The "Elsinore Project" Thread

I will try to post that graphic that shows to difference between NRXC and MFC version of EL-6:

EL-6_BlueMFC_RedNRXC.gif


Red is NRXC and Blue is MFC.

I have now been living with the MFC for at least a month and quite pleased with it. The change in driver is not a downgrade in performance and to my ears works just as well. I have a bias for fibre cones and went with the NRXC originally, but I have always appreciated a top quality Polycone driver, neither is perfect and both have their advantages.

I have to use the MFC Polycone version, as I get visitors who wants to listen to something that is available, so I have no choice - but I don't feel cheated either, that I now have to do with the second-best driver. It has not worked out that way at all. So I feel quite pleased and the Elsinore Mk6 marches on. :)

PS: The graphic above works - for the computer techies out there, I had to get an SSL Certificate, that solved the problem of new posts, but not the older ones. I hope there is someone that can help me out with that.
 
Last edited:
Starting Construction!

Well, I started working on the cabinets for my Elsinores this weekend. Since only 1/2" and 3/4" MDF is readily available (Home Depot 2 miles away), I bought this material, rough cut the panels and now laminating them to the correct thickness. In preparation, I built a torsionbox clamping press to be able to keep even pressure across the panels during glueup, and it seems to be working just as planned. I will have to upload some pictures as it progresses further.

David
 
I will try to post that graphic that shows to difference between NRXC and MFC version of EL-6:

EL-6_BlueMFC_RedNRXC.gif


Red is NRXC and Blue is MFC.

I have now been living with the MFC for at least a month and quite pleased with it. The change in driver is not a downgrade in performance and to my ears works just as well. I have a bias for fibre cones and went with the NRXC originally, but I have always appreciated a top quality Polycone driver, neither is perfect and both have their advantages.

I have to use the MFC Polycone version, as I get visitors who wants to listen to something that is available, so I have no choice - but I don't feel cheated either, that I now have to do with the second-best driver. It has not worked out that way at all. So I feel quite pleased and the Elsinore Mk6 marches on. :)

PS: The graphic above works - for the computer techies out there, I had to get an SSL Certificate, that solved the problem of new posts, but not the older ones. I hope there is someone that can help me out with that.

Good to hear, have you done some distiortion testing as well?

I use both these drivers in some speakers i have, no Elsinore though, and i can not decide which one i like best. I think the Nrxc has a slightly crispier sound, but not by much, but Mfc measures better in some respects in my tests
 
Good to hear, have you done some distortion testing as well?

I use both these drivers in some speakers i have, no Elsinore though, and i can not decide which one i like best. I think the Nrxc has a slightly crispier sound, but not by much, but Mfc measures better in some respects in my tests

Yes, I would go along with that comment. As far as distortion is concerned, I have done tests on individual drivers in the past, but I have not compared NRXC and MFC in that regard. But they are clearly very good. Where I have paid most attention in the overall design, that comes to the tweeter. To me that has often been the Achilles Heel. I have done comparative tests on the HDS tweeter versus a top-of-the-line Scan-Speak tweeter (the HDS is also made by Scan-Speak, but that was not well-known then).

Take a look at this link:

LINK: TWEETER COMPARISON

The Waveguide and the use of the null technique, ensures very low distortion of the tweeter, along with the Waveguide. I use the Waveguide to lower distortion, where other use it to extend bandwidth and cross over at a lower frequency. I resisted that temptation, I wanted less distortion, period!

The use of a single inductor on the other drivers, here I am willing to trade imperfection, in order to keep the current straight and also because a single inductor suppresses the internal non-linear inductance of the driver. But clearly, to be able to do that you have to be very driver selective.

It's been near twenty years ago that Lynn Olson (who has agreed that we have some similar approaches) wrote the article, link below. It is quite long and very detailed, but still surprisingly current and applicable to even today.

LINK: The Family of Direct Radiators by Lynn Olson

For those interested, that will keep you occupied for some time. :)
 
I will try to post that graphic that shows to difference between NRXC and MFC version of EL-6:

EL-6_BlueMFC_RedNRXC.gif


Red is NRXC and Blue is MFC.

That may be a silly question, but what does this plot really show? I assume it's some kind of SPL, but... anechoic? In-room? At which drive voltage? Where's the microphone (distance, level, on/off-axis)? Why is it "stepped"?
 
That may be a silly question, but what does this plot really show? I assume it's some kind of SPL, but... anechoic? In-room? At which drive voltage? Where's the microphone (distance, level, on/off-axis)? Why is it "stepped"?

This is what is called an RTA measurement and it is definitely not anechoic. It is an in-room measurement.

The measurement is about 2.5 metre and about 15 degrees off axis. The Elsinores should be listened to off axis. If the room is a bit dead, then toe them in, if the room is a bit bright, do the opposite. For the most time, when sitting down in the optimum position, you should see some of the insides of the boxes.

Most RTA measurements are 1/3rd Octave and hence the steps. This is a 1/6th Octave and gives a lot more resolution. The stimulus is Pink Noise.

Below was posted earlier on and you can see all the room modes:

763743d1560897148-elsinore-project-thread-el6-mfc-axis-gif


The peaks and the valleys are OK, so this is a pretty good result. The further you go below 1KHz, the more influence room modes will have.

Above 2KHz the result becomes closer to what we would expect in an anechoic chamber.

Definitely voltage drive, but the Elsinores will not deviate even with current drive. It would look the same because the Elsinores draw the same current at all frequencies and that takes the high impedance of current drive out of the FR equation.
 
Last edited:
This is ... an RTA ... in-room measurement.

Ah, ok. I was thinking I might be able to extract some information about the efficiency of these loudspeakers, but I guess these RTA curves are not very meaningful for this. I found it a bit surprising that the Elsinores are > 90 dB down to about 45 Hz (as shown here), but then I realised that this plot does not say what voltage/power was applied, and at which distance the measurement was done. Can you provide some insight?

I have recently read a few things about in-room SPL curves. The gist seems to be that smooth, but sloping curves with lower SPL at higher frequencies are preferred. How does that relate to the Elsinore?
 
Progress!

Over the long weekend, finished cutting my panels and then routed out the baffles, baffle brace and main brace for the cabinets. Instead of using a jigsaw, made an adjustable router template to make the rectangular cuts in the baffle brace and main brace. Worked great, but it did take alot of time. Very close to starting glue-up of the cabinets.

David
 
Last edited:

More pictures showing progress, as well as my press for laminating two pieces of MDF together.

Still learning on posting images. All I seem to be able to do is post links.

20190818_121015.jpg - Google Drive

20190818_120950.jpg - Google Drive

20190818_120940.jpg - Google Drive

Instead of using links its best to upload to diyA server so images will stay forever and never disapear should that link ever change into future, example is below : )

Pick the advanced dialog to make posts because it have will have a clip icon at top middle where one can upload pictures, prepare that images is physical avaiable on one of local computers fixed or removeable harddrives because upload dialog can only point to drive letters, one can upload up to 10 attachments per post in one go, but if image size is bigger than 976.6 KB then don't use one go but instead upload one at a time using the "Upload" button seperate for each new image string, this behavour is because it looks diyA server can downsize if they too huge but works only picking one at a time and fails picking multiples.
 

Attachments

  • 20190902_154121sm.jpg
    20190902_154121sm.jpg
    419.6 KB · Views: 526
  • 20190818_121015.jpg
    20190818_121015.jpg
    671 KB · Views: 497
  • 20190818_120950.jpg
    20190818_120950.jpg
    604.5 KB · Views: 485
  • 20190818_120940.jpg
    20190818_120940.jpg
    546.3 KB · Views: 480
Yes, I would go along with that comment. As far as distortion is concerned, I have done tests on individual drivers in the past, but I have not compared NRXC and MFC in that regard. But they are clearly very good. Where I have paid most attention in the overall design, that comes to the tweeter. To me that has often been the Achilles Heel. I have done comparative tests on the HDS tweeter versus a top-of-the-line Scan-Speak tweeter (the HDS is also made by Scan-Speak, but that was not well-known then).

Take a look at this link:

LINK: TWEETER COMPARISON

The Waveguide and the use of the null technique, ensures very low distortion of the tweeter, along with the Waveguide. I use the Waveguide to lower distortion, where other use it to extend bandwidth and cross over at a lower frequency. I resisted that temptation, I wanted less distortion, period!

The use of a single inductor on the other drivers, here I am willing to trade imperfection, in order to keep the current straight and also because a single inductor suppresses the internal non-linear inductance of the driver. But clearly, to be able to do that you have to be very driver selective.

It's been near twenty years ago that Lynn Olson (who has agreed that we have some similar approaches) wrote the article, link below. It is quite long and very detailed, but still surprisingly current and applicable to even today.

LINK: The Family of Direct Radiators by Lynn Olson

For those interested, that will keep you occupied for some time. :)
Hi Joe,
I was wondering. Why higher (around 3 kHz) x-over point has lower distortion than lower (around 2 kHz). I thought that if tweeter has lower distortion than midwoofer, widening tweeter frequency response will lead to overall lower distortion. Is it because tweeters Fs is not that far away from 2 kHz x-over point?
I was also taking into consideration fact that a crossed higher midwoofers breakup would beam more even though it has RLC trap. What are your thoughts about crossing around 2 kHz?
Best regards,
Bartosz