New 4" driver from Markaudio

[HareBrained]

Mark,

An Alpair 6 was recently tested by ZaphAudio.com and the frequency response didn't look like the published specs. With the CHR, how consistent have they been? Do you measure random units from the production run? I ask only because I don't have measuring equipment and need to know I can depend on the data that's available.

I saw the dual driver measurements you posted previously. The freq response is slightly different from the brochure specs in that the upper midrange dip is not there. Is this indicative of the dual driver arrangement or could this be considered a single driver shifted 3db higher?

I've been looking at dual chamber reflex enclosures. Primary benefit is that excursion is controlled better than a standard BR. Do you think this driver could benefit from this type of enclosure?

Thanks.
John

[markaudio]

Hi john,
I'm not sure why you automatically assume that Zaphs results are accurate.

Listed below is a summary of the lengths we go to in making drivers.

Summary of some of the component testing.
1- All finished cones are weighed prior to sub-assembly. Mass tolerances are +/- 0.05g. We usually reject around 10% from each batch.
2 - All emitter sub assemblies are weighed before entering the production line. These sub assemblies are Fo tested prior to entering the production line. The Fo for this sub assembly is testing using a custom machine. The pass/fail tolerance to +/- 3Hz. We usually reject a further 7% to 9% at this stage of the operation.
3 - Voice coil winding and sub assemblies, 25% of each batch is tested. Pass rate must exceed 90% in order for the batch to be used. Below this figure, the whole batch is rejected and binned.
4 - All rear dampers are stretch tested and them weighed. Tolerances must be within +/- 2% against the Gold samples.

Production tests:
1 - We use a purpose built large walk-in anechoic chamber for our testing.
2 - Tests are conducted on infinite baffle using a calibrated lab-class 1 mic rated at 15Hz-50-kHz (a very expensive piece of kit).
3- Software used in Linear-X (anechoic) and CLIO (production)

In all, there are 43 tests carried out on components prior to sub assembly and line assembly, 17 tests during final sub assembly, line and post assembly.

On batches of 500 units, 15 units are selected at random, anehocic chamber tested, results are averaged and published. Batches of 1000, we test 30 units. Results are averaged and published. Takes quite a while to get all this done.

Most commercial driver makers work on +/- 20% on raw driver tested parameter variance in their production. The better makers get to +/- 10%. We get to +/- 5% on 80% of production and stay within +/- 10% on all production.

The dual driver bookshelf speaker was tested in the anechoic chamber so the results are "actual"not software projected. Drivers once in box will behave differently from tests conducted on baffle. Depending on the box type, design and any filtering employed, smoothing of driver output will happen along with other measured increases. So yes, you'll get some dB gain and handle more power with this type of set-up.

I recently received some results from a Japanese DIYer using the Alpair 5. I've attached his results for your interest. Iizma's results and comments are unedited and were independently translated.

Overall, I do my best to be consistent.

Cheers,

Mark.

[_henry_]

Well done Mark!

I think DIY world just makes another step further, once begin by Mr. Nelson PASS who shared his wealthy knowledge to grace us.

To have a designer and manufacturer of the driver share their details for production is a wonderful thing, and talk to us directly.


Cheers

[EUVL]

> An Alpair 6 was recently tested by ZaphAudio.com

Would you care to post the link?
I couldn't find it at Zaph or by Google.

Thx,
Patrick

[planet10]

Originally posted by EUVL
I couldn't find it at Zaph or by Google.

Top item in his blog

dave

[HareBrained]

Originally posted by markaudio
Hi john,
I'm not sure why you automatically assume that Zaphs results are accurate.

Listed below is a summary of the lengths we go to in making drivers.

Summary of some of the component testing.
{deleted for brevity}

Overall, I do my best to be consistent.

Cheers,

Mark.

Thanks Mark. My question was not to imply any issues with your production process but to increase my confidence in your products and that purpose was well serviced.

I do trust that Zaph's measurement are accurate because I have no reason to doubt them, and have no means of supplying evidence to the contrary. But I also know enough about the process to ask questions of those who do. I have observed the certain drivers panned by Zaph have some devoted followers and I'm inclined to see the devotees as more of an authority than this one man's opinion. I also trust you as an authority for your drivers.

I got very excited by the prospects when I saw the drivers on CSS and I've spent a great deal of time gathering as much information as possible. This has added to my knowledge and has bolstered my enthusiasm. When I return to working, I will be buying a set of CHR-70's for my desk and a set of Alpair 6 Rect's for my car.

[markaudio]

Hi John,
Good to hear from you. The debate generally on data accuracy is a challenge for everyone. There's several systems on the market with no agreed measured accuracy standard and test methodology.

Zaph's frequency graph is missing data in the low range. Much depends on the equipment's quality and how its set up. Usually, the major limitation with home set-ups is the quality and range of the mic. Typically, mics used are lower cost (below 2000 USD). Their actual service range is typically 50Hz to 15-kHz @ +/- 2% to 5%. Above and below this range, their accuracy fails off significantly. Most driver makers with decent set-ups spend from $3000 to $7000 on a mic to get the range sufficient for reliable consistent testing. The mic should also be calibrated to the characteristics of the anechoic chamber. With all sincere respect to anyone using a home test set-up, achieving service accuracy in their measured data is a challenging exercise.

Equally, sadly some makers have helped to create a aura of doubt by "massaging" their data to make their products look more commercially attractive. So there's merit in having guys like Zaph introduce an alternative point of view.

The comparison between the Alpair 6 and the JX92, or the W4/1337 isn't practical. The JX92 is a much larger driver. Jim Griffin's recent thread will give you a better appreciation as he contrasts the larger Alpair 10 to the JX92 in one of his build projects.

Mark Audio Alpair 10 MLTL Design

The CHR-70 is a more appropriate driver to contrast with the W4/1337 from their respective price points, relative size and performance envelopes.

My first build! MarkAudio CHR-70 in da house

I designed and built the Alpair 6 as a "connoisseur's" small project driver. It's sweetness of tone, musical nature and depth of range are its features, characteristics that cant' be measured. I've got the feeling you've already seen the other threads on the Alpair 6. Jim Griffin's and Henkjan's comments are well worth a read for an in-depth review of the Alpair 6's character.

Small MLTL for Mark Audio Alpair 6 Drivers

Anyway, I'm rambling on so I'll end here by saying many thanks for having a go with our drivers. I hope that you will have much pleasure from their use.

Enjoy!

Mark.

[CSS/XBL]

Hi,

While not totally relevant to the initiated subject in this thread, let's not forget another purpose of the Alpair 6 was it's use in line arrays (rectangular version) where it clearly is more suitable and outperforms the Jordan or TB mentioned in Zaph's commentary.

In addition, it is more conventional in that it has a spider/damper which the Alpair5s (and others) do not have. Certain OEMs won't touch a driver without a spider and OEM business is important for a manufacturer's existance.

Bob

[HareBrained]

I have a pair on the way. I plan on building a 7.5L dual chamber reflex. I'll post the final box into the Full Range forum after it's built. I'm looking forward to hearing them.

[HareBrained]

So, I built a dual chamber reflex enclosure of 8.1L (total airspace). The ports are slots with the port width the same as the enclosure depth. This allowed me to use the crossover port as the divider. The ports are 3 sq in. and the internal depth is 4". Overall dimensions are 21.5"x8.5"x5" using 0.5" materials. I selected 50Hz for the tuning for the vents, requiring about 9.5" in length. The enclosure type was selected because it provides better excursion control than a single bass reflex, which I thought would benefit the full range driver.

Overall, the sound quality is fantastic, and they're not even broken in. No crossover is used and the EQ on my computer is flat/off. Bass output is very robust (I left the stuff thin on purpose.) The midbass is a little slow but should come in as the drivers loosen up. Midrange is sweet and articulate. The treble is not exhibiting any hiss. It's a little forward of the midrange but I think a little more stuffing will bring it back.

Mark, thanks for offering such a great driver at such a reasonable cost. And as it doesn't need a xover or a tweeter, that makes them a doubly good value.