Troels says that first-order filters are a no-go for both the purifi and the tweeter used…
That’s a good question. The Elsinores use first-order filters which Troels thinks don’t work for the purifies, but I don’t know whether the impedance smoothing network saves the day somehow. I don’t think Troels uses those, he certainly doesn’t in his Purifi speaker.
I don't think Troels actually says that:
"...it is also clear this is not a niche project aimed at high-efficiency SET nerds, rather digital people with high-order filters and tonnes of digital watts available."
Also, this is before the driver we are going to use was available and it was the original PTT6.5W04 driver. The long linear long-throw (massive high Xmax) and making the voice coil this long means low sensitivity. It is rated 88dB but this is a 4 Ohm driver he is looking at, and he rightly then says it is actually, 85dB for a one nominal Watt. Means it needs a big amp and most will use it in 2-Ways, just as Troels says, because in that application they will be near 10dB down compared to the Elsinore. Think about that, that means ten times the power required. Troels is right on here.
If Troels now had looked at a PTT6-5W8-NFA-01, his view would have been different. The extra 3dB can make a big difference! This what is required for the Elsinores and that is what changed my mind: This Purifi driver is suitable!
Also, the crossover used in the Elsinores is not exactly 1st order a la Butterworth, here you are a further 3dB down at the crossover frequency. We are down 6dB and again makes quite a difference.
I am also sending an email to Troels. I want to comment on his latest (I think it is) project, the www.troelsgravesen.dk/SBA-761.htm. He calls it 1st order, it is electrically but it is not Butterworth, look at the tweeter phase and it has reverse phase and it shows he is doing it the same I am.
The step response would look similar to this (ignore the trailing red):
But the one thing that stands out is a single series inductor to the driver that handles the midrange, without any parallel with the driver networks, no Zobels. Bravo! In fact, the bass driver qualifies as the same treatment.
This is already the sort of thing we are doing in the Elsinores. Re the current EQ, this is the cherry on the top. This further suppresses what Lars Risbo has publically called "impedance modulations" and so do series inductors to drivers that now Troels is doing as well. It is very clear that he is happy about the sound that he is getting.
It’s not his latest project, and he has designed other 2.5-ways, e.g.Discovery-861 and Ellipticor-3, and used 6db filters in a few other speakers for at least some of the drivers.
Troels does explicitly rule out 1st order filters in his Purifi 6R write up: “1st order is no-go here as we have to protect the tweeter and the Purifi, although well behaved, doesn't have the most smooth roll-off around. “
The two projects you mention, they are 2020 and not 2021, so they must be earlier. Never mind, no matter.
“1st order is no-go here as we have to protect the tweeter and the Purifi, although well behaved, doesn't have the most smooth roll-off around. “
I am trying to find the page where Troels says that. Do you have the link? Is it this one:
Purifi-6R
Not there. Not here either:
Purifi_PTT6.5WO4
If it is, he has edited it because it does not say that anymore.
Anyway, it does not matter. Troels and I don't look for exactly the same things in drivers and that's fair enough (but what he says about protecting the tweeter, we are in unison). For example, when I used the NRX driver, he questioned me about using that driver in a low-order situation. Maybe he thought I should have used the MFC driver, but I did not ask. I did end up using the MFC as the NRX original got discontinued. But I got excellent results from both.
So when I look at the PTT6-5W8-NFA-01 driver, I look at it through my own eyes and from another set of experiences. Think of it as having different priorities. Another example, I got my hands of a pair of Usher S520 speakers where the drivers were challenged for what I had in mind - but sometimes you have to be brave and willing to fail when you push boundaries. To my pleasant surprise, it worked much better than anticipated. That made me think differently about what things are paramount and what is not. This rearranges your priorities.
Maybe Purifi simply thinks higher order crossovers will be used and never thought it would be used in a different way, but that does not dissuade me. The measurements done by them and others, I see what I like and I like what I see.
Pilk, re mounting, you are right, Troels mounted it the simple way, you can see the shadow in the photo, so no doubt about it. A good thing as DIY option.
“1st order is no-go here as we have to protect the tweeter and the Purifi, although well behaved, doesn't have the most smooth roll-off around. “
I am trying to find the page where Troels says that. Do you have the link? Is it this one:
Purifi-6R
Not there. Not here either:
Purifi_PTT6.5WO4
If it is, he has edited it because it does not say that anymore.
Anyway, it does not matter. Troels and I don't look for exactly the same things in drivers and that's fair enough (but what he says about protecting the tweeter, we are in unison). For example, when I used the NRX driver, he questioned me about using that driver in a low-order situation. Maybe he thought I should have used the MFC driver, but I did not ask. I did end up using the MFC as the NRX original got discontinued. But I got excellent results from both.
So when I look at the PTT6-5W8-NFA-01 driver, I look at it through my own eyes and from another set of experiences. Think of it as having different priorities. Another example, I got my hands of a pair of Usher S520 speakers where the drivers were challenged for what I had in mind - but sometimes you have to be brave and willing to fail when you push boundaries. To my pleasant surprise, it worked much better than anticipated. That made me think differently about what things are paramount and what is not. This rearranges your priorities.
Maybe Purifi simply thinks higher order crossovers will be used and never thought it would be used in a different way, but that does not dissuade me. The measurements done by them and others, I see what I like and I like what I see.
Pilk, re mounting, you are right, Troels mounted it the simple way, you can see the shadow in the photo, so no doubt about it. A good thing as DIY option.
Last edited:
Thanks, got it. Note the reference to "protect the tweeter" and that tweeter I agree with what he says. That ribbon tweeter has a lot of distortion at the limits of its passband - the lower end. This is why I have for the most part avoided them (I am not saying this is a bad tweeter). I have no problem with the roll-off of the Purifi driver. In fact, it is above average here, but in Troels mind, it is still not good as the SB MW16 driver. But then he said the same thing about the original NRX driver, which I think is worse than the Purifi in this respect.
Let me you a clue, I look at what the inductance is doing as much of its roll-off. This is where Purifi does something special that can be exploited.
I have a test that I would like other testers to use. I call it then 3-Man Couch Test. It works especially well when the speaker is designed to be at its best at 15 degrees off-axis, which is how the Elsinore is done. It passes the test of three people seating gets the same frequency response within about 0.5dB - to pass the test it should not be more than 1dB.
Sitting on the couch and at ear height, there must be pretty much zero tonal change in either three seats. Note that none of the three listeners are hearing the speakers dead on axis. As one moves closer to one speaker, you get closer to 15 degrees, but never on axis. Whilst you are doing that, you are getting further away from the other speaker and further off axis. So as one gets brighter, the other gets less bright and the tonal balance in your head hardly changes. The measured deviation should be no more than 1dB. Use 1/6th Octave Pink Noise that is no-coincidental.
I can achieve this also with the NRX driver (the one Troels questioned) and I will do it with the Purifi drivers too.
Let me you a clue, I look at what the inductance is doing as much of its roll-off. This is where Purifi does something special that can be exploited.
I have a test that I would like other testers to use. I call it then 3-Man Couch Test. It works especially well when the speaker is designed to be at its best at 15 degrees off-axis, which is how the Elsinore is done. It passes the test of three people seating gets the same frequency response within about 0.5dB - to pass the test it should not be more than 1dB.
Sitting on the couch and at ear height, there must be pretty much zero tonal change in either three seats. Note that none of the three listeners are hearing the speakers dead on axis. As one moves closer to one speaker, you get closer to 15 degrees, but never on axis. Whilst you are doing that, you are getting further away from the other speaker and further off axis. So as one gets brighter, the other gets less bright and the tonal balance in your head hardly changes. The measured deviation should be no more than 1dB. Use 1/6th Octave Pink Noise that is no-coincidental.
I can achieve this also with the NRX driver (the one Troels questioned) and I will do it with the Purifi drivers too.
Attachments
Last edited:
IMPORTANT:
The email has just been sent to sales@purifi-audio.com and has set out the general interest and use to them and that 24 drivers confirmed, are needed to be sent to Australia alone. The first eight asap. Also, that number is likely to double and more. That should get their interest. It is just before noon there and evening here (ten hours difference), so I wonder if I will hear from them overnight?
The email has just been sent to sales@purifi-audio.com and has set out the general interest and use to them and that 24 drivers confirmed, are needed to be sent to Australia alone. The first eight asap. Also, that number is likely to double and more. That should get their interest. It is just before noon there and evening here (ten hours difference), so I wonder if I will hear from them overnight?
hello joe, no problem, thanks for the feedback. Actually my question is answered. Could you still check the photo of my building again and confirm it briefly. I would only choose 25 ° for the router. best regards Steff
Attachments
Wow I am so glad I caught this. In the Hamlet brace you did not make the place for the tweeter like we have in the Elsinores.
Elsinore Construction I'm not sure what page the drawings for the Hamlets are in this thread, I'll find them and post, but if you can update them, but I'll copy the Elsinore brace and just shop off the bottom half.
Elsinore Construction I'm not sure what page the drawings for the Hamlets are in this thread, I'll find them and post, but if you can update them, but I'll copy the Elsinore brace and just shop off the bottom half.
IMPORTANT:
Our contact name is Claus at sales@purifi-audio.com and despite that there has been a price increase on their website in the last few days, we have received a favourable price.
Note: When you place your order with Claus, indicate that this is for Elsinore Mk-6 "ULD" Project.
They are packed in a box of eight @ 19Kg total, which I suspected because the SB drivers were packed eight at a time as well. So it will be a single box.
The price without VAT is 8 x DKK 1912 each. That is DKK 15,296
The price with Danish 25% VAT is DKK 19,210
The cost of sending 19Kg will be worked out by Claus I am sure. If outside EU, you will need to handle your local customs and charges. I will be doing exactly the same thing here in Australia. If they ask me what it is I will say "just hobby parts."
Purifi's lead time before sending is four weeks.
I will be officially ordering and arranging to pay for my first lot of eight.
Once I have the drivers here, I will need boxes. How exactly that is going to work out due to the time of year, our Xmas and summer holidays coincide and we are getting some boxes made.
There is a possibility of using my own current DIY Elsinore boxes and taking the SB drivers out and fit the Purifi drivers and do the design work. The drivers will not be rebated and a spacer can be fitted into the rebated area of the SB drivers (the SB drivers will likely go back when I get new boxes). We will just see how we go, but I have in mind not to waste too much time. January is likely, but don't hold me to it.
Cheers, Joe
Our contact name is Claus at sales@purifi-audio.com and despite that there has been a price increase on their website in the last few days, we have received a favourable price.
Note: When you place your order with Claus, indicate that this is for Elsinore Mk-6 "ULD" Project.
They are packed in a box of eight @ 19Kg total, which I suspected because the SB drivers were packed eight at a time as well. So it will be a single box.
The price without VAT is 8 x DKK 1912 each. That is DKK 15,296
The price with Danish 25% VAT is DKK 19,210
The cost of sending 19Kg will be worked out by Claus I am sure. If outside EU, you will need to handle your local customs and charges. I will be doing exactly the same thing here in Australia. If they ask me what it is I will say "just hobby parts."
Purifi's lead time before sending is four weeks.
I will be officially ordering and arranging to pay for my first lot of eight.
Once I have the drivers here, I will need boxes. How exactly that is going to work out due to the time of year, our Xmas and summer holidays coincide and we are getting some boxes made.
There is a possibility of using my own current DIY Elsinore boxes and taking the SB drivers out and fit the Purifi drivers and do the design work. The drivers will not be rebated and a spacer can be fitted into the rebated area of the SB drivers (the SB drivers will likely go back when I get new boxes). We will just see how we go, but I have in mind not to waste too much time. January is likely, but don't hold me to it.
Cheers, Joe
Last edited:
Not there yet. When able, I need to update and even reorganise it altogether.
Now has multiple models/variations. I did to organise it differently. I am hoping for a lull to come...
Cheers, Joe