Bagby Mandolin DIY Speaker Review

[amirm]

This is a review and technical measurements of the (late) Jeff Bagby Mandolin DIY kit as built by our member, @Rick Sykora. Looks like the drivers and crossover for a pair cost US $316.42. Rick will have to advise as to the rest of the cost.

As usual, Rick has built this speaker as a tank, weighing far more than commercial speakers its size:

As expected, there is a port in the back:

Measurements that you are about to see were performed using the Klippel Near-field Scanner (NFS). This is a robotic measurement system that analyzes the speaker all around and is able (using advanced mathematics and dual scan) to subtract room reflections (so where I measure it doesn't matter). It also measures the speaker at close distance ("near-field") which sharply reduces the impact of room noise. Both of these factors enable testing in ordinary rooms yet results that can be more accurate than an anechoic chamber. In a nutshell, the measurements show the actual sound coming out of the speaker independent of the room.

I performed over 1400 measurement which resulted in error rate of around 1%.

Testing temperature was around 65 degrees F.

Reference axis for measurements was the center of the tweeter (by eye).

Measurements are compliant with latest speaker research into what can predict the speaker preference and is standardized in CEA/CTA-2034 ANSI specifications. Likewise listening tests are performed per research that shows mono listening is much more revealing of differences between speakers than stereo or multichannel.

Bagby Mandolin Measurements
Acoustic measurements can be grouped in a way that can be perceptually analyzed to determine how good a speaker is and how it can be used in a room. This so called spinorama shows us just about everything we need to know about the speaker with respect to tonality and some flaws:

On-axis response is less even than I like. We have some peaking around 100 Hz but then a broad dip up to 700 Hz. There is another dip around crossover region of 3 kHz. Tweeter is beaming due to lack of waveguide as you will see later.

Efficiency is low averaging around 83 dB (by eye) so you need a lot of power to drive this speaker.

Port resonance is not an issue but woofer seems to be running out of steam before tweeter takes over:

Early window reflections show the results of tweeter narrowing:

And here is the combined, predicted response:

Horizontal beam width shows the uneven directivity of the woofer and tweeter:

I have shown a secondary upward arrow to show the physics of beam width reducing in size as frequencies get smaller (relative to size of the tweeter). Same is seen in contour graph:

Vertical response shows more degrees of freedom than some other 2-way speakers:

Distortion is good at 86 dBSPL:

Finally, impedance at nearly 7 ohm should be easy for any amp to drive (that has the power this speaker needs):

Speaker Listening Tests
Garden season has started and with me spending a lot more time outside during fruit tree blossoming, my ears are fully plugged today. I did a quick listen and could hear the effect of bass boost around 100 Hz since I also have a room mode there. Lowering that peaked not only helped with a bit of boominess but also resulted in more clarity upstream indicating lowering of distortion. Above bass my ears were too clogged to asses anything so not much else to report.

Conclusions
I don't have any since I could not complete my subjective evaluation. You all can post what you think.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150

 

[rockyb]

I appreciate your comments on ears being plugged. I have that frequently.
That is one the reasons that I like to rely on objective measurements like your make. Hearing aids also reinforce that. I would much rather start with a speaker or speakers that measure well and make my choice than start with all the possible alternatives. Then when others with younger ears listen, the probability that they will be favorably impressed is raised.
When the conditions to listen critically, I want focus on the singer, or how that piano concerto sounds with that pianist, not how the speakers sound.
Life is too short, thanks!
Cheers!

 

[GWolfman]

Thanks!

Interesting dip in the woofer output as you noted. Initially I was thinking bad crossover, but then it wouldn't have popped right back up like that.

 

[napilopez]

This looks excellent to me! The on-axis is a little jagged, but the listening window is clean, plus a bass boost at 100Hz is often good for these small speakers. Directivity is good for a cone-and-dome without a waveguide -- wider around the critical upper mids and lower treble. So as usual, it's going to be less clean than a speaker with a good waveguide, but it will be wider. I bet with a bit of radiusing or other cabinet small tweaks could clean it up further.

Also, look at that, a budget DIY speaker with no notable port issues. Who woulda thunk it!

Weird about the woofer dip -- since it doesn't show up in the listening window.

Its performance actually reminds me a bit of both the JBL HDI-1600 and Revel M106 with the scoop in the upper bass and low mids, plus maybe a slightly elevated midrange. A little more jagged in the midrange but unlikely significantly audible so. Listening Windows compared:

And then the PIR suffers a bit from beaming above 6khz, but this is rarely a bad thing in my experience, more 'different'. In my experience, one tends to notice more the added width from 2-6kHz than beaming above that.

Since I'm here before MZKM, I'm gonna guess it's gonna be 7.5 ish for the 'with sub' Olive score?

Edit: PIRs compared:

Looking at the performance through this comparative lens, it seems like an excellent deal for someone looking to make a DIY build. Sure directivity and likely imaging will be a little less clean/precise than speakers with good waveguide execution, but you might also get a slightly more expansive soundstage too.

 

[tktran303]

Nice build Rick,

interesting that the port tuning is higher than specified by the late Jeff Bagby. It’s supposed to be 38Hz, here we’re looking closer to 48Hz. That would create the bass bump.

I wonder whether the new NRX2 driver unit affected the bass tuning, or whether this wasintentional on your part? Some people do like a bit more mid-bass punch than flat.

And of course it depends on your room acoustics and speaker placement and listening preference what kind of bass response is more or less preferable.
But if one wanted flat bass response all it needs is a port slightly longer by a couple of inches.

the midwoofer’s resonance notch around 1.x kHz (“???”); well that’s likely inaudible due to masking by the tweeter, but I find it incredible how the NFS can resolve it at such high resolution.

also the crossover is specified at 1.8Khz; it’s surprising that the tweeter rolls off earlier than that and leaves a dip in the summed response around 3Khz. Jeff’s own speaker measurement doesn’t have that.

All in all, i think I’m nitpicking here; from the lens of an amateur DIY hobbyist.

As a consumer or potential builder with plenty of time on one’s hands, as a first build speaker I think this is an excellent showing- and remarkable for such a low priced speaker.

Look at that midrange distortion at 86dB from 300Hz up! At 55dB down- that is class leading!Literally off (below) the chart!!!

 

[alex-z]

Be interesting to see how the response changes with 3/4" roundover on the baffle. I do that on all my projects and it seems to smooth out the treble.

 

[napilopez]

It's also worth pointing out that vertical directivity is also surprisingly very well controlled for this type of speaker. Just two narrow notches for the most part. Usually the waveguide-less cone and domes are a lot more messy.

 

[q3cpma]

Some remarks:
1. No waveguide => obvious directivity issues.
2. Razor sharp edges + (1) => nice example of diffraction (2 ~5 kHz on-axis dip 'n' peak not showing in the LW).
3. Incredibly low distortion: if you ignore the usually masked H2, you get < 0.3% over 120 Hz and < 0.2% over 600 Hz at 96 dB! LF distortion is also quite nicely controlled for a 6.5" woofer; the high port tuning/modest extension probably helps. Sadly, the not measured IMD is probably not as pretty.
4. Dead simple port without flaring but no issues? How/Why?

Looks like a solid DIY kit, though it'd be nice to have a waveguide version (e.g. DXT or something).

 

[H-713]

Actually looks pretty good.

I'd rather see no waveguide than a poorly executed waveguide, and doing a waveguide right is no mean feat.

 

[Francis Vaughan]

I'm quite interested in these results. I designed my desktop speakers using much the same drivers. I ran into a couple of issues that look as if they are present here as well.
I ended up using a couple of elliptic sections in the crossover to tame the mid-woofer's behaviour above the crossover frequency. (This is basically what the Neville Thiele Method is so far as I can see.)
I suspect that both the "???" on the driver components near field graph and the rise in distortion above the crossover frequency are a result of the mid-bass driver's bad behaviour still being evident. The "???" is not so much a question of the dip as a question about the rise after the dip.
Not rounding the edges is always going to lead to problems. I tend to be quite aggressive about this. Also, mounting the drivers close to one another helps in near listening. I used the plastic surround version of the tweeter, so creating a cut-out to fit the mid-bass was trivial.
Since mine are used for close in listening on the desktop there are a lot of other differences. But mine are sealed and designed with the intent that they be pairs with a subwoofer.
But no matter what, you get very nice results for the money with these drivers. I do suspect that a bit more tweaking of the crossover and rounding the box corners would have lifted this little speaker up a notch.

 

[restorer-john]

I'm quite interested in these results. I designed my desktop speakers using much the same drivers. I ran into a couple of issues that look as if they are present here as well.
I ended up using a couple of elliptic sections in the crossover to tame the mid-woofer's behaviour above the crossover frequency. (This is basically what the Neville Thiele Method is so far as I can see.)
I suspect that both the "???" on the driver components near field graph and the rise in distortion above the crossover frequency are a result of the mid-bass driver's bad behaviour still being evident. The "???" is not so much a question of the dip as a question about the rise after the dip.
Not rounding the edges is always going to lead to problems. I tend to be quite aggressive about this. Also, mounting the drivers close to one another helps in near listening. I used the plastic surround version of the tweeter, so creating a cut-out to fit the mid-bass was trivial.
Since mine are used for close in listening on the desktop there are a lot of other differences. But mine are sealed and designed with the intent that they be pairs with a subwoofer.
But no matter what, you get very nice results for the money with these drivers. I do suspect that a bit more tweaking of the crossover and rounding the box corners would have lifted this little speaker up a notch.

Show us some pictures of your build- I am interested.

 

[tktran303]

Well I don't care for waveguides. I'd rather a ribbon tweeter.

but on closer inspection I think the dip (3Khz) and peak (5Khz) that is visible in the on-axis response, but not present in the listening window; is thus
a) due to the sharp edges of the box causing diffraction effects
b) averages out to to absent in the listening window.
l

I always wondered why my speakers with large round-overs measure better on axis, but after all that effort, I’m not convinced I can hear the difference.

 

[thewas]

I was never a fan of the sound character of such ring radiator tweeters as their beaming increases significantly in the upper octave, would be interesting to see if that effect can be compensated with a good waveguide like for example of the SEAS DXT.

 

[MZKM]

Preference Rating
SCORE: 5.4
SCORE w/ sub: 7.4

Sensitivity: 82.9dB (300Hz-3kHz ; spec: 85dB)
Frequency response: +/-2.5dB 80Hz-20kHz

​

 

[Zvu]

If port was longer and tuned to 38Hz per Jeffs recommendation (not 48hz like shown), there wouldn't have been as large hump at 100Hz but the port resonance would fall around 600Hz and would be as high as woofer response. I know because i've measured it.

Dip in woofer response at 1700Hz is cone/surround termination left untreated characteristic for sb17nrxc.

Distortion under 150Hz is so bad. I really don't like it. This tweeter, being ring radiator, needs to be tilted slightly upward to even out in power response and to sound good.

 

[MZKM]

Wonder how they got the frequency response listed in the write-up:

EDIT: I guess the port tuning like others brought up, the write-up actualiy says:

Designed as a vented speaker with a rear mounted port tuned to 38Hz, the speaker has a -3dB point of 42Hz with a -6dB point of 38Hz. Unlike the Piccolo, the Mandolin is tuned to be exceptional flat to its bass roll-off with no upper bass rise.

 

[MZKM]

Anyone got a suggested price?

The bare-bones version doesn’t come with the port and other stuff, so you gotta choose the full package, then you have MDF costs. Can this be made out of a single 48x96 board (I believe so), and so keeping the options for everything at the default I am thinking of putting it at $380.

 

[Everett T]

Anyone got a suggested price?

The bare-bones version doesn’t come with the port and other stuff, so you gotta choose the full package, then you have MDF costs. Can this be made out of a single 48x96 board (I believe so), and so keeping the options for everything at the default I am thinking of putting it at $380.

According to meniscus it's $348 for the full kit. Your price with a sheet of mdf would be about right.

 

[Everett T]

Some remarks:
1. No waveguide => obvious directivity issues.
2. Razor sharp edges + (1) => nice example of diffraction (2 ~5 kHz on-axis dip 'n' peak not showing in the LW).
3. Incredibly low distortion: if you ignore the usually masked H2, you get < 0.3% over 120 Hz and < 0.2% over 600 Hz at 96 dB! LF distortion is also quite nicely controlled for a 6.5" woofer; the high port tuning/modest extension probably helps. Sadly, the not measured IMD is probably not as pretty.
4. Dead simple port without flaring but no issues? How/Why?

Looks like a solid DIY kit, though it'd be nice to have a waveguide version (e.g. DXT or something).

Looks like the full kit comes with a flared port at 1 end, FWIW.

 

[Jim Matthews]

Well I don't care for waveguides. I'd rather a ribbon tweeter.

Waveguides permit use of less expensive components with lower crossover points. Ribbon tweeters have a long and expensive history of spontaneous combustion under load, sacrificing many an unwary amplifier to their pitiless god, Voltar.