2021-2023 Isiris Update

[Finleyville]

So, between 4kHz and 15kHz we've actually got 2dB of lift.


I may look into seeing if I can come up with a way to make that adjustable too, but probably just two positions- one being the way it is, because it sounds clean and extended but not like a boom-tinkle box, if you know what I mean.

A Boom-tinkle box, heh heh. That's a good one!

That is totally different than a boom-tickle box!! lol. :W

[dar47]

Jon, the new house being built, is there any considerations for acoustic treatments where you be listening long term?

[JonMarsh]

Many considerations, but they'll all be done after the fact. I have materials for DIY panels, and I have some RealTraps RFZ panels (intended for side wall and ceiling reflections in the primary reflection path)-








I might buy some more stuff from them in the future. I'm literally taking this month by month, with all the stuff going on.

[JonMarsh]

A Boom-tinkle box, heh heh. That's a good one!

That is totally different than a boom-tickle box!! lol. :W

High frequency transients? Let's put a little water on that, and dig in for some natural harmonics and percussive transients!


First two cuts... to start.





Andreas is a harpist, but there's a lot more going on that that, on this album. Gorgeous demo or evaluation piece.

[JonMarsh]

And while treble transients are all good and well, let's not forget bass transients- and one of my favorites, Stanley Clarke "School Days"

The only first rate CD version of that is on the Bass-ic Collection CD- the namesake CD releases sounds like they had the compressor turned up to 11.

[JonMarsh]

Remember when J&J stood for Joni and Jaco?

If you've got no place special to go, why not Cotton Avenue? But don't miss the Overture first...

https://youtu.be/sTxSOJKzgl4

As might or might not imagine, these are all cuts used on a demo CD for a past audio meeting in Northern CA. Hard to imaging those happening again.

[JonMarsh]

OK, I thought it might be a good idea to summarize key points about why the crossover is looking the way it is, how it will be used in this configuration, and show a current (as of this morning) measurement in room. This was partly done for MB also, to highlight a recommended design process and philosophy.

One target has been to provide a certain amount of precision adjustability in the crossover, related to the midrange band and tweeter band, in well defined steps- the live room environment here certainly benefits from that.

You can’t enter a switch element in VituixCAD, but you can “open’ any part, which is the same as switching it out. So, for the schematic below, imaging that all the resistors shown below that have light grayed out traces actually have toggle switches in series. They’re used to adjust the midrange and tweeter level relative to the woofer, and for the tweeter implement a “presence” control which provides a subtle adjustment there (not shown in graphs; about 1.5dB in the presence range).




This is mostly potentially useful if the speaker is installed in a larger room with a long ETC (Energy Time Curve), particularly in the midrange and treble (like my current living room). It can also be used for some voicing control for program material and recordings.

Here I used overlays to capture three of the four midrange settings (max midrange is with all switches open; minimum is with all closed, min and max are shown here)





Here, the tweeter range is shown, with the mid cranked to full. Presence boost is not enabled.





In the current room setup, which is pretty reverberant, I’m running full voicing attenuation on the midrange and tweeter. If you sit just 3-6 feet from the speaker, that sounds a little bottom heavy, but farther away it gives the best balance, subjectively. In a treated room with LEDE setup as for a studio, one would probably open most of the switches.


This is a current in room measurement setup that way. Minimum filtering on the measurement, 1/48th octave. Not hiding stuff by smoothing. There are no bass traps or setup effort, so I have a dip at about 36-38Hz and 150Hz that isn’t addressed.

[JonMarsh]

Steve Wilson Remixes

Doing things "Free Hand"... OK, the work on the Osiris led me to thinking about Gentle Giant again, and having seen a listing for a "Free Hand" Steve Wilson remix- and having seen those sold for some Yes albums a couple of years ago.

Well, I thought, what the heck, and ordered the Free Hand Steve Wilson remix not knowing anything.

It came in early this week, and I also discovered this very interesting article.

https://www.soundonsound.com/people/steven-wilson-remixing-classic-albums

It spells out his philosophy and how he does what he does- which is very thorough and meticulous!

And besides, he gets instant points with me for having a lot of the DAW software I have on hand AND having two older Mac Pro's like models I have tucked away in storage!




But, the point is the music, yesterday I received the two copies of Free Hand that I ordered, and listened to it in Living Mono, through my one Isiris in the great room area, and if the job he did on the Yes albums and others is anything like what he did on Free Hand, I see why artists and record companies have been giving him all the business he can handle.



So I tracked down and ordered Tales from Topographic Oceans (one of my absolute all time favorite Yes albums, especially for the early work) and Relayer. Should be here soon. Not cheap, but for a fan boy like me, I expect they will be worth it.

And HDTracks has a special package 26 remastered cuts at 24/96kHz. That's on my shopping list once I get the networked music player back up.

[JonMarsh]

"Proof of life"

it will be a while before I can listen in stereo..... but it's just another hunk of deferred gratification for now.

Realistically, this is probably only for the serious Yes Fanboys. If you are one, you know who you are.

[1Michael]

I used to be a Yes fan and Relayer was my favorite album. I have heard of gentle giant but never listened to them, although my cousin had a couple of their albums.

[JonMarsh]

Probably my top favorite Yes albums are Union, Tales from Topographic Oceans, Relayer, and Close to the Edge. But many others are close, like Fragile.

[JonMarsh]

What's next

OK, there's going to be something of a pause in progress...

February is going to be busy getting ready for the expected closing in early March on new home- that means a variety of preparation activities including re-packing and other one time events, like the 3rd party inspection now scheduled for three weeks from now.

And of course, digging the other cabinet set out from storage, and updating drives in them!

But to go into a little bit of planning detail...

• Follow up on a potential switch with heavy duty silver contacts for the level adjustment controls
• Update the detailed component BOM based on the most recent design modifications
• Complete review of crossover capacitor choices, including recent and pending technical investigations, and price and availability changes of late- many parts have seen 60+% price increases, and some newer capacitor component technologies show promise.
• Order more phenolic boards for the crossover builds
• My interconnect "initiative" for the external crossovers is to convert to SpeakOn connectors and cables, using the four wire connectors for each signal cable- dual signal and ground paths
• I have already purchased all the connectors and assembled SpeakOn cables for the crossover to cabinet connections-
• Next, I need to fabricate new connector plates for the SpeakOn connectors and update the cabinets
• I will also work on a final layout for assembly, and see what I can do about making this easier to build for others- a couple of folks have expressed strong interest in replicating this build.
• Also, I had updated some of the CAD drawing for the cabinets to include metric notation and to clarify some views, for a gentleman in Singapore- I think he's dropped out because his cabinet shop can't handle the build, but I'll complete these updates for any other interested parties.
• I'll probalby go through all my pictures and records and create a PDF guide for construction- at least, some "hints" about how I got some key things done, including the platform update for the DWS780- just sort of put everything in one place for easy reference. It will be far to big to email, but with Apple messages sending a 48MB PDF is no big deal- I've sent the M8 build article to several people that way.
• I'll update this thread once I can get back to this effort, but with the closing and move, I won't be able to restart it until around late March at the soonest.

[Efalegalo]

[*]I'll probalby go through all my pictures and records and create a PDF guide for construction- at least, some "hints" about how I got some key things done, including the platform update for the DWS780- just sort of put everything in one place for easy reference. It will be far to big to email, but with Apple messages sending a 48MB PDF is no big deal- I've sent the M8 build article to several people that way. [/LIST]

^This would be very much appreciated :-).

[JonMarsh]

Fortunately for me and everyone else, I take pictures and notes during projects like nobody's business and save them to my project files on the Macs... so I have a lot of source stuff.

I'll probably find a bit of time to work on the updated BOM, but I also need to create one for the Obi-Wan.

[JonMarsh]

Isiris SIMPLIFY Crossover concept

Do you ever have one of those nights when you wake up at 3:30 in the morning, and a specific idea floats up quickly from your subconscious, one compelling enough that you can't dismiss it and just fall back asleep for another couple of hours? No?

Well, you're lucky. I had one of those nights/early mornings today.

There are a few folks now interested in this new Isiris build, but while I've been pretty successful at improving the driver BOM cost proposition, the crossover remains a bit pricey by normal mortal human standards. (Obviously, excluding ET).

Well, since I don't know any way to violate the laws of physics, and I'm getting some rude shocks this month when getting updated prices from vendors, obviously, we must look at a topology change. It doesn't take a CPA to realize that if, for example, one goes from 3rd order crossover to 2nd order crossover, there should be some component cost benefits- possibly substantial ones.

Now, my own experience looking at other solutions shows that LR2 is a widely used approach for commercial crossovers, likely due to the cost fundamentals. So, the thought I awoke with today was why not take an initial pass at design using that concept? It's a little bit like a Duelund approach, but simplified, and without the ultimate 4th order slopes for the woofer and tweeter.

The question is, can it be simplified enough for a solid cost payback, without compromising performance too much? One issue I immediately hit on is that keeping trap and notch filters for driver behavior will still be necessary. But is their cost benefit still to be had? Significant cost benefit?


Well, the first step of Isiris SIMPLIFY version is to develop a fairly functional crossover; next step is the BOM cost comparison. So let's see what we've got after step one. This could likely be tweaked and improved further, but I doubt the cost structure will change significantly; for this analysis, it's likely good enough. Any complaints or suggested regarding that, please chime in.



Schematic




SPL Plot





Impedance Plot




This is not as few a parts as I hoped for, even though these drivers are fairly well behaved.

In the case of the woofer crossover, savings exist, but not as big as you might hope- the lower crossover point needed to be shifted up but baffle step compensation remains. And furthermore, providing a notch filter for the woofer breakup requires an explicit inductor for a shunt elliptic filter. But it's smaller and cheaper than the second series pass inductor. The main inductor is reduced in value and cost (plus I have some this size from a previous experiment) and the capacitor values are smaller. Overall, doesn't look bad.

The midrange crossover only has one series cap, again, a bonus for the cost reduction- and inductor values aren't a problem; with the excess midrange SPL, ESR is not a big concern. Here, there are significant cost savings, but selecting the optimum crossover points for the topology and driver slopes took several tries.

Tweeter crossover is a big cost break, if you were planing on the high priced spread for the series cap. In fact, I have a couple of Clarity MR's around of the required value, so this helps with the barrier to testing this, as well as the overall BOM cost.


The downside for this version for some is the impedance, which considering each driver branch is nominally four ohms, is not surprising if you've seen data plots for commercial speakers in test such as at Stereophile. My Edge W amplifiers would shrug it off, to say nothing of the Halcro monoblocks, but some amplifiers will not be happy with the 3 ohm load from 500Hz to about 1200Hz, when all three drivers have some loading effects, similar to a Duelund. Impedance interaction could be minimized by adding a shunt Zobel centered at 4kHz; this would not be an expensive addition.

OTOH, it's much kinder than, say, a Wilson Alex V, a Magico M2, Avalon Indra, and some others.

Next, need to find the time for some BOM comparisons

[JonMarsh]

Updated SIMPLIFY crossover

What's that saying, about modern technology success relying on rapid iteration?

well, I just did my normal 2 mile morning walk, and thinking about the LF filter, even though my early morning response was to remove the series inductor (2nd one) and use a parallel shunt elliptic filter, I remembered why I didn't like the filter response characteristic of those, and so figured I'd try my (lately) normal series parallel notch network.


I'm definitely happier with the response changes, can really crush the woofer breakup mode again now... while keeping things nice for an octave above the crossover point. Costs may be improved also- much smaller inductor, but larger cap. And high value relatively low power resistor.

Oh, and just for the record, one of the things that makes things possibly work OK with the lower order crossover is the fairly high Xmax of the BlieSMa T34A, and the sensitivity doesn't hurt things either- gives one morel latitude with the crossover filter design.


Schematic




SPL





Impedance

[technodanvan]

Looks pretty darn nice. I thought your typical target response had a slight downward trajectory though?

Edit: I suppose over 10k there is a slight lilt.

[JonMarsh]

Looks pretty darn nice. I thought your typical target response had a slight downward trajectory though?

Edit: I suppose over 10k there is a slight lilt.

Finished response usually does, but initial development it's not necessary- minor tweaks can take care of that.

[JonMarsh]

The "Isiris Alexxa Edition"

In the spirit of fast iteration and pushing the boundaries a bit, this is what I'd call the "Isiris Alexxa Edition", as it has some nice transfer and phase characteristics, and I've pushed the crossover point for the woofer to midrange down to 300Hz, but the impedance is a bit more brutal, at or a bit below 3 ohms from 60Hz to 600Hz, though not as bad as the Wilson Audio Alexxa (*per Stereophile measurements )

[JonMarsh]

what I haven't mentioned is cost differences- and lets' keep in mind this varies depending on the type of capacitors and inductors chosen for either configuration. But to cite a small example, in the tweeter crossover, the capacitor cost is reduced by 2/3. With current prices on film caps like the Jantzen Alumen series, that isn't chicken feed. Midrange crossover cap cost reduction isn't as large, but it's more than 50%.

Of course, the problem may be, that with this version, the money you saved on the crossover may get spent on the power amplifier instead!

I have started component orders to build and debug a sample crossover set.

[Efalegalo]

Can I ask a question about the downward tilt?

Maybe it was Harman's research that revealed that most study participants preferred a slightly downward sloping response (high side being in the lower frequency) - but I thought this preference was based on listening via headphones (i.e., void of any room reflections). To active the same "downward tilt" in a stereo setup, the speakers were designed to have a flat on-axis response, but then gradually decreasing off-axis response ---- resulting in the overall downward tilting "Total Early Reflection" response.

With that said - I still prefer an ever so slight downward tilt.

[JonMarsh]

One of the more definitive efforts involving quite a few "expert" listeners was undertake by Sigfried Linkwitz, a former HP engineer and the developer of the Orion system. He was never completely happy with the tonal balance until his review of a variety of research pieces resulted in testing a lot of specific voicing concepts based on studies of how stereo music is perceived, which is different from monophonic.


This is discussed in this thread, along with the design of a simple balanced passive module intended to be inserted between preamp and power amp for use with "flat" measuring loudspeakers.

https://www.htguide.com/forum/showthread.php?43337-Report-back-on-stereo-speaker-voicing-investigation&highlight=report+voicing

In general, I preferred a 1.8 dB version of this,







And the former colleague that owns my original Wavecor Ardents has this in his system and is quite happy with the effect.

Other members have adopted this in one form or another for their Wavecor Ardent systems.

I can see this both ways; designing for flat and then implementing a voicing filter that is easily modified has some attraction to me, over building it in to the speaker itself. Gives one an accurate baseline and some flexibility about how to modify it.

This was a PCB I developed and made a few of (hired out through my usual vendor)



It uses metal film resistors and WIMA caps.







A little trimming was needed on the width to fit into the XLR adapter.

[JonMarsh]

Tweeter crossover update

Is this the part where I'm supposed to whine about supply chain issues because my three FedEx orders (from vendors in the US) are all late?

Well, time waits for no one, placed another order today because I took a little break from the Obi-Wan crossover wiring to review the Isiris SIMPLIFY design, and ended up trying out some tweaks that I figure I'll need to also build and measure for the tweeter crossover... no big deal, only a couple of hundred dollar more parts. Spending money like I'm flush with it, which I'm not. Mainly I like the difference in the low end tweeter attenuation, though it's not quite as flat in simulation, and a little more expensive to build. Well, except I DID take out some parts!

[Bear]

Is this the part where I'm supposed to whine about supply chain issues because my three FedEx orders (from vendors in the US) are all late?

Well, time waits for no one, placed another order today because I took a little break from the Obi-Wan crossover wiring to review the Isiris SIMPLIFY design, and ended up trying out some tweaks that I figure I'll need to also build and measure for the tweeter crossover... no big deal, only a couple of hundred dollar more parts. Spending money like I'm flush with it, which I'm not. Mainly I like the difference in the low end tweeter attenuation, though it's not quite as flat in simulation, and a little more expensive to build. Well, except I DID take out some parts!

The series cap for the midrange would hurt a bit, but it's perfectly understandable given the bottom target for the band pass. The 100uF for tweeter's parallel RLC/Zobel isn't too bad, either. Lots of ScanSpeak models need something similar (e.g., AirCircs seemed to need 80uF - 100uF, depending upon design).

Are you still thinking about doing a "lite" version with the RSS265?

[JonMarsh]

Well, as I'm sure you understand, the biggest advantage to this SIMPLIFY concept is reducing the number of primary series components in the crossover from two to 1. In the original midrange crossover, there was an initial series capacitance of 81 uF, with 1.8 mH to ground, followed to the secondary capacitance of about 350 uF. Low impedance drivers need big series caps. So, dropping this to just 150 uf is a big cost reduction, even if you use the same type of caps. And my traditional caps for series midrange application, or combination thereof, have gotten quite a bit pricier over the last six month. Same for tweeter caps.

So, this prompted me to do more component research, and I settled on the Clarity CSA copper connect series, especially since my favorite vendor in the UK has both them and the MR series in good stock depth. And luck of lucks, I have a pair of the 3.9uF MR's left over from another project, which work perfectly here.

And, furthering the schadenfreude or synchronicity, I'm getting away with the 4.3 mH for the series woofer inductor, as bought for the original crossover (have plenty on hand, I got carried away)

And the shunt cap for the midrange crossover can use the left over 2.2 mH from the purchase for the last crossover design in the woofer!

If this wasn't all mostly a happy accident, I would say I was one smart fella, but alas, I know the truth- I'm just lucky. Sometimes.


Of course, the proof is in the pudding, and that will require both measurements and listening tests, in the new place, with better controlled acoustics.


I have the parts for trying out the "light" version, but several folks considering that have switched their preference to the full bore version- not much difference in the driver cost, and higher sensitivity. However, a light version in a shorter cab with the RSS265 might have better WAF for some, so evaluating that is still on the schedule, but further away time wise. There's just an awful lot of stuff that's got to happen this year...

[dar47]

I know you were thinking of a design to shrink the voicing board to fit the XLR better and ordering more of these for future designs, if you ever find time time for this I would love to contribute and get 4 of these for me and Ben.

[Bear]

I have the parts for trying out the "light" version, but several folks considering that have switched their preference to the full bore version- not much difference in the driver cost, and higher sensitivity. However, a light version in a shorter cab with the RSS265 might have better WAF for some, so evaluating that is still on the schedule, but further away time wise. There's just an awful lot of stuff that's got to happen this year...

You definitely have a lot going on! From my end, the driver cost isn't the issue; but the floor area cost is. California real estate didn't get any cheaper when you left. I still expect that I might take a shot at a SW223 version of the Ardent D, though.

[JonMarsh]

I know you were thinking of a design to shrink the voicing board to fit the XLR better and ordering more of these for future designs, if you ever find time time for this I would love to contribute and get 4 of these for me and Ben.

Funny you should mention that, I have the kit of parts and materials from the original voicing board "in house", so to speak, and it's in the queue. Trust me, you don't want to see my Things Task manager and all the stuff "pending" . But this way I tend to not lose track...

Also, I have a very nice PCB shear that I'd bought and taken into the work lab back in 2019 or 2020, because the one they had there was crap, and well, I picked that up from CA in mid 2021 along with the rest of my "contributions" to the office lab setup and brought them home to Idaho. So I do plan to see if it's possible to just trim the ones I have. The shear is somewhere in the middle or back of one of the three storage units, though...

I need to buy a big rubber stamp... like this one:

[JonMarsh]

You definitely have a lot going on! From my end, the driver cost isn't the issue; but the floor area cost is. California real estate didn't get any cheaper when you left. I still expect that I might take a shot at a SW223 version of the Ardent D, though.

That's an interesting idea, but if you're really interested in following up on that, the only SW223BD I know of for sale right now are at Solen in Canada- they have four. I was just checking about that the week; PE just shows them out of stock, no date given for return to stock.


And, BTW, I'm planning to acquire a pair of the BlieSMa aluminum dome mid to evaluate as replacements for the 1302's. I'll probably have to make an adapter plate, but no big deal...

[Bear]

That's an interesting idea, but if you're really interested in following up on that, the only SW223BD I know of for sale right now are at Solen in Canada- they have four. I was just checking about that the week; PE just shows them out of stock, no date given for return to stock.


And, BTW, I'm planning to acquire a pair of the BlieSMa aluminum dome mid to evaluate as replacements for the 1302's. I'll probably have to make an adapter plate, but no big deal...

I'm building some surrounds next using the Bromo design since there are surprisingly few published designs for the SB17NBAC + SB26ADC tweeter. The design looks like it has a lot of room for improvement, but I want something quickly. If I designed it myself, I'd be here a year from now still trying to complete them (I can rip-and-replace the crossover later).

I've got a four-pack of SW223BD03s from Solen from just before PE picked up the line. I've been wanting to use them ever since. I have a pair in use as a second subwoofer currently. Like many folks here, I have a pile of parts waiting to be used in projects (including a pair of T34As). Steve can undoubtedly mill you an acceptable adapter plate for the aluminum mid dome. The performance characteristics look to be outstanding on the Bliesma mid domes, and it also looks like they may have a half-octave to an octave of more grunt (BW2 at 600 Hz?) than the Morel dome.

[JonMarsh]

Valentines Day Treats

FedEx delivers, though a few days late. The new normal, I suppose.








Another Obi-Wan crossover on the left, and a SIMPLIFY Isiris midrange crossover on the right. Well, part of it...

[JonMarsh]

Updates to SIMPLIFY midrange crossover example and starting layout

I'm finding a bit of time here and there to re-examine and polish- and address things like guard banding power dissipation in resistors in the high power midrange, and setting up parallel values to enable some adjustability.

More parts are on order, and I've started a layout for the midrange crossover, which will be the biggest board, in all likelihood.





You may note the addition of the parallel zobel network at the tweeter crossover consisting of C12, L9, and R22. This is for the "perfectionist" version and only serves to flatten the impedance curve in the midrange to tweeter crossover area. The woofer RC zobel has been updated to hit the target response more closely.








A starting point for the layout drawing with updated component locations is ready to have wiring done- the actual BK-2 boards have been received, and the size may be reduced compared to what is dimensioned here.




I think things are moving in the right direction...

[JonMarsh]

Slow work takes time...

No doubt about it.

[wolf_teeth]

The R9/L4 combo is the why R9 is required to dissipate that much heat. (Not that you don't know that, Jon, but maybe others here don't.) This circuit as done places these 2 parts directly across the amplifier, and it will see the full bass bandwidth below that of the Fc of the coil. As we know that series resistors on woofers are bad for a number of reasons, thermal capacity is one of them. This is just how that R9 is feeling being connected as shown.

I've seen this layout more recently than I used to, and while I'm not the designer, I feel that it is not the best way to contour a mid/tweeter and keep the resistors happy. I second guessed myself before posting this as I've contacted designers (not Jon) privately before to voice my thoughts on just this issue in particular only to be pushed aside or ignored. I feel the possible fire hazard that running resistors this way can be, should be acknowledged by both the designer and the prospective builder. Thermal management of it is to be adhered to for best practices, via multiples or heatsinking.

Since this is not a final posting, but a developing project, i thought I'd offer my 2c on this matter.
Best regards,
Wolf

[JonMarsh]

Understood and agreed, but also considering the system sensitivity and the actual RMS power even with loud program material, I think things will be in hand. We'll see once I finish bench testing. At low frequencies it is essentially a 12.5 ohm resistive load connected to the amplifier, in series with the DCR of L4. It has advantages in shaping the cut off slope of the high pass network, also considering what the midrange driver does by itself.

With 20V RMS input to the speaker, just at low frequencies, this will deliver 100 Watts to the woofer array, the LF output from the woofers will be about 108 dB and the dissipation in the 40W R9 array will be 32W, within the nominal 40W capability of the current resistor selection. This is a situation fairly unlikely to be encountered for the intended use in my home environment, as this would be about 110 dB total from the stereo pair at low frequencies.

The intent with this variation was to provide a lower cost alternative crossover- with good in room system response. Of course, the original "high priced spread" crossover design doesn't have this issue, but the component price of admission is considerably higher. That version is playing in the living room currently.

Once we get moved into the new house, I plan to bench both crossovers, including looking at the distortion contributions with my Audio Precision, as well as acoustic measurements, and in the end it will be interesting to see which I prefer and use. I haven't usually considered two such different approaches for the same project, but being retired now, I have a bit more time to investigate and consider.

And BTW, I do have a FLIR camera and can observe temperatures in the crossover while under test. I used this at work for SMPS converter design and looking for component hot spots, especially magnetics. My former employer (international semiconductor firm) had a very low Cap Ex budget with a new local regional VP that I swear got his bonus from minimizing Cap Ex expenditures for the lab and other groups, so much of the equipment I used my last 6 -7 years was purchased personally. I was even loaning out gear to other Application engineers. So I'm pretty well equipped...

[JonMarsh]

More follow up

I keep getting asked about what I'll be driving these speakers with- while the impedance curve is not as bad as many speakers reviewed by Stereophile, for example, it's in a different category from my usual designs, which are more voltage driven, often with woofers wired in series. (Original Isiris, Wavecor Ardent).

In December 2019 I bought a pair of Cambridge Edge W power amplifiers, from a dealer in Canada I work with now and then, at a very attractive price. Because of my past experience using and benching Cambridge amplifiers that use the Class XD offset bias concept, but also a lot of other very sound construction engineering that I have used myself in the past on consulting work (low inductance parallel bus bar power distribution, for example, and separate regulated gain stage supplies, and over sized capacitance reservoirs for the main amplifier stage).





Oh, did I mention the power transformers are wired out of phase and crossfeed both channels, for magnetic field cancellation? As Jack Reacher would say, details are important.

Here's a YouTube blurb about it from 2018.

Also, the Edge W had some very specific characteristics or capabilities that make it very flexible in how it may be used- this includes a buffered input signal pass through on both balanced and unbalanced inputs, and an overall design favoring high current capabilities over high voltage. So, literally, what you're getting is a 180W/channel amp at 4 ohms, 100W at 8 ohm, that is also fully 2 ohm capable. This opens up many possibilities. For example, one is using a phase reversal XLR jumper cable to connect the input from one channel to the input to the other channel, thus creating a balanced mono block amplifier capable of 350W at 8 ohms, and about 600W at 4 ohms.




This review came out after I bought mine, but in reinforces and confirms the performance levels I expected based on older and lower cost models.

https://audioxpress.com/article/fresh-from-the-bench-cambridge-audio-edge-w-power-amplifier

The full review with measurement graphs is here in PDF form:

https://audioxpress.com/files/attachment/2712

I haven't had the opportunity to bench it yet, because the APx555 has been in climate controlled storage since 2018, just like the Halcro DM68's. In a month or two I should be able to start unpacking all that stuff.

Now, of course, with the buffered input/output capability, other configurations are very easy, also, and my test auditioning and crossover work is currently being done with a side patch cable and both channels receiving the same signal, and the woofers running on one channel, and the mid and tweeter on the other channel. I think that would be called bi-wiring, in a fashion. No worries about gain matching, as you would be concerned using two different amplifiers.

Now, interestingly, they have a new version of this, the Edge M, which is simply configured with single inputs and does the mono block signal feed internally, with the outputs looking "normal". Same end result as I can get, but I guess for folks that lack imagination in configuring the "W", which is much more flexible?


Check out the review and measured performance. In my opinion, Cambridge have delivered the goods, and the Edge W is what I'd call the poor man's thinking man's very high end amplifier, with a lot of flexibility to address varied usage cases. I say "poor man" in relation to the asking prices for brands like Acoustic Arts, D'Agostino, Rotel Michi, darTZeel, Gryphon, Classé, Parasound, etc. Which mostly cannot match it's performance.

[technodanvan]

That's a really interesting amp John, you don't see digital inputs like that every day. Very pretty too!

[JonMarsh]

Oops, the picture I snagged was for the nearly identical integrated.. Gotta read things more carefully, and go back and fix that!

In fact, going back and checking it, the link actually says Edge W, but as you can see, it's actually an Edge A. (reading the back panel- mis posted by the web site)

[technodanvan]

Still shiny though!

[JonMarsh]

A lot of house related stuff going on yesterday, but some progress was made, and new parts received from Parts Express...





The test mount of parts shows that everything fits just fine in the planned 9-1/2" by 13-1/2" space. Next step will be trimming the board, then starting the wiring.


This would go a lot faster if I didn't have all these other pesky tasks on the agenda, like researching epoxy coated floor paint methods for the new garage floor, or researching steel storage racks for the garage (some good ones found at Costco) or researching hot glue sticks, because the two nearby Ace Hardware outfits in the Boise are DO NOT carry the good stuff from Ace Hardware! It's in stock at my old Danville CA store, but Ace Hardware doesn't ship this item to consumers. These ones for rubber, nylon, PVC and metal stick like the taint of original sin.






But wait! Farther afield, there's another Ace Hardware store, in Eagle, ID, and they supposedly stock this! I'm going to have to check that out today if I can find the time! And they're just off of Highway 44, which in Middleton is also Main Street. About 30 minutes from here in Nampa, though.

One thing I've repeatedly noticed is that the quality and variety of stores really varies depending on the community in the Treasure Valley. Eagle is pretty much tops, Boise #2, and Nampa is dragging it's tail, as is Caldwell. This applies to grocery stores as well as things like hardware stores or Home Depot's. Fortunately my favorite grocery store by far just opened in Star, about six miles east of Middleton. More information than you need, but the difference is worth driving well out of your way, to me.

I may have found a good replacement on Amazon- one reviewer said he thinks it's the only good hot glue stick product left for demanding tasks these days- order some, will report on that. Have some ordered.

[dar47]

https://www.costcoconnection.com/connection/202001/MobilePagedArticle.action?articleId=1549140#articleId1549140

I have this Costco one it's great!. You can get two of these on one side of your garage. Load them with all you favorite clear plastic bins, they have steel mesh shelves which you can place a sheet of wood on if you like. I have a paint section, extra parts section, motorcycle section, fuel container section, etc.

[JonMarsh]

I ordered 3 of these last night from Costco, in the 72" length.

[JonMarsh]

Ace Hardware glue sticks

Success at the Ace Hardware in Eagle!



This is the good stuff- sticks like the taint of original sin!

[technodanvan]

I think I've been working on a pack of that glue since 2009.

[JonMarsh]

We probably have different consumption rates....