Ardent Speaker Camp

[ColoradoTom]

Although I have never heard what it takes to cut these facets... maybe 3-1/8 is enough? I wonder if ET remembers how much blade was exposed during his force push Ardent cuts. I would expect in the pending completion webpage on the Ardents the topic will be discussed further.

When I refinished my M8ta's (for the first time - second time is pending) my facets had a maximum depth of 4 1/2 inches. I was able to do this using a Festool plunge saw using the included saw guide - which turns out to be pretty accurate method. I originally used a table saw with a jig similar to what Jon is using for the Ardents and I was able to get upwards of 3 1/2 - 3 3/4 inches of depth. Doing this on a table saw will give you one of those "gut check" moments because most people will be running their saws at absolute maximum blade exposure and while 3 1/2 half inches doesn't sound like much, I did have a bit of a feeling like Richard Dreyfuss staring into the mouth of Jaws. Anybody cutting facets like this really needs to make sure they are COMPLETELY focused on what they are doing!!

Colorado "still have all my fingers to play my guitars" Tom

[Hdale85]

Yes I agree Tom, I'm still wondering if I can do this on the CNC mill though. I think it may be possible. My spindle should have ~6" of travel. Who knows though time will tell.

[JonW]

[QUOTE=savage25xtreme;n611590]

Although I have never heard what it takes to cut these facets... maybe 3-1/8 is enough? I wonder if ET remembers how much blade was exposed during his force push Ardent cuts. I would expect in the pending completion webpage on the Ardents th

With regard to cutting large facets, we had a nice discussion, a while ago, in this thread:

https://www.htguide.com/forum/showthread.php?t=28981

I found it to be very helpful.

Obviously there has been another penetration of our security. I must have a little chat with Captain Needa again....


It is actually a bit more than a "color" or paint; it can be tuned to actively absorb all incident EM or vibrational radiation in tunable bands; applied to significant room areas, the tuning capability permits fine adjustment of room acoustics as well as achieving a suitable background color for deep meditation.

Very nice. It kind of reminds me of how laundry detergent can make your clothes “whiter than white.” What we have here is an alchemistic transformation to “blacker than black.” :W

[ColoradoTom]

Yes I agree Tom, I'm still wondering if I can do this on the CNC mill though. I think it may be possible. My spindle should have ~6" of travel. Who knows though time will tell.

CNC would be really cool..... If I didn't have two kids getting near the college years I might be tempted to look at some of the smaller ones. Creating laminated Ardent with curved Magico-like sides would be really fun.

Tom

[Evil Twin]

Is 4" of cutting depth enough to cut the facets in the Ardents? This of course if I can't do it with the CNC

Recall that the BT3100 cuts shown earlier are 3-5/8" depth, with actual cut required being about 3-7/8; a small hand cut is needed to finish. This is accomplished by only assembling and cutting the upper two layers of the panel at first, then gluing to the bottom (inner most) layer.



There is nothing sacred about this process; it's just what worked in the Imperial skunk works labs as currently equipped.

[Hdale85]

CNC would be really cool..... If I didn't have two kids getting near the college years I might be tempted to look at some of the smaller ones. Creating laminated Ardent with curved Magico-like sides would be really fun.

Tom

Well I'll be building one, and with a cheaper Chinese spindle (~5hp) I'm looking at spending about 4k total for an 8'x4' machine. Might be a 4'x4' depends. But it will be capable of decent speed and be accurate to something like .005 mm I believe is what is commonly achieved.

[johngalt47]

Cnc

Are you going to do a DIY project with that CNC (complete with pics and BOM)?

[Hdale85]

We'll see. It'll be a while.

[JonMarsh]

At the Denver airport this AM- flying to Austria shortly on business, then Germany- will be back in Denver the 16th.

All the initial crossover construction was finished- one cabinet wired up and tested, as reported in the RMAF thread. Mid and tweeter level need to come down about 2 dB, otherwise very flat. It's a puzzle palace to get everything in the box, I think I've figured out how to add the midrange zobel back in. :W

If things go according to plan, should have things sorted out by Sunday the 19th. I figure these aren't overbuilt after all (Hank!), as cranking the one cabinet up with some David Garfield, I could just detect the hint of a vibration in the cabinet front near the woofers. Midrange/tweeter area is quite solid. Sides are mostly good, just a hint of vibration in one area.

What a busy week this has been - was great seeing Chuck, Carl and Paul, saw a lot of things at RMAF, some interesting, some just plain bewildering. :rofl:

I scored some interesting recordings, and am looking forward to working with more high resolution recordings in the future. And I'm looking forward to getting back from Germany and finishing these up!

[ColoradoTom]

At the Denver airport this AM- flying to Austria shortly on business, then Germany- will be back in Denver the 16th.

All the initial crossover construction was finished- one cabinet wired up and tested, as reported in the RMAF thread. Mid and tweeter level need to come down about 2 dB, otherwise very flat. It's a puzzle palace to get everything in the box, I think I've figured out how to add the midrange zobel back in. :W

If things go according to plan, should have things sorted out by Sunday the 19th. I figure these aren't overbuilt after all (Hank!), as cranking the one cabinet up with some David Garfield, I could just detect the hint of a vibration in the cabinet front near the woofers. Midrange/tweeter area is quite solid. Sides are mostly good, just a hint of vibration in one area.

And I'm looking forward to getting back from Germany and finishing these up!

Jon (and ThomasW and others):

Sorry I wasn't able to make it over.. it was my daughters first high school dance and I was designated driver for everything all day/night Saturday! 8O Sunday was pretty much a disaster time-wise as well. Hope to visit when you get back. Working on some finishing ideas for the M8ta's and very interested in hearing the Ardents.

Tom

[Hdale85]

I think they said Thomas will have the Ardents for a while so you could probably schedule a visit. Be very interested in your comments if you get a chance.

[ThomasW]

Jon returns Oct16th. He'll be staying here awhile so I'm sure we can get together

[---k---]

Thomas,
Now that Jon is gone, how do the Ardents handle ACDC at "11". :twisted:

[Curly Woods]

Thomas,
Now that Jon is gone, how do the Ardents handle ACDC at "11". :twisted:

LOL!!!! :rofl:

[JonMarsh]

Actually, Thomas does ZZ Top at "11", not AC-DC. I know this for a fact, as I heard it from down in the basement on Sunday afternoon- with the current builds on the Black Whisper. Just one cut from the greatest hits album... :W

Mit freundlichen gruessen vom Villach.

[ThomasW]

All of us have our dirty little secrets..... 8O

One of mine is dancing in the late 1960/early 1970's to ZZ Top at Tulagi's in Boulder, this was before they signed their 1st record deal. They were a great bar band pounding out really raw blues + R&R. I still occasionally enjoy listening to them.

They sound best LOUD!

As for AC/DC, sorry but I find them boring, ....probably a function of my age...

[Curly Woods]

All of us have our dirty little secrets..... 8O

One of mine is dancing in the late 1960/early 1970's to ZZ Top at Tulagi's in Boulder, this was before they signed their 1st record deal. They were a great bar band pounding out really raw blues + R&R. I still occasionally enjoy listening to them.

They sound best LOUD!

AMEN to that!

[CraigJ]

All of us have our dirty little secrets..... 8O One of mine is dancing.........

So let me get this straight, Thomas is upstairs dancing to ZZ Top, and Jon is downstairs playing with solder fumes? (never mind) Oh, and how are the ESS AMT I Great Heils sounding these days?

CraigJ

[ThomasW]

So let me get this straight, Thomas is upstairs dancing to ZZ Top, and Jon is downstairs playing with solder fumes? (never mind)

Yep ... :W

Oh, and how are the ESS AMT I Great Heils sounding these days?

LOUD, really LOUD They're are the best rock monitors I've built.... :B
We need to do a 1st series of measurements when Jon gets back from Germany.

"I'm shuffling through the Texas sand, but my head's in Mississippi".....arty:

Edit to add, a little something for Jon to do this weekend

http://www.gommers.de/index.php?pagePos=11&id=0163

[PewterTA]

So I take it, it's pretty safe to say that we (the general viewing public) are safe in starting to buy some of the components... like drivers?

Wkhanna and myself are going to build the premium set of speakers and we were wanting to start the purchasing process. Starting with the drivers I think, Seas, Accuton, and Scanspeak.

There's no last minute changes in these are there?!

Plus I wanted to bump the thread. :

[JonMarsh]

So I take it, it's pretty safe to say that we (the general viewing public) are safe in starting to buy some of the components... like drivers?

Wkhanna and myself are going to build the premium set of speakers and we were wanting to start the purchasing process. Starting with the drivers I think, Seas, Accuton, and Scanspeak.

There's no last minute changes in these are there?!

Plus I wanted to bump the thread. :

No problems with thread bumps, and no changes in drivers. :B

I've been reviewing measured data and working out some small crossover tweaks in LspCAD to try when I'm back, these are tweaks to the 9/30 design which is not posted anyway, but close to the earlier one. Mostly adjustments to zobel values to provide some small level tuning in the Fs region, and adding level pads for mid and tweeter. Tweaks will be tested on the one speaker with crossover boards hooked up externally, then both sets of boards tested on the one cabinet, then load everything up. Should be done Saturday afternoon.

Came to Nurnberg last night, and starting Sunday morning this is where I'll be the rest of the week...



...took the pic this morning. It's only about 15 minutes walk from the hotel, which is how I picked this hotel!

It will be the "silicon carbide" channel all week long, other than plugging in my iPod now and then for some Hiromi. :rofl:

Image not available

[benchtester]

No problems with thread bumps, and no changes in drivers. :B

I've been reviewing measured data and working out some small crossover tweaks in LspCAD to try when I'm back, these are tweaks to the 9/30 design which is not posted anyway, but close to the earlier one. Mostly adjustments to zobel values to provide some small level tuning in the Fs region, and adding level pads for mid and tweeter. Tweaks will be tested on the one speaker with crossover boards hooked up externally, then both sets of boards tested on the one cabinet, then load everything up. Should be done Saturday afternoon.

Came to Nurnberg last night, and starting Sunday morning this is where I'll be the rest of the week...



...took the pic this morning. It's only about 15 minutes walk from the hotel, which is how I picked this hotel!

It will be the "silicon carbide" channel all week long, other than plugging in my iPod now and then for some Hiromi. :rofl:

Image not available

Interesting venue, surrounding the event with silicon dioxide will no doubt help to focus the Force on the matters at hand.

[JoshK]

Is silicon carbide like what SemiSouth uses in their new JFETs? Or CREE with their high voltage Schottkys?

Have you seen Pass's new J2? It uses the SS jfets and apparently he is able to get really low distortion with a very simple circuit. What's not to love about that, except for maybe max power, which not everyone needs?

[JonMarsh]

We've been doing HV Schottky SiC diodes and SiC MPS (merged PN/Shottky diodes) for some time, since 2001, including used in Modules with IGBTs, and have commercial power JFETs (switches really, not designed for linear use) about a year away from release- technology development is complete. Our JFET switches are designed for very high speed switching, whereas the structure used by Semi-South has much more device capacitance, and avalanches through the gate, not the drain. Additionally, their structure is problematic as far as future development goes, they've kind of painted themselves into a corner, with their current cell pitch and sloped gate structure. What can be publicly discussed is in my 2008 Digital Power Forum paper, "New Generations of SiC Technology Advance Switch and Diode Capabilities".

http://digitalpower.darnell.com/archive/2008/schedule.php

[knifeinthesink]

Obviously a controlled experiment would be needed to see if this "really" works, but not having that kind of time, and having done some simple tests, for the "prime" set of these I elected to try this out. The pair being built with RS180, RS100, and 27TBFCG won't have this construction.

Will the RS seas version work in the same cabinet? I like these cabinets but there is no way i could part with the money for the deluxe versions with a baby in the house. Im eagerly waiting for the econo version.

[JonMarsh]

Yes, that was the point explicitly stated earlier- to the extent that the midrange is even mounted in an aluminum plate which conforms to the outline of the C79- this is done for consistency and also because it solves the rear masking issues of the RS-100 driver.

With regards to the low cost driver version, I am also considering extending the tweeter selection to the former Peerless HDS tweeter, now a Scanspeak D2608, as it has low distortion and rather nice off axis behavior.

[JoshK]

Further OT: I profess to know nill about silicon carbide. It is unfortunate, though, if what you say is true about SS painting themselves in a corner. Unfortunate for the sliver of a niche known as hi-fi, which has a desire to use these in a not intended fashion, as it means they will likely enjoy only a very short lifespan.

However, faster switches could spell better Class D amps in the further, no?

[JonMarsh]

Further OT: I profess to know nill about silicon carbide. It is unfortunate, though, if what you say is true about SS painting themselves in a corner. Unfortunate for the sliver of a niche known as hi-fi, which has a desire to use these in a not intended fashion, as it means they will likely enjoy only a very short lifespan.

However, faster switches could spell better Class D amps in the further, no?

Possibly- but class D amps usually use lower voltage range components than you need silicon carbide for. The blocking field strength of silicon carbide is 10X silicon, so whereas a relatively optimum range for silicon when not using superjunction technologies like CoolMOS does is 30-300V for unipolar devices, for SiC it's more like 300V to 3.3 kV. Of course, typical applications are things like solar PV inverters, high frequency SMPS power supply, ultra compact motor inverter drives, etc. SiC also has the advantage of thermal conductivity about the same as copper (more than 3X silicon), so smaller chips can handle higher power. Our standard 600V MPS diodes have normal operating current density of > 400A/cm2, and surge capability to 3,000 A or more per cm2. What's holding SiC back is the base material quality and some various issues in processing or creating certain types of device structures, like MOS gates. Silicon wafers have essentially perfect crystalline properties, but that's far from the case for SiC- micropipe screw dislocations, threaded edge dislocations, threaded screw dislocations, and basal planar dislocations are among the most common, and though steady process is being made, there's a ways to go. The largest commercial wafers available right now are 100 mm; they're normally fabricated with 2200 C gas vapor crystal growth in high temperature reactors. It's expected that 6" will become available from a couple of vendors before the end of the year; we already have a six inch capable reactor.

Compared with silicon, properly designed devices can have smaller chip sizes, and much better figure of merit considering output capacitance versus conduction resistance. This means higher frequency switching at high voltage with low losses, and greater robustness, as SiC's thermal intrinsic temperature is >500C, whereas silicon is in deep doo-doo by 250 C. (yeah, that's the technical description for it- conduction runaway through thermal carrier generation = deep doo-doo. ).

[Txgrizzly]

HAHA Jon i think you just gave me a headache... i kept hearing computer binary code as i was reading your responses... needless to say i will leave all the silicon talk to you guys... i am always amazed about the amount of stuff i know nothing about...

[Dave Bullet]

HAHA Jon i think you just gave me a headache... i kept hearing computer binary code as i was reading your responses... needless to say i will leave all the silicon talk to you guys... i am always amazed about the amount of stuff i know nothing about...

There are 10 types of people in this world. Those that understand binary and those that don't. :rofl:

Many of us probably thought the inner workings of solid state transistor technologies... but that is deceptively not the case. I think what Jon was describing was the 8th order hyperdrive ET is planning to install on the Death Star.

[Bear]

Silicon wafers have essentially perfect crystalline properties, but that's far from the case for SiC- micropipe screw dislocations, threaded edge dislocations, threaded screw dislocations, and basal planar dislocations are among the most common, and though steady process is being made, there's a ways to go. The largest commercial wafers available right now are 100 mm; they're normally fabricated with 2200 C gas vapor crystal growth in high temperature reactors. It's expected that 6" will become available from a couple of vendors before the end of the year; we already have a six inch capable reactor.

2200C CVD on 100mm susbtrate is $$. What gate sizes are you using? If you are pushing 400A+, then you are probably not using 32nm processes.

Compared with silicon, properly designed devices can have smaller chip sizes, and much better figure of merit considering output capacitance versus conduction resistance. This means higher frequency switching at high voltage with low losses, and greater robustness, as SiC's thermal intrinsic temperature is >500C, whereas silicon is in deep doo-doo by 250 C. (yeah, that's the technical description for it- conduction runaway through thermal carrier generation = deep doo-doo. ).

This is similar to what drove Gang Green + Big Blue to push through SoI processes. I'd be curious to see what some of Intel's "wet" processes could do with this substrate. Perhaps in a few years, though those wafer sizes need to come up radically. What size die block are they growing for the parent?

[JonMarsh]

Well, the seed crystal development is a challenge, one of several for producing boules suitable for 150 mm wafers. Using the best temperature control processes and an optimum convexity for the initial growth, it's generally only possible to add some mm to the high quality crystal width on a given growth pass, so it takes many steps to grow boules of sufficient diameter and quality suitable for seeds for the larger wafer sizer- this is probably why there wasn't an announcement by Cree at this years ICSCRM. Cree and SiCrystal are the closest to having this accomplished; both are base material suppliers for us.

Again, SiC is used mainly for higher voltage applications than logic, and the small defects present in the best of current material rule it out for fine detail lithography such as high density logic requires, but in no way impair the performance for applications like high power high voltage JFETs, where oxide quality and stability are not issues, and cell pitches are in the range of 3.5 - 15 um.

Also, it's possible to make fairly good performing unipolar diodes and switches even in the kilovolt region- examples at this conference were papers reporting on 3.3 and 6.5 kV SBD- Schottky barrier diodes.

Research is ongoing for bipolar devices also, but the short carrier lifetime of SiC, which is so ideal for diode behavior, makes it difficult to get conductivity modulation of substantial drift regions (even though epi's are much thinner than for silicon- for example, 600V = 3-4 um, not 30-40 um), and achieving substantial combined electron and hole current flow in HV epitaxies will often result in BPD grown (basal plan dislocations), resulting in the formation of crystalline stacking faults in the lattice that increase forward onstate votlage substantially. Various techniques are being explored to remedy this, including adjusting the axis of crystal growth, using additional gasses during growth stages, and surface preparation such as more advanced chemical etch and annealing to prevent BPD formation, or convert BPD to TED or TSD which actually trap stacking fault formation rather than seeding them.

It's a complex topic, SiC is a much more difficult material to work with than silicon- development is similar to where Silicon was in the 60's.

Our own company focusses on MPS/JBS diodes in the range of 300V to 1700V, and within the next year a commercial release of VJEFTs, centered in the voltage class of 1200V, but possibly with components in the 650V, 900V, and 1700V area, depending on market demand and application.

If any of you are curious about this, there's a white paper I wrote for our diodes for internal and external use with some of my European colleagues regarding our SiC diodes and the technology; it's all previously published material, but collected in one handy 32 page document- I can send you a PDF. :W

[Johnloudb]

I used 2 Cree 600v Zero Recovery SiC Diodes in my last preamp, for power supply rectification. They sound great! :B

[JonP]

I used 2 Cree 600v Zero Recovery SiC Diodes in my last preamp, for power supply rectification. They sound great! :B

They have some unique properties...

We actually use those in development for an implantable defibrillator, actually a test station that simulates the high voltage switching supply. It tests the transformer assembly. There was arcing over problems and I went through LOTS of those 600v 1A diodes. Funny how (at least a few years back) the only place you could off and buy 1-20 or so was Parts Connexion....

Compared with silicon, properly designed devices can have smaller chip sizes, and much better figure of merit considering output capacitance versus conduction resistance. This means higher frequency switching at high voltage with low losses, and greater robustness, as SiC's thermal intrinsic temperature is >500C, whereas silicon is in deep doo-doo by 250 C. (yeah, that's the technical description for it- conduction runaway through thermal carrier generation = deep doo-doo. )

:rofl: Jon, you continue to crack me up... it's true, though!! Best technical laugh I've had in a while...

[JonMarsh]

He's back!!!

It's good to be back; I'm in Denver at ThomasW's again, working on the Ardents for a few days before I drive back to California. Built midrange zobels today and modified the midrange crossover with connector wiring to interface; will be testing tomorrow AM.

Had done new impedance measurement today, too, for the midrange and woofers. Have also modified the woofer zobel; got to take out two 100 uF caps in each one to optimize the impedance tuning for killing the upper resonance peak. Will just de-solder and test first, of course...

Got a fair amount done today, but then ThomasW fed me a pretty killer latte, with enough caffeine to raise the dead, even for someone that should be suffering some significant jet lag- never even noticed it (the jet lag) during the day today.

Definitely need more stuffing in the cabinets to kill the vertical dimension resonance; can see it in the impedance curve, and in the woofer output (a dip). I'm pondering whether to put the crossovers internal as originally planned, or go "old school" Avalon like the Ascents with an external crossover box. Will probably not decide on that until I have a chance to carefully evaluate the behavior after I get back to CA.

Listened to some more of Thomas's SACDs, including a Rory Block one that's really nice on CD- "Last Fair Deal" - great songs and playing, very good sonics, too- ordered the SACD version today. IF you like slide guitar and blues, it's a pretty cool album.

[Johnloudb]

I got bought that Prokofiev and Ravel piano concerto CD you recommended. Very nice, and my dad really likes it. I listened a bit and it does have good sonics! I'll listen more when I get a chance. Thanks, for the recommend.

[JonMarsh]

You're welcome, John- to me, it seems like a real and welcome turn-around for DG as regards their recording techniques. I rather like this one, too!

[JonMarsh]

Highly Recommended




Rory Block

Last Fair Deal

Rory Block News Page

[Dennis H]

Rory supposedly uses a socket wrench for a guitar slide. I wonder if she gets a different tone depending on whether it's a Craftsman or a Snap-on?

[Curly Woods]

Rory supposedly uses a socket wrench for a guitar slide. I wonder if she gets a different tone depending on whether it's a Craftsman or a Snap-on?

So long as its not a "strap-on" its all good :-)

[mkc]

Hi Jon,

Please - the suspense is killing me .

I'm really interested to see how good the tracking is to the target curve.

Best regards,
Mogens

[JonMarsh]

The suspense has been killing me, too!

After the zobel updates I started yesterday and finished today, post additional impedance measurements, I can say it's pretty dang good!

Film at 11... maybe earlier! Going for dinner now, after listening since 3:30 (it's 5:45 PM).

[Hdale85]

Well I hope they are better for 2 channel then the Neo CC's, other wise I'd feel a bit silly building these over them Obviously different design goals I know but even so just saying as I know how much you rave about the Neo's.

[JonMarsh]

Testing and listening update

Thomas has been cracking the whip the last two days, and Colorado Tom came over this afternoon to check up on things.

Yes, there are two of these.....



And I measured both, but not having quite enough damping material on hand for the right hand cabinet, I'm just showing measurements for the left hand one for now.... :W



Thanks and a tip of the hat to Colorado Tom for the excellent photography aid... :B

[JonMarsh]

Some measurements...

Per request, the individual drivers are measured as well as the summed total, on the midrange driver axis. Gating window width is 60 msec, so room effects are visible, of course. The cabinets haven't been cranked up on the points fully, and cables are coming through the port, so LF output is a bit restricted- though it didn't really sound that way. Ports don't have the flare extension yet, but tuning frequency is ~32 Hz.



LF Measurement



Midrange Measurement




HF Measurement



All drivers summed. Note that midwoofers and tweeter are at the same level at the center frequency of 900 Hz, at ~ -18 dB. The midrange tracking starts falling off at ~180 Hz on the bottom end, but holds up OK to 8 kHz on the top end. I think I can live with that. :W