Jump to content

Iron-core versus air-core woofer inductor intermod test (Continued).


Recommended Posts

Hello:

 

My first post here. Incredibly impressive forum on two counts - some genuinely serious audio (Chris A. and others) and secondly, rare for today, allowing some difficult conversations without censorship.

 

I spent a few decades in pro audio (concert production) with the intent on bringing the best sound in the world to my audiences. Of course I rarely pulled it off, but I enjoyed the pursuit. : )

 

Much of the endeavor involved modifying the weakest link in the chain: loudspeakers. My hero in the realm is Dick Heyser and he did a review on the mighty Klipschorn in a 1986 Audio Magazine article and ever since then I wanted to purchase a pair and have a go at making them all they could be without violating PWK's vision. Now I'm retired and have the time to make a mess of home audio, so I found a very nice pair of 1974 vintage Klipschorns, Polk SDA SRS 1.2TL's, Martin Logan CLS II's and some other stuff. I'm into inverse filtering per Dave Gunness's work and I'm also a huge Tom Danley fan though and may end in his court, we'll see.

 

---

 

What's my first post about? My stubbornness to stay with passive crossovers per the original Klipschorn design and thus have to deal with something I avoided with professional loudspeakers: cored inductors. Something this "simple" ended up taking much of a month to figure out and I'd like to share my findings. Sure, air core is the way to go as long as you keep them away from ferrous materials and magnetic fields. That's not so hard with the smaller values required for mids and highs, but at low frequencies, reasonable DCR is a challenge.

 

I thought it would be easy getting a high quality 4mH cored inductor with less than 0.5Ω DCR for the Crites cast frame woofer (replaced the square magnet K-33P), but even some of the best of these generate too much distortion for my taste.

 

---

 

What do I mean by distortion? To begin with, THD is a poor, but not quite useless distortion measurement for devices that exhibit natural low-pass characteristics; such as inductors and horn loaded bass cabinets. THD is all about comparing the unwanted odd and even harmonics to a fundamental, and the harmonics are largely filtered out! Thus you can get good THD numbers and still have audibly unacceptable levels of distortion with low pass devices. The solution is to measure distortion components within the passband of the device, and that can be done with intermodulation techniques that use two or more simultaneously generated tones within the passband to generate sum and difference frequencies as well as harmonics of the (two or more) fundamentals.

 

I chose Audio Precision's "MOD" IMD measurement technique as well as a high resolution, low frequency Multitone stimulus to compare 4mH inductors loaded by a 4Ω power resistor. 4Ω is what the Klipschorn and a great majority of low frequency systems use, thus 8Ω wattage specifications tend to be marketing ploys IMO. Inductors are current limited animals, not wattage limited. An inductor that is clean at 10 amps into 4Ω behaves identically at 10 amps into 8Ω. Ohm's law states that Amps(squared) times resistance equals watts, thus the 8Ω load will provide the marketing department with twice the wattage specification compared to the more realistic 4Ω.

 

---

 

It has been a personal endeavor to select a low DCR woofer inductor for my 2nd order (electrical) low pass filter. There is a 12awg copper foil air core inductor that boasts a 0.3Ω-ish DCR, but I wanted to learn about cored inductors, thus purchased the following 4mH coils (that's what their nominal spec's were, but the truth turned out to be 3.4mH - 4.2mH).

 

1. ERSE SuperQ 16awg
2. ERSE SuperQ 14awg
3. ERSE IQ 18awg
4. Jantzen C-coil 14awg
5. Madisound Sledgehammer 15awg
6. Crites Steel Core 0.46Ω DCR (don't know the awg yet, either 18 or 20)

 

I won't have inductor #5 until 2 June.

 

Would the results of my testing be of interest?

 

God bless you and your precious families - Langston

Edited by Langston
  • Like 4
Link to comment
Share on other sites

  • Moderators

@Langston

 

Hi Langston, welcome to the Forum. I split this into a new thread because the one you originally posted this in was two years old and your new post seemed worthy of it's own fresh start.

 

You should be able to post that curve in here now.

 

When you do, you can mention @karlson3 and it will alert him that he has a reply.

 

We also have a pro audio section if you haven't already seen it. I think you will find that of interest as well.

Link to comment
Share on other sites

Link to comment
Share on other sites

Thanks Wirruna - that's the primary thread that landed in my online search on the topic of inductor distortion, and after reading that and bunch of other stuff on this forum I decided to join if the moderators could be fooled into letting me in. : )

 

The whole cored inductor mess started when some very reputable coils measured poorly and some inexpensive ones measured well. I figured it would be a piece of cake selecting a cored 4mH inductor to precede the 130µF cap for my Klipschorn passive LP filter. Boy was I wrong. I'm working on the post now and will make the first installment an intro. to the measurement philosophy using a single inductor (edit: 3 inductors) as an example. Then I'll followup with the shootout. Bob Crites, grumpy Al and others revealed some great stuff in that thread you mentioned, they just didn't seal the deal for me.

 

God bless you and your precious family - Langston

 

I claim that it should be possible to measure audio systems and have those measurements correlate with what we hear out of those systems - Richard C. Heyser

Edited by Langston
  • Like 1
Link to comment
Share on other sites

I've chosen (3) measurements for inductors:

 

1. THD up to 0.5% from 10Hz to 20kHz.
2. IMD up to 0.5% using equal levels at 50Hz and 500Hz.
3. Multitone from 10Hz to 20kHz with 48,000 sample points (crazy high res).

 

I'm using a Powersoft K20 amp through the coil under test, then into a 4Ω power resistor rated at 100W. I'm using 2 second THD sweeps that avoid heating the resistor too much even at 110V (3,000W), which is way more than any cored inductor can tolerate that I've come across. The amp is very comfortable at this level, thus doesn't pollute the results with distortion products of its own.

 

THD

 

Not the right tool for analysis due to the low pass characteristic of inductors. I found it useful to ballpark maximum current an inductor can handle prior to severe saturation. 0.5% THD fits the bill although it understates the actual distortion going on within an inductor's limited passband.

 

IMD

 

This has been the go-to for many years. Audio didn't use to have the bandwidth it has today and IMD was the primary way to assess distortion in limited bandwidth devices - it was also one of PWK's favorites. Stimulating the device with multiple tones is also more like what devices are exposed to in use and the harmonics that fall out of these tests have no harmonic relationship to music, thus sound terrible, thus is something you want to avoid, thus test for. IMD measurements can scream bloody murder when THD looks fine. Keeping IMD under 0.5% through the maximum current the inductor will see in use is the bottom line for me. This is what ERSE, Jantzen and others should be doing when they rate their inductors' maximums. I found several inductors that can't go under even 1% at ANY drive level.

 

Multitone

 

This is cool, but nobody really knows how to objectively correlate it among differing devices to how the devices will sound. Nevertheless, it is very telling to look at a bunch of simultaneously generated tones located in specific areas (it's complicated) and seeing the crap that fills the gaps between these tones. I think this is the future of objective distortion measurement once we figure it out.

 

Sorry for all the words. : )

 

Next post will compare (3) 4mH steel plate cores (you'll be surprised):

 

1. ERSE IQ 18awg
2. ERSE SuperQ 16awg
3. ESRE SuperQ 14awg

 

PS: You've heard this, but it's true - air cores are effectively perfect, you just have to deal with magnetic sensitivity, DCR, size and cost.

 

God bless you and your precious family - Langston

 

An experiment is a question which science poses to Nature, and a measurement is the recording of Nature’s answer - Max Planck

Edited by Langston
  • Like 2
Link to comment
Share on other sites

That is a wise statement Deang - nothing offends the transduction from the electrical to acoustic domain as much as the moving coil loudspeaker itself. A close 2nd would be the room. Nevertheless, it's an interesting pursuit to reduce the electrical domain distortions as much as possible. A hundred tiny improvements = a slap in the face + it's fun. : )

Edited by Langston
  • Like 1
Link to comment
Share on other sites

My wife says that! : ) In this case I'm taking a pair of 1974 Klipschorns and trying to do what I think PWK would do with them if he were here. There's stuff I'll get to in the future, but at the moment I'm simply trying to convey what I've learned about the worst offender in passive crossovers; the inductor. My conclusion will be that it's possible to find good ones, but you can't do it by reading advertisements or subjective reviews or assuming price = performance.

 

Bottom line: anything in the signal chain that has ≥ 10dB less distortion than the moving elements of the loudspeaker system cannot be heard and can effectively be ignored (IMO). The brilliance of PWK was his fixation on the horn, which translates into minimal voice coil movement, which translates into less distortion. One of the things we did with concert production was bring "too much rig for the gig", thus we got no where near its limits, thus it sounded clean. What I'm doing here is making dead sure that voice coil/horn distortion is the only audible distortion in the system.

 

Just for fun:

Decibel Addition & Subtraction.pdf

 

God bless you and your precious family - Langston

 

Whether we like to think of it this way or not, an audio engineer shares the professional goal of a magician - Richard Heyser

Edited by Langston
Link to comment
Share on other sites

Once maximum power distortion tests are done, would there be benefit in low-level testing? Hysteresis, or core distortion from the zero-crossing may be of interest.

I've often wondered if there would be any merit in magnetically pre-biasing the core to avoid zero crossing. If you're using a ~400-Watt capable inductor, and a Klipschorn will wreck your hearing with 20-watts, it seems like there's a bit of magnetic headroom that can be traded in.

Link to comment
Share on other sites

I have built and tweaked over a period of time a pair of crossovers for my La Scala’s 

Also use the Crites cast bass driver 


During my journey to nirvana I have used different capacitors, resistors and induction coils 

 

You can read about my journey and results here 

https://community.klipsch.com/index.php?/topic/149952-a-crossover-design-on-single-exposed-board/

 

  • Like 2
Link to comment
Share on other sites

7 hours ago, Langston said:

 What I'm doing here is making dead sure that voice coil/horn distortion is the only audible distortion in the system.

 

Somewhere in all of the posts, audible is probably defined, as ?

 

PWK seemed to have a clear grasp of Declining Marginal Returns in his designs

 

Once Design Goals are agreed, tolerances are defined and component selection and design begin.

- In most designs, Engineers are trained to use the best components available for the prototypes

Then de-rate the components to lower the BOM Cost

I had to fight the de-rating on every product with the Engineers

If nothing else, it would add months to the schedule while new components were tested and approved

 

I frequently reminded the Engineers

We need the Parachute before we hit the ground.....

 

PWK derated his Atlas compression driver to shave cost

and screwed himself in the process

Atlas dropped him.

 

It is possible, that the inductors were de-rated

and that audible performance suffered as a result

 

3 Questions

How do the factory inductors measure, and are there Rev changes

Are there audible or significant measurable changes

How much does the last 1% of improvement cost ?

 

Can any maxims be derived like:

Air core is always better

or

X design is always done for down cost and performance suffers etc ?

 

Is the Scope of your study limited to high efficiency fully horn loaded designs aka LaScala and K-horm.....Jubes

I believe the answer is yes and the "first 2 watts" for home listening.

  • Like 1
Link to comment
Share on other sites

1 hour ago, Bubo said:

Somewhere in all of the posts, audible is probably defined, as ?

That was an amazing post. I'd hate to be an engineer at a for-profit loudspeaker company. : ) Thanks for the history and questions. I'm going to focus on just one thing in this thread - surveying readily available non-air cored inductors to see if it's possible to approach air-cored electrical performance without their downsides. Much (most?) of the fun of this for me is in the learning.

 

A passing audible grade will be defined as no noticeable difference (to me) between an active low-pass and passive system with the same electrical transfer functions. This is probably achievable because I'm dealing with low frequencies where we have much less ability to hear distortion. I'm also standing or sitting and often falling off the shoulders of those who've gone before me with the measurement vs. audibility stuff. Heyser was all about that and in my feeble attempts, so am I.

 

God bless you and your precious family - Langston

 

I happen to believe that a major purpose of technical audio measurements is to assist in creating better subjective perception, and if that's not happening, then we are getting nowhere with the objective measurement - Richard C. Heyser

  • Like 3
Link to comment
Share on other sites

The following are pictures of the inductor measurement setups. I hope this will help along with the discussion in my prior posts in this thread to better understand what's going on. I hope breaking it up like this makes it tolerable. Tomorrow I'll post actual measurement plots.

 

I've gained huge respect for passive crossover engineering. It adds dozens of variables that active crossovers avoid. I'm enjoying it quite a bit. : )

 

Distortion Testing

 

Audio Precision APx515 AES3 out to Powersoft K20 amplifier (DSP bypassed). The ERSE 18awg IQ laminated steel inductor in the picture is in series with the positive amp output and then connected to the 4Ω power resistor, and the resistor's remaining wire is connected to the amp's negative (it is ground referenced, not bridged). Voltage is read across the resistor. Divide the voltage by 4 (the resistor's value) and you get current. Ohm's law.

 

There's also a Benchmark AHB2 amp involved that I use for most stuff because John Siau forgot to design noise and distortion into it.

 

1392403077_DistortionTesting.thumb.jpg.4596996640c43c4069dbc57c31232cec.jpg

 

Magnetic Coupling

 

I also measured magnetic interference sensitivity using the various cored inductors as receivers and a 2.4mH air core as the transmitter. I drove the transmitting inductor with 2V through a 10Ω resistor and I paralleled the output of each receiving inductor with 10Ω to simulate a loudspeaker load. I oriented each receiving inductor for minimum pickup before recording the result. I used 2 second sweeps from 10Hz to 20kHz. The transmitting and receiving inductors are separated by 6 inches, center to center.

 

I've seen many pictures of passive crossovers on this forum with the inductors physically aligned very poorly with each other...

 

The (3) ERSE laminated steel core inductors aligned for least coupling looked like this

ERSE.jpg.0021f9595a744b8126b98fdb4d660813.jpg

 

The Crites E-Core laminated steel inductor

Crites.jpg.9772b73ac25e4c5b9212ecafe54ee2d1.jpg

 

The Jantzen 14awg C-Coil (actually toroid) inductor

2069668163_Jantzen14C-Coil.jpg.de024e959777592b43e6d931c146f74f.jpg

 

My Air Core test inductors, both 18awg and 2.4mH, in proper alignment for least coupling. Not easy to do, 1 degree off will cost you at least 10dB

217942119_AirGoodAlign.jpg.64311342dd8c1e47fb66721e5c96df40.jpg

 

The wrong way, working like a weak transformer

1851022114_AirBadAlign.jpg.8fd9101b4a01d89cfec96793787c20ad.jpg

 

God bless you and your precious family - Langston

Edited by Langston
  • Like 2
Link to comment
Share on other sites

Today's Contenders

1586213445_ERSEInductors.thumb.jpg.3bbe2ee2262e9d28755b5c56fc1993ff.jpg

 

                      L (1kHz)  Q (1kHz)  DCR      Price

1. ERSE 14awg SuperQ: 4.004mH   21.0      0.1563Ω  $18.55
2. ERSE 16awg SuperQ: 3.849mH   18.1      0.2266Ω  $11.80
3. ERSE 18awg IQ:     4.014mH   29.9      0.4004Ω  $5.02

                     (Keysight U1733C LCR meter)

Conclusions

1. Best to worst follows price. Inversely! DCR alone improves with increasing cost. On the first point, I don't know why, I only have guesses. Someone that's into magnetics needs to chime in.

 

2. Bubo's post makes a fundamental point - keep your eye on the big picture when comparing inductors for highly efficient loudspeaker systems. If your inductor will never need to pass peaks exceeding about 2 amps (15 watts into 4Ω), any of these coils will do nicely and you can focus on DCR.

 

THD

Don't be suckered by THD for inductors (see earlier posts for why). It's only a first step in finding maximum current before ugly. I define ugly as >0.5% because (a) it's easy to stay under this hurdle with modern components, and (b) I'm confident that the cumulative ugly will remain inaudible with such components relative to even the best driver/box/room contributions.

 

Secondly, don't be suckered by wattage specs for inductors. That's the marketing dept. talking. Inductors are current animals. They saturate (ugly) at a given current level. Watts equals current(squared) times load resistance. Thus an 8Ω max will be twice the wattage of the 4Ω max. Same current passing through the inductor in either case.

 

Because I'm a sucker, I include wattage plots below to make it easier to compare with mfg. specs. BTW, the ERSE's THD specs using their 8Ω loads are accurate. Good for them. The problems is that THD and wattage are reasonably close to useless for these devices.

 

1337040626_ERSEITHD.thumb.png.44c0980b24e1be2997b188c1e1847e2e.png

 

981229674_ERSEWTHD.thumb.png.66708c94e7d31e7fc20cc14c15d02da3.png

 

IMD

This is the reference metric I've chosen for inductor current capability (see earlier posts for why). Unfortunately there's no easy way I know of to display IMD levels in amps, thus you're stuck with watts into 4Ω. If you want amps, divide watts by 4, then take the square root of that. If you want the 8Ω wattage, multiply the 4Ω number by 2.

 

1994541565_ERSEWIMD.thumb.png.187e5175e2b18ce451f3f5a72e00237f.png

 

Observation

The THD measurements of the 14awg SuperQ inductor indicate it'll pass about 8 amps (256 watts into 4Ω) with acceptable distortion. IMD measurement say's it'll get ugly beyond about 2 amps (15 watts int 4Ω). IMD is counting distortion components in the audible passband of the device and THD is mostly counting distortion that is being suppressed in the rolloff. Which one are you going to trust? We already know which one the marketing dept. chose.

 

Multitone
This is like two-tone IMD on steroids. No one seems to know how to put this into a distortion metric that works across devices and domains the way THD, IMD and others do. You can take a known good and compare it with other like devices with excellent repeatability using the multitone "TD+N" metric. Since I don't do stuff repetitively like a mfg., I normally just inspect the multitone spectrum to see the stuff that piles up between the generated tones. That stuff shouldn't be there. How much until it's audible? Depends on the device (i.e. I have no stinkin' idea).

 

Here's the reference amp's behavior driving a 4Ω resistor with 2V. This includes cabling and adapters and the very low level HF stuff may be due to the latter acting as an EMI antenna. The low drive level is representative of where we listen, so I'm not going to bother you with the high level plots (same horse wins).

 

222297872_AHB2FFTSpectrum.thumb.png.9e4b6d2260b290cdec1b41071f099e17.png

 

Both SuperQ's produce virtually identical responses, the little IQ is highlighted so you can see it in the foreground (pink). With ERSE cored inductors, cheap is cheerful.

 

82031762_ERSEFFTSpectrum.thumb.png.fd476adc0e522d53436b248e7b377d85.png

 

God bless you and your precious family - Langston

 

 

Edited by Langston
  • Like 1
Link to comment
Share on other sites

On 5/29/2021 at 9:15 PM, Langston said:

and secondly, rare for today, allowing some difficult conversations without censorship.

 

 

Let me start by saying all are welcome to this forum, but IMO anyone with your approach to your quest are especially welcome here. I appreciate your candor and approach while providing actual testing and data. This may lead you to what you are looking for and it may lead you somewhere you never expected. That is my favorite part of a scientific approach. As for what I quoted above, I wish it were completely true. You should have been here 10-12 years ago, I think you would have been well embraced by some past members. Unfortunately todays "cancel culture" has crept into this forum too. I wish some people were still around for lively debate, even if their views were perceived as outlandish. I look forward to following along your jouney!

 

Jeremy

  • Like 4
Link to comment
Share on other sites

As kind of an intermission until the last inductor arrives, I'm going to show a couple of magnetic coupling measurements that I already posted the setup pictures of. Nothing ground breaking here, but crosstalk between passbands will compromise a whole lot of time, effort and money. Is it audible? A lot of designs pretend it isn't.

 

1423954626_HeresyIICrossover.JPG.019371c37c455cef451f7c6405827e96.JPG

 

Why not do the best with what we have to work with? EAW in its glory days was famous for excellent passive crossovers, and it wasn't just component quality. They oriented inductors for minimal crosstalk and mounted each passband on different boards displaced throughout the cabinet.

 

The last inductor will be a Madisound Sledgehammer 15awg laminated steel core of very similar design to the ERSE IQ (or according to the pictures, the IXQ). Kudos to Parts Express and ERSE for quick shipping.

 

I drove the transmitting 18awg air core inductor with 2V through a 10 ohm resistor, once again because this is where we listen with efficient loudspeakers. Then I separated the transmitter from the receiving inductor by 6 inches, center to center. Lastly, I spun the receiver around until I found the least pickup orientation and made the measurement.

 

Cored inductors behave very differently than air core. For a given inductance, they transmit and receive less magnetic flux and thus are easier to place. They are also far less affected by nearby ferrous material and fields. I'm no magnetic scientist, this is what I observed fooling around with this stuff. Another thing I noticed was that air cores develop textbook like fields around them (the right-hand rule is a must-know) where there is one exact orientation between a pair of them that yields the least crosstalk. Cored inductors are much more forgiving, on the other hand, they don't have a perfect null location. Audio engineering is all about reasoned compromise, or as Heyser said, playing the part of a magician. That means hiding the ugly from perception while drawing attention to the beautiful. : )

 

All the cored inductors under test did well. The humble little 18awg IQ coil was best. Again. Ignore the power line frequency EMI spikes at very low levels - that's from my busy lab, not the transmitting coil. Now you know how field strength meters work. : )

 

457065046_MagneticCoupling.thumb.png.e1cafb0513380a624ff643b80ae4eb52.png

 

Air core to air core, ¡mano a mano! These things will marry each other if you're not careful.

 

2059409237_MagneticCouplingAirCore.thumb.png.fcff34e8cf58e02bfe035da5c9d57f35.png

 

God bless you and your precious family - Langston

Edited by Langston
  • Like 3
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...