advertisement


Ergo Number Nine - Adventures in Micromonitor Design

I just couldn't find a pair of stands to suit the E-IXs. So I did the next best thing and built a pair for them out of matching Sapele timber. The uprights were cut from 50mm stock and machined to around 41x45mm, then rounded over. I used nothing fancier than dowels for the joinery and doubled-up 18mm MDF for the bases. I guess edging the MDF base with Sapele could be considered a bit fancy.

6572204-lg.jpg


Compared to the E-IIIs, the E-IXs are positively puny. The nines have a midbass the same size as the III's midrange, except the former needs to cover more than 6 octaves compared to the latter's three and a bit octaves. I expect it would have to work a fair bit harder, but offer a touch more coherence than the 3-way, Time will soon tell. They don't have voice yet.

With the stands now made, I can measure the drivers properly supported as they would be listened to. My original plan was to measure them on bar stools, but they proved to be far too unstable. I can't make any excuses with custom-built stands. At least they look the part. Tomorrow, the PC take a trip to meet the calibrated microphone for an intense session of gated pink noise bursts.

James
 
I'm sure I won't be the only one who finds the measurement phase fascinating. It would be really cool if you could document that part just as you have the rest of the project. I guess you're using software like LspCAD, but other details such as test rig, procedure etc would be awesome.

Cracking project :)

Carl
 
Yep, LspCAD it is Carl. It's a cracking little piece of software that comes bundled with JustMLS for taking measurements. I'll have to remember how to hook it all up and try not to blow anything. The first few times I used the test set-up I kept shorting the amp.

I'll wait for the day to warm up a little. Spring is a little chilly today.

James
 
I'll wait for the day to warm up a little. Spring is a little chilly today
Click to expand...

And on this side of the globe the sun has set on a rather chilly autumn day.

Great shot of the E-III & E-IX, James ;)

Mr Tibbs
 
Something is not right. The 15W is showing an fb (resonance peak) at around 120Hz when it should be closer to 60Hz. Its bass response is also tilted up, and looking as if the Qtc is well over 1.2. That would happen only if the T/S parameters were completely different to the published specs, or that I had put the driver in a much smaller box than it was designed for. I have triple checked my calculations and I'm certain the 15W needs only 9L for a target Qtc of 0.7. The box itself has 8.8L of volume, so I'm a bit baffled.

The only thing I can think of is that the driver is brand spanking new and never been run in anger. It is possible that the suspension is still so tight that measurements are somewhat meaningless. What I'll have to do is hook the pair up with the 15Ws and run them full range for a few days to sort them out. I'll then remeasure them again.

On the bright side, whilst I was measuring the E-IX drivers, I took the opportunity to remeasure the E-IIIs with both the original CA15RLY and the much vaunted M15CH001 mids. I now have the data to model a new E-III crossover to use a better midrange and see if I can address Mr Tibbs' concern the midrange performance is not quite as beguiling as the Epos ES14. For the outcomes of that exercise, I'll start a separate thread.

James
 
James said:
Something is not right. The 15W is showing an fb (resonance peak) at around 120Hz when it should be closer to 60Hz. Its bass response is also tilted up, and looking as if the Qtc is well over 1.2. That would happen only if the T/S parameters were completely different to the published specs, or that I had put the driver in a much smaller box than it was designed for. I have triple checked my calculations and I'm certain the 15W needs only 9L for a target Qtc of 0.7. The box itself has 8.8L of volume, so I'm a bit baffled.
James
Click to expand...

James my modeling agrees with you so I suspect it is either completely different T/S than published (do you know of any independent measurements to check?) or it needs running in. With such a huge discrepancy though, I'm thinking the published T/S are incorrect :(

Are you certain it is not the measurement conditions such as floor bounce? Woofer height of around 71cm?

Sorry not much help.
 
James said:
.......address Mr Tibbs' concern the midrange performance is not quite as beguiling as the Epos ES14. For the outcomes of that exercise, I'll start a separate thread.

James
Click to expand...
Had the same feeling with E-III vs Dynaudios. Didn't find that with E-V though but even they have changed (freed up, more bass) since I started using them full time.

As to you E-III plans, all I can say is - about time.

Steve
 
I took the opportunity to remeasure the E-IIIs with both the original CA15RLY and the much vaunted M15CH001 mids. I now have the data to model a new E-III crossover to use a better midrange and see if I can address Mr Tibbs' concern the midrange performance is not quite as beguiling as the Epos ES14. For the outcomes of that exercise, I'll start a separate thread.
Click to expand...

James, good on you, and I'll be very interested to see where this leads.

Actually, not that long ago I heard a pair of big JM Labs Nova Utopia (bloody expensive 3-way) things in full flight. I expected them to beat the IIIs (in the mid especially) but they really were not a lot different IMHO. They did give some more midrange information, but the extra information did nothing to enhance the enjoyment of the music to my ears. This is what sets the ES14 aside from almost any other speaker I've ever heard - the ability to connect you directly to the music. That Robin Marshall bloke is a bloody genius.

Had the same feeling with E-III vs Dynaudios. Didn't find that with E-V though but even they have changed (freed up, more bass) since I started using them full time.
Click to expand...

When did you last hear ES14s, Steve?

Mr Tibbs
 
Mr Tibbs,

The key to bass to mid integration is seamless phase tracking. Based on the fresh data I got with the CA15RLY, it was not surprising to see where they might be a bit wanting. I have since modelled the M15CH001 to track phase-wise with the Scan-speak 25W to perfection. I'll show the comparative graphs when I'm home later tonight.

James

P/S: The magic of the ES14 may also be attributable to the absence of an electrical filter on the midbass.
 
When did you last hear ES14s, Steve?
Click to expand...
Not sure I ever have but possibly some years back at the local Niam dealer when I heard some different Epos creations. My comment was more related to when I tried the E-IIIs at home. I just felt that the Dynaudios were a bit better than the E-III at conveying that final bit of midrange emotion.

The Dyns were not entirely consistent however so where they excelled with a lot of music they were just averagely good on other pieces. The E-V has proved a lot more even handed and to my ear have a more compelling midrange that the E-III. A couple of times I tried to convince James he should build a speaker combining the E-III bass driver and the E-V mid and tweeter.

Steve
 
I ran in my Scanspeak 00s by putting 75Hz into them for 24 hours... Have you tried doing an open baffle impedence measurement?

PS I like the idea of using horn loaded tweeters to space the driver. A side effect should be less cabinet diffraction. Naim use the open cell foam grills to achieve the same effect, I guess.
 
Richard Nichola said:
I ran in my Scanspeak 00s by putting 75Hz into them for 24 hours... Have you tried doing an open baffle impedence measurement?
Click to expand...
No, I didn't think about that until I had unrigged my measurement set-up. I played them tonight with just the 15W connected unfiltered, and there was a distinct change in character within the first 10 minutes. There is also a surprising amount of low bass coming from the little boxes. I don't have a signal generator, unfortunately. I'll juice them with a drum recording over the weekend to take them close to their 6.5mm Xmax, just before I measure them again.

PS I like the idea of using horn loaded tweeters to space the driver. A side effect should be less cabinet diffraction.
Click to expand...
I chose the horn-loaded driver precisely to ensure close to zero driver offset. The downside is a bit of a rising response around 1kHz, which I hope to eliminate with a bit of contouring. In the worst case scenario, I'll machine up another pair of baffles to take a conventional tweeter. MDF is cheap.

James
 
I chose not to recess my midbass 15Ws (they are 00s as I was going for a true Kan alike in a smaller cabinet, so you get a bigger Q uplift. The 01s are 00s with a smaller magnet.)

My starting point was the B110:

Vas 23.6 Ltrs
Qts 0.3288
Fs 37 Hz
Eff 87 dB
Xmax 3 mm

So in a 6.5 litre box a B110 gets a 0.707 Qb. and an Fb of 79Hz.

If you put a 8531K01 in a 6.5Litre box, you get a Qtc of 0.8, which is too high for a Kan, so that meant switching to am 8531K00 which gives a Qtc of 0.6ish, which I raise by using a cheaper inductor. (Much easier to raise Q than reduce it...)

If I was starting over again, I'd use a bigger box and possibly 01s though.
 
Richard,

The 9-ish litre for the -01 variant is about right for a Qtc of 0.7 and Fb of 60Hz. I tried the trick of using lower Q drivers and higher DCR inductors in the E-Vs, and that seemed to do the trick in terms of getting the driver to behave like a higher Q driver. However, I'm not entirely convinced if the extra impedance added is helpful from an amplifer damping perspective. But this is just a gut feel.

Did you have any difficulty keeping the sticky cone clean? I understand Scan-speak has released an updated 15W/8531-01 to replace the 8530 variants. The only difference is an uncoated cone, which looks better in my view.

James
 
The key to bass to mid integration is seamless phase tracking. Based on the fresh data I got with the CA15RLY, it was not surprising to see where they might be a bit wanting. I have since modelled the M15CH001 to track phase-wise with the Scan-speak 25W to perfection. I'll show the comparative graphs when I'm home later tonight.
Click to expand...

I've just spotted the new thread - will study it closely!

The magic of the ES14 may also be attributable to the absence of an electrical filter on the midbass
Click to expand...

I think that's got a lot to do with it. Add in the two-way advantage, plus the custom designed drivers and you've got a very tough act to beat.

The Dyns were not entirely consistent however so where they excelled with a lot of music they were just averagely good on other pieces. The E-V has proved a lot more even handed and to my ear have a more compelling midrange that the E-III. A couple of times I tried to convince James he should build a speaker combining the E-III bass driver and the E-V mid and tweeter.
Click to expand...

James, Steve,

I guess the point I was (rather clumsily) trying to make is that it just might not be possible to get ANY three-way to match what the ES14s deliver in the midrange. The E-IIIs are not particularly lacking in the midrange - comparing very favourably with those big JM labs speakers and they are no duds AFAIK. If you can develop them further then please do, but it's no shame on them if they can't approach the 14 midband.

Anyway, time to hand this thread back to the E-IXs!

Mr Tibbs
 
Mr Tibbs said:
I guess the point I was (rather clumsily) trying to make is that it just might not be possible to get ANY three-way to match what the ES14s deliver in the midrange.
Click to expand...
I can hear a gauntlet falling to the ground. :D

The fundamental reason why multiway loudspeaker systems fail to engage in the same manner as a full-range, single driver design is - in my view - a phase problem. A 3-way has two areas of overlap, a 2-way has one and a 1-way has none. If the phase alignment is spot on (at least between woofer and midrange), then the 3-way should have a huge advantage with lower IM distortion, better bass control and improved dispersion.

In the past, I had sought to achieve a ruler-flat frequency response using CAD software. I've now come to realise that precise phase alignment is possibly far more important. In a perfect world with perfectly behaved and aligned drivers, both FR and PR would be perfect. Reality is somewhat less than perfect, but that shouldn't stop me from trying.

OK, back to the E-IXs...

James
 
That's quite significant, but not to the degree that James is experiencing.

Sounds like he's got a pair of mids by mistake.:eek:
 
You must log in or register to reply here.


advertisement


Back
Top