Tony_J

Tannoy 8" Dual Concentric active build

Recommended Posts

Moderator
Finally hooked the new speakers up to their respective amps. Just using the old Dirac settings at the moment - but with the low range output of the DDRC-24 split at 300 Hz using the miniDSP 2X4 so the new bass units are now doing duty below 300, the 8" cone of the DC driver taking over up to 1500. Haven't done any measurements at all so far. However, they already sound significantly better than the old single driver setup. I somehow think that the gap in the bass performance has now been fixed - playing my go-to test album, Yello Touch, the bass is definitely deeper and cleaner. 
 
Probably ought to do some proper measuring at this point...but I may have to listen to a bit more Yello first :rockin:
  • Like 3
  • Upvote 2

Share this post


Link to post
Share on other sites
Super Wammer

Speechless admiration.

  • Upvote 2

Share this post


Link to post
Share on other sites
Moderator
On 22/11/2019 at 17:28, Tony_J said:

the 2 pairs of bass drivers will be powered from an additional Icepower 125 ASX2. Should be more than enough power for the job.

Grrr. The 125 ASX2 was bought used on Fleabay, and works absolutely fine for the first 1/2 hour, then the thermal protection kicks in and the amp cuts out...then it cuts back in a bit later...then a few minutes later, out again. Not good. So a refund request is winging its way towards the seller - we'll see. May have to bite the bullet and order a new board. Ho hum.

So, as a temporary measure, I dusted off my Audiolab 8000A to take on the bass duties. The first thing that is noticeable about the Audiolab amp is that its noise floor is significantly higher (worse) than the three Icepower amps - the latter are much cleaner - must be the nasty old linear PSU in the Audiolab, whereas the Icepower amps all have nice clean on board SMPS... :whistle:

Anyhoo, off we went again. Until I noticed that I was getting A LOT of background noise through the Audiolab amp. While pondering why this was, and debating whether to toss the whole lot out of the window, I was idly scrolling the web page that was up on the browser, and noticed that the nature/volume of noise coming from the Audiolab through the bass units changed according to what was on the screen. Weirder and weirder. My lappy is connected to a larger desktop monitor - an old-style VGA - via a HDMI to VGA adapter so I get to use 2 screens. The only connection between the audio system and the lappy was via a USB cable to control (one of) the miniDSP units - the audio path is from SBT via optical interface. So, unplug the USB to the miniDSP and hey presto, the noise goes away. Plug it back in, and it reappears. Unplug the HDMI-to-VGA adapter (which is powered direct from the HDMI interface) and the noise goes away...

Intriguingly, the noise was appearing only on the analogue amp, not on the two Icepower 50's that were doing mid/high duty. So I dusted off one of the diminutive Nobsound amps and its SMPS and re-jigged everything - Icepower 50s for  bass and mid, and the Nobsound for treble - and tried again. No nasty HDMI noise on any of the amps, regardless of whether the miniDSP and/or the HDMI-to-VGA adapter were connected.

Go figure.

Anyhow, looks like I have an amplifier combo that will at least allow me to see what this setup is capable of.

Watch this space...

  • Upvote 1

Share this post


Link to post
Share on other sites
Moderator

OK...had time now to set the crossovers properly and do a quick Dirac setup.  Definitely an improvement in the bass department, which isn't entirely surprising given that I have doubled the bass cone area and separated off the midband from it. Currently crossing over to the bass at 300, and from mid to treble at 1500. Overall impression is of a smoother result - the bass in particular isn't having to strain so much to produce the desired result,so it sounds much cleaner and more precise. The mids also seem to benefit from the change - I guess attempting to make one cone do bass and midrange duty is always going to be a compromise. I will do some more playing with Dirac target curves, etc. to see if further gains can be made, but even if they end up staying as they are now I will still be well pleased with the result.

The other change that I need to make at some point is to load up Dirac V2.0 which I have been putting off doing, but Ed says that it makes his system sound better, so it has to be done!

  • Upvote 2

Share this post


Link to post
Share on other sites
Super Wammer

Try moving the upper crossover point in 100Hz chunks to 2500, searching for a sweet spot? Then refine in 25Hz chunks.

  • Upvote 1

Share this post


Link to post
Share on other sites
On 26/11/2019 at 13:37, Tony_J said:

Grrr. The 125 ASX2 was bought used on Fleabay, and works absolutely fine for the first 1/2 hour, then the thermal protection kicks in and the amp cuts out...then it cuts back in a bit later...then a few minutes later, out again. Not good. So a refund request is winging its way towards the seller - we'll see. May have to bite the bullet and order a new board. Ho hum.

So, as a temporary measure, I dusted off my Audiolab 8000A to take on the bass duties. The first thing that is noticeable about the Audiolab amp is that its noise floor is significantly higher (worse) than the three Icepower amps - the latter are much cleaner - must be the nasty old linear PSU in the Audiolab, whereas the Icepower amps all have nice clean on board SMPS... :whistle:

Anyhoo, off we went again. Until I noticed that I was getting A LOT of background noise through the Audiolab amp. While pondering why this was, and debating whether to toss the whole lot out of the window, I was idly scrolling the web page that was up on the browser, and noticed that the nature/volume of noise coming from the Audiolab through the bass units changed according to what was on the screen. Weirder and weirder. My lappy is connected to a larger desktop monitor - an old-style VGA - via a HDMI to VGA adapter so I get to use 2 screens. The only connection between the audio system and the lappy was via a USB cable to control (one of) the miniDSP units - the audio path is from SBT via optical interface. So, unplug the USB to the miniDSP and hey presto, the noise goes away. Plug it back in, and it reappears. Unplug the HDMI-to-VGA adapter (which is powered direct from the HDMI interface) and the noise goes away...

Intriguingly, the noise was appearing only on the analogue amp, not on the two Icepower 50's that were doing mid/high duty. So I dusted off one of the diminutive Nobsound amps and its SMPS and re-jigged everything - Icepower 50s for  bass and mid, and the Nobsound for treble - and tried again. No nasty HDMI noise on any of the amps, regardless of whether the miniDSP and/or the HDMI-to-VGA adapter were connected.

Go figure.

Anyhow, looks like I have an amplifier combo that will at least allow me to see what this setup is capable of.

Watch this space...

Interesting about the 8000A there. Probably not relevant (and apologies if this sounds a bit like how to suck eggs), but the PS caps in the 8000A can degrade badly over time. I replaced the old 10,000uF Elna caps in mine (I think they were Elna) with Mundorf Mlytic 15,000uF, and there was a big difference. I tested the old ones after they were removed, and they measured at only around 4,000uF each.

Anyway, I love this thread and the project, thanks for sharing it - my cap is doffed to you. I really do like the finish on those cabinets too, and the oak edging does finish it off beautifully.

Edited by Oscroft
  • Like 1

Share this post


Link to post
Share on other sites
Moderator
55 minutes ago, Oscroft said:

Interesting about the 8000A there. Probably not relevant (and apologies if this sounds a bit like how to suck eggs), but the PS caps in the 8000A can degrade badly over time. I replaced the old 10,000uF Elna caps in mine (I think they were Elna) with Mundorf Mlytic 15,000uF, and there was a big difference. I tested the old ones after they were removed, and they measured at only around 4,000uF each.

Anyway, I love this thread and the project, thanks for sharing it - my cap is doffed to you. I really do like the finish on those cabinets too, and the oak edging does finish it off beautifully.

Thanks for the kind words! I'm very pleased with the finish - amazing what you can do with some spare building material!

I re-capped the Audiolab amp about 7 years ago, so I doubt that's the problem. I suspect it is just less immune to noise than the class D designs - after all, they have to deal with all sorts of noise issues as a consequence of how they work.

  • Like 1

Share this post


Link to post
Share on other sites
16 minutes ago, Tony_J said:

Thanks for the kind words! I'm very pleased with the finish - amazing what you can do with some spare building material!

I re-capped the Audiolab amp about 7 years ago, so I doubt that's the problem. I suspect it is just less immune to noise than the class D designs - after all, they have to deal with all sorts of noise issues as a consequence of how they work.

I guessed you probably had the cap thing sorted B|, and I'm sure you're right about the noise. My Audiolab-based system is in the same room I work, which means it's in close proximity to a bunch of computing devices powered by various bricks and warts, and I'm convinced they're introducing noise back into the local mains ring. I now have a cheap Tacima mains strip for the audio components, and I was surprised by the improvement it made.

  • Upvote 1

Share this post


Link to post
Share on other sites
Moderator
8 hours ago, Oscroft said:

I guessed you probably had the cap thing sorted B|, and I'm sure you're right about the noise. My Audiolab-based system is in the same room I work, which means it's in close proximity to a bunch of computing devices powered by various bricks and warts, and I'm convinced they're introducing noise back into the local mains ring. I now have a cheap Tacima mains strip for the audio components, and I was surprised by the improvement it made.

Of course, back when the Audiolab was made, there weren't any SMPS or PCs to generate mains-borne noise, so they probably didn't design it with that possibility in mind.

Edited by Tony_J
  • Like 1

Share this post


Link to post
Share on other sites
Moderator

OK...back to the main event. I've been playing with crossover frequencies for the Mid to High crossover, with interesting results. For these tests, I chose frequencies of 1500 (that was what I used in the old setup), 2000, 2500, and 3000 Hz. I briefly tried 4000 and 5000 Hz, but discarded them because I just couldn't get it to sound at all decent. Mid to bass crossover was 300 Hz in all cases.

With just the "raw" crossovers set, i.e., no additional filters, and with a bit of playing with levels on the bass and mid amps, the best sounding combination was crossing over at 2000, and sounded pretty reasonable for a first attempt. I decided to see what effect the Dirac filtering would have on the four settings, so with the mic positioned mid-way between the speakers, and not changing its position for all 9 measurements that Dirac needs to take (very tedious!) so the results would be comparable, I applied Dirac correction to all 4 crossover settings, using the Harmann target curve (this gives bass boost and treble cut). End result: Pretty much impossible to tell the difference between the four crossover settings. Measuring the Dirac-corrected settings gives a family of 4 curves that are also pretty much indistinguishable.

Before/after traces with 2000 Hz corssover:

Effect of DIRAC - 1

All 4 curves showing the performance after Dirac correction:

Effect of DIRAC - 2

Same as above but with all four "before" curves added:

Effect of DIRAC - 3

Not surprisingly, what little difference there is in the "after" traces is greatest over the region where the crossover frequency is changing. The green (2000 Hz) trace seems to give the flattest result in that region, so for the moment, I will go with that. However, what this seems to show is that the choice of crossover frequency becomes much less critical if you are applying Dirac correction

Anyhow, general impression so far is that the overall sound is much improved from the 2-way build. More/cleaner bass is the main difference (and was the reason for doing it in the first place); not surprisingly, the mid/high performance is similar to the 2-way build. There also seems to be significantly more headroom! :rockin:

Edited by Tony_J
  • Like 1
  • Upvote 1

Share this post


Link to post
Share on other sites
Moderator

More playing with the filter settings today. I finally took the plunge and loaded up Dirac V2.0 - its a free upgrade from miniDSP if you have one of their Dirac-enabled products that supports 2.0. It involves installing a completely new Dirac Live app on the PC, a new miniDSP DDRC-24 "plugin" app on the PC to communicate with the DDRC-24 board and the Dirac app, and new firmware in the DDRC-24. The order of processing is a little different to V1.0, and the Dirac app has had a complete makeover (for the better). Interestingly, the crossover/filter settings and Dirac V1.0 filter sets that I had loaded into the DDRC-24 seemed to survive the upgrade, but on the Dirac side, my old Dirac V1.0 projects are incompatible with V2.0, and they have given the target curve files a new extension (used to be .txt, now .targetcurve) and V2.0 stores them in a different folder under the user name of the user currently logged into Windows. Oh yes, and V2.0 is only supported under Win 10. 

V1.0 had a handful of pre-defined target curves, including the Harman curve that I have used a fair bit in the past; V2.0 doesn't seem to come with anything other than its in-built default curve, which is a slight pain. After poking around to find where the old target curves were stored, I opened one up - the structure is pretty simple, the curve is defined by a list of frequency and dB pairs ordered by frequency, along with a header and a couple of hi/low limit values, all in text. So, I copied the old files into the new location and renamed their extensions, and hey presto, they are now all available as target curves for use with V2.0.

I ran a measurement set using the Harman curve and loaded it up into slot 1 of the DDRC and had a listen; all sounded pretty good. Unfortunately, by this point I had cleared down the old config in the DDRC-24 so I was unable to compare old vs new directly, but hey ho. Next step will be a bit of playing with target curves to see where I can get to with different profiles. It occurred to me that it ought to be possible to take the FR curve of a speaker from one of the online reviews and translate that into a target curve to see whether the end result resembles the other speaker in any way - I have set up a profile based on the FR curve of the Quad ESL 2912 & I now have slot 2 of the DDRC-24 populated with filters based on that - not surprisingly, audibly different to the results with the Harman curve (look online for the Stereophile review - the FR curve resembles parts of the Himalayas!). I also have a Proac Response 2.5 FR plot that I plan to convert into a target curve tomorrow and see how that sounds. All good fun.

Share this post


Link to post
Share on other sites
Moderator

Had a good session this evening, listening at decent volume levels. :rockin:

Very happy with the way they are sounding now - punchy bass, lots of detail, and the imaging seems to have been improved as well. Hearing stuff in familiar tracks that I hadn't noticed before - always a good sign. Will be interesting to see what people make of them at Kegworth!

  • Like 1
  • Upvote 3

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.