Measure once, cut twice

October 7, 2021
 by Paul McGowan

Sorry for the opening humor.

Measurements are essential. Without them, we could never design products. Especially speakers with their complex crossover and cabinet interactions.

Sometimes you have to do what you have to do to make things work. Take for example the problem of measuring speakers without getting reflections from walls and floors. If you’re Harman International you probably have an anechoic chamber. If you’re not, you improvise.

I just wanted to share this with you. Speaker designer Chris (on the right) is operating the measurement system while mechanical engineer, Chet is manually pivoting the speaker to a specific degree mark on this table built by the team.

This arrangement allows extremely accurate measurements of polar responses down to about 60 Hz.

I am always impressed by the ‘get-er-done’ attitude of our team.

Subscribe to Paul's Posts

64 comments on “Measure once, cut twice”

  1. I have no problems with improvisation when done correctly. However, Paul, you always stressed the superiority of line-source designs concerning reflections from floor and ceiling and the special voicing of loudspeakers designed for being placed directly at the front wall. As the coming FR30 is not a line source design I wonder how it is compensated for early floor-reflections of the subwoofer part? The photo shows a most non-realistic measurement avoiding early floor reflections.

    1. Hi Paul, you are correct but that’s not what they are measuring. Instead, they are measuring higher frequency polar responses.

      Indeed, this is not a line source and couldn’t effectively be one given its height.

      I will have Chris make a few guest posts and videos to explain how he figured all this out and what he accomplished. It’s best he shares that with us.

        1. In the meantime, here’s a litte rough draft to digest.

          FR-30 Speaker

          FR-30 is the flagship loudspeaker offering from PS Audio and the blueprint for our future offerings. A clean, modern, elegantly slim 3-way tower speaker that is our modern take on our favorite planar magnetic hybrid loudspeakers of the past. The clarity and detail of the most modern reference design with an ease and musicality and power that you’d expect from PS Audio.

          Premium Custom Transducers:
          Push-Pull Planar Magnetic Midrange and Tweeters:
          • Inherent linearity though symmetrical “push-pull” neodymium motor structure and directly driven ultra-low mass diaphragm
          • None of the cone breakup, inductance modulation, hysteresis distortion than plagues traditional cone drivers
          • High efficiency and durable – high strength N52 magnets and Teonex diaphragm with premium aramid fiber damping.
          • Extremely fast clean decay
          Low Distortion Woofers
          • State of the art woofer design with massive 12-pound motor structure matches the effortlessness and transparency of the planar magnetic drivers
          • Unique split magnetic gap for linear motor force versus excursion
          • Double faraday rings for low, linear inductance and wide bandwidth.
          • Symmetrical double spider with integrated lead wires and low loss inverted rubber surround

          Passive Radiators
          • High compliance, long excursion design for maximum linearity and extended deep bass extension without the compression and distortion of ports

          Enclosure Technology: Extensively braced MDF enclosure with high grade “piano gloss” lacquer finishes. Front baffle is formed from a dense, rigid and well damped high-tech thermoset fiberglass resin composite material with integrated high frequency acoustic waveguides.

          Modular design with removable head module (containing mid-hf drivers and an acoustically isolated crossover network.

          Tower speakers feature a “floating” design with curved bottom and sturdy anodized aluminum stand. Large milled brass spikes with removable tip and integrated plastic feet for hard surfaces and to air in speaker placement

          General Specifications

          Color Options Charcoal grey
          Enclosure Type Two piece – Sealed upper cabinet mid-tweeter module, Passive radiator lower cabinet (4 10” side-firing passive radiators)

          High Frequency Transducer 2 x 3” planar magnetic with Teonex diapgragm
          Mid Frequency Transducer 10” planar magnetic with Teonex diaphragm
          Low Frequency Transducer 4 x 8” woofers, cast frame, aluminum cone, 12lb advanced magnet structure
          Crossover Frequency 400 Hz, 2500 Hz – LR4
          Sensitivity (2.83V @ 1M) 88 dB
          Nominal Impedance 4 ohm (3.2 ohm minimum impedance)
          Recommended Amplifier Power 100-600W
          Frequency Response 28Hz-20kHz (-6dB)
          Dimensions (HxWxD) 60.5” x 16” x 25.75” with base (10” width not including base)
          Net Weight 230 lbs (104.5 kg)

          1. 104.5kgs…holy sh!t; it’s definitely an ‘FR’
            …it even weighs the same as me.
            When do we get the internal construction
            & bracing schematic drawings 🙂

          2. Just a correction, I’m certain you didn’t mean “air in speaker placement” but “aid in speaker placement” concerning the feet/spikes.

            My grandfather said “anyone who doesn’t make mistakes doesn’t do anything”.

  2. Does this mean the claims concerning the discussion what’s „full range“ (and therefore below 60 Hz) are open to interpretation as less accurately measurable and extremely room depending? 😉

    1. Jazznut. No, it doesn’t mean that. The low end response is easy to measure by close miking techniques. What Chris is looking for in these measurements are what’s called a polar response: how the speaker, above a certain frequency, distributes its energy on an off axis.

      It’s really important in a properly designed loudspeaker to have even amplitude (loudness) off axis. Many speakers do not. This gives you a waterfall response like those you often see in Stereophile.

        1. CtA,
          Given that this is PS Audio’s first commercially available loudspeaker & given that CB is genius enough to know how to accurately measure what needs to be measured in loudspeaker design & production without having to over-spend on measuring equipment (said Klippel), it makes perfect sense that PS Audio don’t have their own…yet.
          I suspect that once PS Audio goes into full production of maybe 3 or more loudspeaker models *then* it would make sense to spend money on a dedicated Klippel & possibly even a dedicated anechoic chamber (measuring room)

          You could, of course, cut a big cheque & donate enough money to PS Audio for them to buy their own Klippel, if you feel that owning their own is so important 🙂

          1. Given that their first speaker appears to be over $25k US dollars, and given Chris is a “genius” of measurements, I would have assumed that Paul would have invested in R&D. You can even rent time for a Klippel NFS test.

            It is fascinating that Chris is “science driven” as Paul stated, but you still continue to deny or ignore that modern speakers are designed using science. Even the link that Paul provided from Audioholics is based on a lot of the science from Toole. Did you take the time to read it?

            A Klippel NFS is roughly $100K. A properly designed speaker is “priceless”. And, in addition, accounting rules tell you that you can amortize the investment over time. Just look at the current prices of PS Audio gear. Paul is trying to upmarket his products, thus a starting $25k speaker. You would want to back up your commitment.

      1. Duh, I feel like an idiot. I did not click on the link in your comment. As soon as I did and saw the waterfall graph I knew immediately what you are measuring.

  3. Why are they doing the measurements with the speaker and the microphone ten feet above the floor? Is this to eliminate some type of reflection from the floor?

    1. I think Paul needs to do a follow up post to explain exactly how they performed this speaker test. Otherwise we can only use our imaginations.

      Looking at this photo again I would title it “Brute Force Speaker Measurements”

      1. No, not at all. What Chris is working to test is the polar measurements of the speaker above a certain frequency (around 200Hz). This involves mainly the upper end of the woofers, midrange, and tweeter.

        The whole goal is to get the speaker enclosure up off the floor and away from boundary walls, which is why it’s being done in the warehouse.

        We could just as well have hauled the speaker up on chains, except we don’t have a crane.

    2. Exactly, Tony. It’s just a way to get it off the floor and we’re in the warehouse to get it away from any walls. That way, only the pure output of the speaker is being measured. Down to a specific frequency, it’s an anechoic response.

    1. Yes, JB4, to mimick an anechoic chamber. The test equipment he’s using to measure with is pretty fancy and has a gated response so he can see where any reflections of sound happen and ignore them. It is a perfect anechoic response down to a specific frequency.

  4. Y’all get an “A” for improvisation!

    For those who don’t know this, open the pic in it’s own tab and you can see the rest of the photo. On Windows machines right click the pic and select open image in new tab. It makes more sense when you can see the whole pic with the sound absorbing material at the left. Maybe it’s only an issue for those of us using the Chrome browser?

    Must be a cool place to work, I’ve had plenty of “shirt and tie” jobs but never a “t-shirt and shorts” job.

    1. Thanks for the tip on opening the image in a new tab. I wasn’t aware of it. I’ve been working with Windows for years and I keep learning new things about it.

      1. Tony, that’s just to negate any reflected sound or more precisely reduce it to a point where it’s not a factor. Intensity of sound waves drops off at a rate equal to the inverse square of distance so by the time the sound waves make it back to the microphone they probably won’t even be picked up, or noticeably affect the readings as constructive (or destructive) interference. That’s my take anyway, I’m certain someone will correct me if I’m wrong.

        Glad to know the tip about opening images in a new tab helped!

  5. I suspect there will be several Hernia diagnoses in the next several weeks.

    I understand that Chris is a real innovator and a virtual Encyclopedia of Speaker designer but how can you be certain with a make-shift setup like this that the resultant measurements are going to be accurate? No insult intended but it sure looks like a hell of a mess.

    1. Well, that’s the thing. It may look makeshift but, in fact, all one needs is the appropriate distance between the speaker and any boundary walls. Doesn’t matter how you get there. The actual measuring equipment he’s using is state of the art. It’s juts the contraption to get the speaker off the floor and away from the walls that looks crude.

      Measurements like this are as perfect and accurate as if he were using a fancy lift and wearing a white lab coat. 🙂

          1. That makes sense.

            You are a virtual encyclopedia of everything my friend.

            I used to practice Vedic meditation which I should have never stopped but that was not my mantra.

        1. Good morning Neil!
          For what it’s worth, I’m with you on that man!
          Granted that I can’t see the picture that Paul posted in the early morning posting, I both get and understand what Old Hardware Tech is talking about.
          Sure, I can perform that same trick, that he brought up, but why bother if you can’t see it anyway?
          Now to the meet of my comment.
          When I was in my late teens and early twenties, I needed a pare of speakers, that could play really loud, but give me the really low bass notes too.
          After shopping around for PA speakers, I found that none of them suited me.
          Because of the fact that I DJed at a lot of clubs and also, I also played block parties, I needed the speakers to sound the same, rather I was either inside or outside of any building.
          I ended up building my own speakers to take to work with me.
          But after I built them, I took them outside of my house and tested them.
          The only wall reflections I had if any, were off the front wall of the old ladies’ house across the street from me.
          But sense my back yard was open with a running stream going threw the center of it, I took the speakers from the front yard to the back yard.
          Ok, all well and good right?
          I wanted to see and hear what those speakers would do, without any walls to reflect off of.
          I once owned a sizeable piece of property on the south side of Jacksonville.
          And so, I took them out there, and sat them up.
          In spite of the fact that, there were no walls anywhere to be seen, my Super Blasters not only played really loud, they also made the ground feel like it was moving.
          How does this sound for a pare of kick but speakers?

            1. Hi again Neil!
              In spite of the fact that I used a pare of 1800 watts per channel Peavey amps that I bridged to get 3600 watts out of both of them to drive the speakers, they sounded a little too powerful.
              And so, I had to turn the mane volume on my DJ mixer just about all the way down, so that it wouldn’t blow all the people out of the clubs I played in.
              But for block parties, I razed the mane volume a third of the way up.
              Some of the block parties I played at, the next door Nabors, called the police on me.
              They were complaining about the music being too loud.
              But the people I played for, were digging the earth shaking bass that my Super Blasters produced.
              Some of them asked me, “why did you turn it down?
              We were enjoying the bass that makes you shake your romp.”
              I said, “the cops are here.
              The Nabors are complaining about the music being too loud.”
              A young lady said to me, “oh man, that’s no fun!
              Who the hell called the cops?”
              I couldn’t answer that question for her, because I didn’t know who called the police.

  6. “distributes its energy on an off axis”

    Wow, I guess. It appears the mic is positioned several feet below the bottom of the speaker! What value is that since it would never be listened to from that position? Why not beside or above the tweeter for “off axis”?

  7. The “measure twice, cut once” rule works for most rough carpentry, but for precision work you have to measure three or more times if the first two measurements don’t exactly match. On precision fit pieces I often cut twice, the first cut being intentionally a tiny bit longer than the measured mark, and then, after a test fit, a second cut to shave off the tiny amount of material required for a perfect fit. Alternatively I use a sanding block to remove the last bit of material. With only one cut, no matter how carefully you measure and cut (unless you have ultra-precise measuring and cutting equipment) the piece can end up slightly too long or short.

      1. Wood putty on a nice piece of furniture or hardwood floor? No thank you. Wood putty is the mark of poor craftsmanship. It discolors, shrinks and swells at a different rate than the wood. Even under a paint finish, wood putty can absorb and release gases, and expand and contract at a different rate than the wood, causing the paint to crack, blister or peel.

  8. The more I look at that photo the more nervous I get. It looks like the center of gravity of the speaker and table is dangerously close to the edge of the supporting box below. That worker to the right, who resembles Doug Heffernan, looks a little apprehensive, as though he’s thinking “Careful Buddy! Don’t expect me to run over there and catch that thing if it falls.”

    Are the FR-30s available two-toned as shown in the photo, or would that cost extra? 🙂

    1. Yes, vertical polar measurements are a little tricky to do with a speaker this large. The design axis (tweeter in this case) needs to be in the center of the turntable, which is why we needed such a large structure.

      The guys at NWAA labs use a decommission nuclear reactor to do their measurements on big professional speakers 🙂

    1. gws,
      I suspect that that is exactly why the dude in the salmon coloured shorts & blue/grey T-shirt is hanging on to the aluminium framed rig that the FR-30 is resting on 🙂

      1. Good morning FR. Funny thing, I had a strong feeling you were going to reply to my worried comment – thank you.

        Truely, if it were our shop, we would have put a strong and substantial wood framed lip around the edge so there would be NO worry about a possible overstep of those casters. It is just our work ethic and surprised it was not implied here. (or there) (O;

        It’s all about the safety in the workplace, for the employee AND the product. Ranting concluded.

  9. I have always been fortunate enough to have an anechoic chamber when I was concerned with loudspeaker response. I used to do the “third party” measurements for some major manufacturers of loudspeakers as well as tests for other manufacturers (how many plays does it take before the frequency response of our new cassette tape formulation degrades as compared to our old formulation and the competition’s cassette tape formulation, for example).

    I don’t have an anechoic chamber now, but I do have a mannequin for testing earphones.

    In any case, most anechoic chambers are barely functional below 100 Hz because of the nature of the wedge design. So, while frequency response curves are useful, polar plots are iffy below 100 Hz anyway. The chamber that Harmon has and the one at NRC in Ottowa have the same low-frequency limit issues.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram