Tweeter Quest

No I do believe it's possible to make the three ways sound even more disjointed and unbalanced- amazing as that might sound.

 

Then show me the light S&M. That will forestall anything I can say.

 

I'm not sure if its an ATC OEM design or not, looks very similar but then it would, eitherway its an ATC copy so I'd assume similar characteristics.

If you've got the facilities then measuring the drivers will ensure you get the best performance. You really need distortion, amplitude, phase and impedance measurements of the drive units. From here you can get an excellent starting point to work from. After this, try to mate the drivers at a point where they are most similar in respect to measurements whilst avoiding any driver anomolies or deficiencies, if you've got sufficient data, including TS specs, impedance/phase and amplitude plots then it may be a good idea to use a loudspeaker CAD package such as LspCAD, Speakerworkshop etc. I personally use LEAP5 and find it fairly comprehensive. After this you jump into the crossover work armed with the data you have, then listen, tweak, measure and repeat until happy.

This is fairly simplified for brevity but I'm not exactly sure what you can and can't do in terms of measuring the drivers. You should be OK with the tweeter since it will likely have plenty of manufacturer or 3rd party data. The main prob is the mid, if you've taken a look at those ATC measurements and then consider the 'oversights' that ATC have made in their implementation of the unit in their own speakers, then they're could be room for improvement with your mid. Eitherway, working blind with a driver and basing decision on what was done before with a completely different treble, isn't the way to go for the best implementation you could achieve, infact there's a good chance it will sound like crap unfortunately.

 

Orignally PMC used the atc dome - I think the current unit is there own interpretataion but it differs significantly in details of execution. However I would expect the choice of materials and similiar shape to lead to significant commonalities - at least from my inexperienced perspective. It doesnt suffer from the apparent problems with the atc driver.

Shin - thanks for the info, sounds pretty sensible. I should be able to get the info from PMC to go with the tweeter info. I know one of the tech guys there. It might also be useful to get details for the existing tweeter. With something like the raal options are limited as it only goes down to 2khz so I guess 4khz would be about the min crossover point. I will check the viability there first to see how it looks. I wil also follow T's advice and check out the cheaper isoplanar unit he recommended. It might integrate better.

 

No problem.

I should also mention that if you can get your hands on offaxis response plots for both treble and mid that would very useful for ensuring a smooth power response and a more seamless treble-mid integration. A 3" dome starts to become directional at around 2-2.5Khz by 4Khz your looking at a 45degree off-axis response thats around 5-6dB down. I'm not familiar with the ribbons being discussed but my guess is that they'll be fairly directional at that frequency, it may be better cross lower at 2.5Khz where I can virtually guarantee the ATC and the ribbon will have well matched offaxis performance. Remember to try and closely match dispertion characteristics at the XO point to ensure a smooth power response and minimise nasty offaxis lobing.

I don't won't to confuse things further but any self respecting design also takes into the drivers natural roll off and how this contributes to the overall acoustic slope when paired with the high/low pass filter. The mid should be OK at 2.5Khz since it has something resembling a 6dBOct roll off here. The treble is probably going to working a little at 2.5Khz so I'd recommend a steepish slope - if your active then this shouldn't be a problem. Again try to match the driver rolloff characteristics for improvements in intergration and imaging amongst other things.

 

I think Nick wants to use his current crossovers which means it must cross at 3.8KHz and with a 4th order slope. A tweeter that extends to 2KHz flat should be alright Nick. The Raal looks like it has a slight roll-off towards lower frequencies but when listened too off-axis it will be more flat and the size of baffle will change the amount of low frequency reinforcement anyway. If you are worried though Raal do a big brother (mother?) that goes to 800Hz or so but should be crossed at or above 1.5KHz.

I think at this point you need to just get something cheap and have a play. You can only plan so much, things vary in practice a fair bit usually.

I can lend a hand with measurments and stuff if you need me.

 

Didn't realise that was the case.

Well I'd feel awkward about being locked into such a situation but something serviceable is entirely possible. Its very likely that you won't be getting the best from them though but maybe that doesn't matter and providing its an 'improvement' in subjective quality it could be enough.

A good loudspeaker is usually one that has the design details selected and built around a set of compromises and well matched components that are symbiotic as a whole and in agreement to personal preferences. To do this effectively you cannot work around such constraints no matter how large the shoehorn.

I'd strongly recommend ditching the analogue active XO's and running a digital XO such as DCX2496 or DEQX. If not then do as Simon suggested and try out something fairly economical but ultimately I'd abandon the idea, sell what I had and then start auditioning for something I really wanted.

 

I don't really see it being a problem to be honest. The current tweeter extends down to just below 1KHz so we know that there is no acoustic slope from the tweeter taken in to account in the crossover. Any tweeter that extends down about as low will work in that regard. The crossover has the ability to adjust for the drivers sensitivity so we are okay there as well. We just need to match the other things reasonably well, which is pretty easy. Any tweeter with pretty wide horizontal dispersion will work as well as the current one.

If I remember correctly when I took some measurements of Nicks system from his listening position there was a slight dip at the 3.8KHz Xover point suggesting that the directivity of the mid is setting in a little bit. So a tweeter with a slightly more narrow dispersion than the normal dome will probably match even better. The Neo3 PDR has a well controlled off-axis that is a bit narrower than an average dome at about the frequencies of interest. I’m sure many ribbon / planars do as well.

One thing I would note is that your system has the tweeter a bit above ear height doesn’t it Nick? So something like a ribbon with a very narrow vertical dispersion might need aiming downwards a little bit. Not a problem to do with the face plate though.

 

The adjustment range is constricted and you need design by driver selection. I believe in giving anything a go once, so please by all means try it out.

Designing from measurements, theory and convention works well as a foundation, the really good stuff comes from the tweaking.

 

I believe you can adjust the crossovers further by pulling out the modules if need be and fiddling the internal pots.

The tweeters are at ear height or thereabouts. I might buy both the neo and the hivi and give them a go. I dont really want to start again as what I have is already very good and I dont like that many speakers on the market - as an all round package I havent heard anything better (yet) but I am open to suggestions!

I see this as really just a tweak though - at the moment I suspect my system is most constrained by the phono. Everything else is pretty much sorted.

I should add I have a deq as well for fine tuning.

 

Nick I meant try the Neo to test the idea out. If that works then you can decide between then Hi-Vi or the Raal. I expect the larger Hi-Vi (not the round one) will sound better than the Neo considering the price so I wouldn't buy both at the same time.

 

Neo and hivi look like similiar prices so why not go straight for the hivi?

 

Hi

Virtually all analogue active crossovers will have fixed value components and I hope/expect PMC went that route. Achieving a variable frequency analogue active 4th order LR XO that is accurate and sounds decent is far far more expensive than one with fixed values built around a driver set that is never intended to be changed.

Might be a good idea to check so you know where you stand before you swing the axe so to speak.

 

You can adjust level externally on the Bryston crossovers - the external slope adjustments appear to have gone on these pmc custom versions so they may be fixed at 24db / octave.

 

Neo3 is $44 the Hi-Vi RT2-II is $93.. unless you meant the round Hi-Vi which I think will give you a suckout around the crossover point of a few dB, especilly off-axis. Thorsten thinks if that does happen it will help with the type of music you listen to anyway

 

Oh does he now!!!

He hasnt heard any of my record collection yet

Yes I meant the round one - or should I look at other one as well?

 

Well Shellac has become part of my collection too, which I played at his place

I think the round one will work but I would expect a slight dip at xover.

The Hi-Vi I thought you were looking at was the same one as Thorstens, or the one I would choose which is the same but without the waveguide. That’s the one which is $93.

 

I think T meant the round one.

I didnt realise he had heard shellac already - what did he make of "orchestral punk"?

 

Is that a track or the genre? I don't have the track names!

Ask him what he thought, I don't think it appealed.

 

Thats the genre. I suspect not