6 minutes ago, andyr said:

Interesting, Con.  But just which of your new caps brought about this "dramatic deepening of the sound stage"?

 

The brown ones ... or the black ones?

 

And where did you buy these 'Miflex' caps - I've never heard of them! 

The brown ones went in as (substantially sized) bypass caps in the tweeter, and that was the only change I made. The black ones.. I haven't even put in at all as the Mundorfs are just around the corner. They were an "in-case" purchase while I was still undecided about what other caps to try. The miflex are available from hificollective but I found a secret supplier -TME - in Poland (where they're made) that sells them for about 2/3 the price! They're built similar to the PIO/Copper from Duelund except they use paper and poly as their dielectric and are rated to 630V instead of just 100V and cost a truckload less (but still a truckload.) They got a decent review at humblehomehifi, and I was curious to compare caps with similar technology. Miflex have been making nothing but caps for 60 years!

Edited August 12, 20187 yr by Ittaku

The Mundorfs I decided on were the silver gold in oil EVOs for the midrange, for completeness. I decided against the supremes as their main advantage is lower inductance which is pretty much irrelevant when there is a big fat inductor in the crossover going to a midrange already. I had bought the miflex black metallised polypropylenes actually for the parallel arm of my woofers by the way, though I ended up ordering Mundorf oil EVOs for there as well. It's very hard to get large enough values in extremely high quality for a midrange without extremely high price tags and this was the best compromise I could come up with - I have a rule that I don't spend more on the crossover components than the driver they go to as that makes no sense.

8 hours ago, Ittaku said:

Narrowing down on the design that fits all my criteria best and sounds the best now. Will probably build a more substantial version of this soon in pine for now. Will consider final timber once I confirm this one works well.

My comment on the baffle would be that the rear support should be as small as possible, to reduce reflections, otherwise I like the overall design.

 

What angle are you proposing for the front panel?

This angle may negate the following, it may not, a more vertical angle might be better(?).

 

Have a read of this http://soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf to get an idea of setting up the drivers to merge their individual sounds at the listening position.

Something to try before you build the new baffle with your current setup,  I found it an interesting exercise.

You will find just by listening where the drivers sound/s merge (if at all with multiple drivers in the mid/high area) and the sound may not be directed at the listening position, it was not in my situation.

 

I found I had to move my tweeter forward about 6 mm to get it and the mid range to "sound together" at the listening position.

My mid range is at ear level, directed at the listening position, not the tweeter, which is mounted as close as possible above the mid range and is movable in multiple directions.

If you are using DSP then you may be able to do that "movement (delay)" via software.

8 hours ago, Ittaku said:

Narrowing down on the design that fits all my criteria best and sounds the best now. Will probably build a more substantial version of this soon in pine for now. Will consider final timber once I confirm this one works well.

My comment on the baffle would be that the rear support should be as small as possible, to reduce reflections, otherwise I like the overall design.

 

What angle are you proposing for the front panel?

This angle may negate the following, it may not, a more vertical angle might be better(?).

 

Have a read of this http://soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf to get an idea of setting up the drivers to merge their individual sounds at the listening position.

Something to try before you build the new baffle with your current setup,  I found it an interesting exercise.

You will find just by listening where the drivers sound/s merge (if at all with multiple drivers in the mid/high area) and the sound may not be directed at the listening position, it was not in my situation.

 

I found I had to move my tweeter forward about 6 mm to get it and the mid range to "sound together" at the listening position.

My mid range is at ear level, directed at the listening position, not the tweeter, which is mounted as close as possible above the mid range and is movable in multiple directions.

If you are using DSP then you may be able to do that "movement (delay)" via software.

8 hours ago, Ittaku said:

Narrowing down on the design that fits all my criteria best and sounds the best now. Will probably build a more substantial version of this soon in pine for now. Will consider final timber once I confirm this one works well.

My comment on the baffle would be that the rear support should be as small as possible, to reduce reflections, otherwise I like the overall design.

 

What angle are you proposing for the front panel?

This angle may negate the following, it may not, a more vertical angle might be better(?).

 

Have a read of this http://soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf to get an idea of setting up the drivers to merge their individual sounds at the listening position.

Something to try before you build the new baffle with your current setup,  I found it an interesting exercise.

You will find just by listening where the drivers sound/s merge (if at all with multiple drivers in the mid/high area) and the sound may not be directed at the listening position, it was not in my situation.

 

I found I had to move my tweeter forward about 6 mm to get it and the mid range to "sound together" at the listening position.

My mid range is at ear level, directed at the listening position, not the tweeter, which is mounted as close as possible above the mid range and is movable in multiple directions.

If you are using DSP then you may be able to do that "movement (delay)" via software.

9 hours ago, Ittaku said:

Narrowing down on the design that fits all my criteria best and sounds the best now. Will probably build a more substantial version of this soon in pine for now. Will consider final timber once I confirm this one works well.

My comment on the baffle would be that the rear support should be as small as possible, to reduce reflections, otherwise I like the overall design.

 

What angle are you proposing for the front panel?

This angle may negate the following, it may not, a more vertical angle might be better(?).

 

Have a read of this http://soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf to get an idea of setting up the drivers to merge their individual sounds at the listening position.

Something to try before you build the new baffle with your current setup,  I found it an interesting exercise.

You will find just by listening where the drivers sound/s merge (if at all with multiple drivers in the mid/high area) and the sound may not be directed at the listening position, it was not in my situation.

 

I found I had to move my tweeter forward about 6 mm to get it and the mid range to "sound together" at the listening position.

My mid range is at ear level, directed at the listening position, not the tweeter, which is mounted as close as possible above the mid range and is movable in multiple directions.

If you are using DSP then you may be able to do that "movement (delay)" via software.

 

Edited August 12, 20187 yr by soundbyte

Haha you really really wanted to make that post, looks like you must have tried to commit your post a few times since it came up multiple times. Thanks for your comments.

 

Anyway the angle is 10 degrees. I experimented with tilting the speaker baffle up and down and came up with that angle based on the "direction" the soundstage took at my ears. If it was tilted too far up, the soundstage would start at tweeter height and go downwards as it goes backwards, and vice versa. 10 degrees had about the right staggering for it to remain horizontal (if that makes sense.) One issue I have with the existing baffle is it's too short, and the soundstage "starts" too low overall since it seems to extend from my tweeter's height. We tried to bring the drivers time aligned as much as possible, bearing in mind the arc from listener's ear, however bringing the mid and high frequency drivers as close together as possible ended up being prioritised to minimise lobing - since there are two midranges, the furthest distance from the lower mounted midrange to the tweeter ends up being quite a lot. I am not entertaining an MTM array because that really only works with an odd order crossover and I'm using 12dB throughout, and such a layout would make the speaker firstly much taller, and then any angle I have would have to be removed entirely, or have a staggered front baffle. Oh and I haven't read that article you linked yet, will do soon thanks.

 

I tried to limit the size of the rear support but I want it to be very strong. The large rear support is much smaller than the shelf I currently have but is extra wide because I'll be using it to mount the midrange and tweeter crossover components on top as well as they won't all fit in the base! The components are massive and heavy. I'll consider making it a little narrower again though but the biggest tweeter crossover component is almost 10cm wide.

 

The DSP I have is relatively automated and all the drivers are treated as a continuum since the crossover is passive so I can't do time delay between the separate main drivers, only between the main speakers and the subwoofer. I can't entertain the idea of going active and spend another 100k on equivalent amplification ?

 

Edited August 12, 20187 yr by Ittaku

Any particular reason for 12dB?

WOW, sorry for the multiple posts.

Problems of being in Birdsville with a slow internet connection as well as a slow computer I suppose!

MODS please feel free to remove the multiple posts.

15 minutes ago, Happy said:

Any particular reason for 12dB?

Fair question. There're actually quite a number of strong reasons I'm only using 12dB these days.

It allows perfect phase alignment at the crossover point (with one driver out of phase).

With a Linkwitz-riley alignment there is no summation at the crossover point leading to flat frequency response from that.

I find that the lower frequency you can run the higher frequency drivers, the better it sounds - the quality of high midrange from a tweeter is better and more omnidirectional than that from a midrange, and the quality of high bass from a midrange is better than a woofer and so on. However to do this, the slope needs to be steep enough to prevent the driver from being overloaded from excursion and 6dB is inadequate. 18 or more would seem to be okay then but the higher the crossover order, the more inaccurate they become as smaller and smaller inaccuracies in the passive components and drivers become magnified. Additionally 3rd order (or more) has two other major disadvantages (in my eyes at least) - one is that there will now be two passive components in serial with the drivers instead of one, and the magnitude of capacitance or inductance required starts precluding the use of extremely high quality components without exorbitant cost because they only ever come in smaller values, thus requiring lower quality caps, inductors with iron cores and so on which are all compromises.

 

Bear in mind these are issues more to do with it being a passive crossover and how to find the best compromise between all the requirements, but this is how I find I'm able to spend much more on amplification (instead of having 3 or more cheaper amps), and spare (almost) no expense on the crossover components I end up using whilst still sticking to my rule of never spending more on the crossover than the driver it goes to.

Ahhhh was under an impression u were doing active multi amping fair enough 

Nice!

How much difference in the off-axis response did you see from moving to a baffle that is narrow(er) vs. wavelength?

I would expect very large (improvement) depending on whether you needed to change the crossover/EQ filters in sympathy.

Depending on how high you operate the woofers, and assuming it’s low, (especially given your comment) then I don’t expect reflections from it will be a problem.

The new shape looks cool.

Santa arrived at the door with some gifts. Gonna have to rewire everything to make room with some more cable and remount the speaker terminals.

 

Won't be finished till tomorrow whilst I wait for the speaker terminal silicon I'm using as glue to seal.

26 minutes ago, Ittaku said:

Santa arrived at the door with some gifts. Gonna have to rewire everything to make room with some more cable and remount the speaker terminals.

 

Won't be finished till tomorrow whilst I wait for the speaker terminal silicon I'm using as glue to seal.

WOW!!  Cool! 

 

Andy

 

17 hours ago, andyr said:

Interesting, Con.  But just which of your new caps brought about this "dramatic deepening of the sound stage"?

 

The brown ones ... or the black ones?

 

And where did you buy these 'Miflex' caps - I've never heard of them! 

As a follow up before changing the crossover dramatically, after 3 days of continual burn in and/or playing of music, the miflex going to the tweeters had other noticeable improvements too. The top end seemed to just "clean up" with it getting quieter backgrounds, it was slightly sweeter and had the added effect of not drawing attention to the treble, "blending it in" with the harmonic structure of every note better. I'm sure when I can get the big ones that it will be better again. Warehouse at the moment says it's normally 9 weeks when on back order though ? I've pinged them directly to see what they say though; hopefully they were planning to restock them anyway. Fortunately I have plenty of toys to play with anyway in the meantime

Edited August 13, 20187 yr by Ittaku

Mundorfs were burning in with white noise for a bit over half a day before I finally cranked some music through them. The bulk of the change is in the midrange for reference. So far I'm not exactly blown away. They have a little more detail and add some more information such as the sound of the violin box itself, and the core sound of cymbals and brass, and they seem to have improved image focus slightly, but it's all a bit too much on the cool and slightly analytical side for me at the moment. With so few hours on them it would be unfair to make a call on their final sound, and other reviews talk about a long burn-in time of at least 50 hours. They certainly have promise if they get a little warmer and can keep the extra information. I'm also considering adding a bypass cap to the midrange, using another miflex copper PIO cap at about .47uF, however the amount of musical energy that will likely get through to the midranges on such a small cap is probably going to be negligible. However I'll hold off till the Mundorfs fully burn in and maybe order them whenever the big values come in to minimise transport costs. One thing that's striking about the Mundorf EVO (non-supreme) range is how small the caps are! I've gotten used to caps getting heavier and bulkier when I've gone up in quality, but these are very light and compact for their respective values. If weight was a good predictor of quality, these would fail such a test (nonsense of course I know.) I guess the fact is I spent a lot of time (and money) trying to use film and foil as much as possible since they always sounded better than any metallised films I'd tried to date so this is a very different direction for me.

On 13/08/2018 at 1:52 PM, davewantsmoore said:

then I don’t expect reflections from it will be a problem.

I thought the same about H frame subs, I could hear reflections from the back wall, that's why I now have them in a V frame.

Hopefully that is not the case here, but you never know.

Now have 30 hours or so of burn in on the Mundorfs, and the sound is delightful. They are still slightly on the cool side, but that gap seems to be diminishing with every passing hour so I don't believe it will be a problem. The extra information in the instruments, filling in the sound of the resonant cavities of string instruments and the volume of leading edge on brass is something, but there's added focus to the sound as well now. It's simply a richer midrange now without it actually being an artificially warm and smoothed midrange. Great buy. I'm tempted to buy some for the tweeter to fill in the gap until the big Miflexs become available. Hrm. Wish I'd thought of that before to avoid me creating yet another order and its attendant delivery costs... The problem is I'll be tempted to buy their highest range instead since they come in those values and for a temporising measure it's rather an expensive venture. In fact, I'd need to get supremes for the tweeter since their advantage is non-inductive wiring over the EVOs, and supreme silver in oils are not much less expensive than the Miflex copper PIOs... Okay cancel that idea. I just need to be patient.

Edited August 15, 20187 yr by Ittaku

Good work with the caps. Its fascinating with DIY how you can tune the sound to exactly what you want.

Edited August 16, 20187 yr by Nada

On 14/08/2018 at 6:03 PM, soundbyte said:

I could hear reflections from the back wall

Heh, we've had this conversation before (while I was listening to it) 

 

 

You couldn't hear the "reflections" themselves....  your hearing just doesn't work like that.    BUT, having them like you did before vs now, will change the overall response quite a lot - and that could be audible.

 

Semantics, perhaps... but the point is that Con won't hear a reflection off it assuming the LPF is low enough .... but if he makes changes to it, it may cause the response of the woofer itself to change  (including the polar response and direct/reverb balance) .... so changing the support and hearing a difference != hearing reflections.

Contrary to my expectations, the sound being on the cool side from the midrange did not go away entirely. The caps probably have about 100 hours on them now and sound incredible, but the sound is still too cool for my liking. I've not heard polypropylene caps change dramatically after the first 100 hours so I suspect that's their final sonic signature. I'll definitely be grabbing some miflex bypass caps and see how that goes at tempering it. I'll do some experiments first with caps I already have to see if small value bypass caps in the midrange crossover actually make an audible difference at all since it's a bandpass filter, unlike the tweeters.

This has become more of a blog lately. Anyway I tried putting in some 0.1uF "bypass" caps - the best I had available which were multicap RTX polystyrene and tin foil - on the midrange Mundorfs, and there is a slight tempering of the cool sound that the Mundorfs give. However it's only a very subtle improvement and it's not like it's gone away. The Mundorfs amount to 80uF so it is clearly not going to be bypassing very much music information, and I was sceptical in the first place. I hope with a higher quality copper PIO and larger value it will be more substantial so I may just order myself some 0.47uF Miflexes to put in there.  I was thinking I'd like to try their other range as well, they have copper foil polypropylene caps as well, but only up to 1uF so I might get both and try them. They're actually slightly more expensive than the equivalent sized copper PIO ones which says something (not sure what, just something) but they do have better tolerance at only 2%.

Frustration with the lingering cold sound from the Mundorfs made me do a swag of experiments switching caps around. First I tried the much cheaper Miflex metallised polypropylene (in oil) caps in their place. This was disappointing because they sounded a bit woolly compared to the Mundorfs and lost that detail, focus, and leading edge info and some of the air as well, BUT they were much warmer sounding. I then tried bypassing the Miflex with a decent sized film and foil and the tonality remained warm but the air returned and focus improved! They still sounded woolly though. So I tried using only the one Mundorf 68uF and lots of film and foil bypassing it and then it was a mess - it had the extra detail and leading edge but was both cold and lost the focus though this is probably a function of just too many similar sized caps in parallel which I've experienced before. Either way, the cold sound returned every time I put the Mundorf back in, in whatever capacity. No amount of bypassing seems to be able to take that coldness away where I'm using them. So I think I'm going to retire the Mundorfs after my brief stint with them to the midrange; they simply didn't work the way I wanted them to in that position. The more I listen to the big miflexes in the main position the better they're sounding to me as well. That's pretty sad considering they're about 1/10th the cost of the Mundorfs, but interaction with the rest of the system may be the reason so I can't say openly they're bad or anything like that. Maybe with an overly warm sounding system they may temper it. Bear in mind these are the Mundorf EVO silver gold in oil, and not the supremes which are said to have a nicer sonic signature, but getting 80uF out of supremes is hella expensive, and I'm a little put off the Mundorfs now.

 

So what I'm going to do now is order some more miflex metallised in the proper sizes to be able to bypass them without getting to size differences that will resonate, and bypass it with the miflex copper PIO when they arrive. This means I now have over a thousand dollars worth of caps sitting on the shelf just for experimentation. Helpful I guess but seems wasteful.

4 minutes ago, Ittaku said:

Frustration with the lingering cold sound from the Mundorfs made me do a swag of experiments switching caps around. First I tried the much cheaper Miflex metallised polypropylene (in oil) caps in their place. This was disappointing because they sounded a bit woolly compared to the Mundorfs and lost that detail, focus, and leading edge info and some of the air as well, BUT they were much warmer sounding. I then tried bypassing the Miflex with a decent sized film and foil and the tonality remained warm but the air returned and focus improved! They still sounded woolly though. So I tried using only the one Mundorf 68uF and lots of film and foil bypassing it and then it was a mess - it had the extra detail and leading edge but was both cold and lost the focus though this is probably a function of just too many similar sized caps in parallel which I've experienced before. Either way, the cold sound returned every time I put the Mundorf back in, in whatever capacity. No amount of bypassing seems to be able to take that coldness away where I'm using them. So I think I'm going to retire the Mundorfs after my brief stint with them to the midrange; they simply didn't work the way I wanted them to in that position. The more I listen to the big miflexes in the main position the better they're sounding to me as well. That's pretty sad considering they're about 1/10th the cost of the Mundorfs, but interaction with the rest of the system may be the reason so I can't say openly they're bad or anything like that. Maybe with an overly warm sounding system they may temper it. Bear in mind these are the Mundorf EVO silver gold in oil, and not the supremes which are said to have a nicer sonic signature, but getting 80uF out of supremes is hella expensive, and I'm a little put off the Mundorfs now.

 

So what I'm going to do now is order some more miflex metallised in the proper sizes to be able to bypass them without getting to size differences that will resonate, and bypass it with the miflex copper PIO when they arrive. This means I now have over a thousand dollars worth of caps sitting on the shelf just for experimentation. Helpful I guess but seems wasteful.

 

Very interesting blog, though, Con! 

 

Real shame about the Mundorfs ... but would you like to try some Jupiter 'copper foil, paper and wax' 600v caps for bypasses?  I have a stash of 1uF that I can lend you, if you want to experiment (they're 34m diam & 55mm long).

 

I was interested to read your comment "this is probably a function of just too many similar sized caps in parallel".  I had always thought it was a good idea to use multiple same-value caps, rather than a biggie and a littlie ... but you're saying I thought wrong?

 

Andy

 

 

 

4 minutes ago, andyr said:

Real shame about the Mundorfs ... but would you like to try some Jupiter 'copper foil, paper and wax' 600v caps for bypasses?  I have a stash of 1uF that I can lend you, if you want to experiment (they're 34m diam & 55mm long).

 

I was interested to read your comment "this is probably a function of just too many similar sized caps in parallel".  I had always thought it was a good idea to use multiple same-value caps, rather than a biggie and a littlie ... but you're saying I thought wrong?

Thanks. That sounds like fun, the copper foil in paper ones I'm currently using do amazing things to the tweeter just as a bypass.

 

I think it's both the same size that's a problem, as is more than 10x size difference. I can't find any reference conclusive about the former, but bear in mind the caps I'm using are self-bypassing caps (10 internal) so it's like I've got about 30 caps in parallel!

Edited August 25, 20187 yr by Ittaku