Midrange design study...

[Evil Twin]

This is a very limited and focused "Design Study", principally investigating the feasibility of using a planar midrange in an existing proposed system design, and determining what system parameters may be necessary to "extract" acceptable results from what is not a simple driver to work with.

This is the published response data for the driver under investigation, the GRS PT5010.




In the manufacturer data sheet, many standard specification items are listed as N/A, even when as detailed examination of other data indicates relatively simple derivation.

This is an impedance curve measured on the first test sample...




This effort has involved developing a driver model and investigating theoretical efforts to produce more satisfactory results in a system build than the data sheet hints at. Due to the mandates of SMJ Imperial management, the details are considered proprietary at this time, though that status may change in the future.

This is an interim report, as another update to the test configuration is already in preparation. But the initial results from the first pass exhibit some useful and encouraging characteristics worth discussing in passing...


The impedance curve in the test configuration...




And the measured response...





It might be worthwhile to note that the measured HD3 at this level of output was at ~-70dB from 700Hz to 5kHz. A most satisfactory result. No electrical networks are involved at this point.

Further suppositions have been made from this first test configuration, and a second one is in construction which if insights are correct, should further improve the results.

[Evil Twin]

Two new design paths to evaluate have been defined, and materials ordered- it is expected that new data may be available by the end of the week, as construction has already commenced on the test configurations, and a previously planned trip under consideration to the outer rim next weekend has been cancelled, due to pandemic issues with the target group.

Of course, I always wear a mask when around others, but for reasons I cannot fathom, there seems to be some perceived social stigma about that still... in spite of new trends and a much wider range of infection including vulnerable populations.

[Dave Bullet]

Two new design paths to evaluate have been defined, and materials ordered- it is expected that new data may be available by the end of the week, as construction has already commenced on the test configurations, and a previously planned trip under consideration to the outer rim next weekend has been cancelled, due to pandemic issues with the target group.

Of course, I always wear a mask when around others, but for reasons I cannot fathom, there seems to be some perceived social stigma about that still... in spite of new trends and a much wider range of infection including vulnerable populations.

Dark Lord, as you have been wearing a mask for quite some time (color coordinated of course), it is not a surprise the humans of this planet resist such necessary devices. Maybe if they had a trip to Mustafar, they would understand.

[Evil Twin]

Further experiments and measurements have provided increased levels of insights to the challenges and methods to match them...

Another update will be constructed shortly, perhaps finished before the next run to the Outer Rim this weekend, but some progress may be worth sharing at this point...






Distortion values show considerable promise when placed in perspective with output level...





Calculations are complete for the next iteration, based on test data collected today.

[Evil Twin]

Perhaps it should be noted for any interested, that the current focus is to build on the characteristics of two recent experiments, and if feasible, merge their qualities and impact in the lower part of the driver range and the upper part, as shown here... this is the current target of development operations.

[Evil Twin]

GRS is now making things a bit easier for those interested in exploring the potential of their versions of the Neo drivers.

They are now shipping mounting kit accessories, which, though they come with no instructions and have hardware that might be considered a bit less than robust, still offer a clear step forward for implementing system designs.


Here is a sample view of the hardware kit for the PT-5010 driver...





Note that in this view, the top plate is actually upside down- the bevel edge is designed to meet the slope of the driver stamping.


This kit appears meant to be used in conjunction with the 3M adhesive gasket supplied with the driver, but it appears that now this gasket should be used on the front side of the driver.





This is how the front of the 5010 will look with the gasket installed.





Placing the mounting plate on the driver, it is ready to be bolted up with the supplied hardware, which seems to be a metric size close to 4-40.

[Evil Twin]

The supplied nylon lock nuts should eliminate the likelihood of the screws coming loose...






This side view highlights how flush the driver faceplate and the mounting panel are after assembly.





It seems likely that a dual rebate depth profile will be desired for the driver and plate installation.

[JonMarsh]

The next step, logically, in this investigation, is the "matching" tweeter and optimizing it's characteristics, if possible, for the intended application.

This is not a trivial effort, but the clues for how to proceed are once again to be found in the factory data sheets, especially that for the "dipole" version. This characteristic reveal is quite similar to that of the midrange panel.




Here, one can see the affect of the fundamental resonance on the relatively lower frequency range, peaking between 500Hz and 700Hz approximately, with a soft slope downwards with increasing frequency until the natural radiation resistance and output starts climbing at around 1500 Hz. Note that the nominal playback level as shown in the PT-2522c closed back version of 93 dB is not reached until above 4kHz. The closed back version data sheet curve looks relatively flat, but the fundamental resonance appears to be pushed up to 2kHz, with a 12 dB/octave roll off below that commencing immediately. This characteristic is not favorable for a preferred crossover point of 2500Hz, possibly manageable only with an LR4 acoustic target and 2nd order electrical network. An LR4 network is not preferred due to the lower power response in the crossover region and more limited vertical beam width.

The impedance curve of the PT2522 in dipole mode correlates and is revealing of the physical behavior also.



What might be desired is a characteristic tuning below the frequency point of the PT-2522c, and with a low frequency roll off curve amenable to preferred crossover options such as the LR3 I have historically employed with most satisfying results.

A prototype construction was prepared and evaluated with modified physical characteristics, and mounted with the new faceplate kits.





Of course, an initial check was made with just the impedance plot, to see if what it reveals about the driver characteristics was on track for the proposed development.





This showed the desired "plateau" effect impedance curve (as per experience with the PT-5010 panels) in the preferred frequency range.


After this confirmation, the test article was mounted on a 15" x 9.5" front panel enclosure, and some tests were conducted at roughly 100 dB output level, including distortion checks.







At roughly -60 dB for HD3 distortion, this behavior is acceptable.


15 degrees and 30 degrees showed similar behavior; 30 degrees off axis is shown.






45 Degrees still shows extended response, better than most 1" dome tweeters. But this must be considered in light of the higher level at 0 Degrees...





All in all, for now, this shows the desired progress, though perhaps not the final target yet. But clearly, a Proof Of Concept crossover design check is feasible with the current data, after the midrange enclosure update test is completed.

[JonMarsh]

And, yes, for those curious, I have been accepted tentatively as a Sith Apprentice to ET, pending my progress and future achievements under his guidance...

[Evil Twin]

Additional professional drivers of moderate cost but claimed high performance are now under consideration for the implementation of the proposed SMJ-40. Pricing is in the range of moderate cost Scanspeak tweeters.


These were sourced from another quadrant, but delivery was impressively fast- Imperial shuttle direct could not have bested their service...

A suitable acoustic test panel is being prepared- initial impedance measurements are already complete.

[technodanvan]

Geez those are purdy...

[Evil Twin]

These Beyma parts look and feel (weight and massive construction) like the professional components they they are reputed to be- like all of Beyma’s other components. Published 3rd parity test data looks promising. The smaller tweeter may be a candidate for a PuriFi system configuration under consideration also.