• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Speakerlab S7s Rebuild and Modding (Ribbon Tweeter Model)

DSC07349.JPG DSC07353.JPG DSC07356.JPG DSC07357.JPG

The first listening session will start shortly.

Also, I've been so happy with the results of the last cap mod in the mid section that I started wondering if there is anything to be had by improving the quality of the series resistors. So today I ordered some Mundorf M-Resist Classic and Supreme, enough to change out the series resistors in the mid section. I've also been intrigued with the thought of changing some of the coils to ribbon types. So I also bought some Mundorf ribbon coils for the tweeter and the series coil in the mid section.

LOL. I've been tinkering with these things for over 7 years now. My wife keeps telling me that I am more stubborn than a mule. Actually she uses the name of a very similar animal for some reason. ;) I'd like to think that I am persistent and patient. :)

Cheers,
James
 
My first impressions after listening to 4 cds last night are that the speakers are a little bit tighter in bass/midbass. The difference is much more subtle than what happened when I first placed them on the spikes and granite. I have to admit that I just like the look of this setup a lot more.

One thing is for sure, I have reached the point where there is absolutely no need for more bass. More would be too much. I like the balance of the sound a lot now, and that balance is limited by the sensitivity of the tweeters. I find myself wishing that the tweeters had perhaps 1db more sensitivity, that would give me a bit more room to play with the padding in the mids and go with a woofer coil that has a bit lower resistance. As it is when the new tweeter and mid coils get here I will have to play with the mid paddding a bit because the coils have less resistance.

It's interesting that going from 3cm thick / 35x35cm granite to 4cm thick 40x40cm increased the weight from 8kg (17.6lb) to 19.5kg (42.9lb). The speakers are rock solid on the floor now, not the slightest bit of movement when I try to shake them a bit by hand.

Cheers,
James
 
Tomorrow I'll be replacing the crossiver coils in the tweeter section with some Mundorf ribbon types. Also coming up soon after will be replacing the series coil in the mid section with a ribbon type and all of the Mox resistors in the mid section with Mills and Mundorf wirewound types.

DSC07358.JPG

Cheers,
James
 
The ribbon coils have a lower DCR than the air-core coils even using the smallest AWG that Mundorf offers (16awg). Xsim showed that that would extend the tweeter crossover point down lower than I wanted. After trying out several different values I settled on a 0.27mH coil with 0.16 ohm DCR (compares to 0.31 ohms for the old 0.31mH coil).

Another factor is the boards themselves. I'm coming to the point where I really want to etch (or have etched) some PCBs for these things. They are stable now in terms of what and how many components they will have and the layouts won't be changing from here on. So the best was to solder the ribbons into my existing boards without damaging or shortening the leads is important. With a PCB I can just wrap the ribbon down around an edge and solder it flat onto a trace. With my point-to-point boards I really didn't have anything to solder to.

So after removing the old coils I soldered in some short lengths of 1.5mm² solid core wire and then soldered the ends of the ribbons to them. They have a solid full width connection and should be easy to de-solder without loosing any lead length when it comes time to move them over to a PCB. I haven't the slightest idea if that is "good practice" or not but it did have the positive attribute of working.

DSC07360.JPG DSC07363.JPG DSC07365.JPG DSC07371.JPG

The coils have a wood bobbin/former in the center. Eventually I'll be using some M5 brass screws to attach them but for now cable ties are doing the job. There is a small round of foam mat under them as a cushion.

Cheers,
James
 
And installed in the crossover boxes:

DSC07375.JPG

I have only listened to 1 CD so far. Wife wanted to listen to her Christmas CD by Unheilig so on it went. LOL, a highly ironic subject matter for a band with that name but we like them and it's a pretty good CD.

First impression is that I have gained a bit of volume on the tweeter and a little more presence. Obviously I need to do a lot more evaluating but I'm thinking that the system is now slightly out of balance, the highs are a little too much now. And that is actually a good thing. That gives me a little bit of wiggle room concerning the mids. The new series coil for the mids also has less DCR than the old one and the mids are already padded down by 2db. I'm still waiting for the est of the new resistors to get here before I can do anything more.

Cheers,
James
 
Hi James,
Your thread has enough length now that digging out your final configuration is pressing my ADD buttons. Any chance you might update the first post with links to the later results, i.e., which drivers you chose, the crossover schematic, any sims you did, etc.
Thanks,
John

And Merry Christmas
 
nerdorama said:
Hi James,
Your thread has enough length now that digging out your final configuration is pressing my ADD buttons. Any chance you might update the first post with links to the later results, i.e., which drivers you chose, the crossover schematic, any sims you did, etc.
Thanks,
John

And Merry Christmas
Click to expand...

:) Yes. :)

I have done that to a couple of long threads now, amending the 1st post with what amounts to an index. It is definitely planned to do that here but probably won't happen until after next week. 3 more days to work this year and then I have off until the 7th of Jan.

Merry Christmas to you and your family also John.
James
 
This might help also. A screenshot of the current state of the crossover as shown in Xsim:

Xsim_SpeakerlabS7s_04d1a.jpeg
Note that the blanked out caps in the mid and woofer section are not installed in the crossovers, they are only entered in the schematic so that I could switch them between open/normal to see if Xsim showed and change in response if they were installed.

Cheers,
James
 
So I've done a bit more listening. I have never before been so happy that the crossovers are out of balance. :)

The tweeters are without a doubt now playing a bit too loud. Good and smooth and crystal clear, but in fact too much. The midrange magic is now somewhat overshadowed. And I have already changed my mind and will for sure be changing the series midrange coil tomorrow. I now have a little room to maneuver in that dept and need to explore it immediately.

My understanding of "conventional wisdom" concerning parallel shunt components in crossovers has been somewhat shattered. It started after recently reading some comments by rothwellaudio concerning the value of such components and the reality of what is "in the signal path". I thought for sure that I had saved a link to the thread but can't remember exactly the thread name and context and can't find it to link here. Nonetheless I am now convinced that I must pay just as much attention to the quality of what I put in the parallel positions as to what goes in the series positions.

This is an interesting development.

Cheers,
James
 
Bratwurst7s said:
So I've done a bit more listening. I have never before been so happy that the crossovers are out of balance. :)

The tweeters are without a doubt now playing a bit too loud. Good and smooth and crystal clear, but in fact too much. The midrange magic is now somewhat overshadowed. And I have already changed my mind and will for sure be changing the series midrange coil tomorrow. I now have a little room to maneuver in that dept and need to explore it immediately.

My understanding of "conventional wisdom" concerning parallel shunt components in crossovers has been somewhat shattered. It started after recently reading some comments by rothwellaudio concerning the value of such components and the reality of what is "in the signal path". I thought for sure that I had saved a link to the thread but can't remember exactly the thread name and context and can't find it to link here. Nonetheless I am now convinced that I must pay just as much attention to the quality of what I put in the parallel positions as to what goes in the series positions.

This is an interesting development.

Cheers,
James
Click to expand...
I liked very much this idea James, everything should be important. Looking forward to "hear" your new developments.
Cheers,
Rodolfo
 
nerdorama said:
I really want to get a handle on Xsim. You are inspiring me to get at it.
Click to expand...

https://www.diyaudio.com/forums/software-tools/259865-xsim-free-crossover-designer.html

To get the most out of it you will need a speaker measuring tool like WooferTester2 or such. Something that can do a frequency & impedance sweep and generate a .zma file. It defaults to an 8 ohm driver and I don't see a way to manually adjust that. But it is a really nice tool and lets you real-time adjust component values and see the results in the response window. You can use the scroll wheel to sweep through different values also and see rapid changes. I would not have made the progress of the last year without it.

Cheers,
James

ps: there isn't really that much of a learning curve to use it. If you can sketch out a schematic you can get started with the program straight away.
 
morfeora said:
I liked very much this idea James, everything should be important. Looking forward to "hear" your new developments.
Cheers,
Rodolfo
Click to expand...

The mid mods are now finished. I'm about to do a listening test right now. Comments and photos will follow a little later.

James
 
Hi James,
I have the measuring tools, Arta with the interface box and a cal'ed mic. Just haven't put the effort into Xsim. I have measured many drivers for TS parameters but haven't made many acoustical measurements which are trickier in the setup.

I also have all the parts to build Pete Millett's sound card interface box but need to get the pcb assembled.
http://www.pmillett.com/ATEST.htm

John
 
My universe is back in balance. :) But not at first.

After listening a bit I felt that surprisingly enough there was a bit too much midbass at first. I had already considered 2 scenarios, the need to pad the mids a bit more, and the need for a bit less. So I had ordered extra resistors and thankfully enough they were part of the shipment that did arrive.

First plan was for increasing the resistor value in the notch filter, second plan was to raise or lower the L-pad by a quarter db. In this case I pulled the boards and replaced the 2x 0.82 ohm Mox resistors that make up what I am calling the notch filter with 2x 1 ohm Mundorf M-Resist Classic 10w wirewound. At the same time I went ahead and swapped the 2x 1.8 ohm series part of the L-pad with Mundorf's of the same values.

This has the speakers sounding very balanced again but with more clarity. I have a number of thoughts about what is going on but first some photos.

DSC07397.JPG DSC07399.JPG DSC07401.JPG DSC07413.JPG

The new ribbon coils dwarf the air-core coils. But thankfully there was enough room. I used the same method to solder them in as I did with the tweeter coils, 1.5mm² solid-core wire with the end of the ribbon wrapped around the length of the wire. Also, once again I have not snipped off the ends of the resistor wires, choosing rather to bend them out of the way and leave them full length. This way if they get replaced at a later date they will still be useful for other projects and if they do stay I will still have full flexibility later if/when I make PCB's and want to migrate the parts to the new boards.

Cheers,
James
 
I have read so many times in the forums that a few tenths of an ohm here or there will not make a hearable difference in the speaker crossovers. Same for a few tenths of mH. And the quality of the parallel shunt components in the crossovers will have little impact on sound quality. And yet here I am. A few tenths of an ohm along with a change in conductor type has proven to be the difference in the speakers being balanced or not. As I said a couple of posts ago, my understanding of conventional wisdom is no longer the same. I am hearing that in reality 0.15 ohms is the knife edge between sounding OK and sounding great. A 0.04mH change in a coil value has made an improvement. And the change from wire air-core coils to ribbon foil coils in a parallel shunt position has proven to make a hearable improvement in SQ.

At the moment in some ways I feel so very stupid to have made so many mistakes and then again happy to have been able to start to learn something from them. In any case I feel that I have mad more progress in the last 6 months than in the last 5 years. At this point other than swapping the remaining Mox resistors in the mid section with Mills & Mundorf wirewound I will be simply enjoying the speakers for a while now and evaluate things. The last few weeks have been seriously fruitful.

One last thought for tonight. After 3 hours of listening at spirited volume levels, just for the heck of it, I reached into one of the crossover boxes and touched all of the resistors. And all of them were cold to the touch. So I don't think that I have any issues with resistor wattage. At least, I would think that they would have to get hot first before they could burn up. :)

Cheers,
James
 
I think your observations are true. I think we CAN hear small differences. A friend, who designs crossovers, did an experiment several years ago. He had two pair of identical speakers but in one he matched the crossover components between both in the pair to a fraction of a percent and then we switched back and forth between the highly matched and the pair with off the shelf tolerance components. They sounded the same in general but the imaging and clarity of details in the image was clearer in the pair with the very closely matched components. Transients seemed sharper and clearer as well. I was surprised. And no, I don't think he had the drivers matched in tolerance as much as the crossover components.
 
nerdorama said:
Hi James,
Your thread has enough length now that digging out your final configuration is pressing my ADD buttons. Any chance you might update the first post with links to the later results, i.e., which drivers you chose, the crossover schematic, any sims you did, etc.
Thanks,
John

And Merry Christmas
Click to expand...

Ok John, it took a while but I finally made good on the promise to make some sort of thread index. It has now been edited into the 1st post in the thread. Very simple, just a list of subject headers with post numbers following. Not perfect but hopefully at least a little helpful.

Cheers,
James
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom