diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Planars & Exotics (http://www.diyaudio.com/forums/planars-exotics/)
-   -   Audiostatic baffle step filter (http://www.diyaudio.com/forums/planars-exotics/228821-audiostatic-baffle-step-filter.html)

AAaF 28th January 2013 11:02 AM

Audiostatic baffle step filter
 
2 Attachment(s)
I have a Audiostatic DCM. I bought it some years ago, and have often asked my self what I was thinking. You can hear that there are some qualities, but overall sound is rubbish to be honest. When I started measure, I could understand why. Tried digital correction without too much luck. Then, when I looked at it, it seemed like some baffle step filters might help:

AAaF 28th January 2013 11:07 AM

4 Attachment(s)
I had a old Rotel preamp laying around, who I started to modify.

C9 consist of two 10nF capacitors in series, this makes me able so easily change C9 value between 5nF and 10 nF. 5n gives the lowest dip, while 10 gives a lower center frequency.

AAaF 28th January 2013 11:13 AM

1 Attachment(s)
Here's the result with C9=5nF, this makes the loudspeaker totally new:

MJ Dijkstra 28th January 2013 11:16 AM

Quote:

Originally Posted by AAaF (Post 3345514)
I have a Audiostatic DCM. I bought it some years ago, and have often asked my self what I was thinking. You can hear that there are some qualities, but overall sound is rubbish to be honest. When I started measure, I could understand why. Tried digital correction without too much luck. Then, when I looked at it, it seemed like some baffle step filters might help:

Audiostatics also produced esl with the so called 'mirror drive' which boost the response below 300 HZ. They sound a lot better imho compared with the models without mirrordrive. Technically it is possible to change the esl to mirrordrive by adding an additional transformer. Mid/high can be modified by its segmentation.
So you might be able to avoid any external filters.

AAaF 28th January 2013 11:52 AM

4 Attachment(s)
I remember opening the speaker once, and spotted a resistor in series with signal transformer. I suspected this was to increase response around Fs, to give a false impression of more low end. So I opened speaker and shorted the RC filter inside. This actually rised response at high frequency, while there were no effect at mid and low freq. Red is original and blue is modified with only a 6R8 or nothing at all. As you can see, capacitor inside speaker gives a small rise after 5kHz. If you are to use an active filter, I do think you would benefit from bypassing filter.

With active filter it gets even better. I'm not very experienced in measuring, have one measurement where it looks even better, but I'm not sure if this is valid.

AAaF 28th January 2013 12:00 PM

Quote:

Originally Posted by MJ Dijkstra (Post 3345526)
Audiostatics also produced esl with the so called 'mirror drive' which boost the response below 300 HZ. They sound a lot better imho compared with the models without mirrordrive. Technically it is possible to change the esl to mirrordrive by adding an additional transformer. Mid/high can be modified by its segmentation.
So you might be able to avoid any external filters.

OK, thanks for the tip. Is it possible to obtain any info about how to do this on this older model, and with some measurements to verify?

By the way, tested input impedance by applying a voltage with a signal generator and measure voltage before and after a series resistor of 6,8R. Already at 2Hz impedance was close to 8 ohm, dispute DC resistance of approx. 0,5ohm. So if your amp doesn't put too much DC at terminals, or is very weak, it should be fine to bypass filter.

etalon90 30th January 2013 05:41 PM

I have Audiostatic and feel your pain.
After intensive mods, mine started to really shine and I wouln't let them go anytime soon.
They put a some expensive system to shame.

1- Audiostatic do not use damping behing the mylar and it's very bad.
Acoustat used a kind of 3/8 thick felt pad glued behind stator.
I think Quad did the same but not 100% sure what it was.
I used foam (maybe one day I'll try felt but I had foam around) to cover maybe 70% of the back. I just used pressure fit (no glue or anything)
It help the sound quite a bit.

2-For the dropping response, building wings on both side help tremendously too.
More weight, less vibration from panel and lower rolloff.
My U baffle audiostatic have 12" wing on each side with felt and I used few bracing too.
Impressively, high benefited very much from the side wings.

3-The input resistor in your static is 6.8ohms, mine was factory 2.2 ohms.
I use 0.22ohms now (as a fuse) for a little more spark on the top but still fall short on the 20khz...
I just started to experiment with a supertweeter but still in early stage of experimentation


4-Feets are important too.
I used to use nothing for the worst (concrete floor covered in linoleum)
Spike are a bit better but not much
Soft hockey puck are very good.
I use superball (stupid high bouncing ball from dollarstore) cut in half to do the feets
3 per side and theses feets are yet to be beaten in my system...
I've tried some more feets but nothing sonicaly worth mentionning.

5-For bass, do not dream, you need a sub.
I cross mine at 100hz and do only use a capacitor on the amplifier input.
My subs are much better moving air than the smallish mylar running out of excursion...

6- a bit more hardcore but the audiostatic are segregated panel. Kind of 2 ways if you wish. Method and resistor value seem to change from version to version. I redid mine in 3 parts instead of 2 and got much better (more even) frequency response.
http://www.diyaudio.com/forums/plana...ertweeter.html



Martin

moray james 30th January 2013 06:22 PM

Quote: Acoustat used a kind of 3/8 thick felt pad glued behind stator.

Acoustat only used these felt pads because the diaphragm material that they used DuPont HS65 was heat shrunk to tension and was unable to generate quite enough tension for the s/d spacing and the width of the panel used. The felt added enough resistive air load to help stabilize the panel. With sufficient tension these pads are not necessary. Best regards Moray James.

AAaF 30th January 2013 08:35 PM

1 Attachment(s)
Found the reason why removal of resistor increased output at higher frequencies. Reason is of course that impedance start decreasing after a certain point, then goes to a minimum value of 3,3ohm just above 10kHz. Here are the values if it should be of any interest, sorry for the big hole between 400 - 2000Hz.

Calvin 31st January 2013 05:39 AM

Hi,

Audiostatic had the Bass Boost technique patented. They used a autotransformer with two inputs, one for feeding Bass and one for feeding MH. The inputs were joineded together via coupling parts that shaped amplitude response such that bass got a boost. As every other detail in the construction, the quality of the audio transformer as well as the auto transformer was poor and cheap. The audio transformers hardly reached the quality of standard EI-core power trannies as you can see from any impedance plots (impedance minimum around just 10kHz with a segmented panel!). See US Patent 4,461,931 for the auto transformer patent.

jauu
Calvin


All times are GMT. The time now is 03:34 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2