the Stax apparent load at the step up xfmr pri as ~5uF at audio frequencies from the impedance plot
After looking closer I would choose ~3 Ohm series R to help tame the 20 KHz >6 dB on axis SPL peak in the manual’s FR plot
The 3 Ohms series R would waste ~27% of the circulating power at 3 KHz, the expected R loss with music decreasing proportionately to f^2 below the music spectra breakpoint
that roughly equals Class B 30% loss at max power
I used to have the STAX ESL-F81. The only amp that can drive them is the Futterman OTL3 which I still have them, but sold the STAX. The OTL3 can played them louder than any transformer coupled amp that I have tried (200-300W range). STAX sounded great on guitar, violin and vocal, even on the OTL3 it can not play as loud as the ESL57, that's why I sold it.
You also can go with any large quality SS amp. Conrad-Johnson SS amps are particularly good with big ESL's. I haven't had the pleasure of using them with Stax, but they are a very good match for Acoustats, and Soundlabs. Any of the CJ 200watt or 250 watt models should do you well. Older Classe 25's will work also, but in my opinion did not sound as good. Counterpoints would also drive them fine, I just think the above mentioned amps sounded nicer.
In tubes VTL's will do quite well also.
Maybe it is just me, but I was not favorably impressed with Atma-sphere OTL's on big Soundlabs. It has a distinctive sound. Was pleasant in some ways, but rather colored. It also seemed not to have any guts. ARC or VTL amps just under 100 watts were much better on these speakers than the OTL.
I think any of the good, high power SS amps would work. If they didn't I would try adding a series one ohm resistor.
In tubes VTL's will do quite well also.
Maybe it is just me, but I was not favorably impressed with Atma-sphere OTL's on big Soundlabs. It has a distinctive sound. Was pleasant in some ways, but rather colored. It also seemed not to have any guts. ARC or VTL amps just under 100 watts were much better on these speakers than the OTL.
I think any of the good, high power SS amps would work. If they didn't I would try adding a series one ohm resistor.
Kent from electrostatic solution restore them personally uses the ARC REF300
It's 300 watts can achieve 98-100dB peaks before clipping about the limit of the F81 design max 300 watts.
I sent my F81 panel from Australia to him in the US to get them restored.
He advised me that my Primaluna amp can drive them at decent levels.
It's loaded 8x KT150 100 watts, I get about 92dB.
Basically these speaker aren't design to play loud and if you keep trying to clip them.
It means early panel death.
It's 300 watts can achieve 98-100dB peaks before clipping about the limit of the F81 design max 300 watts.
I sent my F81 panel from Australia to him in the US to get them restored.
He advised me that my Primaluna amp can drive them at decent levels.
It's loaded 8x KT150 100 watts, I get about 92dB.
Basically these speaker aren't design to play loud and if you keep trying to clip them.
It means early panel death.