Diy planarspeakers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Now I have a new and better model!![URL=http://img339.imageshack.us/my.php?image=dsc01581d.jpg]
An externally hosted image should be here but it was not working when we last tested it.
[/URL]
This one has 2 "coils",and is 2 way.
I have a coil in the largest "coil",And it works .
my.php

The membrane is app. 15x118 cm.I cross at 100 hz. but could go lower.
 
Here pictures
URL]

URL]

I am now working on crossover ,and i look at Magnepans models.They are crossing at very different values.Some at 600 Hz.others at 1500 Hz.And not the same for bass and tweeter.
I have to tune the membrane to app.45-50Hz.,but it seems that overnight it is stretching.It might be the tape slipping .For the next i will use gaffatape or simular types
Have your any exprience with tape,evt.for ESL models.
I am crossing electronicl at 80Hz.24db.

Bernt "båndsei"
 
Last edited:
Looks real good.
I think most of the larger Mag's have an additional ribbon for tweeter. Are all the two way models without ribbon? (Otherwise that might be the difference in crossover freq. as the ribbons will not go down to 600 Hz).

As for tape, I think many people uses 3M's double sided tape.
Have you considered to glue it down, and maybe use something addtional in the construction to clamp it down further?
/Baldin
 
The rubber is a good idea, as it will dampen the resonanced in the diaphragm. But using a double sided foam adhesive ... something like this would do a double job at both fastening and dampening.

Good overview of the different Mag's .... I think what they call quasi ribbon is the ribbon placed on the side as tweeter.
I think I would either make such a ribbon for the tweeter, or make a separation between the tweeter section and the bass section. What Mag also does is run a panel like yours with two separate windings, one for bass and one for mid, and a separate ribbon for tweeter.
Does you panel go all the way up to 20 kHz, in the current configuration?

I think you are getting near a really usable result. How does it sound?

Happy New Year
Best regards Baldin

PS. Just found this picture .... one of the most revealing :)
 

Attachments

  • Magnepan 001.jpg
    Magnepan 001.jpg
    786.8 KB · Views: 148
Last edited:
That looks really good and promising.
Is the dip around 3.5 kHz a sign of the crossover region?
Then you should try to make a bit more overlap ... e.g. add a little to the high pass cap.
Also I would try to lower the cross over freq to something like 1 to 2 kHz.
How close do you make the measuremtenst ... 1m?
If so, try to move the mic to some 2-3 m distance and make an in room measurement.
I think it is quite difficult to get a good measurememt on a big planar.

Super work. Congratulation. :up: :up: :up:
How about the sound??
Best regards Baldin
 
My messurements are made 1.5-2m.distance,and my latest crossover is like this:
URL]

I have tried to alter the values and it does not affect the dip much.At other gating settings it looks different regarding the dip at app.3,5 khz.
I realy like the sound .There is a lot of details and room.My subbs helps from 80 hz.and they contributes to the"roomy" sound .
I have ordered mylar 6my.and this is not flexibly,so i might keep the tension to maintain f.res.at 50 hz.It has now dropped to 35 and 25 hz.
 
Last edited:
Hi båndsei

I think you have a big phase problem.
Make some measurements from each driver individual without filter and post back here.
Then make measurements from low and high individual with filter and the combined measurement and the combined should always be above the individual lines.
My suggestion is to make 12dB filters and change the polarity from the tweeter.
If you can make a txt file from the individual speakers with magnitude and phase you can sent them to me and I do some simulation for you.

Rob
 
My strange correction was to get more levet at the top.
This one is made mith 0.85mh. in series with basspart and 10uf in series with 3,3 ohm and tweeterpart.
[URL=http://img8.imageshack.us/i/billede425.jpg/][IMGDEAD]http://img8.imageshack.us/img8/9712/billede425.jpg[/IMGDEAD][/URL][/IMG
Bernt.
 
Well now the 3 k Hz region looks better.
The tweeter section still looks a little high in level.

One question you could ask is maybe whether it is necessary to make it two way!
Try to make a measurement of the tweeter only, without filter .... would it make it down to your sub?
Running two different voice coils on the same membrane, must cause some unwanted mecanical interferences between the "two driver"!

Mag, uses this two way design as they want it to be planar in the full range and not a hybrid.

Best regards Baldin

PS. I have used the same kind of tweeter lift (cap//res) in my Seas center speaker ;)
 
Last edited:
I have been following this project from now and then, very interesting!

The rubber is a good idea, as it will dampen the resonanced in the diaphragm. But using a double sided foam adhesive ... something like this would do a double job at both fastening and dampening.

If you need more tension, I think that tape will not hold for long. Magnepan glue their Mylar without any dampening.

Good overview of the different Mag's .... I think what they call quasi ribbon is the ribbon placed on the side as tweeter.
I think I would either make such a ribbon for the tweeter, or make a separation between the tweeter section and the bass section. What Mag also does is run a panel like yours with two separate windings, one for bass and one for mid, and a separate ribbon for tweeter.

That Magnepan overview is not all correct. Magnepan has real ribbons only in the most expensive three way models. The Quasi Ribbon is just a normal Magneplanar style planar driver where the round wire is replaced by a flat one. The cross sectional area is the same in most cases, just some small variation for different models as with the round wire ones. Yes, Magnepan often uses a gap between the two drivers that are sharing the some piece of Mylar, but it is only an electric gap not acousticly. The Mylar is not uniformily tensioned and is also sectioned with the help of several supports (where the diaphragm is held tight), the resonances are spread.

That picture of Jim Winey behind the Mylar seems to be from a swedish magazine. Do you what magazine? The picture is not showing a working Magneplanar driver. They are manufactored in a different order. They begin with glueing the magnet bars to the sheet metal, glue the spacers on the sides, streches the Mylar and glue it to the spacers, glue sheet metal on top where the spacers are and finally, they glue the wire to the Mylar. It never looks like in the picture of Jim Winey!


Roger Gustavsson
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.