XPA-1 and Magnepan CCR Center

[wrk24wheel]

Does anyone have this setup?  I have two XPA-1 Gen1 running my 3.7s for years with no problems.  Just got another XPA-1 Gen1 and a CCR.  Every time I try to run Dirac Live the center channel XPA-1 goes into protection 1/2 way through the setup of the channel.  I see half the wave on the screen and then a click and the amp goes off and starts blinking red.  Watching the front LED's it never gets past the middle before it shuts down.  I tried swapping everything from amps and all cables and it all follows the center channel.  If I use a different amp other than the Emotiva, it works just fine.  All 5 amps are on their own dedicated 20amp circuits so I know power is not a problem.  I am just curious if these amps are just running out of steam from being too old or if the speaker is just dropping to too low of an ohm load for this amp.  The weird thing is that they all drive the 3.7's with no issues, should be the same ohm load.  The only thing I haven't been able to switch out is the XLR cables as they run through the walls and are only long enough to reach the amps behind each speaker.  I don't think that could cause the problem as they play fine at low volumes.

[ttocs]

Maybe it's possible that the volume level during the sweep is too high for the CCR Impedance at that point in the sweep. In other words, yes, the amp runs out of steam.

Try using a lower level setting in Dirac Volume Calibration with the idea being that you want to set it as low as possible to successfully run a full sweep without the amp shutting down. While you may get a Dirac error for Signal To Noise being too low, again, the target here is to complete a sweep without the amp going into protection.

[wrk24wheel]

Unfortunately, if I lower the level low enough for it not to shut down the amp, I get the low signal to noise error with Dirac. This amp just can't seem to get any volume level above -35db on dirac without shutting down.

[ttocs]

Jun 3, 2023 18:59:10 GMT -5 wrk24wheel said:

Unfortunately, if I lower the level low enough for it not to shut down the amp, I get the low signal to noise error with Dirac.

Yes, but does the sweep complete?

Which microphone are you using?

[wrk24wheel]

Yes. But it gives me the low singal to noise error and then I have to do it again. I am using the Emotiva mic that came with my XMC-2.   I found that at -27db on the main, it will complete without shutting down the amp.  I have tried going up and down on the mic gain 3db each way and every time I continue to get the low signal to noise error.  I have been unable to get it to complete.  The room is quiet at -48db.

[ttocs]

I keep the Mic Gain at 100%. I don't know a reason to go lower, but if it's raised higher than 100% it is done with Digital Gain, so this is something to avoid under normal circumstances. However, in this case I think it's preferred to raise it all the way to check what's going on with your Center speaker.

To make it easy to run Dirac sweeps with the minimum of time involved, I'd setup the system for 3 channels. Just the Fronts and Center. This'll make running sweeps quicker.

So, to keep the sound level from the speakers lower but also so Dirac "hears" the speakers as being louder, raise the Mic Gain all the way.
Then run more sweeps as you have been, adjusting as needed to try to find a point where you get no errors.

[leonski]

How loud when playing MUSIC?


I would personally forget all this 'calibration' nonsense.   If it drives amp into protection when playing music?

That's another more serious matter and than you start into troubleshooting mode.....

I don't see how or why that amp would have trouble with that speaker unless it were simply a very low impedance load.

Magnepan calls the CCR a 3 ohm speaker.....which means possible excursions to 2ohms or maybe a tick lower.....

[wrk24wheel]

Magnepan said down to 1.5 ohms. Still waiting on response from Emotiva tech support on this issue.

[ttocs]

Jun 6, 2023 7:47:25 GMT -5 wrk24wheel said:

Magnepan said down to 1.5 ohms. Still waiting on response from Emotiva tech support on this issue.

At what frequency does the 1.5Ω impedance kick in? I can't find any impedance plots for the CCR.

The XPA Gen1 amps are said to be pretty robust, but a continuous sweep, being a pure tone, might exhibit more stress on the amp than normal music would. Is this the first time you've run sweeps on the CCR? I'm pretty sure that the XPA isn't rated for below 2Ω, and would be surprised if Emotiva even mentioned 2Ω.

I called Emotiva about 9 years ago to quiz them on using XPA amps for my Martin Logan speakers I had at the time, and they said it wasn't a good match being the the impedance of those speakers went down to 1.2Ω.

Also, the age of the amps should probably be considered. They might (probably, if no recapping has been done) have some capacitors out of spec due to age.

I did an Impedance test on my current electrostatic speakers to find out where the impedance drops and how quickly. At just below 9kHz impedance drops below 2Ω. I only tested to 16kHz because I didn't want to stress my speakers at all, but at 20kHz they drop to 0.7Ω according to spec. Notice that the Current needed goes up substantially as frequency rises and impedance drops. Notice that at 1kHz the Current is at 34mA, but at 16kHz it's 491mA.


The ML 13A is a hybrid speaker with self-powered cone woofers, so below the XO of 300Hz the internal amps power the woofers while the external amp powers the stat panel above 300Hz. I created this plot using REW and an imported txt file using SPL as the Impedance number, so that's why you'll see on the left that it's in SPL, but it's really Impedance (I don't know of an easier way to do this).

[KeithL]

I'm going to put my two cents in here...

First of all, to be somewhat pedantic, any speaker whose impedance drops below 2 Ohms is NOT "a well behaved speaker".
While they may be informal... there are standards...  and "a well-behaved speaker should not have an impedance that drops below 75% of its rating for a significant portion of the audio spectrum".
So, for a 4 Ohm speaker, the impedance should not hang around below 3 Ohms for much of its frequency range.
To be quite blunt, the guys who designed the speaker knew this, but they also knew that there's not much musical content at very high frequencies...
(And high-fidelity speakers with a 2 Ohm impedance rating just aren't a thing.)

Now, in real life, there is very little high level musical content at very high frequencies.
For one thing it tends to make your head hurt...
And, for another, it might melt your tweeters...
And, incidentally, it would also cause overload issues for both record mastering lathes and vinyl playback, and saturation issues for analog tape...
Therefore, when playing actual music, this tends to not really be an issue.
And, since our amps are pretty much unconditionally stable, they won't do anything funny either...
However, they may well go into protect if you just plain draw too much current, which could happen if the impedance goes very low.
(And, obviously, while music generally won't do this, it's a different situation with test tones.)

If you were to do a manual calibration, using the manual PEQs, and play the test tones at sane levels, none of this should really be a problem.

It's also worth mentioning that those certain few electrostatics that are widely considered to be "problematic" have an impedance that is both very low and very reactive at high frequencies...
However, I very much doubt that Magneplanars have this issue, or to anywhere near the degree that certain large electrostatic panels do...

However, unfortunately, Dirac Live insists on doing a full-range sweep, and at a relatively high level.

WE DO HAVE A SUGGESTION THAT MIGHT ALLEVIATE THIS ISSUE (NOTE THAT WE *HAVE NOT* TRIED THIS).
The suggestion is to place a resistor in series with each speaker ONLY WHILE RUNNING THE DIRAC CALIBRATION.
A good value to use would probably be 1 Ohm and they should be power resistors of sufficient power rating (10 watts sounds reasonable).
This will limit the low impedance excursions of the speakers... and should prevent the amplifier from going into protect.
(It will also have a slight effect on the overall results - but that effect will be similar for all of your speakers - and you can adjust the target curve to compensate for it if you like.)
(The resistors should be removed after running the test because they will affect the damping factor and efficiency of the speakers - but these factors should have little effect on the calibration Dirac Live does.) 

 

[marcl]

I'm late to the conversation (off pursuing my other expensive hobby the past week in the Shenandoah mountains), but since I have 3.7's and have had a CCR for about 7 months I want to comment a bit.

The CCR is rated 3 ohms. The rated frequency response goes down to 200Hz but I can confirm it goes to 150Hz and rolls off quickly. I have not heard of the CCR or any other Maggie going down to 1.5 Ohms ... they tend to be flat resistive load. The CCR's efficiency is about the same as the 4 ohm 3.7.

How are you doing bass management? Are you crossing the CCR to subs at 150 or 200Hz? Are you using a DWM or Speaker Stand for bass from 200 down to 50, with the CCR connected to the high output, then crossed to subs at 50Hz?

[leonski]

Jun 6, 2023 10:32:18 GMT -5 KeithL said:

I'm going to put my two cents in here...

First of all, to be somewhat pedantic, any speaker whose impedance drops below 2 Ohms is NOT "a well behaved speaker".
While they may be informal... there are standards...  and "a well-behaved speaker should not have an impedance that drops below 75% of its rating for a significant portion of the audio spectrum".
So, for a 4 Ohm speaker, the impedance should not hang around below 3 Ohms for much of its frequency range.
To be quite blunt, the guys who designed the speaker knew this, but they also knew that there's not much musical content at very high frequencies...
(And high-fidelity speakers with a 2 Ohm impedance rating just aren't a thing.)

Now, in real life, there is very little high level musical content at very high frequencies.
For one thing it tends to make your head hurt...
And, for another, it might melt your tweeters...
And, incidentally, it would also cause overload issues for both record mastering lathes and vinyl playback, and saturation issues for analog tape...
Therefore, when playing actual music, this tends to not really be an issue.
And, since our amps are pretty much unconditionally stable, they won't do anything funny either...
However, they may well go into protect if you just plain draw too much current, which could happen if the impedance goes very low.
(And, obviously, while music generally won't do this, it's a different situation with test tones.)

If you were to do a manual calibration, using the manual PEQs, and play the test tones at sane levels, none of this should really be a problem.

It's also worth mentioning that those certain few electrostatics that are widely considered to be "problematic" have an impedance that is both very low and very reactive at high frequencies...
However, I very much doubt that Magneplanars have this issue, or to anywhere near the degree that certain large electrostatic panels do...

However, unfortunately, Dirac Live insists on doing a full-range sweep, and at a relatively high level.

WE DO HAVE A SUGGESTION THAT MIGHT ALLEVIATE THIS ISSUE (NOTE THAT WE *HAVE NOT* TRIED THIS).
The suggestion is to place a resistor in series with each speaker ONLY WHILE RUNNING THE DIRAC CALIBRATION.
A good value to use would probably be 1 Ohm and they should be power resistors of sufficient power rating (10 watts sounds reasonable).
This will limit the low impedance excursions of the speakers... and should prevent the amplifier from going into protect.
(It will also have a slight effect on the overall results - but that effect will be similar for all of your speakers - and you can adjust the target curve to compensate for it if you like.)
(The resistors should be removed after running the test because they will affect the damping factor and efficiency of the speakers - but these factors should have little effect on the calibration Dirac Live does.) 

 

Gee, Keith.....just what I said.   
Magnepan SHIPS many speaker (like my 1.6) with 1 ohm resistors.     given the tweeter is about 4ohms, resistive?    I/5th the resistence is now in the added part.
Which if it is 10 watt rated?    You'd need 50 to the tweeter end which isn't gonna happen......

CCR uses a True Ribbon tweeter.    In the 3.6, this is maybe 3 ohms....   But for detail?   I'd look at the STEREOPHILE test of this speaker.    
The 1.6, which is a 2-way crossing about 600hz uses a 4 amp fuse in that part of the panel.   Bass is unfused.....  No ribbon in this model.

Curious about the 3.6 crossover?    Go to MUG....Magnepan Users Group under tweaks there is a section of schematics.....You'll find it there.
As it turns out?    Bass low-pass is in an external box.   The mid-tweet crossover is INTERNAL to the panel...

[405x5]

Jun 4, 2023 15:05:45 GMT -5 leonski said:

How loud when playing MUSIC?


I would personally forget all this 'calibration' nonsense.   If it drives amp into protection when playing music?

That's another more serious matter and than you start into troubleshooting mode.....

I don't see how or why that amp would have trouble with that speaker unless it were simply a very low impedance load.

Magnepan calls the CCR a 3 ohm speaker.....which means possible excursions to 2ohms or maybe a tick lower.....

I would not go as far as to call all of it “calibration nonsense“,……… but in the context of being over reliant on it, you would be absolutely correct. IMHO. Once you lose the ability to trust your ears, the game is over.
this kind of an issue where a speaker is being overdriven would be easily picked up in the absence of automated room correction.

[leonski]

You can Listen OR Calibrate.

I'd listen and adjust on the fly......But I'm old-school.....

Calibration to the exclusion of listening is nutty......

I'm certain that for the Nth degree, a good cal CAN be useful........

But I also suspect that background of the user is just as important.

[KeithL]

And, at 15 kHz, there is VERY little content in music... and almost never at anything near a high level.

Jun 6, 2023 9:26:54 GMT -5 ttocs said:

Jun 6, 2023 7:47:25 GMT -5 wrk24wheel said:

Magnepan said down to 1.5 ohms. Still waiting on response from Emotiva tech support on this issue.

At what frequency does the 1.5Ω impedance kick in? I can't find any impedance plots for the CCR.

The XPA Gen1 amps are said to be pretty robust, but a continuous sweep, being a pure tone, might exhibit more stress on the amp than normal music would. Is this the first time you've run sweeps on the CCR? I'm pretty sure that the XPA isn't rated for below 2Ω, and would be surprised if Emotiva even mentioned 2Ω.

I called Emotiva about 9 years ago to quiz them on using XPA amps for my Martin Logan speakers I had at the time, and they said it wasn't a good match being the the impedance of those speakers went down to 1.2Ω.

Also, the age of the amps should probably be considered. They might (probably, if no recapping has been done) have some capacitors out of spec due to age.

I did an Impedance test on my current electrostatic speakers to find out where the impedance drops and how quickly. At just below 9kHz impedance drops below 2Ω. I only tested to 16kHz because I didn't want to stress my speakers at all, but at 20kHz they drop to 0.7Ω according to spec. Notice that the Current needed goes up substantially as frequency rises and impedance drops. Notice that at 1kHz the Current is at 34mA, but at 16kHz it's 491mA.


The ML 13A is a hybrid speaker with self-powered cone woofers, so below the XO of 300Hz the internal amps power the woofers while the external amp powers the stat panel above 300Hz. I created this plot using REW and an imported txt file using SPL as the Impedance number, so that's why you'll see on the left that it's in SPL, but it's really Impedance (I don't know of an easier way to do this).

[leonski]

Keith Said::
'And, at 15 kHz, there is VERY little content in music... and almost never at anything near a high level.'


This is how ElectroStats get away with that wacky electrical performance. Not even 10% of musical energy is above 15khz.....Maybe even only 5%?

And while that impedance data does smell of Rat, the rest of the data? REACTANCE is what baits the trap. 1.5ohms or whatever AND at 60 Degrees? (take your pick of Capacitive OR Inductive!)
and you are SUNK.
Many amps simply rebel at such treatment.....

[vcautokid]

The funny thing with many room correction schemes, many times you end up manually tweaking the settings for one reason or another. I wouldn't call it nonsense per se, but maybe a starting point good, or otherwise depends on many factors. There are plenty of posts just on this forum of people tuning with the room correction scheme in play. So your ears are still required regardless. But with that resistor in play you should be good.

[AudioHTIT]

I don’t remember exactly but isn’t the XPA-1 a bridged amp? Which isn’t really saying anything different than what everyone else is saying, that is, it’s an impedance issue. Just that with bridged amps, a low impedance speaker is an even a bigger deal because they see half what an un-bridged amp sees. If the XPA-1 isn’t bridged then … never mind.

[wrk24wheel]

Jun 6, 2023 10:32:18 GMT -5 KeithL said:

I'm going to put my two cents in here...

First of all, to be somewhat pedantic, any speaker whose impedance drops below 2 Ohms is NOT "a well behaved speaker".
While they may be informal... there are standards...  and "a well-behaved speaker should not have an impedance that drops below 75% of its rating for a significant portion of the audio spectrum".
So, for a 4 Ohm speaker, the impedance should not hang around below 3 Ohms for much of its frequency range.
To be quite blunt, the guys who designed the speaker knew this, but they also knew that there's not much musical content at very high frequencies...
(And high-fidelity speakers with a 2 Ohm impedance rating just aren't a thing.)

Now, in real life, there is very little high level musical content at very high frequencies.
For one thing it tends to make your head hurt...
And, for another, it might melt your tweeters...
And, incidentally, it would also cause overload issues for both record mastering lathes and vinyl playback, and saturation issues for analog tape...
Therefore, when playing actual music, this tends to not really be an issue.
And, since our amps are pretty much unconditionally stable, they won't do anything funny either...
However, they may well go into protect if you just plain draw too much current, which could happen if the impedance goes very low.
(And, obviously, while music generally won't do this, it's a different situation with test tones.)

If you were to do a manual calibration, using the manual PEQs, and play the test tones at sane levels, none of this should really be a problem.

It's also worth mentioning that those certain few electrostatics that are widely considered to be "problematic" have an impedance that is both very low and very reactive at high frequencies...
However, I very much doubt that Magneplanars have this issue, or to anywhere near the degree that certain large electrostatic panels do...

However, unfortunately, Dirac Live insists on doing a full-range sweep, and at a relatively high level.

WE DO HAVE A SUGGESTION THAT MIGHT ALLEVIATE THIS ISSUE (NOTE THAT WE *HAVE NOT* TRIED THIS).
The suggestion is to place a resistor in series with each speaker ONLY WHILE RUNNING THE DIRAC CALIBRATION.
A good value to use would probably be 1 Ohm and they should be power resistors of sufficient power rating (10 watts sounds reasonable).
This will limit the low impedance excursions of the speakers... and should prevent the amplifier from going into protect.
(It will also have a slight effect on the overall results - but that effect will be similar for all of your speakers - and you can adjust the target curve to compensate for it if you like.)
(The resistors should be removed after running the test because they will affect the damping factor and efficiency of the speakers - but these factors should have little effect on the calibration Dirac Live does.) 

Keith - Thank you for the info. The CCR is designed and listed as a 3ohm speaker to start with. Therefore, they are stating only a 50% drop in impedance. Magnepan told me to do the same thing, but the amp still went into overload with the supplied 1.2ohm resistor. One thing that I do not quite understand is that I have an older MIT Terminator 2 Speaker cable. I hooked that up and was able to actually raise the gain another 6.5db before clipping. I think the internal crossover network on the wire must be changing the impedance. I am getting a new mic to try and see if that will put me over the edge to complete Dirac. Hopefully, that does the trick.

[wrk24wheel]

Jun 6, 2023 12:34:39 GMT -5 marcl said:

I'm late to the conversation (off pursuing my other expensive hobby the past week in the Shenandoah mountains), but since I have 3.7's and have had a CCR for about 7 months I want to comment a bit.

The CCR is rated 3 ohms. The rated frequency response goes down to 200Hz but I can confirm it goes to 150Hz and rolls off quickly. I have not heard of the CCR or any other Maggie going down to 1.5 Ohms ... they tend to be flat resistive load. The CCR's efficiency is about the same as the 4 ohm 3.7.

How are you doing bass management? Are you crossing the CCR to subs at 150 or 200Hz? Are you using a DWM or Speaker Stand for bass from 200 down to 50, with the CCR connected to the high output, then crossed to subs at 50Hz?

I have tried several different configurations.  Setting the CCR at Small (200Hz) and running the Speaker stand off a separate amplifier as a sub.  I have also tried setting the CCR as large and going straight to the stand.  Right now the best configuration even sound wise, is using the MIT bi-wire cable and running the lows to the stand and highs to the CCR with the XMC-2 set as large.  I currently run my sub at 35Hz as I prefer the bass from the 3.7's and really only use the sub for the impact during a movie.