Little woofer tower for litte panel?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Little woofer tower for little panel?

Hello folks,

I am considering getting one of the ER Audio mini electrostat panels, or maybe some other small electrostat panel (if you can name one) for use as desktop speakers. These panels about about 18 inches tall, and I figure that instead of making them even taller by putting a woofer underneath, why not make a mini woofer tower on the side.

In this case, I might use a bunch of little 4" speakers and put them in a vertical array off to the side. I only need this to play down to 80 Hz, since I can cross over to a sub.

One idea was an open baffle, but I think I simply don't have any more room to the side. The speaker will already be up against the edges of my two LCD monitors, and I could probably only have the total speaker system be 12" wide or so. I don't know if it "counts" to have the electrostat panel or monitor be a "baffle" on one side. On the other side, I suppose I could have the baffle extending backwards? But in general it seems I don't have enough room to make baffles to support 80 Hz.

How about a sealed vs ported box? Front or rear port?

Or maybe just have one larger woofer after all, but where could be another space-saving place to put it, besides directly below the panel? (It would make the speaker overall too tall.

Thanks
 
Last edited:
I've built two bass towers to complement my Acoustat panels. I use four Dayton Audio RS180-4 7" drivers in each tower. The Dayton drivers are pretty cheap considering the build quality and the low distortion (and they look great :)). The towers use an U-shape and about 8x30x9" (WxHxD). They operate in the range 40-250 Hz and in this range they have enough max SPL for my living room which is about 180 ft2. Here are some pictures and measurements.

If you're going to XO to subs at 80 Hz and use at close range, two 7" drivers in each tower will give you more than enough SPL. The towers would then be about the same height as the mini panel. Going with a smaller woofer means you could make a slimmer tower but you will lose max SPL because of the lower effective driver area. I would then chose a driver with higher xmax like the Tang Band W6-1139SIF for example. Note that you might need a digital EQ like the miniDSP or similar to fix the frequency response of the open baffle (see measurements in my photo gallery above). So bi-amping and active XO is recommended.

A sealed box would give you higher max SPL of course, but also a different radiation pattern compared to the mini panel. The OB towers integrate beautifully with the panels as they have a similar radiation pattern in the XO region. The sealed boxes will also be much larger if you're using the same number of drivers. I wouldn't even consider a ported box if you XO to subs at 80 Hz. Linkwitz has lots of information about OB speakers, including an Excel sheet for calculating max SPL of an OB speaker.

Btw, I'm really interested to hear what you think of the mini panels as I've considered getting a pair myself.
 
Last edited:
Hi,

OB woofers mate the better, the lower the Xover frequency.
Its not easy to achieve a decently flat amplitude response with OBs as they tend to exhibit a comb-filter response with rising frequencies.
I'd rather use the array of small drivers in CB, as this assures not only the highest prescision at lower frequencies but also the more compact casing.
As desktop speaker You probabely are quite restricted regarding depth and width of the casing.

jauu
Calvin
 
Member
Joined 2004
Paid Member
Probably CB is closed box = sealed box.

Sometimes (like today) on craigslist one sees old Janszen models that use square ESL sections. Perhaps those can be re-purposed?
 

Attachments

  • Janszen Z-412HP.jpg
    Janszen Z-412HP.jpg
    15.8 KB · Views: 346
Hi There !!

I used a 5.25" CB woofer for use with my 9.75" tall desktop DIY ESL and the two worked together seamlessly!!

I could not locate either driver in any of the material I had sampled on them, They just worked as a whole like there was only one driver.

The First FR chart shows the interaction of the two drivers as measured from my listening position as showed in the First picture.

The next 3 FR charts are more precision measurements I did with less room reflections and are calibrated in true SPL levels at a nominal 95db or so at 1 meter.

If I run this at a more normal lower level of 85db (nearfield at 2' to 3') then the bass response becomes much flatter and more extended.

This was about the maximum that I can get out of the one little woofer, and if I had maybe 3 or 4 of them it would suit this little panel quite nicely as it would do in excess of +105db all by itself.

At this level (95db) the woofer was at its Xmax mechanical limit for the lowest frequency's (40-60Hz) but it still sounded clear and non-distorted.

Kick drums and Bass guitars did sound amazing, very clean and punchy even though I thought that the cone was going to come lose and fly out at me!!! ;)

I was measuring somewhere around 1cm of p-p excursion and maybe slightly more by a few mm or so on the very same tests that I made these FR charts.

FWIW

jer :)
 

Attachments

  • old esl.jpg
    old esl.jpg
    128.7 KB · Views: 338
  • fr4.jpg
    fr4.jpg
    71.9 KB · Views: 338
  • fr1.jpg
    fr1.jpg
    88.1 KB · Views: 329
  • fr2.jpg
    fr2.jpg
    85.5 KB · Views: 325
  • fr3.jpg
    fr3.jpg
    84.6 KB · Views: 60
Last edited:
Ahhhhh,ha,ha,ha,ha,ha,ha,ha!!!!!!

Ya, Well.........I did get nailed by a full 10Kv from my power supply off of that red wire on the top of the woofer box that day!!!! ;)

That is the one feeding the diaphragm and I am quite lucky to be able to be typing this today!!!! He,he,he,he :)

I was reaching across the top to re-position it and I was busy watching the SPL meter and some how in that situation I manged to graze upon the open connection of the twisted wire ends!!!!

Any other time I can handle the panel without getting shocked, but I guess I missed the panel and caught the wire........Eeeeeeeeeouwie!!!

:eek: :yikes: :wave2s: :confused:

:cheers:

jer :)
 
He,he,he.....Thanks Calvin !! :)

Not to go off topic but for those reading this, I have been zapped before from this supply.

But, Not at full output as it can deliver about 30 to 40 watts (input power) continuously at 13.6Kv and I didn't have a current limiting resistor hooked up!!

This is was a perfect example of why you should have one hand in your pocket when working with such voltages, and, I do this religiously and with reverence when working with such voltages.

Or else the fun stops!!!!

Voltages such as these come out of the step-up transformers too and I have been bitten by them as well!!

So this instance is not just pertaining to the just the Bias Supplies.

I did feel the charge go down the side of my body and not across my chest.......Thankfully!!

But that did knock me back a bit!! ;)

FWIW

jer :)
 
gerald and Andersonix:
Do you know what frequencies these little panels you mentioned play down to?

Phazer: Thanks for the info in your post. I started another thread where I ask about making a full size tower for my larger electrostat speakers.
I can't tell due to the foreign language, but it seems like the speakers without EQ drop about 15 db from about 50-400 HZ? It seems your speakers have a Q of 0.5 whereas it is recommended to have a Q around 1. I saw somewhere that a U frame boots Q but a little bit, but so does adding some mass on the cone. Did you try that?

Calvin: A few years ago, you were promoting dipole woofer towers for electrostat speakers. Did you change your mind?
 
"The panels are quite sensitive at 88 dB/W/m but are only useful to say 200 Hz"
As quoted from here,

The Electrostatic “Sonic Panel” – Interview with Charlie van Dongen and Rob Mackinlay

These panels are very similar to my own design only I made the panel in the picture that I posted in 2003!!

I could get down to the 200Hz range with my panel and was about the limit for my design as the diaphragm starts clipping into the stator's at that point using a D/S of only .072".

At that kind of displacement they are still very very loud even at 200hz to 300hz!!!

The resonate frequency of my diaphragm was about 70Hz to 90hz depending on how much heat I gave it for tensioning and about 110Hz for a resonance was the highest I ever got for the 3.75" x 9.25" panel.

The resonance in this range was a bit annoying when I listened to them with no filter and over accentuated the bass signals at moderate levels causing the clipping to occur very easily at the low end.

They just aren't designed for bass.
I ended up increasing the D/S to about .090" or so to combat the issue and they never clipped again, but I didn't do any experimenting with taming the peak (Q).

I get a similar efficiency as well at about 89db.

Typically 88db to 91db as measured but room reflections had a lot to do with the peaks at certain frequency's as well.

Once those room reflections where solved and eliminated they are ruler flat as shown in my FR charts.
This is typical for any type of ESL really, from a nearfield position.

I have some photos of this in a thread back in 2010 using Audacity for measurement.

These charts here were done at 1 meter I believe and with only a slight EQ'ing as I don't have an electronic crossover made for them yet.

I will be building some new ones very soon as those are now DOA and EOL due to very heavy testing and are beyond any capability of being repaired.

For high powers step up transformers are the limiting factor for the lowest frequency's at about 300hz to 360hz at very high SPL's.

Although there are ways around this and one of them being that if you are using lower input voltages than you can get down lower.

But it is not necessary as the actual physical dimensions are working against you at the lowest frequency's anyhow.

These theory's and rules are discussed in great detail in other threads.

You can pretty much determine most of an ESL's performance by its dimensions and knowing what voltage levels that it is running at.

Believe it or not, But this calculator is quite accurate!!

Electrostatic Loudspeaker (ESL) Simulator

I have many spreadsheets as well that I work with, that can be found in these threads, to help determine the performance of ESL and Dipole systems.

Hope that helps you. ;)

jer :)

P.S if you would rather save some money and try building something similar, Take a look at this thread as I also go into the performance predictions of such a system as well,

A Segmented Stator Desktop ESL
 
Last edited:
gerald and Andersonix:
Phazer: Thanks for the info in your post. I started another thread where I ask about making a full size tower for my larger electrostat speakers.
I can't tell due to the foreign language, but it seems like the speakers without EQ drop about 15 db from about 50-400 HZ? It seems your speakers have a Q of 0.5 whereas it is recommended to have a Q around 1. I saw somewhere that a U frame boots Q but a little bit, but so does adding some mass on the cone. Did you try that?
The measurement of two towers with 4 drivers each shows that the SPL drops about 15 dB from 250 Hz to 40 Hz. However that was when I had the towers placed on stands (as shown in the pictures). I now have them placed on the floor and the curve then flattens out with bass boost from room acoustics (I don't have any measurement result of that though). I haven't bothered much to try to flatten the curve as I always use XT32 calibration anyway which fixes the pretty much all frequency anomalies. I'm considering connecting my miniDSP though to manually flatten the frequency response before applying XT32.

The driver I use was recommended by Calvin on this forum and he mentioned that the Qts should be between 0.4 and 0.6. Maybe he can elaborate further about that :) I took a lot of advice from him before the build and a some inspiration from one of his projects.

A quick translation of the text above my measurement graphs:

First graph, title: Mätning: frontbaffel vs låda(Measurement: front baffle vs box)

This measurement shows the difference in SPL for two drivers in just a front baffle (purple curve) compared to in a U-box with 20 cm walls (torquise curve). In general there is 5-7 dB more SPL with the U-box. The downside is a nasty resonance around 333 Hz, so the crossover frequency should be a bit below this.

Second graph, title: Mätning: dämpning (Measurement: damping)

You might wonder if it's worth putting damping material inside a U-box. The measurement shows the difference between no damping material (torqiuse curve) and damping with low density material (sheeps wool) (purple curve). There is a substantial damping of the resonance around 335 Hz (about 6 dB), but below that the effect of damping is small.

Third graph, title: Mätning: dämpning och lågpassfilter 250 Hz (Measurement: damping and low pass filter 250 Hz)

This measurement shows the effect of damping (sheeps wool) when using a steep low pass filter at 250 Hz. Because the box resonance is about 335 Hz the damping doesn't give any improvement. I crossover at 250 Hz to the electrostatic panels so I don't use any damping in my boxes.

Fourth graph, title: Mätning: harmonisk distorsion(Measurement: harmonic distortion)

The measurement is performed on two boxes with 4 drivers and shows about 3% distortion at 40 Hz. Above 70 Hz the distortion is below 1%. I crossover to the subs at 40 Hz (the lowest crossover possible in the receiver).

Fifth graph, title: Mätning: efter XT32 kalibrering(Measurement: after XT32 calibration)

This is how the frequency curve looks before (purple) and after (blue) calibration using Audyssey MultEQ XT32. It's easy to build speakers when you have such good calibration software :)
 
Last edited:
Member
Joined 2012
Paid Member
Tone Tubby Red for my OB Build?

Thanks to Lynn Olson for this wonderfully thorough review of the Tone Tubby Red http://www.nutshellhifi.com/MLS//MLS6.html I’ve yet to hear any open baffle speakers but I’m very anxious to build a pair of them. But how compatible the Tone Tubby will be for my particular situation is the big question.

It’s really sad and just as surprisingly that there’s next to no one making reliable, good sounding high and midrange band ESL drivers for DIYers. There’s certainly none in the USA that I know of, at least who do it on a continual commercial basis. Please correct me if I’m wrong on this. Otherwise, thank goodness for
Rob Mackinlay’s 505 mini panels Mini Panels There are happy 505 users at audiocircle.com and some at diyaudio.com. And most of them report excellent results with the 505s in open baffles. The trouble is that I’ve never found anyone at these or other forums who’ve used 505s in OBs with Tone Tubbys. Sure, there are woofersthat work well in OBs with the 505s, but none that I know of are Alnicos, and its clear to many that they do have a unique and pleasing sonic signature.

But would Rob’s 505 ESLs and a Tone Tubby woofer actually be compatible?

Why Rob doesn’t post their impedance at website is strange. But assuming the 505s’ nominal impedance is 8 ohms, and using the 8 ohm of the TTs, the only other parameter required to insure their compatibility would be their sensitivity, correct? Well, there we do have the numbers, but not very good ones for me. Rob says the 505s are 88db. But Lynn Olson said that the Tone Tubby Red is 97db, though I read 96db here Tiny Alnico Open Baffle and 100db here Red Alnico 12" 8 Ohm

Thus, even with an averaged Tubby sensitivity of 97.6db, unless two-way active crossovers are used, there would appear to be a big problem with blending the levels of these two drivers-a 9.6db difference.

On the other hand, as pointed out in the wiki report on ESLs Electrostatic loudspeaker - Wikipedia, the free encyclopedia , being line source drivers, the 505s’ SPL decreases by 3db per meter, while-comparatively-the SPL of point source drivers like the TTs decreases by 6db per meter.

So even if they are far from electrically compatible, does this mean that the 505s and the TTs are at least more acoustically compatible? Perhaps by an acoustical difference of only ~ 4.8db rather than an electrical difference of ~ 9.6db?

However, as that ~4.8db acoustical/9.6db electrical level difference between the 505s and the TTs would still be quite audible-and since I don’t want to bi-amp these speakers-what manner of attenuation do I impose on the TT’s that will cause the least damage to overall sound quality?

Are L-Pads the only way? And what penalties will I pay for using them?

I also should point out that my amplifier is only 25 wpc into 8 ohms and 40 wpc into 4 ohms. Damping factor is 40. On the other hand, my room is only 12 ft x 13 ft. And my ears can’t tolerate average SPLs more than 77 or 78db-if that much- at ~11 ft, at least in a room that small. Finally, if whatever OBs I build won’t do down below 65Hz, my pair of Jim Salk’s version of Brian Ding’s Rythmik servo subs will take over from there.

Please understand, that I do have a basic technical school understanding of passive filter design and I can follow about 70% of what most diyaudio.com members converse over. But I am no design engineer and have no math skills above precalculus.

I’m really hoping that Rob’s 505 ESLs and the TTs can be a great match for my first venture into the OB speaker world. But in case there is not a sonically “non-destructive” solution that you can suggest for my situation, please try to suggest another 12 or 15” 8 ohm woofer with an Alnico magnet. And one with a high Qts, making it a good bass response performer in an OB. Thank you.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.