The trick with Gradient Helsinkis is that bass frequencies from the dipole get reflected from nearby walls and we hear the final room response at spot. They are working only below room's Schröder point. But positioning is work of art and perhaps not succeeding in every room.
https://www.audiocircle.com/index.php?topic=68553.0 All links are not working anymore
Back when I was building fancy Bluetooth speakers I dreamed of making a pendant speaker, maybe with a lamp, or with a matching lamp. I never could come up with anything that would sound decent and wasn't huge. If you could do it, it should be a hit.
Yes! I took them with me to a friends house and both him and his dad were really impressed. Especially since the speakers don't look that impressive 😛 They had both not heard open baffle speakers before. The highs are very acceptable now. Very wide soundstage, not sharp or piercing at all, and nothing's missing there. We added a subwoofer for the lows. Instrument separation is on a whole other level!
I'm not that active on here at the moment but my interest in dipoles is growning and these speakers get more use than I expected.
I took a difficult decision yesterday and removed the electrostatic panels and electronics from a 15yr old pair of ML Aeons.
The panel output had fallen to the point where even with the amp's treble turned fully up, the sound was bassy and dull.
I couldn't afford a new pair of panels (£2k).
However I wanted to preserve the open baffle sound and came across the GRS 6825 mid/treble dipole planar speakers. The impedace curve is a horizontal line at 8ohms so crossover design is simpllfied and initially I have chosen 500Hz.
The Aeons sealed box woofer/mid driver is still using the existing 2nd order 450hz crossover. (2.5mh and 50uf cap)
As the GRS' are very sensitive (94db) I need to experiment with attenuating them. I left space for a second one
I also have a pair of TDL RTLSE but I keep swapping back to the MLs - there is indeed no going back to a box speaker.
https://www.soundimports.eu/en/grs-pt6825-8.html
The panel output had fallen to the point where even with the amp's treble turned fully up, the sound was bassy and dull.
I couldn't afford a new pair of panels (£2k).
However I wanted to preserve the open baffle sound and came across the GRS 6825 mid/treble dipole planar speakers. The impedace curve is a horizontal line at 8ohms so crossover design is simpllfied and initially I have chosen 500Hz.
The Aeons sealed box woofer/mid driver is still using the existing 2nd order 450hz crossover. (2.5mh and 50uf cap)
As the GRS' are very sensitive (94db) I need to experiment with attenuating them. I left space for a second one
I also have a pair of TDL RTLSE but I keep swapping back to the MLs - there is indeed no going back to a box speaker.
https://www.soundimports.eu/en/grs-pt6825-8.html
Attachments
Are you going to wash the ML panels to rejuvenate them? Seems to make some like new. Just read the "how to" first so they aren't damaged. Nothing to lose anyway.
The resistor increases the Q of the driver. Since OB does not significantly increase system Q, which you want to be around .7, and since most drivers have a lower Q than this, adding the resistor might be a good idea. (Low system Q leads to a down sloping low end, hence the shrill sound you mentioned)
Yes indeed, that's a good target. My first successful OB build used 15" Selenium woofers with a Qts of 0.65-0.7 IIRC. I had always been told that 0.7 was accurate sounding because it is critically damped, not too tight with truncated bass, not too loose with flabby resonant response. But at some point I found an article by forum member John K that argued that a Q of 1 is the most accurate in the bass. John backed it up with good examples and a lot of math. Hmmm.....
Thinking about it later, the woofers that I loved so much - in combination with a series inductor, the baffle and of course the room, likely pushed the over all Q up near 1. And it was perfect. So starting with a woofer Q of ~0.7 then adding the other gains can sound very good. Full, strong and accurate all at once
I tried a bigger series resistor but that didn't make it sound better. Series resistor is at 1,1 Ohm and I've added an RCL filter in front of the series resistor & driver.
(+) - (RCL) - (1,1 Ohm) - (Driver)
The inductor was 0,56 mH at first and has now been changed to 0,68 mH. Cap was at 6,8 uF first, then 4,7, now 11,5 uF. The resistor is at 3,3 Ohm (so only a slight correction). I need to get a 0,60 mH inductor. 0,56 is the tiniest bit too shouty, 0,68 is a bit too dull...
Sounds like you could use a miniDSP; why not use a SOTA device, instead of fooling with fractional inductance values?
I know, the passive elements will sound better than chopping up an audio signal, doing math on the numbers, then putting it back together again. Still better to "dial up" a transfer function that sounds good, then figure out what needs to happen in analog to implement the same.
Because a MiniDSP is expensive and an inductor is cheap. Besides, I like working like this. It gives me some intuition on crossover components, hearing how minor changes can have an impact on the balance and sound of a speaker. Also, I don't have any measuring equipment because I don't have a lot of money... So I don't even know what the specs of these drivers are 🤐
Last edited by a moderator:
See https://www.diyaudio.com/community/...e-speaker-quick-assessor.359272/#post-6322655. You can get what a driver does analytically for not a lot of money. It wont show the peaks and dips as sharp as a sine sweep, but for pocket change, not bad.