Hi
today my amp quit working , the fuse is black an a part called sanken 2sa1216
is black too it was a sony 550 es amp I mean it still is I hope replacing the parts should fix it.
so this amp isn't the right one i think.now i have the sub a TL with a sls10 and a active crossover behringer and the esl panels work well but i need now a amp for driving the esl.
I also tried a resistor 1 ohm at primary winding but they get hot and smoky.
My question what amp work fine for an esl and what specs must it have.
what amps do you have?
(Ps: I accidentally unsubscribe from my last thread )
thanks hironimo
today my amp quit working , the fuse is black an a part called sanken 2sa1216
is black too it was a sony 550 es amp I mean it still is I hope replacing the parts should fix it.
so this amp isn't the right one i think.now i have the sub a TL with a sls10 and a active crossover behringer and the esl panels work well but i need now a amp for driving the esl.
I also tried a resistor 1 ohm at primary winding but they get hot and smoky.
My question what amp work fine for an esl and what specs must it have.
what amps do you have?
(Ps: I accidentally unsubscribe from my last thread )
thanks hironimo
Hi hironimo,
you didn't give some details 'bout your ESL's.
In my installation, I use a Vincent SP331. It has good control over my ML Sequel2 and doesn't sound stressed. Tight bass and silky highs.
I have also good experience with a tube amp (KT88 PP), but only with impedance correction to flaten the impedance.
Most ESL are a difficult load to amps, not because low efficiency, the reason is more the twist in phase combined with a capacitive load.
Regards
Olaf
you didn't give some details 'bout your ESL's.
In my installation, I use a Vincent SP331. It has good control over my ML Sequel2 and doesn't sound stressed. Tight bass and silky highs.
I have also good experience with a tube amp (KT88 PP), but only with impedance correction to flaten the impedance.
Most ESL are a difficult load to amps, not because low efficiency, the reason is more the twist in phase combined with a capacitive load.
Regards
Olaf
Hi Olaf
my system is:
2 toroidal transformers 1:65 ,
regulated bias supply
the tall speaker has 4 panel each 6 inch x 10 inch ( 15 x 25 cm)
the grey panel is 16 x 20 inch ( 40 x 50 cm)
a tl with a peerless sls10
coated with elvamide.
behringer crossover
behringer amp reference500
I'm experimenting how o to put all things together it sounds ok but not so good it's hard to match the tl with the esl and daming the tl more or less makes not much difference to me
my priority is now to get another amp.
thanks!
hironimo
my system is:
2 toroidal transformers 1:65 ,
regulated bias supply
the tall speaker has 4 panel each 6 inch x 10 inch ( 15 x 25 cm)
the grey panel is 16 x 20 inch ( 40 x 50 cm)
a tl with a peerless sls10
coated with elvamide.
behringer crossover
behringer amp reference500
I'm experimenting how o to put all things together it sounds ok but not so good it's hard to match the tl with the esl and daming the tl more or less makes not much difference to me
my priority is now to get another amp.
thanks!
hironimo
Attachments
If I could presume you could read German :
FAQ
under the section "Können Sie einen geeigneten Verstärker empfehlen ?"
Naturally they are meant for Capaciti products, but most ESLs have similar impedance behaviour, so you might wish to consider what he says. He did recommend using Class D amps in one of the German HiFi forums, if I remember correctly.
PS I have no link with the company whatsoever.
Patrick
FAQ
under the section "Können Sie einen geeigneten Verstärker empfehlen ?"
Naturally they are meant for Capaciti products, but most ESLs have similar impedance behaviour, so you might wish to consider what he says. He did recommend using Class D amps in one of the German HiFi forums, if I remember correctly.
PS I have no link with the company whatsoever.
Patrick
Hi,
as Taotao wrote low resistance are typically not the major problem for a decent amplifier. The major problem is rather the complex load which You can read from the phase-response of the impedance plots.
Amplifiers with global feedback which is connected from their very outputs to the inputs lets the amplifier react on this complex load. The negative feedback, which under normal conditions reduces failure artefacts within the signal, can change towards a positive feedback that enhances failure artefacts and can lead to oscillation of the amp. I assume this happened to Your amp.
There are some measures against this effect, some relate to the speaker itself and some relate to the amplifier.
Toroid transformers typically are lowloss devices which means that the amplifier ´sees´ more from the ESL´s capacitive nature. The phase response shows exaggerated values which can reach values of >80°. This is alot more than many amps can handle, hence they are prone to oscillate. Transformers with different cores like EI typically present the amp with lower values of phase shift and the amps life is much easier and it might remain in a stable working condition.
Series connected resistors (they need to be power resistors rated at several W) also reduce the burden on the amplifier by reducing phase shift and increasing the resistive part of the load impedance. Because this measure also influences the amplitude response of the speaker, resistor tuning works only within tight limits.
The amplifier related measurements mean modifications of an existing amplifier -without the guarantee of success- or the choice of a capable amplifier. Tube power amps coulod be a choice. Their output transformer counters the capacitive ESL load to a degree and Tubes show a more relaxed behaviour under large signal or overdrive conditions. The copper resistance of the transformer works like a seriel resistance, so in some cases You don´t need a dedicated series resistor or a lower value (otherwise there could be a high frequency loss). With transistor amps it may be useful to choose an amp which has been tested to work stables into loads with vast levels of phaseshift. It may also be helpful to break the direct output-input connection of the negative feedback. Most amps feature sourcefollowers or emitterfollowers as power output stage. This confirguration has a gain of ~1 and a tight internal feedback. Because of this one can place the output feedback node right in front of this last stage, thereby ´isolating´ the feedback from the load. Mr. Erno Borbely had some amps on his website where You could choose between different feedback-nodes. Probabely the best would be to omit with global feedback altogether. Single Ended Triode amps wo global feedback seem to be the best solution not only from the point of stability but also from the point of sound. ESLs are very revealing and I must say that I still have to hear the globally fed back amp that doesn´t sound grainy and flat with a good panel.
Class-D amplifiers are a different class of amplifiers that may give good results with regard to stability. But this depends similar to class-A or class-AB amps on their topology. Class-D which don´t incorporate their output filter into the feedback loop (type A) are rather stable working, class-D with global feedback (type B) have their feedback loop modified by the load. Class-D amps have problems similar to classA/AB and problems of their own. In case of type A the output inductor forms a resonant circuit with the ESL. As a result a quite strong peak in the amlitude response develops (I measured up to +12dB!) which lies within the audio band (10kHz-20kHz) or the low ultrasonic range. Amplitude response above the peak frequency drops sharply. This must be countered with an equalizer that notches the peak. Still tough the bandwidth limit of the system may be as low as 10kHz. Class-D of the type B show a better amplitude response and higher bandwidth. Sonically on both types the ESL sounds very tightly controlled and high level dynamics can be truely outstanding. But the tonal character is rather technical -some may say ´cold´. With ´processed´ music this can sound awesome with natural sounds and classical music its sounds rather uninspiring. You never doubt any second that You listen to a technical reproduction of music. Sadly ESLs are very unforgiving in this quality of an amplifier which ha meade me very supspicious over the years about the use and good or bad of global feedback in audio amplification.
More unfortunately are good and powerful SE-triode amps rare and expensive. If Your budget can afford the abuse You might give such a tube amp a chance. But beware You might be affected by a music virus ;-) ....and what the heck...building a concert hall in Your garden and hiring musicians will be even more expensive in the long term
jauu
Calvin
as Taotao wrote low resistance are typically not the major problem for a decent amplifier. The major problem is rather the complex load which You can read from the phase-response of the impedance plots.
Amplifiers with global feedback which is connected from their very outputs to the inputs lets the amplifier react on this complex load. The negative feedback, which under normal conditions reduces failure artefacts within the signal, can change towards a positive feedback that enhances failure artefacts and can lead to oscillation of the amp. I assume this happened to Your amp.
There are some measures against this effect, some relate to the speaker itself and some relate to the amplifier.
Toroid transformers typically are lowloss devices which means that the amplifier ´sees´ more from the ESL´s capacitive nature. The phase response shows exaggerated values which can reach values of >80°. This is alot more than many amps can handle, hence they are prone to oscillate. Transformers with different cores like EI typically present the amp with lower values of phase shift and the amps life is much easier and it might remain in a stable working condition.
Series connected resistors (they need to be power resistors rated at several W) also reduce the burden on the amplifier by reducing phase shift and increasing the resistive part of the load impedance. Because this measure also influences the amplitude response of the speaker, resistor tuning works only within tight limits.
The amplifier related measurements mean modifications of an existing amplifier -without the guarantee of success- or the choice of a capable amplifier. Tube power amps coulod be a choice. Their output transformer counters the capacitive ESL load to a degree and Tubes show a more relaxed behaviour under large signal or overdrive conditions. The copper resistance of the transformer works like a seriel resistance, so in some cases You don´t need a dedicated series resistor or a lower value (otherwise there could be a high frequency loss). With transistor amps it may be useful to choose an amp which has been tested to work stables into loads with vast levels of phaseshift. It may also be helpful to break the direct output-input connection of the negative feedback. Most amps feature sourcefollowers or emitterfollowers as power output stage. This confirguration has a gain of ~1 and a tight internal feedback. Because of this one can place the output feedback node right in front of this last stage, thereby ´isolating´ the feedback from the load. Mr. Erno Borbely had some amps on his website where You could choose between different feedback-nodes. Probabely the best would be to omit with global feedback altogether. Single Ended Triode amps wo global feedback seem to be the best solution not only from the point of stability but also from the point of sound. ESLs are very revealing and I must say that I still have to hear the globally fed back amp that doesn´t sound grainy and flat with a good panel.
Class-D amplifiers are a different class of amplifiers that may give good results with regard to stability. But this depends similar to class-A or class-AB amps on their topology. Class-D which don´t incorporate their output filter into the feedback loop (type A) are rather stable working, class-D with global feedback (type B) have their feedback loop modified by the load. Class-D amps have problems similar to classA/AB and problems of their own. In case of type A the output inductor forms a resonant circuit with the ESL. As a result a quite strong peak in the amlitude response develops (I measured up to +12dB!) which lies within the audio band (10kHz-20kHz) or the low ultrasonic range. Amplitude response above the peak frequency drops sharply. This must be countered with an equalizer that notches the peak. Still tough the bandwidth limit of the system may be as low as 10kHz. Class-D of the type B show a better amplitude response and higher bandwidth. Sonically on both types the ESL sounds very tightly controlled and high level dynamics can be truely outstanding. But the tonal character is rather technical -some may say ´cold´. With ´processed´ music this can sound awesome with natural sounds and classical music its sounds rather uninspiring. You never doubt any second that You listen to a technical reproduction of music. Sadly ESLs are very unforgiving in this quality of an amplifier which ha meade me very supspicious over the years about the use and good or bad of global feedback in audio amplification.
More unfortunately are good and powerful SE-triode amps rare and expensive. If Your budget can afford the abuse You might give such a tube amp a chance. But beware You might be affected by a music virus ;-) ....and what the heck...building a concert hall in Your garden and hiring musicians will be even more expensive in the long term
jauu
Calvin
Hi
Thank you for your replies I'm getting slowly smarter.
so a transistor amp should work with 2 ohm speaker
min 100 w and it's better it has a huge powersupply ?
and a class D amp B type should be fine Also 100 W ?
and when I want to get a tube amplifier , how much W should it have?
Can i just buy one and it will work fine with my panels (sound and quality of the amp is another topic)
(and what about "TaoTao :I have also good experience with a tube amp (KT88 PP), but only with impedance correction to flaten the impedance.")
hironimo
Thank you for your replies I'm getting slowly smarter.
so a transistor amp should work with 2 ohm speaker
min 100 w and it's better it has a huge powersupply ?
and a class D amp B type should be fine Also 100 W ?
and when I want to get a tube amplifier , how much W should it have?
Can i just buy one and it will work fine with my panels (sound and quality of the amp is another topic)
(and what about "TaoTao :I have also good experience with a tube amp (KT88 PP), but only with impedance correction to flaten the impedance.")
hironimo
Hi hironimo,
I'll give you an example, concerning impedance correction.
If I measure the impedance over frequency at my ML Sequel2, I get the diagram as you see in the first left pic.
After adding an impedance correction, I get the curve as shown in the second pic.
Tube amps in general have lower dampening factor in comparison to solid state amps,
as a result, the peaks in impedance response produce a peak in frequency response
(in my case up to 4 dB more in the midrange with the Sequels if impedance is uncorrected).
The corrected impedance gives a nicer load for the amp,
also the phase twist is lower. All in all the sound is better.
Even SS amps can benefit from this correction
(although, with my Vincent it's not necessary/ audible).
You need only three parts (if you don't flaten the imp peak in the bass region).
This parts don't have to be high quality parts.
You need a coil, a resistor (cheap cement type, 5-10 watt) and a cap
(electrolytic is sufficient).
The values I needed were: a coil with 0,15mH (0,3ohm),
a 4ohm resistor and 47 mF cap (not milli, but micro, I can't put the right sympol in this editor
This three parts have to be connected in serie from the minus to plus terminal of the amp (or at the speaker terminals).
Of course, the values are for the sequel, but may work with your combination, too
(if you should decide for a tube amp).
Concerning power: My amps have EL34 and produce around 38 watts per channel.
This is plenty enough for my needs and I can play very loud (OK, very subjective, of course)
One addition: I didn't calculate the values, I used ARTA for imp. measuring (ARTA Home)
with a measuring box and some parts I had at hand.
Luckily they fitted well.
Regards
Olaf
I'll give you an example, concerning impedance correction.
If I measure the impedance over frequency at my ML Sequel2, I get the diagram as you see in the first left pic.
After adding an impedance correction, I get the curve as shown in the second pic.
Tube amps in general have lower dampening factor in comparison to solid state amps,
as a result, the peaks in impedance response produce a peak in frequency response
(in my case up to 4 dB more in the midrange with the Sequels if impedance is uncorrected).
The corrected impedance gives a nicer load for the amp,
also the phase twist is lower. All in all the sound is better.
Even SS amps can benefit from this correction
(although, with my Vincent it's not necessary/ audible).
You need only three parts (if you don't flaten the imp peak in the bass region).
This parts don't have to be high quality parts.
You need a coil, a resistor (cheap cement type, 5-10 watt) and a cap
(electrolytic is sufficient).
The values I needed were: a coil with 0,15mH (0,3ohm),
a 4ohm resistor and 47 mF cap (not milli, but micro, I can't put the right sympol in this editor
This three parts have to be connected in serie from the minus to plus terminal of the amp (or at the speaker terminals).
Of course, the values are for the sequel, but may work with your combination, too
(if you should decide for a tube amp).
Concerning power: My amps have EL34 and produce around 38 watts per channel.
This is plenty enough for my needs and I can play very loud (OK, very subjective, of course)
One addition: I didn't calculate the values, I used ARTA for imp. measuring (ARTA Home)
with a measuring box and some parts I had at hand.
Luckily they fitted well.
Regards
Olaf
Attachments
Hi,
the proton is a typical complementary symmetrical class-AB Amplifier with global feedback, running on 50ma-100mA idle current.
What makes it special is, that for high voltage demands the supply voltage for the output stage is switched to app. double the value.
There´s considerable capacitance in the power supplies and the supply for Input - and driver stage is nicely regulated. The load current has to run over traces on the PCB though which might be small. In comparison a Rotel 980BX features massive copper rails which look more trustful for the claimed current-/power capability. The compensation cap parallel to the feedback resistors is quite small in value (3pF).
Apart from that a nice looking amp. ESL capability?? Needs testing.
jauu
Calvin
the proton is a typical complementary symmetrical class-AB Amplifier with global feedback, running on 50ma-100mA idle current.
What makes it special is, that for high voltage demands the supply voltage for the output stage is switched to app. double the value.
There´s considerable capacitance in the power supplies and the supply for Input - and driver stage is nicely regulated. The load current has to run over traces on the PCB though which might be small. In comparison a Rotel 980BX features massive copper rails which look more trustful for the claimed current-/power capability. The compensation cap parallel to the feedback resistors is quite small in value (3pF).
Apart from that a nice looking amp. ESL capability?? Needs testing.
jauu
Calvin
I am going to give what may be an unpopular answer here.
99% of all solid state amps sound like doggie poop running an ESL.
Not all. 99%
I make solid state amps, so I don't say that lightly.
Tube amps tend to sound MUCH better on most ESLs.
ESLs that run down to 2 ohms, imo are perhaps not properly designed - but IF you are going to run an ESL that drops to sub 4 ohm impedance then you are more or less stuck with a big solid state amp with multiple paralleled output devices. Or you can build one which is not hard given the info on this site... I do not like that option much, the big solid state amp.
An OTL + Step down transformer is a bit silly, because the entire point of an OTL is to NOT have an output transformer. I'd say just build an amp with a low output Z transformer would be my advice.
Probably less expensive
In general ESLs want to see a big voltage swing.
That mandates a "higher power" amp because that is what has the voltage swing.
In my experience using receivers and cheap amps yields rather inferior sonic results compared to what is possible from ESLs. But any amp is better than no amp I guess... there aren't that many inexpensive commercial amps that will sound decent on an ESL and also provide sufficient output voltage swing.
On that last point, I'd rather be hitting the rails on a typical Mosfet output amp than a typical bipolar amp because the bipolar "hits hard" and makes ugly sounding artifacts on clipped peaks. Fwiw, you will be clipping your amp unless you listen pretty darn low in SPL. Sad fact.
Of course these are merely my opinions.
_-_-bear
Edit: once you step down using a transformer with that OTL, what has happened to the voltage swing?
99% of all solid state amps sound like doggie poop running an ESL.
Not all. 99%
I make solid state amps, so I don't say that lightly.
Tube amps tend to sound MUCH better on most ESLs.
ESLs that run down to 2 ohms, imo are perhaps not properly designed - but IF you are going to run an ESL that drops to sub 4 ohm impedance then you are more or less stuck with a big solid state amp with multiple paralleled output devices. Or you can build one
which is not hard given the info on this site... I do not like that option much, the big solid state amp.An OTL + Step down transformer is a bit silly, because the entire point of an OTL is to NOT have an output transformer. I'd say just build an amp with a low output Z transformer would be my advice.
Probably less expensive
In general ESLs want to see a big voltage swing.
That mandates a "higher power" amp because that is what has the voltage swing.
In my experience using receivers and cheap amps yields rather inferior sonic results compared to what is possible from ESLs. But any amp is better than no amp I guess... there aren't that many inexpensive commercial amps that will sound decent on an ESL and also provide sufficient output voltage swing.
On that last point, I'd rather be hitting the rails on a typical Mosfet output amp than a typical bipolar amp because the bipolar "hits hard" and makes ugly sounding artifacts on clipped peaks. Fwiw, you will be clipping your amp unless you listen pretty darn low in SPL. Sad fact.
Of course these are merely my opinions.
_-_-bear
Edit: once you step down using a transformer with that OTL, what has happened to the voltage swing?
Last edited:
The one best way: direct high voltage Sanders-like amp.
Before losing my nerve (or my life) I ran an amp direct, no caps, to the stators with a B+ of 2400 volts and an ESL bias of minus couple thousand to the diaphragm.
No question in my mind or in some loose A-B testing against D-W fancy transformers, that direct drive is way better. But the transformers are way better than shaking heavy cardboard object to create sound.
Before losing my nerve (or my life) I ran an amp direct, no caps, to the stators with a B+ of 2400 volts and an ESL bias of minus couple thousand to the diaphragm.
No question in my mind or in some loose A-B testing against D-W fancy transformers, that direct drive is way better. But the transformers are way better than shaking heavy cardboard object to create sound.
the problem with "direct drive" tube amps may relate to the output impedance of the tubes then being in some relationship (depending on feedback or lack of) to the impedance (capacitive) of the cells... in other words that impedance "matches" the cells at some frequency. Peter Walker teaches about this relationship which is one of bandwidth vs. apparent sensitivity.
So, it seems that in some cases with DD amps that some sort of EQ is required?
_-_-bear
So, it seems that in some cases with DD amps that some sort of EQ is required?
_-_-bear
Sorry if I post and then claim ignorance.
Since the drivers use hardly any power, my DD amp is essentially driving a big chain of low-inductance power resistors (and a tiny capacitance from my wide-spaced elements). Toronto is blessed with good surplus electronics stores. Nice load, eh.
Even if some EQ is needed, it is likely to be an orderly situation and so easily addressed, as compared to the usual intractable bumps and dips with speakers and crossovers. Also, no big trick actually measuring the voltage the DD amp is putting out, as compared to the measurement challenge of measuring the output of a 1:100 transformer connected to an ESL speaker.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.