|
|
Post by boomzilla on Jan 14, 2019 19:40:06 GMT -5
...A quick look through some of the on-line reviews reveals a number of different integrated and power amps were used and nobody had problems driving them...I really think something must be wrong with the amp or the speaker. I disagree. Yes, many (most?) amps will drive electrostatic speakers without issue, but that's scant comfort to someone who has an incompatible amp. After all, Martin-Logan doesn't say "our speakers are compatible with most amps" but rather "our speakers are compatible with all amps." This is patently untrue. And no, neither the amplifiers nor the speakers are defective. It's just that some amplifiers are more sensitive to low impedance loads than others. |
|
|
|
Post by donh50 on Jan 14, 2019 19:56:31 GMT -5
...A quick look through some of the on-line reviews reveals a number of different integrated and power amps were used and nobody had problems driving them...I really think something must be wrong with the amp or the speaker. I disagree. Yes, many (most?) amps will drive electrostatic speakers without issue, but that's scant comfort to someone who has an incompatible amp. After all, Martin-Logan doesn't say "our speakers are compatible with most amps" but rather "our speakers are compatible with all amps." This is patently untrue. And no, neither the amplifiers nor the speakers are defective. It's just that some amplifiers are more sensitive to low impedance loads than others. That may be but it is pretty disappointing if Emotiva's DR-3 cannot drive an 11A... Their reference amp can't handle what many lesser amps can? I only skimmed a dozen or so; what review(s) did you find wherein they could not drive the speakers? My XPA-3 and XPA-2 had no problem driving Maggies that dipped to ~2 ohms at HF but are not so reactive. I have friends driving bigger ML ESLs with various XPA (and other) amps without problem (but older gen1 and gen2 XPAs, no DR designs). I just can't believe that amp can't drive those speakers but you may be right. I'd like to see what Emotiva says, and wonder if other folk have that problem. Yes, ESLs are a nasty HF load, but overall not that bad to drive especially with the active woofer and a small'ish panel. If the DR-3 cannot it is less stable than I would have expected. I would double/triple-check all the cables to and from the amp and preamp. A stray speaker wire strand would do it, and my XPA-5 was shutting down because one of the input cables was loose after one of my boys got behind the console to plumb their game box to the system. |
|
|
|
Post by exnihilo1031 on Jan 14, 2019 20:29:01 GMT -5
I used the word "impedance" intentionally and the power at 20 kHz and above is very low for any reasonable source. The problem I see is that the ML 11A is in the middle of ML's speaker line and many people are driving them with much less than a DR-3 with no problem. Sensitivity is rated at 91 dB/2.83 Vrms/m so not horrible (better than my current conventional speakers and much better than my old Magnepans). A quick look through some of the on-line reviews reveals a number of different integrated and power amps were used and nobody had problems driving them. If amps were routinely shutting down there would be all sorts of posts on the various audio fora and ML would be dealing with it. I really think something must be wrong with the amp or the speaker. I disagree. Yes, many (most?) amps will drive electrostatic speakers without issue, but that's scant comfort to someone who has an incompatible amp. After all, Martin-Logan doesn't say "our speakers are compatible with most amps" but rather "our speakers are compatible with all amps." This is patently untrue. And no, neither the amplifiers nor the speakers are defective. It's just that some amplifiers are more sensitive to low impedance loads than others. That may be but it is pretty disappointing if Emotiva's DR-3 cannot drive an 11A... Their reference amp can't handle what many lesser amps can? I only skimmed a dozen or so; what review(s) did you find wherein they could not drive the speakers? My XPA-3 and XPA-2 had no problem driving Maggies that dipped to ~2 ohms at HF but are not so reactive. I have friends driving bigger ML ESLs with various XPA (and other) amps without problem (but older gen1 and gen2 XPAs, no DR designs). I just can't believe that amp can't drive those speakers but you may be right. I'd like to see what Emotiva says, and wonder if other folk have that problem. Yes, ESLs are a nasty HF load, but overall not that bad to drive especially with the active woofer and a small'ish panel. If the DR-3 cannot it is less stable than I would have expected. I would double/triple-check all the cables to and from the amp and preamp. A stray speaker wire strand would do it, and my XPA-5 was shutting down because one of the input cables was loose after one of my boys got behind the console to plumb their game box to the system. One of the tests I originally ran was to run speaker wires directly to the speakers (to avoid possibility of unseen wiring issues), using a different wire for each speaker, testing each speaker individually against the amp (In series, Left channel alone, Center channel alone, Right channel alone), on 2 different processors (old Pioneer Eliter VSX-21TXH and RMC-1). Each speaker individually triggered the same shutdown on the DR-3, so I am leaning toward the DR-3 having the issue itself. I haven't gotten around to testing with thinner gauge wire or a different circuit yet. However, Martin Logan did reach out to Emotiva for me to ask "Do you know how much signal from the binding posts to earth ground will cause the protection to trigger? Is there a time period below which the current may be ignored?" after Emotiva asked me to check if there was a DC current spike during speaker wakeup. Still waiting to hear back from Emotiva. |
|
|
|
Post by donh50 on Jan 14, 2019 20:54:12 GMT -5
I used the word "impedance" intentionally and the power at 20 kHz and above is very low for any reasonable source. The problem I see is that the ML 11A is in the middle of ML's speaker line and many people are driving them with much less than a DR-3 with no problem. Sensitivity is rated at 91 dB/2.83 Vrms/m so not horrible (better than my current conventional speakers and much better than my old Magnepans). A quick look through some of the on-line reviews reveals a number of different integrated and power amps were used and nobody had problems driving them. If amps were routinely shutting down there would be all sorts of posts on the various audio fora and ML would be dealing with it. I really think something must be wrong with the amp or the speaker. That may be but it is pretty disappointing if Emotiva's DR-3 cannot drive an 11A... Their reference amp can't handle what many lesser amps can? I only skimmed a dozen or so; what review(s) did you find wherein they could not drive the speakers? My XPA-3 and XPA-2 had no problem driving Maggies that dipped to ~2 ohms at HF but are not so reactive. I have friends driving bigger ML ESLs with various XPA (and other) amps without problem (but older gen1 and gen2 XPAs, no DR designs). I just can't believe that amp can't drive those speakers but you may be right. I'd like to see what Emotiva says, and wonder if other folk have that problem. Yes, ESLs are a nasty HF load, but overall not that bad to drive especially with the active woofer and a small'ish panel. If the DR-3 cannot it is less stable than I would have expected. I would double/triple-check all the cables to and from the amp and preamp. A stray speaker wire strand would do it, and my XPA-5 was shutting down because one of the input cables was loose after one of my boys got behind the console to plumb their game box to the system. One of the tests I originally ran was to run speaker wires directly to the speakers (to avoid possibility of unseen wiring issues), using a different wire for each speaker, testing each speaker individually against the amp (In series, Left channel alone, Center channel alone, Right channel alone), on 2 different processors (old Pioneer Eliter VSX-21TXH and RMC-1). Each speaker individually triggered the same shutdown on the DR-3, so I am leaning toward the DR-3 having the issue itself. I haven't gotten around to testing with thinner gauge wire or a different circuit yet. However, Martin Logan did reach out to Emotiva for me to ask "Do you know how much signal from the binding posts to earth ground will cause the protection to trigger? Is there a time period below which the current may be ignored?" after Emotiva asked me to check if there was a DC current spike during speaker wakeup. Still waiting to hear back from Emotiva. That sounds interesting... I assume the the DR-3 I assume uses differential (balanced) output modules, creating essentially a bridged amp, and that will not tolerate either speaker terminal being grounded. Because the ESLs require a power supply for the panels (and the woofer amps), it could well be that the (-) terminal is connected to ground, perhaps via a small resistor. That would make it incompatible with most any bridged amplifier configuration. If true, then boomzilla is correct, though perhaps not for the reasons I and others were thinking. Any similar configuration using a common ground could be problematic, including some in-wall or ceiling installations where the installer used a common ground wire among the speakers. It may also be that ML is grounding, or nearly, the input until the power supply is up and running for the 'stat panels. Keep us informed, this one is intriguing (yes, not the word you would use right now, I am sure). |
|
andyo
Minor Hero
Posts: 21 
|
Post by andyo on Jan 14, 2019 23:38:09 GMT -5
Directly from Martin Logans website. IMPEDANCE: 4 Ohms (0.6 at 20kHz), compatible with any amplifier Martin-Logan prevaricates. Their nominal and minimum impedance numbers are correct, but what they don't say (and it's equally important) is that the impedance typically CONTINUES dropping above 20 KHz, and that at turn on, the overall impedance is lower until the electrostatic stator becomes fully charged. And now for the misdirection (to put it politely) - The Martin-Logans are definitely NOT compatible "with any amplifier." I've had an Emotiva XPA-1L blow out on startup when hooked up to ML speakers. The OP of this thread is having continued problems with his amps tripping off due to the ML speakers. These are not unique or even unusual occurrences with electrostatic speakers (of any brand). The amplifiers most compatible with electrostatic speakers are: Those with output transformers in series with the speakers' impedance (the McIntosh "autoformer-coupled" amps are in this category) Those with delayed turn-on, allowing the electrostatic stators to stabilize while the amp warms up (tube-rectified amps are in this category) Those specifically designed to be stable into low-impedance loads (Krell and most "pro" amplifiers are in this category) Note that in the Best-Buy/Magnolia kiosks nation-wide, their Martin-Logan speakers on display are ALWAYS paired with McIntosh amplifiers. This isn't to say that many (maybe even most) amplifiers won't work just fine with ML speakers - but it IS to say (emphatically) that ML speakers are not compatible with just "any" amplifier. This isn't an opinion - it's a fact (despite what Martin-Logan claims). Boomzilla I have ML Theos and have used them with XPA-1Ls for over 2 years. I have never had any issue. The amps become slightly warmer than when used with "regular" speakers. I even tried ML speakers with a cheap Pioneer HT receiver for some time and the receiver had no problem driving them. There isn't much signal energy above 16kHz with music anyway.
|
|
|
|
Post by donh50 on Jan 15, 2019 15:08:27 GMT -5
I really doubt it is the low impedance at HF, especially for a turn-on problem... I am very curious about the potential for a ground fault causing a bridged amp to fail. But I have not heard of that, it would rule out a lot of amplifiers, and there is this note on the ML website:
I wonder if there is some protection circuit in the ML that the DR-3 is having problems with, or the amp or speaker may be bad.
If only connected to one speaker at a time, rotating through them, does it still fail?
|
|
|
|
Post by exnihilo1031 on Jan 15, 2019 16:48:22 GMT -5
I really doubt it is the low impedance at HF, especially for a turn-on problem... I am very curious about the potential for a ground fault causing a bridged amp to fail. But I have not heard of that, it would rule out a lot of amplifiers, and there is this note on the ML website: I wonder if there is some protection circuit in the ML that the DR-3 is having problems with, or the amp or speaker may be bad. If only connected to one speaker at a time, rotating through them, does it still fail? Yah I ran that test. It failed during the transition from standby on each speaker, on each channel, one by one. |
|
|
|
Post by leonski on Jan 15, 2019 19:17:14 GMT -5
doh50, It is interesting to compare magnepan with nearly any 'Stat. Maggies are very low reactance and don't have much in the line of wacky impedance swings. And yes, they are fairly low in sensitivity but NOT a bad load for any reasonable partnering amp. They would appear to like current..... While stats CAN have very crazy swings in impedance, all the way to near-short, which regardless of 'demand' will still put a Real crimp on Current. And at the same time as impedance is plummeting, you'll have large dislocations between Current and Voltage peaks. Something not all amps are comfortable driving. The main take-away from my viewpoint is that if you have 2 amps which both are say.....200watts into 8 ohms and 'near double' into 4 ohms, that says nearly NOTHING about the ability to drive highly reactive loads. I think it was the Krell KSA-100 which could do the deed with 'stats and not break a sweat. No standards exist which are actually USED by the amplifier industry. They LIKE the resistive test since even less than capable amps can look good doing it. That is not to say that Realistic Speaker Dummy Loads don't exist. They do. Just not as a 'standard' with which to compare across different amp manufacturers. I've had this chat with the folks emo without getting any traction. A company called Audio Graph has a system for stressing an amp with both inductive and capacitive reactive loads and taking actual measures at down to maybe 2 ohms. This data CAN be compared across different amps. Have alook at the link if you are curious. www.audiograph.seIs there something wrong with the OPs setup or gear? That's a good question and I wish I knew. I may look into the Specific ML speakers talked about in the first post... |
|
|
|
Post by donh50 on Jan 17, 2019 19:05:43 GMT -5
doh50, It is interesting to compare magnepan with nearly any 'Stat. Maggies are very low reactance and don't have much in the line of wacky impedance swings. And yes, they are fairly low in sensitivity but NOT a bad load for any reasonable partnering amp. They would appear to like current..... While stats CAN have very crazy swings in impedance, all the way to near-short, which regardless of 'demand' will still put a Real crimp on Current. And at the same time as impedance is plummeting, you'll have large dislocations between Current and Voltage peaks. Something not all amps are comfortable driving. The main take-away from my viewpoint is that if you have 2 amps which both are say.....200watts into 8 ohms and 'near double' into 4 ohms, that says nearly NOTHING about the ability to drive highly reactive loads. I think it was the Krell KSA-100 which could do the deed with 'stats and not break a sweat. No standards exist which are actually USED by the amplifier industry. They LIKE the resistive test since even less than capable amps can look good doing it. That is not to say that Realistic Speaker Dummy Loads don't exist. They do. Just not as a 'standard' with which to compare across different amp manufacturers. I've had this chat with the folks emo without getting any traction. A company called Audio Graph has a system for stressing an amp with both inductive and capacitive reactive loads and taking actual measures at down to maybe 2 ohms. This data CAN be compared across different amps. Have alook at the link if you are curious. www.audiograph.seIs there something wrong with the OPs setup or gear? That's a good question and I wish I knew. I may look into the Specific ML speakers talked about in the first post... I realize I am going against the majority here but remember this is happening at turn-on. I just can't see it being a load impedance issue at turn-on with no signal applied. I am familiar with reactive loads and the difference in impedance characteristics between ESLs and Maggies. The Power Cube has been around a long time but never caught on. Probably a combination of too difficult to explain to the masses and too many amplifiers would fare badly. I would throw in with the crowd if it was shutting down whilst playing but that is not the case here. |
|
|
|
Post by leonski on Jan 17, 2019 20:19:53 GMT -5
Totally agree about the power cube. Imagine some Mega-Buck amp doing worse than an inexpensive one? It just never got any traction. too bad, IMO. I was just reading another thread which gave me kind of an idea. The transient 'thump' produced when turning some gear on or off? Somehow, the self-protect circuit of the amp seems to be triggered.....  Also, the OP tested one speaker at a time and apparently different amp channels, all to no help. Heck, the 'Problem with DR3' aspect may actually be the cause of this. And SOME speaker wires are expecially capacitive which may make matters worse.....trying another speaker wire, especially that massively cheap stuff from the home store, couldn't hurt.... |
|
|
|
Post by exnihilo1031 on Jan 18, 2019 15:01:48 GMT -5
No solution yet, but just wanted to update everyone.
I spoke with Lonnie at Emotiva, and they are going to do some lab testing with the DR-3 and the particular Martin Logan's I have (A11 and C34A).
Hoping to hear some good news in the next couple of weeks.
|
|
|
|
Post by leonski on Jan 18, 2019 21:02:43 GMT -5
That's a neat idea.
TEST it! That should pretty much settle it.
If EMO can't duplicate the problem using same speakers and amp than it's a long shot that speaker wire is involved.
Some speaker wire / amp / speaker combinations can lead to the kind of problems the OP is having.......
Another solution? Without adding speaker wire? Get a 1 ohm / 5 watt resistor and plug that in...series with one leg of the speaker wire.
If it gets past the turn-on problem, do NOT turn it up too loud. The resistor will be sinking a LOT of the power at higher frequencies where
the impedance is lowest.
|
|
|
|
Post by donh50 on Jan 19, 2019 10:06:39 GMT -5
Totally agree about the power cube. Imagine some Mega-Buck amp doing worse than an inexpensive one? It just never got any traction. too bad, IMO. I was just reading another thread which gave me kind of an idea. The transient 'thump' produced when turning some gear on or off? Somehow, the self-protect circuit of the amp seems to be triggered..... Also, the OP tested one speaker at a time and apparently different amp channels, all to no help. Heck, the 'Problem with DR3' aspect may actually be the cause of this. And SOME speaker wires are expecially capacitive which may make matters worse.....trying another speaker wire, especially that massively cheap stuff from the home store, couldn't hurt.... That seems like a reasonable possibility... I have had problems when input coupling caps were bad and leaked current from a bad output (or to a bad input). And once or twice when the mod craze was big and some 10 uF input caps were replaced with 1000 uF -- the extra charging current made a much bigger "thump" and tripped the protection circuit. In this case since it is apparently all channels it may be a fundamental mismatch between amp and speakers, hopefully readily solved. I know some preamps have a fairly nasty turn-on transient. |
|
|
|
Post by boomzilla on Jan 19, 2019 10:39:46 GMT -5
OK - One MAJOR factoid here - Magneplanar speakers are NOT electrostatic speakers. The Maggie impedance may drop briefly to about 2 ohms on some models, but it never drops below that. Electrostatics, on the other hand, can appear as DEAD SHORTS at some frequencies and under some conditions. The two are NOT the same. Just because an amp can drive Maggies without issue does NOT mean that it can drive 'stats.
Second - The ability to drive electrostatic speakers (or not) has absolutely NO relationship to how good the amplifier is (or isn't). A welding machine could theoretically drive electrostatic speakers, but that doesn't make it a good high-fidelity amplifier. A $300 AVR might be able to drive electrostatic speakers, but that wouldn't make it a good high-fidelity amplifier either. Most any amplifier with fully differential circuitry (or any "bridged" stereo amplifier) will often (almost always?) have issues with electrostatic speakers. But some can do it fine IF they're specifically designed to handle exceptionally low impedances. The Emotiva amps are specified down to the two-ohm range, if I remember correctly, but that's just not good enough for some electrostatic speakers. Despite their inability to drive "virtually shorted" loads, the Emotivas remain very good amplifiers. But they're only that if used within their limitations.
Third - I still contend that the simplest solution to amplifier-speaker compatibility (with electrostatic speakers) is to introduce a very small series resistance between the amp and the speaker. This "buffers" the impedance when the speaker starts up and is charging, and also pads the treble at 20 KHz and above (where there's no music anyway that you can hear). With even a fraction of an ohm series resistance, almost any amplifier should be able to drive electrostatic speakers without issue. 15 feet of 16 ga. copper wire would add 0.0602 Ω to the system, and even this seems sufficient to ensure amplifier stability in my experience.
|
|
|
|
Post by leonski on Jan 19, 2019 14:06:09 GMT -5
OK - One MAJOR factoid here - Magneplanar speakers are NOT electrostatic speakers. The Maggie impedance may drop briefly to about 2 ohms on some models, but it never drops below that. Electrostatics, on the other hand, can appear as DEAD SHORTS at some frequencies and under some conditions. The two are NOT the same. Just because an amp can drive Maggies without issue does NOT mean that it can drive 'stats. Second - The ability to drive electrostatic speakers (or not) has absolutely NO relationship to how good the amplifier is (or isn't). A welding machine could theoretically drive electrostatic speakers, but that doesn't make it a good high-fidelity amplifier. A $300 AVR might be able to drive electrostatic speakers, but that wouldn't make it a good high-fidelity amplifier either. Most any amplifier with fully differential circuitry (or any "bridged" stereo amplifier) will often (almost always?) have issues with electrostatic speakers. But some can do it fine IF they're specifically designed to handle exceptionally low impedances. The Emotiva amps are specified down to the two-ohm range, if I remember correctly, but that's just not good enough for some electrostatic speakers. Despite their inability to drive "virtually shorted" loads, the Emotivas remain very good amplifiers. But they're only that if used within their limitations. Third - I still contend that the simplest solution to amplifier-speaker compatibility (with electrostatic speakers) is to introduce a very small series resistance between the amp and the speaker. This "buffers" the impedance when the speaker starts up and is charging, and also pads the treble at 20 KHz and above (where there's no music anyway that you can hear). With even a fraction of an ohm series resistance, almost any amplifier should be able to drive electrostatic speakers without issue. 15 feet of 16 ga. copper wire would add 0.0602 Ω to the system, and even this seems sufficient to ensure amplifier stability in my experience. See my post @ 4:17, above. I noted diffs between Maggies and 'Stats, while not specifically saying 'they are vastly different in operating principle'......which they of course ARE. Older Maggies with the original Ribbon DO have that low impedance you speak to but the '2nd model' of ribbon dealt with that and it's not been an issue for a long time. As for some additional series resistance? maggies are Issued with a pair of 1 ohm / 5 watt (10 watt?) power resistors (non-inductive) to 'tame' what some perceive as a hot high end. That's what all the EXTRA input terminals are used for, to insert in bridge fashion on the tweeter end. I'm not certain about the new X.7 models which ALL have a series crossover but anything prior should have the extra terminals for resistor insertion. Not that it matters, but if you think you need such a resistor, try SETUP first. Maggies and presumably 'Stats can be sensitive to placement. Spacing from wall and toe are just the beginning. It should be noted that as far as the OP, some additional resistance MIGHT help, especially if the original speaker wires are very capacitive. Some speaker wires simply do NOT like 'Stats. Without replacing / testing, maybe the addition of a 1 ohm resistor would be a work around FOR TEST only. Make it a 5 watt non-inductive.....all you need is MORE reactance! |
|
|
|
Post by donh50 on Jan 20, 2019 13:52:37 GMT -5
OK - One MAJOR factoid here - Magneplanar speakers are NOT electrostatic speakers. The Maggie impedance may drop briefly to about 2 ohms on some models, but it never drops below that. Electrostatics, on the other hand, can appear as DEAD SHORTS at some frequencies and under some conditions. The two are NOT the same. Just because an amp can drive Maggies without issue does NOT mean that it can drive 'stats. Second - The ability to drive electrostatic speakers (or not) has absolutely NO relationship to how good the amplifier is (or isn't). A welding machine could theoretically drive electrostatic speakers, but that doesn't make it a good high-fidelity amplifier. A $300 AVR might be able to drive electrostatic speakers, but that wouldn't make it a good high-fidelity amplifier either. Most any amplifier with fully differential circuitry (or any "bridged" stereo amplifier) will often (almost always?) have issues with electrostatic speakers. But some can do it fine IF they're specifically designed to handle exceptionally low impedances. The Emotiva amps are specified down to the two-ohm range, if I remember correctly, but that's just not good enough for some electrostatic speakers. Despite their inability to drive "virtually shorted" loads, the Emotivas remain very good amplifiers. But they're only that if used within their limitations. Third - I still contend that the simplest solution to amplifier-speaker compatibility (with electrostatic speakers) is to introduce a very small series resistance between the amp and the speaker. This "buffers" the impedance when the speaker starts up and is charging, and also pads the treble at 20 KHz and above (where there's no music anyway that you can hear). With even a fraction of an ohm series resistance, almost any amplifier should be able to drive electrostatic speakers without issue. 15 feet of 16 ga. copper wire would add 0.0602 Ω to the system, and even this seems sufficient to ensure amplifier stability in my experience. I do not think anyone was confusing the two types of speakers; the debate was whether low impedance was causing the DR-3 to shut down. Pretty sure we noted reactance/phase angle is worse with ESL's. They (ESLs) do not normally drop low until very high in frequency where power is very low (unless the amplifier goes unstable and oscillates). And I am not sure about below two ohms, but the ribbon tweeter on 3.x and 20.x dips below 3 ohms for much of its range, and if you use Magnepan's center channel scheme or add their bass panel to the center it also dips very low (3 ohms or less) for about an octave or so around the crossover. Yes phase angle is low. I never debated the "goodness" of amplifiers best I recall but didn't go back to look at my posts. Most designers strive for "unconditionally stable" and that includes dealing with reactive loads and momentary (or even long-term) shorts. At least for my designs and the designers with whom I have spoken over the years. I find it surprising that a company's "reference" amplifier line can't handle one of the bigger audiophile lines around when so many other amplifies are fine with it. I just don't think shutting down at turn-on is related to the HF load of ESLs. Of course I could be wrong, but the only time I have seen that is when the amplifier was unstable into the load, which is possible here but there are usually other signs of that. I still think there is a funky interaction at turn-on or a defect someplace. Oh, Quad ESLs did (and may still do) short the amplifier inputs to protect the panels when overdriven. Pretty harsh and more than one amp died from that treatment... Adding the resistor might resolve the shutdown but will also increase the distortion in the panels. They exhibit modes that are reduced with lower driving impedance. That said, IME modes are worse in planar dynamics like Maggies than ESLs or ribbons. Whatever, don't think we'll ever agree on this one. I am most interested in the resolution and then we'll know. It could be the amp is oscillating and shutting down as you argue; that would be vexing but may be. - Don |
|
|
|
Post by leonski on Jan 20, 2019 15:47:28 GMT -5
don, i don't think either of us are overly concerned with Agree or Disagree. You have great input and a VERY valid point of view. Personally? I'd LOVE to know how this shakes out. The problem might yet be something we haven't mentioned, being the INPUT electronics on the 'Stats. Couple 'a bad caps would be all it would take. And Of course, SOME (few) speaker wires are notorious for incompatibility with 'Stats. Too much capacitance added to an already capacitve load? No Bueno. But at the level of the EMO purchaser? You'd have to be simply unlucky to happen on that combo.... The resistor in the amps output? Might tend to tip it from voltage source to current source? Pass did such experiments with full range drivers and Carver in one of the sunfire amps had a switchable or maybe a 2nd output terminal with a fixed resistance in series. Ideal voltage source delivers same voltage wtihout regard to load while ideal current souce delievers same current, also without regard to load. I doubt either exists 'in pure form' except as a lab demo. www.stereophile.com/content/magnepan-magneplanar-mg36r-loudspeaker-measurementsThis is a reasonable measured data from the 3.6r Magnepan. Doesn't drop below 3 or so ohms and doesn't really have a reactance problem.......Easy load, provided the amp can provide the current. Tube or SS seems equally 'good'... |
|
|
|
Post by donh50 on Jan 20, 2019 16:47:54 GMT -5
don, i don't think either of us are overly concerned with Agree or Disagree. You have great input and a VERY valid point of view. Personally? I'd LOVE to know how this shakes out. The problem might yet be something we haven't mentioned, being the INPUT electronics on the 'Stats. Couple 'a bad caps would be all it would take. And Of course, SOME (few) speaker wires are notorious for incompatibility with 'Stats. Too much capacitance added to an already capacitve load? No Bueno. But at the level of the EMO purchaser? You'd have to be simply unlucky to happen on that combo.... The resistor in the amps output? Might tend to tip it from voltage source to current source? Pass did such experiments with full range drivers and Carver in one of the sunfire amps had a switchable or maybe a 2nd output terminal with a fixed resistance in series. Ideal voltage source delivers same voltage wtihout regard to load while ideal current souce delievers same current, also without regard to load. I doubt either exists 'in pure form' except as a lab demo. www.stereophile.com/content/magnepan-magneplanar-mg36r-loudspeaker-measurementsThis is a reasonable measured data from the 3.6r Magnepan. Doesn't drop below 3 or so ohms and doesn't really have a reactance problem.......Easy load, provided the amp can provide the current. Tube or SS seems equally 'good'... Thanks Leon, no worries, we're on the same page (I think). And I am with you 100% in wanting to find out root cause of this one, don't really care if I am right or wrong, just want to keep the options alive. I have certainly seen some very highly-regarded amplifiers croak on an ESL (not just the Quads). I've been a planar guy for ages and often enough get asked what to use, often by folk who know I have Emotiva amps, so I want to be able to suggest them with confidence (or not). Adding a series resistor probably isn't enough to turn it into a current source but may limit inrush current enough to fix the problem. (Carver did it to emulate a tube amp's output, as well as debiasing the output stage to add more 2HD IIRC.) Or decouple the load enough to make the amp stable. I consider that a band-aide, however, and would still like to know what exactly is the issue. Is it an unstable (or marginally stable) amp driving the ESLs, or something about the amp/speaker interaction causing a low-Z condition at turn-on that the DR-3 senses as a fault? I was at the CES when the Cobra Cable guy convinced the rep to put them on the new class D amp (Infinity, I think? -- it was not the HCA, I think it was just called the DSP? -- extremely new at the time). The amps went "ziiiiiiip!" and they were gone. Most audio amplifiers are designed to be as close as possible to pure voltage sources, at least the SS amps. There have been a few notable exceptions, including Carver's "T" amp switch and the Threshold Stasis (pretty sure that is the one), but by and large they strive to be an ideal voltage source. My old MG-IIIa's dropped to 2.7 ohms (looked up my old notes) and the CC3/DWM about the same. Magnepan has played with the tweeter circuit over the years so some are lower, some higher, and they've almost always provided that power resistor to attenuate the tweeter a bit. Mine came with two pair, 0.5 and 1.0 ohm. Note Maggie 3.x/20.x/30.x, and most ribbons, look inductive at HF whilst an ESL panel is primarily capacitive (after the transformer, natch). Below that they are essentially resistive except around the crossover. Yah, easy, just a bit low and pretty insensitive so need power. The old Apogee Scintillas were the amp-killers of their time; almost purely resistive but 1 ohms nominal impedance made life hard on a a lot of amps. They got rid of the transformer from the first model (4-ohm looking into the transformer) in favor of direct drive. Here's the Infinity amp, I think: davidsaudio.com/Infinity_DSP.pdf -- doesn't look quite like I remember, but nothing does anymore.  |
|
|
|
Post by leonski on Jan 20, 2019 17:30:34 GMT -5
Similar thing happened to me, back a long time ago.
Friends were selling some MG-1 which I'd NEVER heard or heard OF. What a revelation and I bought them without bickering.
These same guys were very DIY oriented and had some 'Stats and a home-brew amp. One Ultrasonic chirp later.......OOOPS!
Sounds like ziiiiiiip!" to me. My MG-1 lasted 2 decades or more AND made a trip to White Bear Lake for a rebuild.....
I'm not certain that the MG-IIIa didn't use the ORIGINAL ribbon which was a lower impedance than the one currently in use.
Funny you should bring up the Scintilla. You needed an amp labeled 'Lincoln Electric' or maybe 'Miller' to be sure.
The pair of resistors you have? Individually, or In series at 1 1/2 ohms or in Parallel at 0.33 ohms.
I love the infinity amps ADVERT. They totally skip over, after the mere mention, of the phase problems and power limits brought
on by the output Zobel. and STILL insisting on calling it a 'Digital' amp, which has spawned a generation ++ of people STILL calling
it the wrong thing. I'd love to see phase / frequency data for this amp....as well as it's REAL long term power output.
|
|
KeithL
Administrator 
Posts: 9,919
|
Post by KeithL on Jan 21, 2019 9:54:48 GMT -5
Just to clarify one detail that many people seem to find confusing.
A pure resistance remains the same at all frequencies (so, for example, a "4 Ohm speaker" that doesn't exhibit any reactance will be 4 Ohms at 20 Hz, 4 Ohms at 1 kHz, and 4 Ohms at 20 kHz). A load that exhibits capacitive reactance will have an impedance that gets lower as the frequency goes higher; a load that exhibits inductive reactance will do the opposite. Virtually all speakers are somewhat reactive, and most have an impedance that zigs and zags up and down because they actually have a rather complicated combination of inductances and capacitances. However, most "well behaved" speakers are mostly resistive - and most 4 Ohm speakers never fall below about 2.5 Ohms anywhere in the audible frequency range.
Most amplifiers are designed to work best with a purely resistive load. Most amplifiers also work just fine with a well-behaved load (see above).
What we're talking about here is how tolerant a particular amplifier with difficult load.
Asking an amplifier to drive a load that drops to 0.5 Ohms at 20 kHz is about equivalent to towing a small house using a consumer pickup truck.
Martin-Logan prevaricates. Their nominal and minimum impedance numbers are correct, but what they don't say (and it's equally important) is that the impedance typically CONTINUES dropping above 20 KHz, and that at turn on, the overall impedance is lower until the electrostatic stator becomes fully charged. And now for the misdirection (to put it politely) - The Martin-Logans are definitely NOT compatible "with any amplifier." I've had an Emotiva XPA-1L blow out on startup when hooked up to ML speakers. The OP of this thread is having continued problems with his amps tripping off due to the ML speakers. These are not unique or even unusual occurrences with electrostatic speakers (of any brand). The amplifiers most compatible with electrostatic speakers are: Those with output transformers in series with the speakers' impedance (the McIntosh "autoformer-coupled" amps are in this category) Those with delayed turn-on, allowing the electrostatic stators to stabilize while the amp warms up (tube-rectified amps are in this category) Those specifically designed to be stable into low-impedance loads (Krell and most "pro" amplifiers are in this category) Note that in the Best-Buy/Magnolia kiosks nation-wide, their Martin-Logan speakers on display are ALWAYS paired with McIntosh amplifiers. This isn't to say that many (maybe even most) amplifiers won't work just fine with ML speakers - but it IS to say (emphatically) that ML speakers are not compatible with just "any" amplifier. This isn't an opinion - it's a fact (despite what Martin-Logan claims). Boomzilla It's actually WORSE than Boom indicates. Not only is the impedance at higher frequencies nearly a dead-short, but the REACTANCE is a dilly, too. www.stereophile.com/content/sound-lab-1-electrostatic-loudspeaker-measurementsWhile this is ONE electrostat, these measures follow the typical pattern. The 75 degrees at the highest frequencies is something that FEW amps will work well with and virtually NO HT receivers. Just read the first paragraph of the review. Than have a look at the graph of impedance and phase data. Where the 2 lines cross, if at a low impedance AND high phase angle, spells TROUBLE. Get an old Krell and be done with it. Some persons dont' agree with the Mc house sound. Another POSSIBLE solution would be to use an impedance matching transformer between amp and speaker. I can't find the link right away, but this article might be of some help. www.electronics-tutorials.ws/transformer/audio-transformer.html |
|
