Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
220.240.101.96
In Reply to: RE: Going active on Tympani 1Ds posted by Satie on January 17, 2017 at 11:13:48
Satie,
The intrinsic setup with a pair of Tympani 1D's is a 2 way driver system.
You have two bass panels which work as one. This is an important point which I think was getting missed in some of the earlier points.
This means that there is no lower bass panel coupled with a mid bass panel at another XO frequency.
When I had these bass panels rewired with QR, with a passive setup the capacitor range was approximately halved and the inductor coil was doubled (approximately) to reflect the fact that the panels were now 4.2 ohms and when linked in series gives you 8.2 ohms.
So the bottom end of the bass panels starts anywhere from 25-35Hz. Remember this is not a conventional Tympani 1D panel and simply listening to these versus a conventional setup most assuredly resulted in multiple listeners claiming how much lower the speakers were getting to at that bottom end.
The stated XO point between the bass panels and EXISTING tweeter panel in the conventional setup is 1100Hz. So that tweeter panel covers the spectrum FROM 1100 TO the end of the HF range ~ 20kHz.
I have gone from that 2 way system to the addition of a pair of version 2 ribbons for the tweeter. The version 2 ribbons is what you will find in the later maggies - 3.6,3.7,20.1,20.7. They are physically different in size compared to the previous true ribbons although it is a very small difference.
So I now have the option of going 3 way AND active.
My feeling at the current minute is to go from 80Hz TO 1100Hz for the bass panels where the previous tweeter panel which is NOW the midrange to kick in. Now are you saying that the midrange panel would cover the spectrum FROM 1100Hz where it crosses over to the ribbons at say 5000Hz? Is that what you are saying.
Can you see now Satie what I am trying to do?
Follow Ups:
Yes, I understand what you are doing. You are going to make this into a 3way +subs. The original tweeter is going to become the midrange.
I offered 2 approaches to doing this. 1, as you are now using the DSP with LR4 crossover (I would suggest you try Bessel 4th on the DSP at some point). Where I suggest that you can run the T1D "mids" down to appx 500hz and the tweeter down to 5khz and if you are not really playing hard perhaps even 4khz.
2. leave the DSP out of it for the "mid" and tweeter, using it only for bass low pass to subwoofers while the bass panels are kept 1st order high pass and low pass done at stock XO values and to match the subwoofer (which probably would mean that the subs would be run at Butterworth 3rd order for LP) - run the high pass at 1khz 1st order and the ribbon at 6-8khz 1st order.
2a. Alternately.you can let the mids run without a low pass and kick in the ribbons at 12-14 khz 1st order in order to reduce "head in a vice" problems.
Satie,
Correct. At the minute the original tweeter is NOW the mid range.
So what I was going to do then is basically follow the previous setup whereby the the TWO bass panels deal with frequencies FROM 80Hz to 1100 where it crosses over to the NEW mid(formerly tweeter driver) and then cross that over to the true ribbon at 3000.
From what I can see you are saying then to let the TWO bass panels cover from say 80Hz(this can vary depending upon where I cut the subs off at) and then cut them off at 500Hz where the mid kicks in.
My question here is will the mid(old tweeter driver) be able to deal with going down to 500Hz? Clearly it had no issues from 1100Hz to 20000Hz as that is what it was primarily doing before.
So if I read correctly just thinking about XO points only and disregardind slopes it would read something like this:
2 subs(mono)- 20Hz-80Hz
2 Bass panels - 80Hz-500Hz
Mid range panel(old tweeter) - 500Hz - ~5000Hz
True Ribbon - Any frequency above the cutoff in the midrange so from 3000-5000Hz.
Is that a correct summary Satie just focusing on XO points for now only.
Yes those are my suggestions for LR4 or B3 crossover slopes. (not for 1st order as the strain on the mid would be too much)
500 Hz for tweeters may be a hard task, they have the thinnest wire of any Magneplanar type of driver. My old Tympani IIIA had the slightly thicker AWG 35 wire and came without fuses. As far as I remember they had resonances below 400 Hz or so. You could use the thinnest foil as used in the midrange drivers in the 20-series. That would increase the impedance to 8 Ohm for the tweeters. I use that foil on my Tympani IVa midrange drivers, which are even lower moving mass than the Tympani I-D tweeters. The T-IVa mids are far better to fill in between the basses and the ribbon driver than your I-D tweeters.
I would try to use the Tympani I-D without subs. I-D is really not that bad down to 30-35 Hz.
Roger,
Did some experimenting last night. Will start to plot some more graphs from REW.
I tried 500 for the mid and it just didnt sound good at all so went back to 1100.
I then tried 2200 for the ribbon and found it was a bit bright which was especially noticeable when it came to females singing at a higher pitch. It also sound tinny. It just didnt sound good.
When I bumped it up to 3000 it was a lot lot better. I am still getting a drop off from 10k onwards.
There are a lot of things one can play with on the minidsp plugin especially with the PEQ.
So when you lowered the XO to 500 hz the mid sounded strained and distorted or was it something else?
Satie very much so mate. Did not sound good at all.
I played around a bit and set the XO at 1200 for the mid.
For the ribbon I set it at 4000.
I didnt do any listening at these levels but rather relied upon REW and graph reproduction.
This is by far the most level graph I could get.
XO points are accompanied by 36 L/R.
I tried Bessel, 12db, 24db, 48dB and the "best" graph was with this configuration.
Could I get it smoother? Who knows. But the sound was quite pleasing.
Very clear and very revealing compared to before. Disregard sub 80Hz.
So you ended up with 36db LR as the final choice (for now)? at 1.2 and 4khz?
Results look very good. See if the Bessel sounds better or worse than the LR at the same order. The Bessel should give a less phasey sound - which either does or doesn't matter to you.
Did you run impulse response?
Okay this is the latest graph.
A big big difference from previous testing.
We identified a critical issue which was the ribbon falling down badly.
As a thought we decided to swap out the Quasimodo on the ribbon with the Plinius 8200 which has a lot of muscle.
Surprise surprise it resolved a big issue in that region. Put simply the Quasimodo wasnt dealing with a 3 ohm load
as well as the Plinius.
Whilst its not a perfect representation its certainly far better than where I was at before.
And it sounds damn good as well.
Changed from 36dB to Bessel. Much better outcome. Adjusted the gain in the line levelling and the result got better.
Didn't expect the amp to have that much of a problem driving the tweeter, though the Bryston 4B NRB I used on it for a while ran hot. Calsse DR 9 - closer design to the Plinius - did well and was just a bit warm.
Results look even better now. Did you run the Bessel filter at 6th order as before with the LR filter or did you lower the order?
Anyway, glad you like it.
Satie,
Look I spent many many hours trying every conceivable option I could muster to resolve that issue with the ribbon BUT the reality is it was falling away. Thats clear and unambiguous as indicated by the graphs I was getting.
When I swapped the amp out it made the difference you can see right now.
I am now convinced that I could do even better and across the board with a bit more grunt which clearly maggies are renowned for wanting.
I'm sorry I cant answer your question about Bessel filtering.
Are you familiar with the mindisp plugin? It gives you the option of what you want and all I sighted was 12,24,36,48 L/R and Bessel. How do I run Bessel at 6th order? Or is that assumed when you click onto it??
Appreciate the suggestions. They have helped.
XO for mid was 1200 and ribbons 4000. That seemed to get the best graph.
And trust me I tried every variant in between.
Bot expressing doubt or criticism on the amp/tweeter match just being surprised by the extent of the mismatch you had.
I am not familiar with the miniDSP interface, so from what you said I understand that their Bessel alignment option is for a single order option, I would guess 4th order but that is just a guess.
If you want more "grunt" what amplification is your baseline from which you want to upgrade? Any particular candidates in mind?
Satie,
In answer to your question I was using a custom built an amp called the Quasimodo.
Its rated at 40watts into 8 ohms and about 75 into 4.
It works really well but as we all know maggies need a tonne of power.
Bessel appears as a single order position on minidsp plugin.
I think I can get more out of the midrange with an amp with more muscle as well.
Low end is working quite well.
At the moment the system sounds better to my ears than before.
I have just levelled and adjusted the 2 Dayton Audio 12" drivers into mono. It has made a huge difference at the bottom end.
I have them cutoff at 80Hz but obviously I can play around with this.
My advice to anyone who wants to go active with maggies needs to start with a clear page and try everything. I did. Despite all the advice I was getting in the end that didnt translate to positive audible experiences and the graph reproduction was evidenced of that.
I ended up cutting the bass off at 1200 and then cut off the mids at 4000.
I tried every setting from 2000 to 5000. I tried everything from 12dB to 48dB and different variants in between.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: