Impedance Anyone?

[geeji]

Standard Purifi struggles a bit.

I do have the Purifi 1ET400A (25A) and although the latest Purifi 1ET7040SA may be slightly superior, the original Purifi performs superbly on my Summit X. Highly recommended in either version.

 

[Nuri58]

If I'm not mistaken then the wattage in music is quite limited, hence I'd think that vertuallu no amp would face a big issue. But maybe someone can explain what wattage is nornal for the different frequencies.

 

[ttocs]

If I'm not mistaken then the wattage in music is quite limited, hence I'd think that vertuallu no amp would face a big issue.

I would say that most amps would have difficulty driving low impedance such as the ML stats present. I first started testing the frequency response specifically because I knew there were instruments in the recordings I was listening to but wasn't hearing them when using the tube amps like when I used the solid state Krell. I won't go into a review of why I used one amp or another at the moment as this is just to demonstrate how an amp is affected by frequency and impedance.

Here is a test I did two years ago when comparing a couple different driver tubes for my M-125 mono amps and using a Krell S-1500 for comparison overall. The M-125 amp slumps as soon as the impedance drops below 3Ω, using the 4Ω tap on the amp and ML 13A. The Krell on the other hand hangs on to 15kHz before taking a dive, which is the point at which the impedance falls below 1Ω.

For fun this weekend I plan to compare 5 solid state amps for demonstration purposes. I expect one of them, an old Onkyo stereo receiver from the 1990's, to go into oscillation.

edit: I should've pointed out, for those who might be confused what is being shown, that the plot above is REW full frequency sweeps using a mic at the MLP.

 

[Robert D]

I would say that most amps would have difficulty driving low impedance such as the ML stats present. I first started testing the frequency response specifically because I knew there were instruments in the recordings I was listening to but wasn't hearing them when using the tube amps like when I used the solid state Krell. I won't go into a review of why I used one amp or another at the moment as this is just to demonstrate how an amp is affected by frequency and impedance.

Here is a test I did two years ago when comparing a couple different driver tubes for my M-125 mono amps and using a Krell S-1500 for comparison overall. The M-125 amp slumps as soon as the impedance drops below 3Ω, using the 4Ω tap on the amp and ML 13A. The Krell on the other hand hangs on to 15kHz before taking a dive, which is the point at which the impedance falls below 1Ω.
View attachment 24326

For fun this weekend I plan to compare 5 solid state amps for demonstration purposes. I expect one of them, an old Onkyo stereo receiver from the 1990's, to go into oscillation.

Wow, that graph is awesome and really shows us what happens at high frequencies.

 

[Fidji99]

I can serve with Class D Amp measurements - violet ones, down there 25A, 225W/400W/425W at 8/4/2Ohm respectively. Graphs are offset by 10dB. Measured at 80dB 2.5m distance.

 

[Fidji99]

I would say that most amps would have difficulty driving low impedance such as the ML stats present. I first started testing the frequency response specifically because I knew there were instruments in the recordings I was listening to but wasn't hearing them when using the tube amps like when I used the solid state Krell. I won't go into a review of why I used one amp or another at the moment as this is just to demonstrate how an amp is affected by frequency and impedance.

Here is a test I did two years ago when comparing a couple different driver tubes for my M-125 mono amps and using a Krell S-1500 for comparison overall. The M-125 amp slumps as soon as the impedance drops below 3Ω, using the 4Ω tap on the amp and ML 13A. The Krell on the other hand hangs on to 15kHz before taking a dive, which is the point at which the impedance falls below 1Ω.
View attachment 24326

For fun this weekend I plan to compare 5 solid state amps for demonstration purposes. I expect one of them, an old Onkyo stereo receiver from the 1990's, to go into oscillation.

edit: I should've pointed out, for those who might be confused what is being shown, that the plot above is REW full frequency sweeps using a mic at the MLP.

Interesting - I thought dip between 300-500Hz and 1-2kHz is my acoustics/placement, but I can see similar pattern in your measurements.
Now, it looks like this is actually typical in-room response.

 

[ttocs]

Interesting - I thought dip between 300-500Hz and 1-2kHz is my acoustics/placement, but I can see similar pattern in your measurements.
Now, it looks like this is actually typical in-room response.

Could be room response is similar, dunno. When I did those sweeps the mic was placed in the MLP and was pointed up. This weekend I want to take more of the room out of the equation and put the mic closer to the speaker and aim the mic at the speaker.

 

[Leporello]

I would say that most amps would have difficulty driving low impedance such as the ML stats present. I first started testing the frequency response specifically because I knew there were instruments in the recordings I was listening to but wasn't hearing them when using the tube amps like when I used the solid state Krell. I won't go into a review of why I used one amp or another at the moment as this is just to demonstrate how an amp is affected by frequency and impedance.

Here is a test I did two years ago when comparing a couple different driver tubes for my M-125 mono amps and using a Krell S-1500 for comparison overall. The M-125 amp slumps as soon as the impedance drops below 3Ω, using the 4Ω tap on the amp and ML 13A. The Krell on the other hand hangs on to 15kHz before taking a dive, which is the point at which the impedance falls below 1Ω.
View attachment 24326

For fun this weekend I plan to compare 5 solid state amps for demonstration purposes. I expect one of them, an old Onkyo stereo receiver from the 1990's, to go into oscillation.

edit: I should've pointed out, for those who might be confused what is being shown, that the plot above is REW full frequency sweeps using a mic at the MLP.

I think I've discussed with you before my experience with the M125's. I had a blast building mine, but found them entirely unsuitable for the CLS II's, for the reasons you describe. I had never heard of, let alone tried, the autotransformer doohickey you used, but proceeded instead to acquire a pair of Parasound JC-1's, and never looked back.

 

[ttocs]

I think I've discussed with you before my experience with the M125's. I had a blast building mine, but found them entirely unsuitable for the CLS II's, for the reasons you describe. I had never heard of, let alone tried, the autotransformer doohickey you used, but proceeded instead to acquire a pair of Parasound JC-1's, and never looked back.

Can't go wrong with a pair of JC-1 amps! Fantastic choice!

I too had lots of fun building my M-125 amps. Learned a lot in the process. There is one bit of magic in the soundstage, there is a realism that I haven't witnessed with other amps. But that is really the only thing lacking with my current setup, and it's not a huge difference, but it's a difference none the less.

 

[Leporello]

If I'm not mistaken then the wattage in music is quite limited, hence I'd think that vertuallu no amp would face a big issue. But maybe someone can explain what wattage is nornal for the different frequencies.

It's not about wattage, it's about voltage. Say an amp is putting out 1 watt into 4 ohms, on average. That translates to 2V RMS. You want that 2 volts to hold at any frequency regardless of how the impedance fluctuates, for a flat frequency response.

All other things being equal, you want an amp to behave as closely as possible as an ideal voltage source, hence the doubling of power from 8 to 4 ohms, and again from 4 down to 2. An ideal voltage source would put out infinite wattage into a short circuit, obviously an impossibility. But an amp suitable for a ML 'stat, operating into a short circuit, would do great damage somewhere, if not shut down by some sort of protection.

 

[geeji]

Can't go wrong with a pair of JC-1 amps!

Well, I am glad some people on this thread do not care about prices, but for myself, if I had that amount to spend, I would invest it into a pair of brand new ESL15A. I very much doubt anybody in a blind test could make the difference between the outrageously priced JC1+ and the Purifi 1ET400A I currently use with my old Summit X.
Furthermore, I very much prefer a 6 kg amplifier to a 75kg one, especially when I have to carry them...

 

[Russr]

Well, I am glad some people on this thread do not care about prices, but for myself, if I had that amount to spend, I would invest it into a pair of brand new ESL15A. I very much doubt anybody in a blind test could make the difference between the outrageously priced JC1+ and the Purifi 1ET400A I currently use with my old Summit X.
Furthermore, I very much prefer a 6 kg amplifier to a 75kg one, especially when I have to carry them...

I was with you until the 6 kg bit. While D-Class (in pretty much any configuration) are fine for the low frequencies/woofers, I'm not at all sold on D-Class switch-mode power supplies handling ESL Panel "antics" at high frequencies. To my (old) way of thinking, gimme big ol' transformers and massive capacitors, with so much reserve the MLs could drop to 0.1Ohm @ 20KHz, and I'd still be a happy camper. And THAT's why I won't be getting any 6kg amps for my MLs any time soon!

 

[Nuri58]

It's not about wattage, it's about voltage. Say an amp is putting out 1 watt into 4 ohms, on average. That translates to 2V RMS. You want that 2 volts to hold at any frequency regardless of how the impedance fluctuates, for a flat frequency response.

All other things being equal, you want an amp to behave as closely as possible as an ideal voltage source, hence the doubling of power from 8 to 4 ohms, and again from 4 down to 2. An ideal voltage source would put out infinite wattage into a short circuit, obviously an impossibility. But an amp suitable for a ML 'stat, operating into a short circuit, would do great damage somewhere, if not shut down by some sort of protection.

That sounds logical. But does it not also translates to that the voltage you supply for hig frequency is less than for low frequency? Meaning tjat the.load on the amp is kept relatively down?

 

[Fidji99]

That sounds logical. But does it not also translates to that the voltage you supply for hig frequency is less than for low frequency? Meaning tjat the.load on the amp is kept relatively down?

Blue trace is speaker measurement - so teoretical capability in room for speaker, measured by sweep.
Black trace is real time analyzer of one relatively complex music - complete orchestra, averaged over 1 min.
There is virtually nothing above 15kHz,and quite not a lot happening above 8kHz [which happends to be second harmonic of highest piano key] On top most of the DACs will have some filters and attenuation around 22kHz

So it is important to have amp that is technically capable to run 2ohm and stay stable, in case of ML panels there is added layer of phase twist, but long story short - you can get current limited, more than voltage limited. So look at something with enough Amperes. There some D Class modules with over 40A, that shoud allow to play at high SPL levels, to fill even larger room.

 

[Nuri58]

I see. I have deceided to go for electromotion X (1.8 ohm at 2o kHz) and hearig is down to between 12 and 15 khz anyway...

 

[Leporello]

That sounds logical. But does it not also translates to that the voltage you supply for hig frequency is less than for low frequency? Meaning tjat the.load on the amp is kept relatively down?

If you're talking about the power spectrum of typical music signals, true. But a drop in impedance at higher frequencies that results in attenuation will still be perceived as a loss of highs.

 

[Robert D]

If you're talking about the power spectrum of typical music signals, true. But a drop in impedance at higher frequencies that results in attenuation will still be perceived as a loss of highs.

Reminds me of how when a panel gets old, it's the highs you lose. I wonder, does impedance of the panel go up when it ages?

 

[Nuri58]

If you're talking about the power spectrum of typical music signals, true. But a drop in impedance at higher frequencies that results in attenuation will still be perceived as a loss of highs.

That's what i suspected. And of course it will be persived as a loss. I'd think this will be compensated by room correction sw. This might explain why the sound level with/without Direc is more than 10db different. The curve I get is quite flat still rolling off at the high end but not dramatically.

 

[Robert D]

That's what i suspected. And of course it will be persived as a loss. I'd think this will be compensated by room correction sw. This might explain why the sound level with/without Direc is more than 10db different. The curve I get is quite flat still rolling off at the high end but not dramatically.

Yes, I used Audyssey on my Prodigy speakers from the start and didnt really realize anything was wrong with the panels. If I turned off Audyssey, the panels sounded extremely dull and lifeless. Looking at the receiver, the gain on the speakers was maxed out too, at like +15 db or whatever it is that Marantz maxes out at. So Audyssey actually made them sound good believe it or not, but it was requiring my amp to run hard. I think too that the bass on the speakers was way down to compensate. Once I got the new panels, the speaker gain now is only + 1.5 db on both.

 

[Nuri58]

Yes, I used Audyssey on my Prodigy speakers from the start and didnt really realize anything was wrong with the panels. If I turned off Audyssey, the panels sounded extremely dull and lifeless. Looking at the receiver, the gain on the speakers was maxed out too, at like +15 db or whatever it is that Marantz maxes out at. So Audyssey actually made them sound good believe it or not, but it was requiring my amp to run hard. I think too that the bass on the speakers was way down to compensate. Once I got the new panels, the speaker gain now is only + 1.5 db on both.

A solution cheaper than new panels though well I replaced my Cinema i with the Focus - definetly not cheaper, but the repair would apparently add up to the same as a new C model. The sound in the focus is a few levels up and since I'll add the ESL X as replacement of my 18y old Aion i, it ended up a bit pricy but defenitly a real upgrade that make it worth it