Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
184.58.110.124
In Reply to: RE: Neolith has posted Excel Spread Sheet for magnepan speakers posted by ghost4man on April 19, 2017 at 14:00:38
Then look to add the ribbon at least at 5khz and above as that is the lowest freq at which a reasonably priced ribbon can withstand the signal at high levels corresponding to what the T 1D tweeter can do.
In this case, I would go for a 1st order symmetrical XO (passive or active) anywhere in the 5-8 khz range, depending on the power handling of the ribbon you are looking for within that XO range. The other option (what I do right now) is to apply the ribbon for the top half of the top octave without a low pass on the midrange (T 1D tweeter in your case)... That allows more coherence while obtaining the ribbon's fine high frequency extension, and puts down the dreaded "head in a vise" problem of high frequency crossovers. The high pass range in this case would be 12-18 khz, 1st order.
Follow Ups:
I wonder about QR versus regular round WIRE.
It would seem that if you used a certain size wire, that Squishing it flat to a ribbon shape would leave you with the same resistance? Mass of wire per length would be the same as would resistance.
The benefit would be that it sticks so much better to the substrate.
Too much is never enough
Yes
It sticks better damps the membrane better and ages better. All good things.
Thanks guys for that.
2 dedicated 32amp circuits all wired and up.
Will make a big difference as I was having major power issues.
Do we know what the XO and slopes are for the 3.7i and 20.7 as some sort
of guide??
I think Neolith has incorporated the .7 model generation into his spreadsheets or in a separate one, I just don't remember which.
That is not really relevant to your situation because the drivers are very different in structure and size. If you went with the popular Aurum Cantus G2 ribbon then it would be a monopole tweeter that you would be using more as a supertweeter integrated into the top octave rather than the long large maggie ribbon which is a dipole and can be used as a classic tweeter down to 2.5khz.
Hmmmm
A bit of a concern but I isolated both tweeter panels - that is the old tweeters - and
did an impedance test.
I found both panels measured 2 ohms each.
Now I thought that each should be 4.8 ohms.
Can someone clarify what the impedance is on the tweeters on a pair of Tympani 1Ds.
Ozzie
As far as I know, the tweeter of the I-D should be about 4.7 Ohms. The tweeters of the I-C and IIIA, are about 7.6 Ohms. Have your tweeters be re-wired?
Have your tweeters been exchanged for a ribbon or been rewired? That is not the spec impedance. Is your sock off so you can see the driver?
Satie and Roger,
This is a concern to me.
Okay just so where are on the same page I have the following.
The T1DS consisted of 3 panels. There were 2 panels which accounted for the bass and were either linked in series or parralelled. This was cut off at 1100Hz.
The other panel was the tweeter. The tweeter went from 1100 to 20k.
Now in my case the base panels were rewired BUT in QR and NOT the traditional wire. They measure approx 4.25 a piece. This meant they were linked in series.
Now the tweeter panel was rewired. The right hand channel had the mylar replaced as well due to overwhelming damage. So both tweeter panels were rewired.
Unless you have seen the T1D tweeters you wont know that the drivers themselves have to be separated to see the wiring. Magnepan do not recommend doing this. In fact Magnepan do not do any repairs and simply send you out a new one. The repairers however managed to work on them.
I am not aware of the wiring on these panels. So I dont know if they have double runs or not. My thinking right now is that they have not used the right thickness wire or they have not completed double runs etc.
What I can say after repeated testing is that both drivers measure 2 ohms.
Now what are the implications in terms of XO filters.
I have set it up where I can go active or passive.
In a passive setup I have a bank of capacitors which measure in total 35.1 uF.
From what I can see the single inductor has been replaced by one that is double the size of the original. I believe the original meaured .7(?)
My feeling at the moment is that the rewiring on the driver was not done correctly and more importantly the filters are not correct in line with the respective impedances.
Not sure what else to say. Any ideas on what the caps and inductor should be in a passive setup given the nominated impedances and does anyone have a photo of a stripped tweeter panel to show the wiring.
Cheers Ozzie
Did they write the value of the inductor on the bobbin?
The cap was obviously intended for a 4.8 ohm tweeter to produce a 1khz high pass. so I suspect they rebuilt it with a QR wire. You will need to double up the capacitor to 70uF to get to 1.1 khz.
You can measure the bass freq response to determine if it rolls off around 1-1.1 khz as in the original or at 2.2 khz to match the current high pass capacitor.
Alternately you can just go active now rather than when you get your supertweeter project done.
Just be sure your amp can handle the 2 ohm tweeter load. Many can't and many that can will not survive long doing so.
Hi Satie,
Just to further clarify, the bass panels ONLY were done in QR.
From memory they used the QR that you would find on the 20.7 mids.
I will get the exact sizing to confirm.
The tweeter however was rewired. It was not done in QR. That much I do know.
My feeling is that they perhaps they have not done double runs or maybe they used thinner wire.
To date I have NOT been happy with going active. Its probably fair to say that has been exacerbated by amps which simply cant deal with the impedances.
My thinking was that the tweeter should be measuring around the suggested 4.8 ohms. This however is not the case.
Given the inherent problems with amps I am going back to passive with the view to adding the appropriate filters when I add the true ribbon. This will obviously require a HP and LP filter to cut the tweeter off and then kick in the supertweeter.
The existing caps are 35.1uF and the new inductor is double the old one.
How the tweeter has ended up being a 2 ohm load is highly problematic.
Will have to call back the repairer to get clarification.
Thanks for you help.
Ozzie
Satie,
On the basis of the impedance measurements that I have stated, that is:
2 basis panels linked in series @4.25 ohms = 8.5 ohms total
and tweeter panel 2.0 ohms
What should the inductor and caps be if the XO remains at 1100Hz?
Ozzie
Didn't realize the measures were for the individual bass panels. For the 8.5 ohms the inductor would be at 1.12mH at 1.1 khz. 1.35mH for 1khz.
L=R/(2 Pi Fc)
I think Roger's guesses at the wiring choices made in your restoration is very likely. If so the wire in the tweeter is possibly 20.7 QR foil and the wire in the bass is the older 1C equivalent.
The midrange QR foil of the 20-series is 0.84 Ohm/m. If they had them connected as two loops in parallel, it would result in a very low impedance, 1.8 Ohm or so. I have that foil in my Tympani IVa midrange driver. Sure it would be better to have a single loop of that foil in the tweeters. A pity that you have "degraded" your I-D.
Satie and Roger,
I can confirm the following was used.
When the Bass panels were redone they did NOT use the original wire.
Rather they used the following:
MG1.7i MID/BASS Foil - 0.06 x 0.005
Each bass panel ended up being 4.25 ohms a PIECE and when linked in series gives you a total of 8.5 ohms.
Now the intention was to do the ORIGINAL TWEETER in the following:
MG20.7 M/R Foil - 0.1 x 0.0005
This did NOT occur for the tweeter. So the tweeter was redone in WIRE - NOT QR.
I am getting a colleague to come over to test using his multimeter to see if arrives at the same figure as what I am getting.
The only conclusion that I can come to is that the wire that was used was NOT the same as the original OR they have not done enough runs etc.
Either way they do not measure 4.7 ohms and instead at this stage measure 2.0 ohms.
So right now the existing passive XO with a cut off at 1100Hz has 35.1uF and has doubled the inductor compared to the original.
Clearly this isnt what it should be.
Satie,
Reading into what you have said would it be fair to say that at a cut off of 1100Hz the inductor should be 1.12mH.
Furthermore if we assume that 35.0uF was required for a 4.7 ohm driver to cut off at 1100 then would a 2.0 ohm driver require [4.7/2.0] x 35 = 82.25uF?
Do those figures seem right?
Cheers Ozzie
Simply C=1/(2 Pi Fc R) for the high pass
For the 35.1 cap and 4.7 resistance that gives a HP of 965 hz.
For the 2 ohm a 1khz high pass would be 79.6uF and for 1.1khz it would be 72.3uF
have you any idea what wire they used and whether it was double run or single run?
Satie,
Embarrassingly things soured at the end which resulted in the panels and the rewiring taking a different direction.
I know that the tweeter was done in wire. As to what wire I simply do not know. Its difficult because the driver itself is sealed. Magnepan advise against any attempted repairs on the T1D tweeters. They simply replace.
I know for sure it was not done in QR.
I am half tempted to pull the driver apart to see. They got it apart initially to replace the mylar on the RHS driver so I know separating can be done.
A colleague of mine is going to his friends place tomorrow who has a pair of 1Ds and will test the impedance on the drivers and report back.
Its probably fair to say that at the end of the job they just wanted to finish and get it out. Now I have this problem.
Am wondering what the overall impedance would be given the 2 ohm load on the tweeter and 8.5 ohm load on the bass panels.
I still intend upon adding the true ribbon. But this has set me back because a 3 way active system now has 2 drivers with 2 ohm loads to drive which will nose bleed most amps out there.
Currently having some really good success with a Lexicon CX-5 which seems to be really stable at the 2 ohm load.
The power demands for the tweeter are not that great so the Lexicon is well built and well designed and should do well. You can also look at old multichannel Theta amps for the tweeters/mids. Driving the bass panels is a different issue. If you drive them in parallel you get a 2 ohm load and you can use them as a 2 ohm speaker with fine tuning of relative levels. wiring the bass panels in series you would not be able to make a single amping option with a simple crossover.
Before you give up on the renovated tweeters, see what they sound like when biamping and go to your friend's with the factory T 1D and try and compare the performance of the tweeters, They may be wired for 2 ohms but that does not mean they are necessarily done in that heavy a wire. But for your purposes in context of employing a supertweeter you are probably less concerned with the tweeter's high freq performance. You would want to see how well resolving they are for mid/tweeter use and decide where you want the crossover to be in order to spec the ribbon, You might need to buy a more expensive tweeter that can operate effectively down to 4-5 khz and can take a 1st order XO. Like this Aurum Cantus G1 https://www.parts-express.com/aurum-cantus-g1-ribbon-tweeter--276-420
Vs, using a supertweeter like this cheap ribbon https://www.parts-express.com/fountek-neocd10-ribbon-tweeter--296-701
Hi Satie,
During the testing phase with my other 3 amplifiers I actually blew the true ribbon in the 3 way active set up.
At that point I pulled the pin on doing any additional testing until I could source an amplifier that was capable.
It was at that point that I accidentally stumbled across the Lexicon.
One of the good things about it is its claimed stability into a 2 ohm load. Now being stable AND being musical dont necessarily mean the same thing.
Thus far the testing with the Lexicon has been really really impressive.
In the absence of the true ribbon I have been going 2 way active with the Lexicon dealing with the bass duties and tweeter duties.
I havent gone for anything extravagant in terms of XO. By this I mean I have retained the intrinsic XO point at 1100Hz with a 6dB first order slope.
Now one of the bigger differences has been the complete elimination of ALL the PEQs that I had in place from previous testing. In other words they have been bypassed. I have gone for the minimalist approach at the advice of some colleagues.
Results thus far have been very very good.
By comparison, the previous setup with the other 3 amps were dreadful. So much so I reverted back to the passive setup.
This has been a big wake up call to me as to the importance of having an amplifier which isnt just sufficient in terms of brute strength but more importantly can deal with 2 ohm loads.
To my mind there is no point having a 10k amplifier that has all the bells and whistles and is stable into a 4 ohm load but bleeds out when presented with a consistent 2 ohm load.
When you say:
"The power demands for the tweeter are not that great so the Lexicon is well built and well designed and should do well."
I am not so confident about that. There is still a lot of territory that the Lexicon has to deal with in terms of frequencies from 1100Hz and above.
Dont forget my other amplifier only had to deal with the true ribbon from 3-4k and that resulted in a bad outcome.
I have looked at the Theta Dreadnaught which seems to be a weapon of an amp in its own right whether for musical listening or HT.
I dont quite understand your comment about the bass panels.
In the active state they are linked in series which means the amplifier will a single speaker with an impedance of 8.5 ohms.
If I go down the path of keeping the passive option I can either retain the series setup or go parallel. The latter gives me an impedance rating of just over 2 ohms which in itself is another problem because that would mean in the existing setup that I would have the existing tweeter sitting on 2 ohms as well as 2 bass panels presenting a 2 ohm load.
So when you say:
"wiring the bass panels in series you would not be able to make a single amping option with a simple crossover. ?
Can you explain what you mean here because that is precisely what was done.
Just remember the original setup for the T1Ds was 2 way with a first order 6dB slope cutting off at 1100Hz.
If you maintain the intention of keeping the cutoff at 1100 as well as a first order slope then would it not be the case that you adjust your cap an inductor to suit? Or does it get more complicated?
Now if I decide to add the TR into the mix its a game to see at what point I cut off the existing tweeter.
Right now my gut is telling me that 5k would seem appropriate.
I have had a look at the historical values and it seems the earlier versions were kicking into gear around the 3k mark. My understanding is that the 3 series maggies all used the later TR and were still cutting off at 3k??
So in a 3 way active setup I would be looking at those cutoffs. Thats my gut feeling at the moment. I am building a Holton Precision Audio amplifier here in Australia which is rated at 800wpc into 4 ohms with unconditional stability into a 2 ohm load. I could use it and the Lexicon for the drivers and feel very confident about their ability to handle the lower impedances.
Ideally I want the option of passive and active especially for resale purposes. Not everyone is inclined to want to buy a set of speakers which would require 6 channels of amplification.
I am looking at contacting a guy over here is Oz who is highly regarded and has worked on ESLs a lot. I've been told he may have similar experience with maggies but I will have to speak to him to find out. More importantly he has the reputation of being very astute in terms of XO design.
Have to wait and see though. I am finally starting to get some decent results. Its been 12 months of getting constantly pushed back with one drama after the other.
The Lexicon is simply built to deliver the power and has appropriately sized power supplies and cooling capacity as do the older linear Theta amps (they are class D now). Theta amps employ no feedback so can't rely on that to iron out their distortion and departure from linearity so they are built to meet specs with raw power - power supplies with headroom, As many output devices as necessary and then some and ample cooling.
My idea was that using the 2 ohm tweeter with the 8.5 ohm bass is not going to work because both drivers are matched in sensitivity (at least OEM but probably pretty close) but due to differing impedances all the power in a passive would go to the tweeter through the crossover region, so the actual transition to tweeter dominance would be happening much lower, That might explain why your restorer chose to raise the size of the bass inductor while not using the original crossover freq with a bigger cap, but letting the high pass go up to 2.khz. So that left a spaced XO with the LP about 1khz and HP at 2khz.
So that is one sort of solution for the passive speaker level crossover. The other is to run the bass panels parallel and get a giant Crown Macrotech series amp built for the role (rated to 2 ohms stable to 1 ohm.). Or one of the classic heavy metal amps that were designed to drive the Apogee Scintilla.
The PEQ is not exactly the most transparent way to do a crossover. Costs allot in detail and tonal texture and imaging. Not surprised you liked the passive better.
Once your ribbon is repaired you should look at something around 5khz if it is a maggie tweeter. With 1st order slopes it can't handle a crossover below 4khz. The 3.x and 20.x had 2nd order high pass at 2.5khz till the .7 models came up but the latter use nested series crossovers so that the top drivers are both protected from overload with an additional pole besides the 1st order mid/tweeter XO. I do something like that and use a line level 6db high pass for the mid/tweet amp and then a 1st order for the ribbon at speaker level.
So if I understood what your restorer ended up doing, then just keep those values for passive operation and integrate a proper mid low pass and ribbon tweeter high pass to match. But for passive operation I would want the tweeter to have a 2nd order HP in this setup as everything over 1-2khz would be 2 ohms and that is going to restrict the amp options of the potential buyer so I would not want them to use a distressed amp on a tweeter operating at the extremes of its capacity.. .
Satie,
I will undertake some extensive testing via minidsp and REW and do some
frequency sweeps this weekend.
I already have the graphs from previous testing so it will be very interesting to see the comparison.
I want to make the following point re PEQs.
The PEQs were invoked via the minidsp.
Once I inserted the Lexicon into the chain I removed ALL PEQs and simply added the XO point along with the first order slope.
Thats it.
This combination with Lexicon has been a huge improvement to what I was getting before.
This has really confirmed to my mind the importance of an amplifier which has unconditional stability into a 2 ohm load when coupled with maggies.
The difference in sonics between the Lexicon and the other amps couldnt be greater.
I am so far very very impressed with the apparent brutish strength that the Lexicon has. It really punches way above its weight category.
The Lexicon gets very littl air time in Australia but is obviously well regarded in your neck of the woods.
My ideal outcome is however to have the capacity to switch from PASSIVE to ACTIVE. I want that option in case I wish to sell.
I have taken on board your recommendations in terms of XO figures. Its clear having the 2 ohm tweeter along with another 2 ohm super tweeter plus the series linked 8.5 ohm bass panels presents an interesting channel for a XO setup.
I think the original XO setup offered a lot of simplicity and obviously worked given the respect that the 1Ds still maintain to this day.
The Lexicon has offered a lot of potential in both the top and bottom end.
There is far greater extension into the bottom end to the extent that I almost feel like using a pair of subs which I would cut off at 40Hz given just how good the bass is at the moment.
The Lexicon is experiencing no issues in the HF and at greater than normal dB levels does not struggle with the upper frequencies. Absolutely no sibbilance at all. The previous brightness that was evidenced before has disappeared.
I think that the single amp passive crossover setup your restorers came up with is a good solution for resale.
The DSP makes it all very easy to biamp and I presume you got a set with more channels for triamping.
The fortuitous acquisition of the Lexicon amp appears to really have impressed you with the capacity of a solid conservative design that is carefully voiced for sound quality rather than bench test numbers. Though I always found those to make an obvious difference in SQ, many stick to spec reading as the sole guide to amp selection.
Satie,
Absolutely mate.
My focus is ultimately in adding the true ribbon as a supertweeter and I also think that a 5k cutoff on the existing tweeter would be prudent.
To me that makes it easy for the next person, if there ever will be one to acquire my setup, to simply run a single amp for a 3 way passive setup.
My minidsp was custom built by a good friend of mine. Its an 8x8 board which has a whole heap of balanced and unbalanced inputs and outputs.
It can facilitate a 4 way active setup which includes subs.
I think its been done very well and will post a link for it. Its amazing what you can achieve for this type of build.
The Lexicon has been superb. I am so over the smoke and mirrors that seems to pervade this entire industry and which it has to be said has been fueled by audiophiles who only look at data points.
The bottom line is that if you have maggies then an amp which can handle the heat that a 2 ohm load is what I first look at. I have seen so many amps where you pay serious money that bleed out the harder the load becomes.
This is a definite keeper. I am yet to confirm but I think that the current tweeters are actually 1.9 ohms. When you think the type of real estate that is covered by this tweeter from 1100Hz and above that is a mean effort for the Lexicon to cover as a permanent load. It not only hasnt skipped a beat but gives spades in SQ. I would describe it as a very honest amp which should be considerd by any maggie owner out there.
Ozzie
I think that the single amp passive crossover setup your restorers came up with is a good solution for resale.
The DSP makes it all very easy to biamp and I presume you got a set with more channels for triamping.
The fortuitous acquisition of the Lexicon amp appears to really have impressed you with the capacity of a solid conservative design that is carefully voiced for sound quality rather than bench test numbers. Though I always found those to make an obvious difference in SQ, many stick to spec reading as the sole guide to amp selection.
Hi guys,
As some of you are aware there appears to be an ambiguity in terms of the measurements that I am getting for my T1D tweeter panels.
I have put a multimeter across the wires which yields a figure of 2.25 ohms.
This is independent of the caps and inductor.
Does anyone know what the tweeter driver should measure BY ITSELF for this speaker as I am going to go back to the repairer and seek resolution.
I received an email today from him which contained the following information:
"Regarding your concerns with the tympani tweeter impedances , I saw these being rewired and it is nigh impossible to get the number of runs wrong as the magnets dictate where the wires run and that is a fixed parameter .
The wiring used was the same as the original as there is a colour code on these very thin wires indicating the gauge .
Could it be that some other parameter is confusing the reading ?
Also keep in mind that a multimeter - set to Ohms - does not read impedance , it merely gives a DC resistance reading . It is a very confusing aspect of audio theory that both resistance and impedance are measured in Ohms but are quite different parameters .
Impedance is a load consisting of a combination of inputs , ie. Capacitance , inductance , frequency , and , good old Pi (3.141...-.) ."
I understand that each speaker as a whole is rated at 4 ohms impedance.
However what does the individual tweeter driver measure?
I am getting the distinct feeling the repairer is seeking to avoid fixing if there is an issue.
I want to first make sure I have my measurements correct before I go back.
If someone has a pair of T1Ds and can measure the tweeter drivers I would appreciate any feedback.
Cheers Ozzie
As you now have increased the mass of the bass wiring, it is of no surprise that they go lower. The fundamental resonance is lower in freqeuncy and probably the peak in the bass is larger. It means the diaphragm will hit the magnets earlier.
Edits: 04/27/17
Roger,
Absolutely mate.
I've had an opportunity to hear the 20.7s and 20.1s so I know what the bigger maggies are like and I can honestly attest to my personal experience being that the bass on these speakers is superior to the aforementioned ones.
I have listened to 20.7s coupled with a pair of DWMs with some highly rated amplifiers which struggled under the load and was not overly impressed.
I live 5 minutes away from the hifi shop operated by the national distributor of maggies in Australia so I have heard ALL the current maggies that are available.
I think the best way to describe the differences in the range is that as you move up so to speak the next speaker gives you a bit more of the same thing as the previous iteration.
So in other words a 20.7 and a .7 will offer you that wonderful, spatial almost atmospherice imaging that is signature to maggies but the 20.7 will give you so much more of it which is more palpable to your hearing.
I do not believe they used the midrange QR foil of the 20-series. It is very thin and certainly way off for basses. I guess they may have used the bass QR-foil of the 3.7, it is about the value you mention. Tweeters can use the midrange QR foil of the 20-series but I will come out 7.6 Ohm or so. That is a higher mass than the original I-D tweeter and on par of the I-C tweeter. I understand Magnepan have runned out of the AWG 36 wire originally used (4.7 Ohm). If you have 2 ohm on the tweeters, I think they used a far thicker wire, maybe the midrange wire of the 3-series? The I-D should be 4 Ohm for the bass (two drivers in parallel) and 4.7 Ohm for the tweeter. This can easily be calculated from the wires used by Magnepan. Now you seems to have a higher mass version of the I-D. In other words more like an I-C.
The bass was probably easy enough. But from experience, 2 ohm loads make plenty powerful amps bleed.
What do you have on hand in amplification?
I was guessing that they used a heavier QR for the tweeter rather than the original thin gauge wire. The off spec impedance in the tweeter is cause for concern but judge by ear to see if it is ok functionally and do a FR sweep..
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: