Wavecor Ardent Journal - the first Builds

[meb46]

Sweet, great effort(s) lads.

[BobEllis]

Nice, Ben. As I first scrolled down to your veneer it made me think of an H.R. Giger painting. Of course, I recently watched Alien...

[Evil Twin]

I got one completely veneered and ready for finishing. I'm pretty happy with how the veneering turned out. On the next one I may try to do the facets first then put on the front piece.

Clearly the Force is strong with this one...

[JonMarsh]

What is the purpose of the lead?

Two fold- it adds mass, changing the resonant frequency of the panel system, and lead is also very lossy, so it provides damping. It's fairly easy to make enclosure walls stiff relative to bass frequencies - but that pushes the resonance up well into the midrange, and the stiffer you try to make a wood wall, the higher the Q of the ring. The lead knocks that down quite a bit in frequency and q and amplitude. It's not fun to work with, and one needs to be careful not to release particles or dust, and to clean up carefully when handling it. Best to wear nitrile gloves while working with it.

And of course, it's probably a Romper Room No-No in the EU, ROHS and all. :W But trust me, lead is much less toxic than reactor grade Thorium... which is pyrophoric and requires special handling. :W That's why I use a Mr. Fusion in my flying DeLorean instead of a compact reactor designed by Taylor Wilson.

[dar47]

Nice, Ben. As I first scrolled down to your veneer it made me think of an H.R. Giger painting. Of course, I recently watched Alien...

That's so true. :rofl: I keep thinking his poor soon to be wife isn't going to be able to put anything in the room that will even come close to pulling your attention away from these.

[JonMarsh]

That's so true. :rofl: I keep thinking his poor soon to be wife isn't going to be able to put anything in the room that will even come close to pulling your attention away from these.

Now that's an interesting way of putting it!

[TEK]

I got one completely veneered and ready for finishing. I'm pretty happy with how the veneering turned out. On the next one I may try to do the facets first then put on the front piece.

Wow, those sure look nice!
And as Jon mentioned, they seems "just right"

[TEK]

Two fold- it adds mass, changing the resonant frequency of the panel system, and lead is also very lossy, so it provides damping. It's fairly easy to make enclosure walls stiff relative to bass frequencies - but that pushes the resonance up well into the midrange, and the stiffer you try to make a wood wall, the higher the Q of the ring. The lead knocks that down quite a bit in

If there is no electrical shilding reason for using lead, do you think asphalt plates, i.e. 3mm thick, will do the same job with the damping?

[meb46]

TEK - I have used bitumen pads before... same as what the Car Audio guys use to take the rattles and noise out of metal panels. The Bitumen pads I used were just an adhesive Bitumen sheet that you use Contact Adhesive to secure to the internal walls. It works a treat. A crude knuckle test will yield some good results

[JonMarsh]

The bitumen pads will work fairly well- but bitumen with lead will work better. But is it necessary? Audible? Harder to say.

For some of us, if some is good, more is better, and too much is just enough... :W

Note that many products combine a foam material with thin lead sheet as a damper, like VB8, for example. This is widely used in automotive for sound damping.

[JonMarsh]

Curious-





Are you going to go for a gloss or semi-gloss/satin finish on these?

[TEK]

The bitumen pads will work fairly well- but bitumen with lead will work better. But is it necessary? Audible? Harder to say.

For some of us, if some is good, more is better, and too much is just enough... :W

Note that many products combine a foam material with thin lead sheet as a damper, like VB8, for example. This is widely used in automotive for sound damping.

Think I will have to keep the cost a little bit under control, and it so happends that I have quite a bit of this laying around:

5mm thick, 7,5kg/m2. Guess I will use it instead of bying something else.

[GPHammer]

Darrell, you and Ben have done an amazing job, Ben's veneer looks incredible, I look forward to coming by and listening to them.

Gregg

[dar47]

Curious-





Are you going to go for a gloss or semi-gloss/satin finish on these?

I think we were both thinking the same thing but I may do mine manually, build a finish up with high gloss (no flatteners) to get the best look through the finish. The last couple of coats with satin that can be rubbed out without going through to the high gloss. Ben may use some tinting in his finish to create an even tone and highlight the burl. He is taking samples to have them sprayed. The one finish I'm curious a bought is the conversion varnish, this can be none poly based until the top coats. This would give a clear look through as well as the toughness of poly. My last go around with poly created some delamination that I still have to fix on my center so.

A little story of why poly is good on the top, a few years back we hosted a staff Christmas party for my wife's school where she teaches. Her VP put her drink on one of my towers, I picked it up and put it on a coaster on the coffee table while she watched. Later she put her drink back on the same speaker and this time I handed it to her and smiled. Again later she put her drink on one of my surrounds and this time I handed it to her and said "cups don't go on speakers" and smiled. I thought this looks a little passive aggressive and I felt like "drifting her".

[dar47]

Sure thing Greg, we have to do a get together with the locals as I think your all going to love the Masters sound! Sorry I didn't hook up for the CNC work, it went fast and furious. ops:

[JonMarsh]

Think I will have to keep the cost a little bit under control, and it so happends that I have quite a bit of this laying around:
[ATTACH=CONFIG]22986[/ATTACH]
5mm thick, 7,5kg/m2. Guess I will use it instead of bying something else.

As Steven Stills said, love the one you're with...

[benthe8track]

Curious-





Are you going to go for a gloss or semi-gloss/satin finish on these?

Semi-gloss I think. We didn't really get into it I just said make them look like this and well see what they come up with:

[JonMarsh]

I think we were both thinking the same thing but I may do mine manually, build a finish up with high gloss (no flatteners) to get the best look through the finish. The last couple of coats with satin that can be rubbed out without going through to the high gloss. Ben may use some tinting in his finish to create an even tone and highlight the burl. He is taking samples to have them sprayed. The one finish I'm curious a bought is the conversion varnish, this can be none poly based until the top coats. This would give a clear look through as well as the toughness of poly. My last go arround with poly created some delamination that I still have to fix on my center so.

A little story of why poly is good on the top, a few years back we hosted a staff Christmas party for my wife's school where she teaches. Her VP put her drink on one of my towers, I picked it up and put it on a coaster on the coffee table while she watched. Later she put her drink back on the same speaker and this time I handed it to her and smiled. Again later she put her drink on one of my surrounds and this time I handed it to her and said "cups don't go on speakers" and smiled. I thought this looks a little passive aggressive and I felt like "drifting her".

Geez, a bit slow on the uptake, wasn't she? The Isiris are poly, and I'm actually pondering sanding them down and re-doing them depending on how the lacquer turns out.

I believe you have the right idea, multiple coats of gloss with just a couple of satin or semi-gloss is my plan.

now I'm dying to unroll my veneer and see what I've got to work with- just haven't figured out where I can lay down a couple of 4'x8' sheets!

[JonMarsh]

Semi-gloss I think. We didn't really get into it I just said make them look like this and well see what they come up with:

Yeah, that's a good target!

[knowledgebass]

What glues do you guys use for applying the veneer? Jon, you hinted at an epoxy top coat, is that also your glue of choice for the veneer?

[JonMarsh]

Actually, what we're both doing here (for the most part) is using veneer with 3M PSA pre-applied (Pressure Sensitive Adhesive). I'm using a more conventional hot PVA process for the facets, or Titebond veneer glue with solid wood veneer instead of backed for that part. That way I think I'll be better able to hide the veneer seams on the edge- this worked very well with the first set of Ardents.

In some corners PSA gets a bad rep, but how well it works really comes down to surface preparation and application- the epoxy is used for surface filling and to get a nice smooth flat NON POROUS surface for the PSA to adhere to. Putting PSA directly on wood gives mixed results, in my opinion. It's typically recommended to at least shellac a surface before using PSA veneer. Using epoxy is just a way to get more fill and durability and a harder working surface for the PSA adhesive. It should be sanded smooth after application and inspection.

It's important to get alignment just right before actually applying PSA backed veneer to a surface- repositioning is NOT possible.

Note- I always use a veneer scraper with PSA veneer, just as with contact cement processes- it assures a high contact pressure for the bond and contributes to durability.

(note, I do NOT represent myself as any kind of expert woodworker, just a wires and sparks guy who's learned enough stuff the hard way to have a bit of a handle on what works and what turns into a disaster at the least provocation...)

Check out Joewoodworker.com for a lot of helpful info.

[benthe8track]

Yeah I did the same thing Jon just described but I used the raw on the back and PSA veneer on the facets. Managing the glue/paper lines is a bit tricky.

[knowledgebass]

Thanks gentlemen. Heading over to my local saw mill this morning. I was planning to use raw veneer and UF, but I'll ask if they have anything affordable with PSA. Also considering hammer veneering, but I'd need to get another tool (and more glue).

[cjd]

Cold press with cauls/platens isn't bad at all, just takes a little planning. The top is probably the worst part in the process. I imagine one day I'll get a vacuum bag setup and wonder why I didn't do so sooner!

I had to go to iron-on for the facets on the Nebbiolo, something I could have avoided with bags. Since I was book-matching the face it was a little tricky. The tops (which required a little iron cleanup) ended up with a small gap between the pieces. Oh well - first time, can't complain.

C

[JonMarsh]

Very busy weekend, but distracted by many other tasks- such as moving organ, unpacking and setting up Leslie 3300 bought on eBay, helping GF with some storage issues.

And discovering Sunday AM I'd forgotten to bring the temperature controlled soldering station over along with the other tools last weekend! GF went above and beyond the call of duty, making a run to Radio Shack to buy their clone of a Weller soldering gun... which wasn't a very good clone, but was good enough to strip the insulation on the Cardas SE-11 wire and do some soldering and mounting of drivers.



I used Private Messages stainless steel philips head sheet metal screws for mounting drivers; 3/4" for the tweeter, 1" for the mid, and 1-1/2" for the woofers.









Major stumbling block #2, after getting everything wired up for one cabinet, discovering I'd also left my HP scope at home, and had no way to set reference levels for driver testing (so that results would be directly comparable with all other testing I do). Normally 4V P, which is ~2.83VRMS; adjusted when I want to test distortion at higher levels.




Well, next Friday I managed to squeak out a PTO day, originally for Apple shareholder meeting and getting my windshield replaced in my Element; with the timing necessary for that appointment, it looks more likely it will be Wavecor Ardent testing and getting the windshield replaced!

[5th element]

If you've got a multimeter that should be good enough for setting the level. Even crappy, non true-RMS versions are pretty good at getting 50Hz right. It's annoying when your excitement gets curbed by such silly oversights.

[JonMarsh]

Alas, I had my really good portable LCR meter at GF's but no multi-meter- the good Fluke one was at home, too. Right next to the scope... which I forgot to load up this AM when I left my home. Sigh... Well, I'll get it together eventually! Maybe Friday morning I'll just glue up the other baffle...

[sdl2112]

Haven't visited in a while...what progress!...looks great everyone :T

[JonMarsh]

Though I don't have new data yet, I've been thinking about the design and reviewing and revising the concept test crossover. Mostly this has meant nudging the LF to MF crossover a little higher in frequency, and altering the LF voicing, and double checking and optimizing the reverse null behavior. (we don't care about flat response so much, many ways to get that at one measurement point (and not at others) but we just like to see really deep nulls in the reverse wired response... indicates high degree of phase tracking in crossover region).





Smaller component values, less expensive crossover, less volume in cabinet.

Hmmmm, nice reverse nulls.




Smoothed out the bottom end and brought up the mid and tweeter level.




Impedance doesn't look bad, either... nothing out of sorts, and pretty smooth.

[dar47]

Your doing great especially with all the irons in the fire your trying to deal with. It's tough to remember stuff when your not working at your usual work space, I know I did when we were at the factory down town. Jon don't feel rushed on our account, Ben's working on getting his finish finalized with samples while doing a new school term and planning a move scheduled for this week. My veneer isn't getting here till maybe Thursday, so you have plenty of time to tweak things to your hearts content. I know those cabinets must be tugging at you, Man those drivers look sweat mounted in that baffle. :T

That impedance is looking like an easy load so far.

[JonMarsh]

I know those cabinets must be tugging at you, Man those drivers look sweat mounted in that baffle. :T

That impedance is looking like an easy load so far.

Thanks for the kind words and understanding. But I really want to get the measurements and crossovers moving along, so I can get to the finishing stage and possibly have these ready for the 2014 Northern CA DIY meeting in early May. Yes, with all the other things going on, it's a bit crazy... we're slowly getting things figured out for what of my stuff will go where over at GF's.

Yeah, the impedance curve is another "indicator" that things are playing out correctly- hopefully it will stay in this range, but it IS looking quite sweet, and MUCH nicer to drive than an Indra...




Especially considering the punishing LF impedance and phase shift!

With this smooth an impedance curve, there should be less interaction with the output impedance of an amplifier, such as a zero feedback solid state design or low feedback tube design. That's not a critical factor, IMO, but still desirable.

[5th element]

A tip that might interest you Jon, that little blip that you've got in the phase at around 3500Hz is obviously an anomaly and these, as I'm sure you've noticed, crop up from time to time. If you open the frequency response file in a text editor you should be able to see the phase and magnitude data points vs frequency quite easily. If you scroll down and find the correct frequency it should be possible to manually edit a couple of entries so that the data points for the phase are in line with what you'd expect and thus eliminate the problem. I mean it's not like it's a huge issue, but it only takes a minute or two to fix.

[JonMarsh]

A tip that might interest you Jon, that little blip that you've got in the phase at around 3500Hz is obviously an anomaly and these, as I'm sure you've noticed, crop up from time to time. If you open the frequency response file in a text editor you should be able to see the phase and magnitude data points vs frequency quite easily. If you scroll down and find the correct frequency it should be possible to manually edit a couple of entries so that the data points for the phase are in line with what you'd expect and thus eliminate the problem. I mean it's not like it's a huge issue, but it only takes a minute or two to fix.

You're on the right track, but the data is currently good, even after manual editing and spacing alignment of the FRD files this was the best I got (should have seen it before I massaged the data a bit...) it's more an issue with having a CRAP-TON of data points from Fuzzmeasure and it kind of spazzes out LspCAD. Sometimes I have to manually clean up alignment of the FRD files just to avoid more of these artifacts. Since this is just a "concept test" I'm not too concerned about that at this point.

What makes things easier for LspCAD but which I sometimes forget to do is use a shorter gating window that cuts down on the bandwidth resolution and minimum LF point- particularly for the midrange and tweeter- this lowers the number of data points, as the LF resolution directly ties in to the data density at higher frequencies. I'll be doing that fairly carefully for the measurements for checking the final design this weekend.

Because Fuzzmeasure and Praxis can both get long gating measurements with reasonable control of reflections via Half Hamming or Half Hanning windows, I tend to use that when measuring drivers just to understand the drivers and enclosures- but for collecting the data for designing the crossover, I should use a different approach with a minimum length gating window to support the needed bandwidth without killing resolution too much.

[JonMarsh]

Making progress

Busy day yesterday - a scheduled day off work- got lots of errands done, including a new windshield for the Element.

I also glued the second front baffle up in the AM before heading out for all those necessary errands...





This morning we moved things around in the family room and got setup for the driver measurements. That seems to alway have minor glitches with sorting out th signal path on the TC Electronic Konnekt 8, but nothing that I can't eventually figure out. One of the advantages of bachelor life is being able to leave things just setup- can't do that here! :W


Things look pretty much as I expected...

SB223 Wavecor woofers wired in series



Note that the breakup mode is at about 3.7 kHz, and that even with HD3 resonance amplification of distortion products, the distortion behavior for a 3 way will be quite nice even with a 400 Hz crossover point.

It's also interesting how similar in level HD2 and HD3 levels are - I'd say the efforts for the balanced drive motor design have paid off,



C79 midrange




SS 6640 tweeter




This measurements were done more for getting clean data for crossover design, not investigating the drivers per se.


I didn't have time to setup with a lot of additional wall damping and synthetic comforters draped to knock down reflections, so there's a bit more "ripples" in the measured response than usual, but I don't think this will impact checking / developing the crossover design. Just won't look quite as idealized and pretty, unless I use more smoothing of the data.

Since the weather is sort of decent at the moment, the rest of the afternoon will be cabinet sanding!

[benthe8track]

Awesome! Glad there weren't any surprises

[JonMarsh]

Awesome! Glad there weren't any surprises

Me too! Working on sanding now, taking advantage of the day's weather- am close to finish pre-sanding of the unmounted cabinet. Tomorrow I'll do impedance measurements in the AM, I downloaded the program for woofer tester 2 to use under parallels.

[benthe8track]

Me too! Working on sanding now, taking advantage of the day's weather- am close to finish pre-sanding of the unmounted cabinet. Tomorrow I'll do impedance measurements in the AM, I downloaded the program for woofer tester 2 to use under parallels.

Must be nice. We are hitting -50 again, this is the worst winter I can remember.
Yeah there is no shortage of sanding, I'm glad that portion is finally over. Also hopefully my samples will be finished on Monday and I can get the cabs in for finishing next week.

[JonMarsh]

That is brutal weather, even for you guys!

Well, I got the impedance measurements done this afternoon, no surprises, either, but the woofer looks quite nice in the operating range - mind you, this is the two wired in series as we're using it.



Fb in the low 30's, and a nice flat impedance curve in the midrange where we're using it! :T

[5th element]

Eminently usable. And no sign of any resonances up until its primary cone resonance well out of band.

[JonMarsh]

Some more progress yesterday evening...

This is a preliminary design with the new data, I'm going to look through my part drawers and see what I've got on hand to put together a test configuration- might be able to cover a lot of it, we'll see. It's worth a trip to the storage unit. Don't go buying parts off of this, there will be a detailed BOM when I'm ready to say ready! :W


Nothing should be very surprising here, but of course it's using a higher crossover point than the Isiris for the LF to Mid.




A reminder, no diffraction treatment on the cabinet for these initial measurements, I'm used to seeing certain bumps in the tweeter response as a result and don't try to fix those.



For reference, here's the tweeter on a test baffle without the diffraction issues:




Impedance has a bit of a dip in the 200-350 Hz range due to some peaking entered in to the crossover based on slight compensation for the in room measurements- that may not really be needed, but some more extensive measurements would be required to know for certain...

I didn't take any special measures for the woofer breakup mode at 3.7 kHz; it's about 40 dB down, and that seems like enough- subject to change, would be the addition of a trap cap across L4 to create a notch filter. Votes, anyone?



And for reference, reverse polarity on the mid, which shows some fairly nice notches due to heavy phase cancellation out of phase.



I think we're heading in the right direction.

Oh, and I found pictures of my previous felt construction for the Ardent; it doesn't show clearly, but the tweeter cut out is a vertically oriented diamond slit, as I did for the old M8a book shelf, which worked well in that case, too. I'll make a new test piece, or one to fit in the supplied parts from Ben, and detail it in a drawing after testing.

[5th element]

The only thing that really bothers me is that hump in the tweeters response at 5k, no doubt mostly deffraction related, but it's presence is causing you to push down the rest of the tweeters response to prevent the 5k area from becoming too hot. Also the mid looks like it's running a tad hot too, no doubt to help with the upper end issues.

This is where the felt comes in. If it can smooth out the 5k problem then your issues go away and things will look a lot better. If you aren't going to do this then maybe a traditional lpad on the tweeter (perhaps giving C4 it's proper response shaping around FC), or going 4th order will give you enough attenuation to put a reverse knee around 5k and squash out the hump. I'd have figured that you might have done this already, but if you're banking on the felt to cure the hump, then I can see why you might have ignored it. The issue looks very similar to what I encountered in my last design. 2nd order electrical provided enough attentuation to meet the target slope, but didn't have enough response shaping abilities to pull down the hump around FC. Going third order solved everything very nicely.

[JonMarsh]

My experience with the last Ardent cabinets and the 6640 showed that the hump was "curable" with the felt treatment. Have to confirm this, but for now working under that assumption. Since this is a 3rd order design, there is some flexibility for tweaking. I'm starting with additional focus on the woofer- I have some additional measurements made at different heights, and will take a look at averaging those and see what comes out closer to near field measurements.

Still, it's progress, things are pointing in the right direction, and I'm used to seeing LspCAD sims that look like the above, but with measurements of the finished crossover in system that look like this:




With a longish gate, no excessive smoothing, not perfect, not an Isiris like plot, but I think we can even improve on it this time around.

[dar47]

Great progress, I'm liking how the woof plays nice up top. I glad I resisted doing anything other then sanding the burn edge off the felt, we will wait for your drawing and solution.:T

I'm a little bummed, as I'm off till Tuesday and still no veneer. 2 days USPS to get it to Canada but 5+ and Canada post can't get it here.:M

[JonMarsh]

Great progress, I'm liking how the woof plays nice up top. I glad I resisted doing anything other then sanding the burn edge off the felt, we will wait for your drawing and solution.:T

I'm a little bummed, as I'm off till Tuesday and still no veneer. 2 days USPS to get it to Canada but 5+ and Canada post can't get it here.:M

That's a REAL drag- available time, but missing critical ingredients! I hate when that happens!

Well, I have some time on my hands while doing laundry today, so more simulation work and pondering the design- think I've made some good progress. I did go back and use some additional measurement points lower and higher, and created an average LF woofer response- usually only something I do for balancing the power response of the tweeter on and off axis, but it seems applicable here, as I couldn't get the measurement distance as far away as I prefer (1-1/2 to 2 meters- was just using the "classic" 1 meter yesterday).

Also, based on 5th Element's inputs, I went back and looked at the more conventional tweeter topology, which I don't use so often because the Isiris configuration gives some input impedance advantages- but in this case the conventional approach just may work out better, since the tweeter Fs is so low and isn't interacting with the network. Also, I tried out adding in the notch element for the woofer resonance, C10, and while it could be left out, I'd go ahead and put it in- 2.0uf will do, but 2.1 (2 + 0.1) will tune it more accurately.

So, this is the net results after some tweaks, but with less smoothing on the imported data, so it won't look as pretty in theory (a reminder, this is essentially a modified LR3, an alignment which is not normally used, but can be very useful with the time delay in real box systems; as used first in Modula MT and NatlieP):





The simulated response; could possibly argue for bringing down the Tweeter by a dB or so- but need to be measuring and listening to make an assessment:



The reverse polarity on mid response, checking for quality of reverse nulls:



Impedance:



This I'll check the parts bin for soon, (have most of the tweeter parts already, I know...) and see what I can raid from an older Isiris Crossover for a test build.

[JonMarsh]

Pair of woofer crossover parts ordered. Tweeter crossover parts already on the way; will see if I have enough parts bin stuff to modify an Isiris Midrange crossover to this spec for test next weekend. Ordered the updated inductor and resistors for tweeter crossovers this AM.

In the semiconductor business, this is what we would call a hot lot- (wafer lot- getting priority in the fab) but there's no wafers here, just pricey caps, inductors, and resistors!

It's going to be a challenge doing crossover work and finishing... but for now I'll keep one cabinet back for driver testing, and then hopefully catch it up when I get to the spraying stage. Was taking a look at GF's bench grinder, for trimming down the trim bits- I think we need to talk about an upgrade! her's is a museum piece, and not in the best of shape- don't know that the stone will get the job done. I suppose I should give it a try first..